article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
a growing tree - like network can model different processes such as a technological or biological systems represented by a set of nodes , where each element in the network can create new elements .
innovation and discovery @xcite , artistic expression and culture @xcite , language structures @xcite and the evolution of life @xcite can naturally be represented by a branching process in a tree @xcite describing a wide range of real - life processes and phenomena @xcite .
the general branching process is defined mathematicaly as a set of objects ( nodes ) that do not interact and , at each time step , each object can give rise to new objects .
in contrast , interacting branching processes are much more interesting and difficult for analysis @xcite . a generalized tree with one ( or more ) ancestor(s )
have been used to depict evolutionary relationships between interacting nodes such as genes , species , cultures . besides the interaction among nodes
, one can consider spatially embedded nodes .
the evolution of networks embedded in metric spaces have been attracted much attention @xcite . in this work
we study the evolution of a population , i.e. , the number of nodes in the network , influenced by the interaction among existing nodes and confined to a limited area , representing a competition of individuals for resources .
we assume that the growing tree is embedded in a metric space and we consider that spatially close nodes , previously placed in the network , will suppress their ability to born new nodes . in other words , overcrowding of nodes will drain the resources and supress the offspring . in our model
each node lives for three generations .
the evolution of the population of nodes is actually determined by two parameters : the minimum distance between any pair of nodes @xmath0 , and the area in which the network is embedded , namely the linear size of the area , @xmath1 . for simplicity , we assume that this area does not change in time .
the population evolves in two different regimes . at the initial generations ( time steps )
, one can see an exponential evolution , followed by a saturation regime , after a crossover time . in the saturation regime
, the size of the network will finally approach some limiting value .
the network has even a chance to extinguish if at some moment all its nodes occur in a small area .
we investigated this possibility of complete extinction .
the term extinction for our model implies the end of evolution and the absence of new generations .
the interaction among the nodes inside the radius is defined by a parameter @xmath0 and the value of @xmath1 regulates the population dynamics .
our results show that , under certain conditions , the entire population can be led to extinction .
this paper is organized as follows . in sec . 2 we present our model details and obtain simple estimates for its growth . in sec .
3 we describe the populational evolution .
the possibility of extinction for the model embedded in a bounded space is discussed in sec . 4 , and , finally , in sec .
v , we summarize the results and present our conclusions .
in our model , the population consists of interacting nodes spatially separated by some distance .
we start our process from a single root node at time @xmath2 , as one can see in fig .
the single root node ( black circle in fig . [ fig1 ] ) , can branch to produce up two new daughter nodes ( dark gray circles ) at future generation , i.e. , at the next time step .
the position of each new node is randomly chosen inside a circle with a given _ radius _ ( @xmath3 ) centered in the parents positions .
the attempt to add a newborn node is refused in the case the chosen position is closer than at distance @xmath0 from other nodes .
the attempt to generate offsprings takes place at the next time step after the introduction of a new node in the network and each node can produce daughter nodes only at this time . at the next time step , after three generations , the node is removed from the network . and
new attempts are made each time step . in @xmath4 ,
one can see a refused attempt ( blue circle ) due to the proximity to other nodes ( closer than a distance @xmath0 ) . in @xmath5 ,
the oldest node is removed and new nodes are created . ] at each time step , each of the nodes previously introduced , attempts to branch , so at each time step a new generation of nodes is born .
the nodes are chosen uniformly at random one by one and during a unit of time we update the entire network .
the total area of the system is limited , considering that it is natural the introduction of a spatial restriction into the model .
the first node is settled as the origin of the space and from the origin we set a maximum length for each spatial coordinate of a two - dimensional space .
in other words , the geometric position of each node in the network , for our model , is restricted in the range @xmath6 , @xmath7 .
the linear size of the area , @xmath1 , is introduced as a parameter of the model and we assume that this area does not change in time . in our simulations we used open boundary conditions .
if one lets the population dynamics evolve embedded in a infinitely large system ( @xmath8 ) , the population always increase in size .
the number of new nodes grows very fast as @xmath9 for initial times , and , after certain crossover time @xmath10 , the growth is slower than exponential , as one can see in the fig .
[ fig2 ] . .
the behavior for the initial time steps , @xmath11 , is also exhibited .
data are averaged over 50 samples . ] at this regime the total population as function of the time is @xmath12 , for @xmath13 greater than @xmath10 .
we can estimate , very roughly , @xmath10 from @xmath14 and @xmath15 , we have @xmath16 which leads to the estimate @xmath17 at small @xmath0 . our numerical results are considering that @xmath18 , for the estimates of the total population in the saturation regime .
we should emphasize that in our model the population is confined into a limited area and it is not possible to grow indefinitely .
the general result of our simulations for this model is exhibited in fig .
[ fig3 ] , where we consider a two - dimensional space and a sufficiently small value of @xmath0 , in comparison with the linear size of the system @xmath1 .
initially , the population grows exponentially and , after certain crossover time , one can see that the population reach an steady state .
after the crossover time , @xmath19 is nearly constant .
the maximum value of the population is @xmath20 , since we are considering a two - dimensional space for the simulations .
the growth of the populational density of _ paramecium _ in laboratory , for instance , is reported to have the same behavior of our model @xcite .
one can see that in the saturation regime , the total population is smaller than @xmath21 .
this is due to the fact that the interaction among the nodes does not allow that each possible offspring may be created at some generation , keeping the total population below this limit . and
the inset shows that even for one sample , the population reaches a constant value after some time . ]
for a small population , the possibility of extinction is higher .
the network has even a chance to extinguish if the offspring created are too few and , at some moment , all its nodes occur in a small area .
it is a well known characteristic that the smaller a population , the more susceptible it is to extinction by various causes @xcite .
figure [ fig4 ] demonstrates an example of the evolution of the population , which in this case has @xmath23 and @xmath24 .
the population rapidly increase and the system enters in the fluctuation regime , for which the population fluctuates around a mean value for a few generations . after some time ,
the population decreases and is extinguished . and @xmath24 .
] the picture which we observe agrees with traditional views on extinction processes which show `` relatively long periods of stability alternating with short - lived extinction events '' ( d. m. raup ) @xcite . the competition for resources combined with the restricted space limits the total population and , for some values of the parameters in the model , the population should finally extinct . in real situations ,
extinction may require external factors , an environmental stress @xcite or an internal mechanism , such as mutation @xcite .
one can see an example in the case of extinction of reindeer population in st .
matthew island @xcite .
coast guard released 29 reindeer on the island during world war ii , in 1944 .
the reindeer population grows exponentially and , in 1963 , it was about 6,000 animals on the island , 47 reindeer per square mile .
the overpopulation , limited food supply and the exceptionally severe winter of 1963 - 1964 significantly affect future offspring . the reindeer population of st .
matthew island drops to 42 animals in 1966 and dies off by the 1980s .
this kind of extinction may also occur in branching annihilating random walks and other related processes studied in refs .
@xcite , in which the random processes play the role of an external factor , internal mechanism or an environmental stress that may lead to extinction . in our model ,
if we choose a large enough value of the parameter @xmath0 , the population will be small and the number of new nodes after some generations can decrease and , sometimes , vanishes . when the nodes competition increases , the population may decays or even vanishes , as we can see by the rapidly decreasing of the new nodes in fig .
[ fig4 ] .
we investigated the state of the branching process after a long period of time , @xmath25 generations ( i.e. , time steps ) . for the case when @xmath26 the population always dies off , since no offspring is allowed , for any value of @xmath1 .
we simulated our model for @xmath27 different samples and for various values of @xmath1 and @xmath0 . from this data
we obtain the probability of extinction @xmath28 , i.e. , the fraction of samples in which the population dies off before the 10@xmath29 generation , for different l , as one can see in fig . [ fig5 ] .
generations versus @xmath0 for 100 samples and different values of linear size l of the system . ] in fig .
[ fig6 ] one can see a diagram where an extinction ( below the curve ) and non - extinction ( above the curve ) regions are shown .
each point is fig .
[ fig6 ] is defined as follows . for a given value of @xmath1 and considering @xmath30 generations
, the value of @xmath0 for which the probability of extinction goes to one defines one point in the graphic of fig .
our results show that high values of populational density , represented in our model for small @xmath1 and large @xmath0 , can be lead to extinction .
this picture is a different representation of the probability of extinction in which we are considering the values of the parameters @xmath1 and @xmath0 corresponding to @xmath31 .
plane , where one can see extinction and non - extinction regions . ]
we studied the evolution of a population embedded into a restricted space in which the interaction among the population is determined by the relative position of nodes in space .
our model generates a competition between species or individuals ( represented by the nodes ) . starting from a single root node and , at each time step
, each existent node in the network can branch to produce up to two new daughter nodes at the next generation .
the new nodes are not allowed to emerge closer than a certain distance of a pre - existent node , defined by a parameter @xmath0 , i.e. , overcrowding suppresses the `` fertility '' of population .
evolutionary processes are usually considered in low dimensions and , for this case , our results do not depend qualitatively on the system s dimension for @xmath32 .
we have demonstrated that the embedding of the network into a restricted area , which is natural for general populational evolution , set limits to growth and , for some values of the model s parameters , can result in complete extinction .
the simple model we studied can schematically describe a real process in nature .
f. l. forgerini would like to thank the fct for the financial support by project no .
sfrh / bd/68813/2010 .
n. c. would like to thank the brazilian funding agencies capes and cnpq for the financial support .
this work was partially supported by projects ptdc / fis/108476/2008 , ptdc / sau - neu/103904/2008 , and ptdc / mat/114515/2009 .
30 s. n. dorogovtsev and j. f. f. mendes , _ evolution of networks : from biological nets to the internet and www _ , clarendon press , oxford ( 2002 ) ; s. n. dorogovtsev , _ lectures on complex networks _
( oxford university press , oxford , 2010 ) . | we study the competition and the evolution of nodes embedded in euclidean restricted spaces .
the population evolves by a branching process in which new nodes are generated when up to two new nodes are attached to the previous ones at each time unit .
the competition in the population is introduced by considering the effect of overcrowding of nodes in the embedding space .
the branching process is suppressed if the newborn node is closer than a distance @xmath0 of the previous nodes .
this rule may be relevant to describe a competition for resources , limiting the density of individuals and therefore the total population .
this results in an exponential growth in the initial period , and , after some crossover time , approaching some limiting value .
our results show that the competition among the nodes associated with geometric restrictions can even , for certain conditions , lead the entire population to extinction . |
bar structure as a major non - axisymmetric feature on all scales is important in studying the morphology , mass and light distributions ( e.g. , freeman 1996 ; elmegreen & elmegreen 1985 ; elmegreen 1996 ; elmegreen et al .
1996 ; eskridge et al .
2000 ; menndez - delmestre et al . 2007 ) , star formation ( e.g. , zurita et al . 2004 ; knapen 2005 ; ondrechen & van der hulst 1983 ; regan et al .
1996 ; sheth et al .
2000 ) , gas dynamics ( e.g. , kormendy 1983 ; bettoni & galletta 1988 ; sancisi et al .
1979 ; benedict et al . 1996 ; downes et al . 1996 ; regan et al . 1999 ) and central activities ( e.g. , ho et al . 1997b ; hawarden et al . 1986 ; knapen et al . 2002 ; sakamoto et al . 1999 ; martini et al . 2003 ; sheth et al . 2005 ) of disk galaxies .
theoretical models , including n - body and hydrodynamic simulations , generally confirm that bar formation is spontaneous and ubiquitous in disk evolution ( e.g. , athanassoula 1992 ; sellwood & wilkinson 1993 ; friedli & benz 1993 , 1995 ; athanassoula & bureau 1999 ) . because of the dissipative nature of the interstellar medium ( ism ) , the streaming motions of the molecular gas in and around bar regions can be different from the stellar orbits ( athanassoula 1992 ; regan et al .
1999 ; sheth et al .
2002 ) . due to the delayed star formation after the clouds have been triggered ( @xmath430myr ; vogel et al .
1988 ) , the locations of gas / dust in galaxies can often be offset from that of young stars ( e.g. , sheth et al .
2002 ; phillips 1996 ; martin & friedli 1997 ) .
the molecular gas can be transported from galactic disk toward central region by the gravitational torques from bars ( e.g. , sakamoto et al .
1999 ; sheth et al .
2002 , 2005 ) , and the condensation of gas leads to subsequent circumnuclear star formation ( e.g. , ho et al . 1997b ; knapen et al . 2002 ; martini et al . 2003 ; jogee et al . 2005 ; fisher 2006 ) .
observationally , the gas / dust patterns can often be seen as dust lanes , atomic and molecular gas concentrations , or isophotes of h ii regions with active star formation ( martin & friedli 1997 ; sakamoto et al .
1999 ; regan et al .
1999 ; rand et al . 1999 ; crosthwaite et al .
2000 ; sheth et al .
2002 , 2005 ) . as predicted by theoretical models ( athanassoula 1992
; friedli & benz 1993 , 1995 ) , there is a small position angle misalignment between the gas / dust distribution and the stellar bar , usually of a few ( and up to 10 ) degrees , in the sense that the former is _ leading_. kenney et al . ( 1991 ) found the gaseous pattern is offset from the major axis of the stellar distribution by 24@xmath0@xmath16@xmath0 in m 101 .
crosthwaite et al .
( 2000 ) found that the central gas distribution as indicated by h i map leads the stellar bar by almost 10@xmath0 in the late - type galaxy ic 342 .
similarly , rozas et al .
( 2000 ) identified a large sample of h ii regions in barred galaxy and showed a position angle misalignment of a few degrees exists in h@xmath2 and i - band images .
they also pointed out that the @xmath5-band image of this galaxy shows a bar pattern more aligned with h@xmath2 , further suggesting massive star formation `` at the leading edge of the bar '' .
sheth et al . (
2002 ) found offsets between molecular gas ( co ) and star formation ( traced by h@xmath2 ) in bars of six nearby spirals , which were caused by the gas flow dependent star formation .
understanding the misalignment between stellar and gas / dust patterns and their formation scenarios is crucial for studying the ism properties and star formation processes taking place in environments where gas dynamics are strongly perturbed ( e.g. , regan et al .
1996 ; martin & friedli 1997 ; sheth et al . 2000 ; zurita et al . 2004 ) , and also offers a good opportunity to study dynamical properties and secular evolution of barred galaxies ( e.g. , kormendy 1983 ; benedict et al . 1996 ; regan et al .
1999 ; kormendy & kennicutt 2004 ; sheth et al . 2005 ; kormendy & fisher 2005 ; fisher 2006 ; regan et al .
2006 ) .
the _ spitzer space telescope _
s ( werner et al . 2004 ) observations in the mid - infrared , with its higher sensitivity and better angular resolution than previous observations ( e.g. , _ iso _ ) , provide a new opportunity to study both stellar and gas / dust structures in galaxies ( e.g. , pahre et al 2004 ; wang et al .
2004 ; cao & wu 2007 ) .
in particular , the four infrared array camera ( irac ; fazio et al .
2004 ) bands from 3.6 to 8.0 @xmath3 m probe both stellar continuum and warm dust emissions ( of the so - called polycyclic aromatic hydrocarbon , or pah , and dust continuum emissions ) with identical spatial sampling , thus enabling a powerful probe to compare non - axisymmetric features such as bar structures involving gas / dust and stellar mass .
recently , _
observations of nearby galaxies have demonstrated the importance of using mid - infrared images for studying galaxy secular evolution driven by bar instabilities ( e.g. , fisher 2006 ; regan et al .
2006 ) . in this paper
, we present an analysis of data from _ spitzer _ and sdss of the late - type barred spiral galaxy .
previous studies show that , with an estimated distance of 39.9 mpc ( at this distance , 1@xmath6 corresponds to @xmath4193 parsecs ) and a total infrared luminosity of @xmath7 @xmath8 4.6@xmath910@xmath10 @xmath11 ( bell 2003 ) , [ hubble type sb(s)c ] has a weak bar ( @xmath41.5 kpc ) , with spiral arms beginning at the bar s end but without an inner ring .
this is consistent with the conventional view that bars in most late - type spirals are relatively weak ( erwin 2005 ; menndez - delmestre et al .
2007 ) , and that weak bars tend to produce a sb(s ) type response ( in which the spiral arms begin at the ends of the bar ; kormendy & kennicutt 2004 ) .
the data reduction is presented in @xmath122 , and results on the bar structures in with multi - wavelengths analysis are described in @xmath123 .
possible explanations of the large misalignment between the bar and dust patterns are discussed in @xmath124 .
broad - band infrared images of were acquired with irac on board _
spitzer_. the basic calibrated data ( bcd ) were part of the lockman hole field in the _ spitzer _ wide - field infrared extragalactic ( swire ) survey ( lonsdale et al .
2003 ) . following the preliminary data reduction by the _
science center pipeline , images of each of the four irac bands ( 3.6 , 4.5 , 5.8 and 8 @xmath3 m ) were mosaicked , after pointing refinement , distortion correction and cosmic - ray removal ( fazio et al . 2004
; huang et al . 2004 ; wu et al . 2005 ; surace et al . 2005 ; cao & wu 2007 ; wen et al . 2007 ) .
the mosaicked images have pixel sizes of 0.6@xmath6 and angular resolutions ( full width at half maximum , fwhm ) of 1.9@xmath6 , 2.0@xmath6 , 1.9@xmath6 and 2.2@xmath6 for the four bands , respectively .
the angular resolution of irac 8@xmath3 m images ( @xmath42.2@xmath6 ) is significantly improved over that of pre-_spitzer _ data at similar wavelengths ( e.g. , @xmath410@xmath6 for isocam lw2 at 7@xmath3 m ; roussell et al . 2001 ) . in order to derive the dust - only 8@xmath3 m component ( pah and dust continuum emissions ) ,
we remove the stellar continuum from the irac 8@xmath3 m image by subtracting a scaled irac 3.6@xmath3 m image ( assuming that the 3.6@xmath3 m emission is entirely due to old stellar population ) : @xmath13 where the scaling factor @xmath14 = 0.232 was calculated based on _ starburst99 _ synthesis model ( leitherer et al .
1999 ) , assuming solar metallicity and a salpeter initial mass function between 0.1 and 120 @xmath15 .
this approach has been adopted in several previous works ( e.g. , helou et al .
2004 ; wu et al . 2005 ; regan et al . 2006 ; bendo et al .
2006 ) for studying dust emissions and the 7.7@xmath3 m pah feature based on broad - band measurements , and shown to be effective .
the five - band optical images ( @xmath16 ) and the fiber spectrum for the central region ( @xmath17 diameter ) of were taken from the sdss data archive ( york et al .
2000 ; stoughton et al .
the background in each band was subtracted by fitting a low - order legendre polynomial to it , after masking out all bright sources ( zheng et al . 1999
; wu et al . 2002 ) .
figure 1 shows the three - color image of derived from the sdss data archive .
north is up , and east is to the left , as denoted by the crosshair .
the four - band irac images of are shown in figure 2 .
we find that has two bar - like patterns that are bright in either stellar emission ( at 3.6 and 4.5@xmath3 m ) or warm dust emission ( at 8@xmath3 m ) .
we took the conventional approach of treating the bar as an elliptical feature for measuring its semi - major axis _ a _ and position angle pa .
they can be measured with a fitting routine such as the ellipse task in iraf .
the pas are approximately 36@xmath0 , 36@xmath0 , 20@xmath0 , 15@xmath0 for the bars in the 3.6 , 4.5 , 5.8 , 8@xmath3 m band , respectively .
uncertainties of the position angles are estimated to be @xmath18@xmath0 , derived from the deviation of pas along the major axis of the bar .
taking the mean of the two shorter wavelength bands as representing the stellar bar , we measured that it trails the dust pattern traced primarily by the continuum - subtracted 8@xmath3 m emission ( fig .
3c ) , by a large position angle difference .
spiral arms with bright knots are also visible in the 8@xmath3 m 40@xmath0 ) in the sdss _
g , r , i , z_-band images , and is also spatially coincident , within the measurement uncertainty of @xmath4@xmath18@xmath0 , with that of the irac 3.6 and 4.5@xmath3 m images , but trails the dust pattern at 8@xmath3 m by 25@xmath0 ( fig .
3b ) . however , the bar in the sdss @xmath5-band is quite different from that in the other sdss bands : at pa @xmath19@xmath0 , it is instead much better aligned with the dust pattern bright at irac 8@xmath3 m ( fig .
the deprojected values of the relative length [ l@xmath20 and the misalignment between the stellar bar and the dust pattern [ @xmath21(i ) ] are calculated using the equations given by martin ( 1995 ) : @xmath22^{1/2}}{d_{\rm 25}}\ ] ] @xmath23 - arctan[cos(i)tan(\phi_{\rm a;stellar})],\ ] ] where @xmath24 is the inclination angle of ( @xmath449.5@xmath0 , calculated using the formula given by bottinelli et al .
1983 ) , @xmath25 is the length of the semi - major axis , @xmath26 is the angle between the patterns ( dust or stellar ) and the node lines ( the major axis defines the pa of the galaxy , @xmath4175@xmath0 ) , and @xmath27 is the diameter of ngc 3488 at a @xmath28-band surface brightness of 25 mag arcsec@xmath29 ( @xmath41.86 arcmin ) . , and @xmath30 of were taken from the third reference catalogue of bright galaxies ( de vaucouleurs et al .
these parameters of the patterns are summarized in table 1 .
the angular misalignment between the optical bar and the infrared dust pattern , @xmath31 , is approximately 20@xmath0 after correcting for inclination effect .
uncertainties of these parameters can be as large as @xmath32 , due mostly to simplistic assumptions concerning projection effects and the difficulty in decoupling the bar from the bulge ( martin 1995 ; martin & friedli 1997 ) , especially in the irac 8@xmath3 m image . nevertheless , this angular misalignment , assuming it represents the misalignment between the dust and stellar patterns , belongs to one of the largest misalignments found in previous observations or in typical numerical simulations .
besides the bright stellar bar , a faint and clumpy bar - like structure appears to be present in the sdss @xmath33-band image ( fig .
4 , left panel ) .
we use two different techniques to enhance the visibility of this structure : unsharp masking ( e.g. , walterbos et al . 1994 ) and deconvolution ( fig .
4 , right panel ) .
unsharp masking was made by using a 2-pixel wide ( 2@xmath34 ) gaussian filter .
the deconvolution was performed with the task lucy ( lucy 1974 ) in iraf , using a point - spread function derived from the associated psfield file of the sdss data ; the total number of iterations is 30 .
the pa of this faint bar - like structure is approximately 15@xmath0@xmath12@xmath0 , so it is spatially coincident with the patterns in @xmath5-band and 8@xmath3 m images .
the bar bright at optical ( sdss ) and irac 3.6 , 4.5 @xmath3 m is known to be dominated by old stellar population .
gadotti & de souza ( 2006 ) showed that the bar color index can be used as an indicator of the bar age : old bars are on average redder than young ones .
the sdss @xmath35 and @xmath36 colors of the stellar bar of were compared with an instantaneous burst model calculated based on galaxev ( bruzual & charlot 2003 ) , adopting a salpeter initial mass function and solar metallicity as initial conditions .
this comparison shows that the stellar bar bright in optical and irac 3.6 and 4.5@xmath3 m bands is evolved , with a mean age of the order of 4 gyr .
the pattern bright at 8@xmath3 m is dominated by the strong 7.7@xmath3 m pah feature and warm dust continuum emission from very small grains , while the pattern shown in the irac 5.8@xmath3 m image is probably a mixture of stellar and dust ( 6.2@xmath3 m pah and warm dust ) components . from the correlation between 8@xmath3 m dust emission and star formation activities ( wu et al .
2005 ; calzetti et al . 2005 ) and the previous result that 8@xmath3 m dust and 24@xmath3 m hot dust emission are well correlated on kiloparsec scalesm emission , which is mainly due to hot dust emission from very small grains , is thought to be a good measure of the sfr in galaxies ( e.g. , wu et al .
2005 ; calzetti et al . 2005 ; prez - gonzlez et al .
( e.g. , bendo et al .
2006 ) , we suggest that the dust pattern seen at 8@xmath3 m is associated with young stars and thus represents recent star forming activity . some authors ( e.g. , regan et al .
2006 ; r. c. kennicutt , priv .
comm . ) , however , argue that pah emission is likely a better tracer of the general ism rather than star - forming regions heated by young , massive stars .
the coincidence of the bar shown in the @xmath5-band with that bright at 8@xmath3 m ( fig .
3a ) confirms that there exists a young stellar population in the region of the dust pattern .
the star formation rate ( sfr ) in the bar region of can be estimated using either the h@xmath2 line flux or the 8@xmath3 m dust emission .
the sdss spectrum , taken through a 3@xmath6-diameter ( @xmath40.58 kpc ) fiber , indicates that the central region can be classified as an `` h ii nucleus '' ( ho et al .
1997a ) , one whose main source of ionizing photons derives from young stars .
the h@xmath2-based sfr is calculated using equation b3 of hopkins et al .
( 2003 ) , after correcting for extinction using the observed balmer decrement ( calzetti 2001 ) : @xmath37 where @xmath38 and @xmath39 are the stellar absorption - corrected line fluxes , derived from the emission - line catalog given by mpa - sdss ( tremonti et al . 2004 ) .
the measured sfr for the central 3@xmath6 based on h@xmath2 is 0.0077 @xmath15 yr@xmath40 . to estimate the sfr along the bar , we use the 8@xmath3 m dust emission , which is thought to be measure of the sfr in galaxies ( wu et al . 2005 ) . from equation 4 of wu et al .
( 2005 ) : @xmath41 an aperture of 3@xmath6 diameter was selected for measuring the sfr centered on the nucleus using the dust - only 8@xmath3 m image , to enable a sensible comparison with the sfr derived from the h@xmath2 flux .
the sfr measured in this way is 0.0079 @xmath15 yr@xmath40 , consistent with that derived from h@xmath2 .
this result indicates that the 8@xmath3 m dust emission can be used as a tracer of the sfr in the central region of galaxies , at least for the case of .
the enhanced 8@xmath3 m emission in the central region of is consistent with the result that barred galaxies tend to have strong central excesses in 8@xmath3 m emission ( regan et al .
2006 ) , which suggests that bars induce gas inflows toward the center of galaxies as an internal process of galaxy secular evolution ( e.g. , sakamoto et al . 1999
; sheth et al . 2000 , 2002 , 2005 ;
jogee et al .
we also estimated the total sfr along the bar after excluding the contribution from the central region .
the photometry was performed on the dust - only 8@xmath3 m image using an ellipse with a semi - major axis of 3@xmath6 and an ellipticity of 0.5 , chosen to match the isocontour of the dust pattern .
the measured sfr for the bar and nucleus is @xmath40.0150 @xmath15 yr@xmath40 , which implies that the sfr along the bar is roughly 0.0072 @xmath15 yr@xmath40 , comparable to the value at the nucleus .
martin & friedli ( 1997 ) found that the h@xmath2 bars ( tracing young stars and h ii regions ) and stellar bars are misaligned by up to 10@xmath0 among 11 barred galaxies .
friedli & benz ( 1993 , 1995 ) , using their n - body + sph simulations of spontaneous bar formation , showed that the h ii regions tend to lead the stellar bar by an angle of several degrees . in these simulations
the misalignments between bars are similar to those observed , and they occur as a result of orbit crossings of gas motion , mostly at early epochs during the formation of a strong , fast - rotating bar in the absence of an inner lindblad resonance ( martin & friedli 1997 ) .
the best example is , in which star formation is completely absent in the galaxy nucleus ( martin & friedli 1997 ) , and the age of the bar is young ( @xmath4400 myr ; martin & roy 1995 ) . but this is unlikely to be the case in , due to the fact that the age of its stellar bar ( @xmath44 gyr ) is much older than bars in the spontaneous bar formation scenario ( e.g. , ) .
alternatively , perhaps the morphology of this galaxy represents an episode shortly after the capture of a small , secondary galaxy .
in such a galaxy merger scenario , the stellar and dust patterns may have different formation histories .
the stellar bar may have formed previously as in a spontaneous formation scenario , and the dust pattern could be tidally induced later from a dwarf galaxy swallowed by .
the different stellar populations in the stellar bar and the dust pattern ( see 3.2 ) also support this scenario .
berentzen et al .
( 2003 ) investigated the dynamical effects of the interaction between an initially barred galaxy and a small companion , using n - body / sph numerical simulations .
they found that the interactions can produce offset bars , nuclear and circumnuclear disks , and tidal arms connected to the end of the bar .
based on their results that the fate of the stellar bar is determined by the impact position , we speculate that in there may have been an impact on the bar major axis when the bar was weak .
in such a case , the tidal force exerted on the bar does not disrupt much the bar structure ( i.e. , the stellar bar survived after the impact ) .
the nearby bright , compact object toward the northeast of the sdss images ( see fig .
3b ) , identified as sdss j110124.04 + 574045.1 , is a plausible candidate for a dwarf galaxy that may have hit .
its colors are very blue ( @xmath42 , @xmath43 , @xmath44 ) and similar to those of irregular galaxies ( fukugita et al .
1995 ) . and its absolute magnitude ( @xmath45 , @xmath46 , @xmath47 , @xmath48 , @xmath49 for _ u , g , r , i , z _ , respectively ) are located at the faint end of the luminosity function of extremely low - luminosity galaxies ( blanton et al .
2005 ) , if we assume that its distance is the same as that of ( 39.9 mpc ) .
however , we can not exclude the possibility that it is a foreground white dwarf star superposed on , since its very blue colors are consistent with those of spectroscopically identified white dwarf stars in sdss ( kleinman et al .
2004 ) .
large misalignments ( @xmath5010@xmath0 ) between stellar bars ( bright at optical and 3.6@xmath3 m ) and dust distributions ( shown at 8@xmath3 m ) are quite rare among nearby barred galaxies with archival _
spitzer _ data . we have examined @xmath450 barred spirals ( sb and sab ) in the _ spitzer _ infrared nearby galaxies survey ( sings , kennicutt et al .
2003 ) and in the mid - ir hubble atlas of galaxies ( a _ spitzer _ gto program , pid : # 69 , pi : g. fazio ; see also pahre et al .
2004 ) , and found that _ none _ of them shows a misalignment as large as that found in .
most of the galaxies only have a single old stellar bar , which is bright at irac 3.6 and 4.5 @xmath3 m but absent at 8@xmath3 m ( e.g. , ) .
others that have younger bars always show good alignment between 3.6 and 8@xmath3 m ( e.g. , , consistent with the previous result that no misalignment was observed between its h@xmath2 and stellar bars ; martin & friedli 1997 ) .
this statistical evidence indicates that the bar structure seen in is quite rare among barred galaxies in the local universe , and that the misalignment between bar and dust patterns may be a short - lived phenomenon in the evolutionary history of the galaxy .
however , any firm conclusion must await a quantitative analysis of a large , well - defined , and unbiased sample of barred galaxies .
using mid - infrared images from _ spitzer _ and optical images from sdss , we show that the late - type barred spiral galaxy contains two misaligned patterns , one composed of old stars and the other young stars and dust .
the angle between the two patterns ( @xmath425@xmath0 ) is among the largest ever reported .
the stellar bar is bright in the optical and in the irac 3.6 and 4.5@xmath3 m bands , and is dominated by old stars .
the dust pattern is more prominent in the 8@xmath3 m band , but also shows up in the sdss _ u _ and _ g _ bands ; it traces regions of recent or ongoing star formation .
the dust pattern could be tidally induced by a dwarf galaxy swallowed by . from examination of mid - infrared images of a large sample of nearby barred galaxies with archival _ spitzer _
data , we find that bar structure such as that found in is quite rare in the local universe . to further test the hypothesis that the large misalignment in was triggered by a recent merger
, it would be desirable to obtain deeper imaging observations of the system in order to search for morphological features suggestive of tidal interactions . obtaining
a spectrum of the candidate dwarf galaxy would help validate its physical association with .
additional numerical simulations will help to validate whether two large - scale patterns can coexist over long time scales , since it is possible that the competing torques will introduce chaos to the system ( i. berentzen , priv .
comm . ) .
+ * acknowledgments * + + we would like to thank the anonymous referee for very constructive comments and suggestions .
we thank x .- y .
xia , s. mao , r. kennicutt , i. berentzen for advice and helpful discussions , and c .- n .
wang , f .- s .
liu for their capable assistance throughout the process of _ spitzer _ data reductions .
this project is supported by nsfc no.10273012 , no.10333060 , no.10473013 , no.10373008 .
this work is based in part on observations made with the _
spitzer space telescope _ , which is operated by the jet propulsion laboratory , california institute of technology under nasa contract 1407 .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/. the sdss is managed by the astrophysical research consortium for the participating institutions .
the participating institutions are the american museum of natural history , astrophysical institute potsdam , university of basel , cambridge university , case western reserve university , university of chicago , drexel university , fermilab , the institute for advanced study , the japan participation group , johns hopkins university , the joint institute for nuclear astrophysics , the kavli institute for particle astrophysics and cosmology , the korean scientist group , the chinese academy of sciences ( lamost ) , los alamos national laboratory , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , ohio state university , university of pittsburgh , university of portsmouth , princeton university , the united states naval observatory , and the university of washington .
athanassoula , e. 1992 , mnras , 259 , 328 athanassoula , e. , & bureau , m. 1999 , apj , 522 , 699 bell , e. f. 2003 , apj , 586 , 794 bendo , g. j. , et al .
2006 , apj , 652 , 283 benedict , f. g. , smith , b. j. , & kenney , j. d. p. 1996 , aj , 111 , 1861 berentzen , i. , athanassoula , e. , heller , c. h. , & fricke , k. j. 2003 , mnras , 341 , 343 bettoni , d. , & galletta , g. 1988 , a&a , 190 , 52 blanton , m. r. , lupton , r. h. , schlegel , d. j. , strauss , m. a. , brinkmann , j. , fukugita , m. , & loveday , j. 2005 , apj , 631 , 208 bottinelli , l. , gouguenheim , l. , paturel , g. , & de vaucouleurs , g. 1983 , a&a , 118 , 4 bruzual , g. , & charlot , s. 2003 , mnras , 344 , 1000 calzetti , d. 2001 , pasp , 113 , 1449 calzetti , d. , et al .
2005 , apj , 633 , 871 cao , c. , & wu , h. 2007 , aj , 133 , 1710 crosthwaite , l. p. , turner , j. l. , & ho , p. t. p. 2000 , aj , 119 , 1720 downes , d. , reynaud , d. , solomon , p. m. , & radford , s. j. e. 1996 , apj , 461 , 186 de vaucouleurs , g. , de vaucouleurs , a. , corwin , h. g. , buta , r. j. , paturel , g. & fouque , p. 1991
, third reference catalogue of bright galaxies ( new york : springer ) elmegreen , b. g. , & elmegreen , d. m. 1985 , apj , 288 , 438 elmegreen , d. m. 1996 , iau colloq . 157 : barred galaxies , 91 , 23 elmegreen , b. g. , elmegreen , d. m. , chromey , f. r. , hasselbacher , d. a. , & bissell , b. a. 1996 , aj , 111 , 2233 erwin , p. 2005 , mnras , 364 , 283 eskridge , p. b. , et al .
2000 , aj , 119 , 536 fazio , g. g. , et al .
2004 , apjs , 154 , 10 fisher , d. b. 2006 , apjl , 642 , l17 freeman , k. c. 1996 , iau colloq . 157 : barred galaxies , 91 , 1 friedli , d. , & benz , w. 1993 , a&a , 268 , 65 friedli , d. , & benz , w. 1995 , a&a , 301 , 649 fukugita , m. , shimasaku , k. , & ichikawa , t. 1995 , pasp , 107 , 945 gadotti , d. a. , & de souza , r. e. 2006 , apjs , 163 , 270 hawarden , t. g. , mountain , c. m. , leggett , s. k. , & puxley , p. j. 1986 , mnras , 221 , 41p helou , g. , et al . 2004 ,
apjs , 154 , 253 ho , l. c. , filippenko , a. v. , & sargent , w. l. w. 1997a , apj , 487 , 579 ho , l. c. , filippenko , a. v. , & sargent , w. l. w. 1997b , apj , 487 , 591 hopkins , a. m. , et al .
2003 , apj , 599 , 971 huang , j .- s . , et al .
2004 , apjs , 154 , 44 jogee , s. , scoville , n. , & kenney , j. d. p. 2005 , apj , 630 , 837 kenney , j. d. p. , scoville , n. z. , & wilson , c. d. 1991 , apj , 366 , 432 kennicutt , r. c. , et al . 2003 , pasp , 115 , 928 kleinman , s. j. , et al .
2004 , apj , 607 , 426 knapen , j. h. , prez - ramrez , d. , & laine , s. 2002 , mnras , 337 , 808 knapen , j. h. 2005 , astronomy and geophysics , 46 , 28 kormendy , j. 1983 , apj , 275 , 529 kormendy , j. , & kennicutt , r. c. 2004 , ara&a , 42 , 603 kormendy , j. , & fisher , d. b. 2005 , revista mexicana de astronomia y astrofisica conference series , 23 , 101 leitherer , c. , et al . 1999 ,
apjs , 123 , 3 lonsdale , c. j. , et al .
2003 , pasp , 115 , 897 lucy , l. b. 1974 , aj , 79 , 745 martin , p. 1995 , aj , 109 , 2428 martin , p. , & friedli , d. 1997 , a&a , 326 , 449 martin , p. , & roy , j .-
1995 , apj , 445 , 161 martini , p. , regan , m. w. , mulchaey , j. s. , & pogge , r. w. 2003 , apj , 589 , 774 menndez - delmestre , k. , sheth , k. , schinnerer , e. , jarrett , t. h. , & scoville , n. z. 2007 , apj , 657 , 790 ondrechen , m. p. , & van der hulst , j. m. 1983 , apj , 269 , l47 pahre , m. a. , ashby , m. l. n. , fazio , g. g. , & willner , s. p. 2004 , apjs , 154 , 235 prez - gonzlez , p. g. , et al .
2006 , apj , 648 , 987 phillips , a. c. 1996 , iau colloq . 157 : barred galaxies , 91 , 44 rand , r. j. ,
lord , s. d. , & higdon , j. l. 1999 , apj , 513 , 720 regan , m. w. , teuben , p. j. , vogel , s. n. , & van der hulst , t. 1996 , aj , 112 , 2549 regan , m. w. , sheth , k. , & vogel , s. n. 1999 , apj , 526 , 97 regan , m. w. , et al .
2006 , apj , 652 , 1112 roussel , h. , et al .
2001 , a&a , 369 , 473 rozas , m. , zurita , a. , & beckman , j. e. 2000 , a&a , 354 , 823 sakamoto , k. , okumura , s. k. , ishizuki , s. , & scoville , n. z. 1999 , apj , 525 , 691 sancisi , r. , allen , r. j. , & sullivan , w. t. , iii 1979 , a&a , 78 , 217 sellwood , j. a. , & wilkinson , a. 1993 , reports of progress in physics , 56 , 173 sheth , k. , regan , m. w. , vogel , s. n. , & teuben , p. j. 2000 , apj , 532 , 221 sheth , k. , vogel , s. n. , regan , m. w. , teuben , p. j. , harris , a. i. , & thornley , m. d. 2002 , aj , 124 , 2581 sheth , k. , vogel , s. n. , regan , m. w. , thornley , m. d. , & teuben , p. j. 2005 , apj , 632 , 217 stoughton , c. , et al .
2002 , aj , 123 , 485 surace , j. a. , et al .
2005 , the swire data release 2 : image atlases and source catalogs for elais - n1 , elais - n2 , xmm - lss , and the lockman hole vogel , s. n. , kulkarni , s. r. , & scoville , n. z. 1988 , nature , 334 , 402 walterbos , r. a. m. , braun , r. , & kennicutt , r. c. , jr .
1994 , aj , 107 , 184 wang , z. , et al .
2004 , apjs , 154 , 193 wen , x .- q . ,
wu , h. , cao , c. , & xia , x .- y .
2007 , chinese journal of astronomy and astrophysics , 7 , 187 werner , m. w. , et al .
2004 , apjs , 154 , 1 wu , h. , et al . 2002 , aj , 123 , 1364 wu , h. , cao , c. , hao , c .-
n . , liu , f .- s . ,
wang , j .- l . , xia , x .- y . ,
deng , z .- g .
, & young , c. k .- s .
2005 , apj , 632 , l79 york , d. g. , et al .
2000 , aj , 120 , 1579 zheng , z. , et al . 1999 ,
aj , 117 , 2757 zurita , a. , relao , m. , beckman , j. e. , & knapen , j. h. 2004 , a&a , 413 , 73 -band , @xmath33-band , irac 3.6@xmath3 m , and dust - only 8@xmath3 m images of the bar region in .
north is up ; east is to the left .
the @xmath5-band image was smoothed using a 2-pixel wide ( 2@xmath34 ) gaussian filter .
contours of the dust - only 8@xmath3 m image are superposed on the @xmath5-band , @xmath33-band , and irac 3.6@xmath3 m images .
a large misalignment by an apparent angle ( @xmath21 ) of about 25@xmath0 is shown between the bright bar at optical ( with a pa of @xmath440@xmath0 ) and the dust distribution shown in the dust - only 8@xmath3 m image ( with a pa of @xmath415@xmath0 ) .
a nearby compact , bright object ( sdss j110124.04 + 574045.1 ) is shown in the northeast of the sdss images .
[ fig3 ] ] notes.@xmath51col.(1 ) : the lengths of semi - major axis of the stellar bar and dust pattern ; col.(2 ) : the relative lengths of the patterns ; col.(3 ) : position angles measured in the conventional manner , from north through east ; col.(4 ) : the measured angular misalignment between stellar and dust patterns ; col.(5 ) : the angular misalignment after correcting for inclination effect ; col.(6 ) : images used for the measurements . | a large position angle misalignment between the stellar bar and the distribution of dust in the late - type barred spiral was discovered , using mid - infrared images from the _ spitzer space telescope _ and optical images from the sloan digital sky survey ( sdss ) .
the angle between the bar and dust patterns was measured to be 25@xmath0@xmath12@xmath0 , larger than most of the misalignments found previously in barred systems based on h@xmath2 or h
i / co observations .
the stellar bar is bright at optical and 3.6@xmath3 m , while the dust pattern is more prominent in the 8@xmath3 m band but also shows up in the sdss _ u _ and _ g_-band images , suggesting a rich interstellar medium environment harboring ongoing star formation .
this angular misalignment is unlikely to have been caused by spontaneous bar formation .
we suggest that the stellar bar and the dust pattern may have different formation histories , and that the large misalignment was triggered by a tidal interaction with a small companion . a statistical analysis of a large sample of nearby galaxies with archival _ spitzer _
data indicates that bar structure such as that seen in is quite rare in the local universe .
galaxies : individual ( ) , galaxies : spiral , galaxies : structure , infrared : galaxies + 98.62.hr |
diffraction patterns produced by grazing scattering of swift atoms and molecules ( with energies in the kev range ) on surfaces are nowadays becoming a powerful surface analysis tool , which is giving rise to a technique known as grazing - incidence fast atom diffraction ( gifad or fad ) winter11,zugarramurdi15 . in recent years
the fad method was successfully applied to very different kinds of materials , ranging from insulators schuller07,rousseau07,schuller12 to semiconductors khemliche09,debiossac14 and metals @xcite , as well as structured films @xcite and molecules seifertprl13 adsorbed on surfaces . however , in spite of the extensive experimental and theoretical work devoted to the research of fad since its first experimental observation @xcite , the complete understanding of the underlying quantum processes is far from being achieved . in particular , the study of the mechanisms that contribute to the coherence or decoherence of the scattered particles is still in its infancy .
the observation of quantum interference effects for fast atoms impinging on crystal surfaces strongly relies on the preservation of quantum coherence @xcite and in this regard , the coherence conditions of the incident beam play an important role . motivated by ref .
@xcite , in this article we investigate the influence of the collimation of the incident beam on fad patterns making use of a recently developed approach , named surface - initial value representation ( sivr ) approximation @xcite . with this goal
we explicitly take into account the experimental collimating conditions to determine the surface region that is _ coherently _ illuminated by the particle beam , using this information to build the initial wave packet that describes the unperturbed state of the incident particle within the sivr method .
the sivr approximation is a semi - quantum approach that was derived from the initial value representation ( ivr ) method by miller @xcite by using the corresponding semi - quantum time evolution operator in the frame of a time - dependent distorted - wave formalism .
this strategy incorporates an approximate description of classically forbidden transitions on the dark side of rainbow angles , making it possible to avoid the classical rainbow divergence present in previous semi - classical models for fad , like the surface - eikonal ( se ) approach @xcite .
such a weakness of the se method affects the intensity of the outermost diffraction maxima when these maxima are close to the classical rainbow angles rubiano13 , i.e. the extreme deflection angles of the classical projectile distribution .
the sivr approach , instead , provides an appropriate description of fad patterns along the whole angular range , even around classical rainbow angles , without requiring the use of convolutions to smooth the theoretical curves @xcite .
therefore , the sivr method can be considered as an attractive alternative to quantum wave packet propagations , offering a clear representation of the main mechanisms of the process in terms of classical trajectories through the feynman path integral formulation of quantum mechanics . in order to analyze the influence of the beam collimation on fad spectra , an extended version of the sivr approximation - including the collimating parameters -
is applied to evaluate fad patterns for he atoms grazingly impinging on a lif(001 ) surface after going through a rectangular aperture .
the paper is organized as follows .
the theoretical formalism is summarized in sec .
results for different sizes of the collimating aperture are presented and discussed in sec .
iii , while in sec .
iv we outline our conclusions .
atomic units ( a.u . ) are used unless otherwise stated .
let us consider an atomic projectile ( @xmath0 ) , with initial momentum @xmath1 , which is elastically scattered from a crystal surface ( @xmath2 ) , ending in a final state with momentum @xmath3 and total energy @xmath4 , with @xmath5 the projectile mass . by employing the ivr method @xcite , the scattering state of the projectile at the time @xmath6
can be approximated as @xcite : @xmath7 where @xmath8is the initial momentum eigenfunction , with @xmath9 the position of the center of mass of the incident atom , and the sign `` @xmath10 '' in the supra - index of the scattering state indicates that it satisfies outgoing asymptotic conditions . in eq .
( [ estado - ivr ] ) the position ket @xmath11 is associated with the time - evolved position of the incident atom at a given time @xmath6 , @xmath12 , which is derived by considering a classical trajectory with starting position and momentum @xmath13 and @xmath14 , respectively .
the function @xmath15 denotes the classical action along the trajectory @xmath16 , \label{st}\]]with @xmath17 the classical projectile momentum at the time @xmath6 and @xmath18 the surface - projectile interaction , while the function @xmath19 \label{j}\]]is a jacobian factor ( a determinant ) evaluated along the classical trajectory @xmath20 .
this jacobian factor can be related to the maslov index @xcite by expressing it as @xmath21 , where @xmath22 is the modulus of @xmath23 and @xmath24 is an integer number that accounts for the sign of @xmath23 at a given time @xmath6 . in this way
, @xmath24 represents a time - dependent maslov index , satisfying that every time that @xmath25 changes its sign along the trajectory , @xmath24 increases by 1 . the functions @xmath26 and @xmath27 , present in the integrand of eq .
( [ estado - ivr ] ) , describe the shape of the position- and momentum- wave packet associated with the incident projectile . in a previous paper @xcite @xmath28
was considered as a gaussian distribution illuminating a fixed number of reduced unit cells of the crystal surface , while @xmath29 was defined as an uniform distribution .
here these functions are derived from the collimation conditions of the incident beam in order to incorporate a realistic profile of the coherent initial wave packet , as explained in the following sub - section . by using the sivr scattering state , given by eq .
( [ estado - ivr ] ) , within the framework of the time - dependent distorted - wave formalism dewangan94 , the sivr transition amplitude , per unit of surface area @xmath30 , can be expressed as @xcite @xmath31where @xmath32 \quad \label{aif}\end{aligned}\]]is the partial transition amplitude associated with the classical path @xmath33 , with @xmath34 the projectile momentum transfer and @xmath35 \label{fitot}\]]the sivr phase at the time @xmath6 .
details of the derivation of the sivr method are given in ref .
@xcite . in this article
we use a frame of reference placed on the first atomic layer , with the surface contained in the @xmath36 plane , the @xmath37 versor along the incidence direction and the @xmath38 versor oriented perpendicular to the surface , aiming towards the vacuum region .
the sivr differential probability , per unit of surface area , for elastic scattering with final momentum @xmath3 in the direction of the solid angle @xmath39 , is obtained from eq .
( aif - sivr ) as @xmath40where @xmath41 and @xmath42 are the final polar and azimuthal angles , respectively , with @xmath41 measured with respect to the surface and @xmath42 measured with respect to the @xmath37 axis
. a schematic depiction of the process and the coordinates is displayed in fig .
1 ( a ) . in eq .
( [ aif - sivr ] ) , the variables @xmath13 and @xmath43 represent the starting position and momentum , respectively , of the classical projectile trajectory , both measured at @xmath44 , while the functions @xmath26 and @xmath45 determine the shape of the initial wave packet , satisfying the heisenberg uncertainty relation .
we discompose the starting position as @xmath46 , where @xmath47 @xmath48 and @xmath49 are the components parallel and perpendicular , respectively , to the surface plane , with @xmath49 a fixed distance for which the projectile is hardly affected by the surface interaction .
we assume that the size of the coherent initial wave packet , at a distance @xmath50 from the surface , is governed by the collimation of the incident beam as given by the van cittert - zernike theorem @xcite . by considering a rectangular collimating aperture placed at a long distance @xmath51 from the surface , the coherence size of the incident beam on the @xmath49-plane , which is located parallel to the surface at a @xmath49 distance from it ,
is defined by the complex grade of coherence , @xmath52 .
it reads bornwolf @xmath53 where @xmath54 is the spherical bessel function and @xmath55 and @xmath56 denote the lengths of the sides of the rectangular aperture , which form angles @xmath57 and @xmath58 , respectively , with the surface plane , @xmath59 being the glancing incidence angle ( see figs . 1 ( a ) and ( b ) ) .
( [ ucoh ] ) the de broglie wave lengths @xmath60 and @xmath61 are defined as @xmath62respectively , this last one being associated with the initial motion normal to the surface plane , while @xmath63 . for most of the collision systems
, the @xmath49 value can be chosen as equal to the lattice constant of the crystal , leading to @xmath64 .
according eq .
( [ ucoh ] ) the spatial profile of the initial wave packet can be approximated by a product of gaussian functions , @xmath65^{1/4}\exp ( -x^{2}/\omega ^{2 } ) , \label{gauss}\ ] ] as follows : @xmath66where the parameters @xmath67 and @xmath68 were derived by fitting the complex grade of coherence , i.e. @xmath69 , reading @xmath70 the lengths @xmath67 and @xmath68 represent the effective widths of the @xmath71 and @xmath72 distributions , respectively , being defined as the corresponding root - mean - square deviations @xcite .
notice that these widths are associated with the _ transversal _ _ coherence _ _ size _ of the initial wave packet , magnitude that is crucial in matter - wave interferometry tonomura86,keller00,barrachina15 .
on the other hand , concerning the momentum profile of the initial wave packet , as we are dealing with an incident beam with a well defined energy , i.e. @xmath73 @xcite , the _ longitudinal coherence length _ does not play any role @xcite .
consequently , the starting momentum @xmath14 satisfies the energy conservation , with @xmath74 , and the integration on @xmath75 can be solved by making use of the change of variables @xmath76 , with @xmath77 and @xmath78 varying around the incidence direction .
the shape of the corresponding angular wave packet is described again in terms of gaussian functions , reading @xmath79where @xmath80 is de solid angle corresponding to the @xmath14 direction and the angular widths of the @xmath81- and @xmath78- distributions were derived from the uncertainty principle as @xcite @xmath82respectively .
replacing eqs .
( [ ffi ] ) and ( [ ggi ] ) in eq .
( [ aif - sivr ] ) , the extended version of the sivr transition amplitude , including explicitly the proper shape of the incident wave packet , is expressed as @xmath83 where @xmath84 is given by eq .
( [ aif ] ) and @xmath85 .
we apply the extended sivr method to @xmath86he atoms elastically scattered from a lif(001 ) surface under axial surface channeling conditions since for this collision system , diffraction patterns for different widths of the collimating slit were reported in ref .
the sivr transition amplitude was obtained from eq .
( [ aif - sivrn ] ) by employing the montecarlo technique to evaluate the @xmath87 and @xmath88 integrals , considering more than @xmath89 points in such an integration . for every starting point ,
the partial transition amplitude @xmath90 was evaluated numerically from eq .
( [ aif ] ) by employing a potential @xmath18 derived from a pairwise additive hypothesis .
the potential model used in this work is the same as the one employed in ref .
it describes the surface - projectile interaction as the sum of the static and polarization contributions , the first of them evaluated incorporating no local terms of the electronic density in the kinetic and exchange potentials .
the potential @xmath18 also takes into account a surface rumpling , with a displacement distance extracted from ref .
details of the surface potential will be published elsewhere @xcite . in this work
we vary the size of the collimating aperture keeping a fixed incidence condition , given by helium projectiles impinging along the @xmath91 channel with a total energy @xmath92 kev and an incidence angle @xmath93 deg . in all the cases ,
the distance between the collimating aperture and the surface is chosen as @xmath94 cm , in agreement with the experimental setup of ref .
@xcite . in figs . 2 and 3 we show two - dimensional projectile distributions , as a function of @xmath41 and @xmath42 , derived within the sivr approximation by considering collimation slits with the same length - @xmath95 mm - but two different widths : @xmath96 mm and @xmath97 mm , respectively .
both sivr distributions reproduce quite well the corresponding experimental ones @xcite , also displayed in the figures .
they present the usual banana shape , characteristic of the axial surface scattering @xcite , with final dispersion angles lying on a thick annulus , whose mean radius is approximately equal to @xmath59 . from the comparison of figs . 2 and 3 it
is clearly observed that the width of the collimation slit strongly affects the diffraction patterns , making the well - defined peaks present in the distributions of fig .
2 , for the more narrow slit , completely disappear when the width of the slit is increased , as it happens in fig .
3 . in the experimental and theoretical intensity distributions of fig .
3 , only maxima at the rainbow deflection angles @xmath98 are visible .
as discussed in ref .
@xcite , this behavior is related to the area @xmath99 of the surface plane that is coherently lighted by the incident beam and it will be studied in detail within the sivr approach . in eq .
( [ aif - sivrn ] ) , by splitting the @xmath87-integral on the area @xmath99 into a collection of integrals over different reduced unit cells , it is possible to express @xmath100 as a product of two factors @xcite : @xmath101 each of them associated with a different interference mechanism .
the factor @xmath102 , named unit - cell form factor , is derived from eq .
( [ aif - sivrn ] ) by evaluating the @xmath87-integral over only one reduced unit cell , being related to supernumerary rainbows @xcite .
while the factor @xmath103 is a crystallographic factor associated with the bragg diffraction , which originates from the interference of identical trajectories whose initial positions @xmath104 are separated by a distance equal to the spacial periodicity of the lattice .
the factor @xmath103 depends on @xmath105 and the area @xmath99 coherently illuminated bythe particle beam , being insensible to the potential model . in eq .
( [ aif - sivrn ] ) the effective area @xmath99 coherently lighted by the incident beam results to be @xmath106 , where the distances @xmath107 with @xmath108 were determined from the @xmath109 values for which the function @xmath110 , given by eq .
( [ ucoh ] ) , vanishes . under typical incidence conditions for fad , the dependence of @xmath111 on the azimuthal angle @xmath42 becomes completely governed by the number @xmath112 of reduced unit cells in the direction transversal to the incidence channel that are illuminated by the initial wave packet , i.e. @xmath113 , where @xmath114 is the length of the reduced unit cell along the @xmath115 direction . for @xmath116 the factor @xmath103 gives rise to bragg peaks placed at azimuthal angles that verify the relation @xmath117 , with @xmath118 an integer , as observed in fig .
2 where @xmath119 .
the relative intensities of theses bragg peaks are modulated by @xmath120 , which acts as an envelope function that can reduce or even suppress the contribution of a given bragg order , while the peak width is determined by @xmath121 , narrowing as @xmath112 increases .
but when the coherently illuminated region shrinks to cover around a reduced unit cell in the transversal direction , only the unit - cell factor is present in eq .
( [ a - factors ] ) . consequently , the angular distribution shows structures associated withsupernumerary rainbow maxima exclusively , as it happens in fig . 3 where @xmath122 . with the aim of studying more deeply the variation of the diffraction patterns with the width of the slit , in fig .
4 we display the differential probability @xmath123 , as a function of the azimuthal angle @xmath42 , for different values of @xmath56 . as given by eq .
( [ sigmax ] ) , when @xmath56 augments , the number @xmath112 of the coherently illuminated cells decreases while the width of the bragg peaks increases , as observed in fig . 4 for @xmath124 mm .
for wider collimating slits bragg peaks start to blur out , disappearing completely for @xmath125 mm , where @xmath126 .
therefore , varying @xmath56 we can inspect two different zoologies : bragg peaks at small @xmath56 values and supernumerary rainbow peaks at large @xmath56 .
we also analyze the influence of the length of the collimating aperture , @xmath127 , on fad patterns . in fig .
5 we display angular projectile distributions derived from the sivr approach by considering a collimating slit with the same width , @xmath96 mm , and three different lengths : @xmath128 , @xmath129 and @xmath130 mm . for a small square aperture ( fig
5 ( a ) ) , bragg peaks are observed like circular spots lying on a thin ring whose radius is equal to @xmath59 , corresponding to an almost ideal elastic rebound @xmath131 .
but when the length of the collimating aperture augments up to @xmath132 @xmath129 mm ( fig . 5 ( b ) ) , transforming the square orifice into a slit , bragg peaks become visible like elongated strips which are placed at slightly different radius .
this effect is even more evident in fig . 5 ( c ) for @xmath132 @xmath130 mm , where the projectile distribution resembles the diffraction charts for different normal energies @xmath133 .
the explanation is simple : from eqs .
( [ sigmax ] ) and ( [ sigmatita ] ) , if @xmath55 is large @xmath134 is also large , enabling a wide spread of the impact momentum normal to the surface plane , @xmath135 .
such a @xmath136- dispersion gives rise to the structures along the vertical axis of fig .
hence , the intensity oscillations along the @xmath41- axis observed for long collimating slits are probing the surface potential for different distances to the topmost atomic plane .
they might be a useful tool to explore different distances to the surface without varying the mean value of normal energy @xmath137 .
additionally , notice that the transversal coherence length @xmath138 ( @xmath68 ) depends on the ratio @xmath139 ( @xmath140 ) , as given by eq .
( [ sigmax ] ) , so that any change of the collimating conditions that kept this ratio as a constant will produce the same interference patterns .
we have derived an extended version of the sivr approximation gravielle14 that incorporates a realistic description of the coherent initial wave function in terms of the collimating parameters of the incident beam .
the model was applied to helium atoms grazingly impinging on a lif(001 ) surface considering a rectangular collimating aperture with different sizes . as it was experimentally found @xcite ,
the sivr interference patterns are strongly affected by the width of the collimating slit , which determines the transversal length of the surface area that is coherently illuminated by the incident wake packet .
the number of lighted reduced unit cells in the direction transversal to the incidence channel determines the azimuthal width of the bragg peaks , making either bragg peaks or supernumerary rainbows were visible . on the other hand ,
the length of the collimating slit affects the polar @xmath141- distribution of scattered projectiles , this effect being related to the dispersion of the component of the initial momentum perpendicular to the surface . as the length of the collimating aperture increases ,
diffraction maxima are transformed from circular spots into elongated strips , where interference structures along the @xmath41- axis arise for the longer slits .
these findings suggest that collimating slits with several millimeters of length might be used to probe the projectile - surface interaction for different normal distances . and @xmath143 , for @xmath144 kev @xmath86he atoms impinging on lif(001 ) along the @xmath91 direction with the incidence angle @xmath145 deg .
the incident helium beam is collimated with a rectangular aperture of sides @xmath132 @xmath146 mm and @xmath147 @xmath148 mm .
upper panel , experimental distribution extracted from ref .
winter15 ; lower panel , sivr distribution . , scaledwidth=40.0% ] , for @xmath144 kev @xmath86he atoms impinging on lif(001 ) along the @xmath150 direction with the incidence angle @xmath151 deg .
the incident helium beam is collimated with a rectangular aperture of length @xmath132 @xmath146 mm and different widths : @xmath152 @xmath153 , @xmath148 , @xmath154 , @xmath155 , @xmath156 and @xmath149 mm , respectively .
vertical lines indicate the angular positions of bragg peaks , as explained in the text.,scaledwidth=40.0% ] | the influence of the collimating conditions of the incident beam on diffraction patterns produced by grazing scattering of fast atoms off crystal surfaces is studied within a semi - quantum approach , named surface initial value representation ( sivr ) approximation . in this approach
we incorporate a realistic description of the incident particle in terms of the collimating parameters , which determine the surface area that is coherently illuminated .
the model is applied to he atoms colliding with a lif(001 ) surface after passing through a rectangular aperture . as it was experimentally observed
@xcite , sivr spectra as a function of the azimuthal angle are very sensitive to the width of the collimating slit .
we also found that the length of the collimating aperture affects polar angle distributions , introducing additional interference structures for the longer collimating slits . |
figure [ toad ] shows two of the 64 different ways of tiling an order three aztec diamond with dominoes .
aztec diamonds are defined in `` alternating - sign matrices and domino tilings , '' by elkies , kuperberg , larsen , and propp @xcite .
each tiling will be referred to as a toad , a tiling of an aztec diamond . in all of the illustrations
, we have rotated the aztec diamonds and drawn spurs on each corner to simplify later calculations .
an alternating sign matrix ( asm ) is a square matrix whose entries are @xmath0 , @xmath1 , or @xmath2 , such that the entries of each row or column must add up to @xmath1 , and all nonzero entries must alternate in sign .
examples : @xmath3 elkies , kuperberg , larsen , and propp showed that there are @xmath4 tilings of an order @xmath5 aztec diamond with dominoes @xcite . as part of their proof , they used a relationship between toads and pairs of comatable asms . _ compatibility _ of asms was defined by robbins and rumsey in terms of the asms themselves , without reference to domino tilings @xcite .
the interior vertices of an order @xmath5 toad are arranged as a @xmath5-by-@xmath5 square matrix inside an @xmath6-by-@xmath6 matrix .
each interior vertex is connected to two or more of its four nearest neighbors .
for the smaller square assign a @xmath0 to each vertex incident to exactly three edges , a @xmath1 to each vertex incident to exactly two edges , and a @xmath2 to each vertex incident to exactly four edges .
for the larger matrix , do the same thing , but reverse the roles of @xmath1 s and @xmath2 s .
the two above examples of asms come from the aztec diamond in figure [ asmtoad ] .
a pair of asms of order @xmath5 and order @xmath7 is said to be compatible if there is a toad that is related to the pair by the above algorithm .
a pair of compatible asms uniquely determines a toad .
it has been shown that an order @xmath7 asm with @xmath8 @xmath2 s is compatible with @xmath9 order @xmath5 asms .
consequently an asm with no @xmath2 s is only compatible with one smaller asm .
an asm with no @xmath2 s is a permutation matrix .
the study of baxter permutations began outside of the field of combinatorics @xcite , and even though they show up many places , it was surprising to find a connection between domino tilings and baxter permutations .
we will give two definitions of baxter permutations .
the latter follows from the former and is described in terms of permutation matrices .
a permutation @xmath10 of @xmath5 objects is a baxter permutation if for each @xmath11 , there is a @xmath12 such that @xmath13 is between @xmath14 and @xmath15 ( allowing @xmath12 to be @xmath16 but not @xmath17 ) and for every @xmath18 between @xmath14 and @xmath13 ( including @xmath13 ) , @xmath19 , and for every @xmath18 between @xmath13 and @xmath15 ( not including @xmath13 ) , @xmath20 @xcite . the easiest way to understand this definition is by looking at at a permutation matrix .
let @xmath21 be a permutation matrix . by definition ,
each row and each column of @xmath21 contains exactly one entry that is a 1 : for each @xmath16 , @xmath22 and the rest are zeros .
take any two adjacent rows in @xmath21 , the @xmath16th and @xmath23th . then @xmath24 .
then , we look at the columns between the @xmath14th column and the @xmath15th column
. if the permutation is baxter then there will be a vertical dividing line between two of these columns , so that every 1 on the @xmath15 side of the line is in a row below the @xmath17th row , and every 1 on the @xmath14 side of the line is in a row above the @xmath16th row .
for example , let us test permutation 45123 between the second and third rows : @xmath25 because there is a place to draw a vertical line between columns 1 and 5 , this matrix passes the test for @xmath26 .
since it passes the test for each of the other rows as well , it is baxter .
now we are ready to present our theorem , which we will prove in section [ proof ] : [ bpth ] an order @xmath6 asm without @xmath2 s is compatible with an order @xmath5 asm without @xmath2 s if and only if it is a baxter permutation matrix . as a consequence of this theorem
, we know the number of order @xmath5 toads with the property that vertices corresponding to the smaller asm never have four edges incident to them , and vertices corresponding to the larger asm never have two edges incident to them .
that number is the same as the number of baxter permutations of order @xmath7 , which is given by chung , graham , hoggatt , and kleiman @xcite .
the number of baxter permutations on order @xmath5 is @xmath27
to prove theorem [ bpth ] , we will present an algorithm that will produce the smaller asm that is compatible with a given permutation matrix .
the following lemma is a formal description of the algorithm .
we will use the relationship between asms and aztec diamonds to prove the lemma by the end of section [ sasm ] .
[ alg ] if @xmath21 is an order @xmath7 permutation matrix , and @xmath28 is the order @xmath5 asm that is compatible with @xmath21 , then @xmath29 if and only if @xmath30 or @xmath31 or @xmath32 or @xmath33 . the nonzero elements of @xmath28 alternate between @xmath1 and @xmath2 .
we will use an example to show how to apply the algorithm .
suppose we are given the permutation 31425 .
first , let us define the matrix @xmath21 which corresponds to our permutation .
@xmath34 now draw the smaller matrix @xmath28 interspersed within @xmath21 .
@xmath35 from the description of the algorithm in lemma [ alg ] , @xmath36 because @xmath37 and @xmath38 because @xmath39 and @xmath40 because @xmath41 .
repeated applications of this method will give : @xmath42 @xmath43 graphically : @xmath44 lemma [ alg ] tells us that the other elements@xmath45 , @xmath46 , @xmath47 , @xmath48 , @xmath49 , and @xmath50are nonzero and alternate between 1 and @xmath2 .
thus , @xmath51 the proof of lemma [ alg ] comes in two parts .
first we show that entries in @xmath28 that have a certain property are equal to 0 .
second , we show that entries that lack this property are nonzero .
the values of the nonzero entries are given by the definition of an asm .
first let us show that @xmath52 by induction on @xmath53 .
let @xmath54 .
figure [ varprop ] shows a situation where @xmath55 .
we need to show that @xmath56 must be zero . to do this
, we look at the positioning of the dominoes in the toad that corresponds to @xmath28 and @xmath21
. the square located between the vertices of @xmath57 and @xmath58 must be part of a domino .
since it is on the left edge of the aztec diamond , the domino must be pointing northeast or southeast .
figure [ varprop ] shows both cases .
the figure also shows the @xmath57 and @xmath58 vertices with three edges incident to them .
either configuration leads to @xmath59 . for the induction step choose any @xmath60 such that all of the @xmath61 s and @xmath62 s to the left of @xmath63 are zero .
we assume not only that all of the @xmath64 s to the left of @xmath63 are zero , but also that the zig - zag shape from the far left edge of the aztec diamond has propagated inward towards the @xmath53th column . since @xmath65 , both of these vertices meets exactly three edges . as figure [ induction ] shows , once we know the shape of the tiling to the left of @xmath63 , and that @xmath65 , we can determine the value of @xmath63 just as we determined the value of @xmath56 .
figure [ prop ] shows that this induction can continue until we encounter a 1 . in other words , 0
s propagate from the edge of @xmath28 until they hit a nonzero entry in the surrounding rows and columns of @xmath21 .
this proof can also be applied to the top , bottom , and right edges of the asms .
then we have proved half of the lemma .
to finish proving lemma [ alg ] , we will need to look at how the tiling patterns must propagate inside the matrix @xmath28 .
we will concentrate on those elements of @xmath28 whose value is still undetermined . to represent those entries
, we will use the symbol @xmath66 .
we will now look at one individual row of @xmath28 , and look at the two rows of @xmath21 that are above and below it . here is an example : the second and third rows from the permutation 34215 .
@xmath67 notice first that all of the unknown entries on any row in @xmath28 will be between the 1 s in the two closest rows of @xmath21 .
as figure [ direction ] shows , a 0 will show up between the 1 s only if there is a vertical path of zeros from either the top edge or the bottom edge . in this example
, the zero is there because of a path from the bottom edge .
we can say that a @xmath66 arises when the columns of @xmath21 to the immediate left and right of the box have their 1 s in opposite directions , and that if these two columns of @xmath21 have their 1 s in the same direction , there is a 0 .
@xmath68 our next example is the ( non - baxter ) permutation 3142 .
we choose this because the second row of @xmath28 will have three @xmath66s .
@xmath69 looking only at the center row , we can reconstruct where the 1 s are in @xmath21 : @xmath70 it can now be seen that there is an odd number of boxes in each row , because each box always corresponds to a switching in the direction of the 1 s , and the 1 on the left of the leftmost box must be in the direction opposite of the 1 in the column to the right of the rightmost box .
consequently , @xmath66s can be in either of two configurations .
the important thing is the direction of the 1 in the two adjacent rows and two adjacent columns of @xmath21 .
there are only two types of configurations , up to various symmetries , as figure [ windmill ] illustrates .
call the configuration represented by the one on the right _ windmilled _ , and the configuration represented by the one on the left _ non - windmilled_. @xmath71 it can be seen that the @xmath66s along a row alternate between non - windmilled and windmilled , starting with a non - windmilled @xmath66 . it is claimed that the left configuration will lead to @xmath72 and the windmilled always lead to @xmath73 . to prove this , we will have to look at the tiling patterns between the 1 s in two consecutive rows of @xmath21 . in this section
, we will assume for a moment that we know the values of each entry in the @xmath16th row of @xmath28 and the @xmath16th and @xmath17th rows of @xmath21 .
the reader might find it useful to generate example rows and try to draw the corresponding tilings .
for example , these rows : @xmath74 will produce this partial tiling : notice that in the interesting area between the 1 s in @xmath21the tiling is uniquely determined by these three matrix rows .
we will prove this , and use the machinery of the proof to show that @xmath66s are always non - zero .
here are examples of how the titling patterns propagate from left to right .
the arrows serve to highlight the change that @xmath1 s and @xmath2 s make to the pattern . * ( a ) * and * ( c ) * show configurations that go unchanged when they encounter a 0 in @xmath28 . *
( b ) * and * ( d ) * show how encountering a 1 or a @xmath2 in @xmath28 changes the configuration . ] figure [ fill ] shows how the tiling patterns that propagate in from the left side of the toad change first when they hit a 1 in @xmath21 and later when they hit 1 s and @xmath2 s in @xmath28 .
notice that in between the two 1 s in @xmath21 , the tiling can take on one of two patterns : one will happen if the next nonzero element to the right in @xmath28 is @xmath1 , and the other will happen if if the next nonzero element to the right is @xmath2 .
figure [ rl ] shows the two patterns .
once we are to the right of a 1 in @xmath21 , the tiling can have two patterns , @xmath75 and @xmath76 .
@xmath75 corresponds to having the closest non - zero @xmath28 entry on the left being 1 , and @xmath76 corresponds to being on the right of a 1 in @xmath28 .
if @xmath76 and @xmath75 fit together one way , there is a @xmath2 in between .
if they are fit together the other way , there is a @xmath1 in between . ] to help us keep this straight , the components in figures [ fill ] and [ rl ] have an arrow that points to the right if the next nonzero element on the right in @xmath28 is 1 , and to the left if the next nonzero element on the left in @xmath28 is a 1 .
for example , look at this row from an asm , with the arrows filled in .
@xmath77 suppose that this row had been determined by a larger permutation matrix it was compatible with .
here is what might look like : @xmath78 + the last picture shows the tiling broken into components . notice that the components between the 1 s in @xmath21 are all the same , up to a rotation .
each rotation corresponds to a nonzero element in the asm @xmath28 that fits between the two components .
this constitutes an algorithm to determine the shape of the local tiling from the two asms .
now we have assembled all of the tools necessary to finish proving lemma [ alg ] .
we know that the @xmath66s alternate between two configurations : windmilled and non - windmilled ; if we were to show that the windmilled boxes can not contain 1 s or 0 s , then the sign - alternation property of @xmath28 would force the non - windmilled boxes to contain 1 s , and the lemma would be proved
. here is what a windmilled configuration looks like , up to reflection .
@xmath79 let us assume that @xmath80 , and find a contradiction .
the contradiction arises when we try to draw the tiling components that correspond to the arrows that point towards the 1 .
@xmath81 but the next figure shows that the tilings that result from looking at the column and the row do nt match , so we have a contradiction . let us now assume that @xmath82 , and find a contradiction . if we have a 0 in @xmath28 , then the arrows on either side point in the same direction .
this is true for the arrows in the row and in the column . in the next figure , we have chosen to make the arrows point up and to the left . because of symmetry , we will get the same result for any equivalent configuration . @xmath83
but the next figure shows that the tilings that result from looking at the column and the row do nt match , so we have a contradiction .
this concludes our proof of lemma [ alg ] .
with an algorithm in hand that will produce the unique order @xmath5 asm that is compatible with an order @xmath7 permutation matrix , we are prepared to show that the conditions in the permutation that produce @xmath2 s in the smaller asm are identical to the conditions that would prevent that permutation from being baxter .
first , assume that a permutation @xmath21 is compatible with a smaller asm @xmath28 that contains no @xmath2 s .
if @xmath28 it contains no @xmath2 s then each row of @xmath28 must contain exactly one @xmath66 .
remember that the @xmath66 represents the place where the direction that the 1 s of @xmath21 are located switches from up to down or down to up .
then for each row , the location of the @xmath66 is the natural place to draw a dividing line between the columns to pass the baxter test .
consequently , @xmath21 is baxter .
assume that a permutation @xmath21 is compatible with a smaller asm @xmath28 that contains at least one @xmath2 .
the row of @xmath28 that contains the @xmath2 has at least three @xmath66s , and consequently , has no natural place to draw the dividing line between columns .
let us show that this will keep @xmath21 from being baxter . from our earlier discussion
, it is apparent that @xmath21 contains a windmilled configuration of 1 s .
@xmath84 it is easily checked that a permutation that looks like this is not baxter , because there can be no vertical dividing line between columns that properly segregates the 1 s into two groups .
this concludes the proof of theorem [ bpth ] .
i would like to thank james propp for counting pairs of compatible permutation matrices for small values of @xmath5 , for making the conjecture about the connection with baxter permutations , and for bringing the problem to my attention .
the other members of the spatial systems lab at the university of wisconsin - madison also deserve thanks for their help and support with this proof and with this paper , especially dominic johann - berkel for finding out what a baxter permutation was . | we present a proof of a conjecture about the relationship between baxter permutations and pairs of alternating sign matrices that are produced from domino tilings of aztec diamonds .
it is shown that if and only if a tiling corresponds to a pair of asms that are both permutation matrices , the larger permutation matrix corresponds to a baxter permutation . there has been a thriving literature on both pattern - avoiding permutations of various kinds @xcite @xcite and tilings of regions using dominos or rhombuses as tiles @xcite @xcite .
however , there have not as of yet been many links between these two areas of enumerative combinatorics .
this paper gives one such link . |
for active galactic nuclei ( agn ) , an important and interesting issue is the duration of their active phase and the time scales for recurrence of such periods of activity .
it is believed widely that the central activity is related to the feeding of a supermassive black hole in the centre of the galaxy .
one of the more striking examples of episodic activity is when a new pair of radio lobes is seen closer to the nucleus in a radio loud agn before the older and more distant radio lobes have faded .
such sources have been named double - double radio galaxies ( ddrgs ; schoenmakers et al .
a ddrg consists of a pair of double radio sources with a common core , where the two lobes of the inner double have an edge - brightened radio morphology . in such sources
the newly - formed jets may propagate outwards through the cocoon formed by the earlier cycle of activity rather than the general intergalactic or intracluster medium . in table 1
we summarise the presently known large - scale ddrgs , including a couple of candidates .
all of these are identified with galaxies ( not quasars ) and the linear sizes of the outer doubles of most ddrgs are larger than @xmath01 mpc . however , in addition to the ddrgs there are other signatures of episodic activity .
for example , the radio galaxy 4c29.30 ( j0840 + 2949 ) , which is associated with a bright ( r @xmath1 ) host elliptical galaxy at a redshift of [email protected] , has an inner double - lobed radio source with two prominent hotspots separated by 29 arcsec ( 36 kpc ) , embedded in a large - scale halo which has an angular size of @xmath0520 arcsec ( 639 kpc ) .
the radio luminosity of the inner double at 1400 mhz is 5.5@xmath310@xmath4 w hz@xmath5 , which is significantly below the dividing line of the fanaroff - riley classes , while that of the entire source is 7.4@xmath310@xmath4 w hz@xmath5 .
it is interesting to note that in some of the ddrgs , the luminosity of the inner double is in the fri category although its structure resembles that of frii radio sources ( cf .
saikia et al .
the spectral age of the inner double is estimated to be less than @xmath033 myr .
the extended diffuse emission has a steep spectrum with a spectral index of @xmath01.3 and a break frequency less than @xmath0240 mhz .
the spectral age is greater than @xmath0200 myr , suggesting that the extended diffuse emission is due to an earlier cycle of activity ( jamrozy et al .
2007 ) .
in addition , diffuse relic radio emission due to an earlier cycle of activity may also be visibile around radio sources which may not be characterised by a ` classical double ' structure with hotspots at the outer edges .
examples of such sources include her a ( gizani & leahy 2003 ) , 3cr 310 ( e.g. leahy , pooley , & riley 1986 ) , and cen a ( e.g. morganti et al . 1999 ) .
giacintucci et al .
( 2007 ) have explored the possibility that the interesting wide - angle tailed source at the centre of abell 2372 might be an example of a restarted radio galaxy .
.the sample of known ddrgs [ cols="<,^ , > , > , > , > , > , > , > , > " , ] column 1 : source name ; columns 2 and 3 : optical identification and redshift ; columns 4 and 6 : projected linear sizes of the inner and outer lobes in kpc ( calculated for @xmath6=71 km s@xmath5 mpc@xmath5 , @xmath7 = 0.27 , and @xmath8 = 0.73 ) ; columns 5 and 7 : largest angular size of the inner and outer lobes in arcsec ; columns 8 and 9 : 1400-mhz flux density of the inner and outer lobes in mjy ; column 10 : references . 1 : saikia et al . ( 2006 ) ; 2 : saripalli , subrahmanyan , & udaya shankar ( 2002 ) ; 3 : perley et al .
( 1980 ) ; 4 : bridle , perley , & henriksen ( 1986 ) ; 5 : clarke et al .
( 1992 ) ; 6 : machalski et al .
( 2006 ) ; 7 : schoenmakers et al .
( 2000 ) ; 8 : brocksopp et al .
( 2007 ) ; 9 : willis , strom , & wilson ( 1974 ) ; 10 : strom & willis ( 1980 ) ; 11 : schilizzi et al .
( 2001 ) ; 12 : owen & ledlow ( 1997 ) ; 13 : marecki et al .
( 2003 ) ; 14 : bondi et al .
( 2004 ) ; 15 : konar et al .
( 2006 ) ; 16 : saripalli , subrahmanyan , & udaya shankar ( 2003 ) ; 17 : safouris et al .
( 2008 ) ; 18 : kronberg , wielebinski , & graham ( 1986 ) ; 19 : leahy et al .
( 1997 ) ; ( * ) a triple - double radio galaxy .
the lobes of the outer double of j1453 + 3308 are separated by 336 arcsec ( 1297 kpc ) , and their radio luminosity is above the fri / frii break although they do not show any prominent compact hot - spots .
the inner lobes have a separation of 41 arcsec ( 159 kpc ) , and their luminosity is below the dividing fri / frii line although they have an edge - brightened structure ( figure 1 ) .
spectral ages have been determined using the standard theory describing the time - evolution of the emission spectrum from particles with an initial power - law energy distribution , corresponding to an injection spectral index @xmath9 ( murgia 1996 ) . the spectral break frequency above which the radio spectrum steepens from the injected power law , @xmath10 , is related to the spectral age . in order to determine a value of @xmath9
, we fitted the ci ( continuous injection ) and jp ( jaffe & perola 1973 ) models of radiative losses to the flux densities of the inner and outer lobes , respectively .
the fits of the models to the flux densities of both the inner lobes as well as the entire outer double are shown in figure 1 .
it is worth noting that both the fitted values of @xmath9 are @xmath11 .
the spectral age obtained for the inner double is of @xmath122 myr , while the maximum ages for the northern and southern lobes are @xmath047 myr and @xmath058 myr , respectively .
this indicates a mean separation velocity of the lobe s head from the radio - emitting plasma of 0.036c .
however , assuming presence of a backflow with a backward speed comparable to the advance speed of the head , an average advance speed would be about 0.018c which gives a maximum age of @xmath0130 myr ( konar et al .
2006 for more details ) .
it is interesting to note that only a small number of radio galaxies appear to exhibit clear signatures of episodic activity ( see table 1 ) .
a deep low - frequency search with the gmrt for evidence of fossil radio lobes which could be due to an earlier cycle of episodic activity showed no unambiguous examples in a sample of 374 sources ( sirothia et al .
most of the sources in this sample are however small , consistent with the trend that radio galaxies with evidence of episodic activity tend to be large , often over a mpc in size .
this suggests that the time scales of episodic activity are in the range of 10@xmath13 to 10@xmath14 yr . in the list of ddrgs ( table 1 ) , there is only one case of a triple - double radio galaxy ( brocksopp et al .
2007 ) , and so far no case of a ddrg associated with a quasar has been reported .
our low - frequency images of known ddrgs to search for an even earlier cycle of episodic activity ( figure 2 ) have not shown any candidate triple - double radio galaxy .
both j1453 + 3308 and 4c29.30 appear to have similar injection spectral indices for both the outer and inner doubles , although they are traversing through very different environments .
a similar trend has also been noted for pks 1545@xmath15321 ( safouris et al .
if the episodic activity is triggered by a fresh supply of gas one might find evidence of cold gas .
saikia , gupta , & konar ( 2007 ) reported the detection of hi in absorption towards the ddrg j1247 + 6723 and suggested a strong relationship between episodic activity and detection of hi gas with complex line profiles .
konar et al .
( 2006 ) and jamrozy et al .
( 2007 ) reported evidence of variability in the weak radio cores of j1453 + 3308 and 4c29.30 .
all these aspects need to be explored for a larger sample of radio sources with signs of episodic activity , to put better constraints on models of these objects .
bondi , m. , march , m. j. m. , polatidis , a. , dallacasa , d. , stanghellini , c. , & antn , s. 2004 , mnras , 352 , 112 bridle , a. h. , perley , r. a. , & henriksen , r. n. 1986 , aj , 92 , 534 brocksopp , c. , kaiser , c. r. , schoenmakers , a. p. , & de bruyn , a. g. 2007 , mnras , 382 , 1019 clarke , d. a. , bridle , a. h. , burns , j. o. , perley , r. a. , & norman , m. l. 1992 , apj , 385 , 173 giacintucci , s. , venturi , t. , murgia , m. , dallacasa , d. , athreya , r. , bardelli , s. , mazzotta , p. , & saikia , d. j. 2007 , a&a , 476 , 99 gizani , n. a. b. , & leahy , j. p. 2003 , mnras , 342 , 399 jaffe , w. j. , & perola , g. c. 1973 , a&a , 26 , 423 jamrozy , m. , konar , c. , saikia , d. j. , stawarz , . ,
mack , k .- h . , & siemiginowska , a. 2007 , mnras , 378 , 581 konar , c. , saikia , d. j. , jamrozy , m. , & machalski , j. 2006 , mnras , 372 , 693 kronberg , p. p. , wielebinski , r. , & graham , d. a. 1986 , a&a , 169 , 63 leahy , j. p. , pooley , g. g. , & riley , j. m. 1986 , mnras , 222 , 753 leahy , j. p. , black , a. r. s. , dennett - thorpe , j. , hardcastle , m. j. , komissarov , s. , perley , r. a. , riley , j. m. , & scheuer , p. a. g. 1997 , mnras , 291 , 20 machalski , j. , jamrozy , m. , zola , s. , & koziel , d. 2006 , a&a , 454 , 85 marecki , a. , barthel , p. d. , polatidis , a. , & owsianik , i. 2003 , pasa , 20 , 16 morganti , r. , killeen , n. e. b. , ekers , r. d. , & oosterloo , t. a. 1999 , mnras , 307 , 750 murgia , m. 1996 , laurea thesis , university of bologna owen , f. n. , & ledlow , m. j. 1997 , apjs , 108 , 41 perley , r. a. , bridle a. h. , willis , a. g. , & fomalont , e. b. 1980 , aj , 85 , 499 safouris , v. , subrahmanyan , r. , bicknell , g. v. , & saripalli , l. 2008 , mnras , 385 , 2117 saikia , d. j. , konar , c. , & kulkarni , v.k .
2006 , mnras , 366 , 1391 saikia , d. j. , gupta , n. , & konar , c. 2007 , mnras , 375 , l31 saripalli , l. , subrahmanyan , r. , & udaya shankar , n. 2002 , apj , 565 , 256 saripalli , l. , subrahmanyan , r. , & udaya shankar , n. 2003 , apj , 590 , 181 schilizzi , r.t . , et al . 2001 , a&a , 368 , 398 schoenmakers , a. p. , de bruyn , a. g. , rttgering , h. j. a. , van der laan , h. , & kaiser , c. r. 2000 , mnras , 315 , 371 sirothia , s. k. , saikia , d. j. , ishwara - chandra , c. h. , & kantharia , n. g. 2009 , mnras , 392 , 1403 strom , r. g. , & willis , a. g. 1980 , a&a , 85 , 36 willis , a. g. , strom , r. g. , & wilson , a. s. 1974 , nature , 250 , 625 | one of the striking examples of episodic activity in active galactic nuclei are the double - double radio galaxies ( ddrgs ) with two pairs of oppositely - directed radio lobes from two different cycles of activity .
we illustrate , using the ddrg j1453 + 3308 as an example , that observations over a wide range of frequencies using both the gmrt and the vla can be used to determine the spectra of the inner and outer lobes , estimate their spectral ages , estimate the time scales of episodic activity , and examine any difference in the injection spectra in the two cycles of activity .
low - frequency gmrt observations also suggest that ddrgs and triple - double radio galaxies are rather rare . |
population genetics is concerned with the investigation of the genetic structure of populations , which is influenced by evolutionary factors such as mutation , selection , recombination , migration and genetic drift . for excellent reviews of the theoretical aspects of this field ,
see @xcite . in this paper ,
the antagonistic interplay of mutation and selection shall be investigated , with mutation generating the genetic variation upon which selection can act .
pure mutation
selection models exclude genetic drift and are therefore deterministic models , and accurate only in the limit of an infinite population size ( for a review , see * ? ? ?
a further simplification taken here is to consider only _ haploid _ populations , where the genetic material exists in one copy only in each cell .
however , the equations used here to describe evolution apply as well to diploid populations without dominance . for the modelling of the types considered , the _ sequence space approach _
is used , which has first been used by @xcite to model the structure of the dna , where individuals are taken to be sequences . here
, the sequences shall be written in a two - letter alphabet , thus simplifying the full four - letter structure of dna sequences . in this approach
, the modelling is based on the microscopic level , at which the mutations occur , hence the mutation process is fairly straightforward to model .
however , the modelling of selection is a more challenging task , as selection acts on the phenotype , and the mapping from genotype to phenotype is by no means simple . to this end , the concept of the _ fitness landscape _
@xcite is introduced as a function on the sequence space , assigning to each possible genotype a fitness value which determines the reproduction rate .
apart from the problem that a realistic fitness landscape would have to be highly complex ( too complex for a mathematical treatment ) , there is also very limited information available concerning the nature of realistic fitness functions .
therefore , the modelling of fitness is bound by feasibility , trying to mimic general features that are thought to be essential for realistic fitness landscapes such as the degree of ruggedness .
a very common type of fitness functions is the class of permutation - invariant fitness functions , where the fitness of a sequence is determined by the number of mutations it carries compared to the wild - type , but not on the locations of the mutations within the sequence .
although this model describes the accumulation of small mutational effects surprisingly well , it is a simplistic model that lacks a certain degree of ruggedness that is thought to be an important feature of realistic fitness landscapes @xcite . in this paper ,
hopfield - type fitness functions @xcite are treated as a more complex model . here , the fitness of a sequence is not only determined by the number of mutations compared to one reference sequence , but to a number of predefined sequences , the _ patterns_. this yields a class of fitness landscapes that contain a higher degree of ruggedness , which can be tuned by the number of patterns chosen .
while this can still be treated with the methods used here , it is a definite improvement on the restriction of permutation - invariant fitness functions .
particular interest is taken in the phenomenon of mutation driven error thresholds , where the population in equilibrium changes from viable to non - viable within a narrow regime of mutation rates . in this paper , a few examples of hopfield - type fitness functions are investigated with respect to the error threshold phenomenon .
section [ the mutation selection model in sequence space ] introduces the basic mutation selection model with its main observables . in section
[ sequences as types ] , the model is applied to the sequence space approach , formulating the mutation and fitness models explicitly .
sections [ lumping for the hopfield - type fitness ] and [ the maximum principle ] present the method , which relies on a lumping of the large number of sequences into classes on which a coarser mutation selection process is formulated .
this lumping is necessary to formulate a simple maximum principle to determine the population mean fitness in equilibrium . in section [ error thresholds ] , this maximum principle is used to investigate some examples of hopfield - type fitness functions with respect to the error threshold phenomenon .
the model used here ( as detailed below ) is a pure mutation selection model in a time - continuous formulation as used by @xcite and @xcite , for instance .
[ [ population . ] ] population .
+ + + + + + + + + + + the evolution of a population where the only evolutionary forces are mutation and selection is considered , thus excluding other factors such as drift or recombination for instance .
individuals in the population shall be assigned a type @xmath0 from the finite _ type space _ @xmath1 .
the population at any time @xmath2 is described by the _ population distribution _
@xmath3 , a vector of dimension @xmath4 , the cardinality of the type space .
an entry @xmath5 gives the fraction of individuals in the population that are of type @xmath0 .
thus the population is normalised such that @xmath6 .
[ [ evolutionary - processes . ] ] evolutionary processes .
+ + + + + + + + + + + + + + + + + + + + + + + the evolutionary processes that occur are birth , death and mutation events .
birth and death events occur with rates @xmath7 and @xmath8 that depend on the type @xmath0 of the individual in question , and taken together , they give the effective reproductive rate , or _ fitness _ as @xmath9 .
mutation from type @xmath0 to type @xmath10 depends on both initial and final type and happens with rate @xmath11 .
these rates are conveniently collected in square matrices @xmath12 and @xmath13 of dimension @xmath4 , where the reproduction or fitness matrix @xmath12 with entries @xmath14 is diagonal . the off - diagonal entries of the mutation matrix @xmath13 are given by the mutation rates @xmath11 , and as mutation does not change the number of individuals , the diagonal entries of @xmath13 are chosen such that @xmath15 , which makes @xmath13 a markov generator .
the time evolution operator @xmath16 is given by the sum of reproduction and mutation matrix , @xmath17 .
[ [ deterministic - evolution - equation . ] ] deterministic evolution equation .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + in the deterministic limit of an infinite population size , the evolution of the population is governed by the evolution equation @xmath18{\boldsymbol{p}}(t ) \;,\ ] ] where @xmath19 is the population mean fitness . the term with @xmath20
is needed to preserve the normalisation of the population .
note that this term makes the evolution equation ( [ evolution equation ] ) nonlinear .
[ [ equilibrium . ] ] equilibrium .
+ + + + + + + + + + + + the main interest focuses on the equilibrium , i.e. , the behaviour if @xmath21 , which is attained for @xmath22 .
all equilibrium quantities shall be denoted by omitting the argument @xmath2 , for instance the equilibrium population distribution is @xmath23 . in equilibrium ,
the evolution equation ( [ evolution equation ] ) becomes an eigenvalue equation for @xmath16 , with leading eigenvalue @xmath20 and corresponding eigenvector @xmath23 .
if @xmath13 is irreducible , as shall be assumed throughout , perron - frobenius theory ( see , for instance , * ? ?
* appendix ) applies , which guarantees that the leading eigenvalue @xmath20 of @xmath16 is non - degenerate and the corresponding right eigenvector @xmath23 is strictly positive , which implies that it can be normalised as a probability distribution . [
[ ancestral - distribution . ] ] ancestral distribution .
+ + + + + + + + + + + + + + + + + + + + + + + similarly to the population distribution @xmath23 , there is also another important distribution in this model , namely the ancestral distribution @xmath24 .
consider the population at time @xmath25 , but count each individual not as its current type , but as the type its ancestor had at time @xmath2 .
thus an entry of the ancestral distribution @xmath26 determines the fraction of the population at @xmath25 whose ancestor at time @xmath2 was of type @xmath0 . in the limit @xmath27 ,
this also approaches an equilibrium distribution @xmath28 .
as shown by @xcite , the equilibrium ancestral distribution can be obtained as a product of the left and right pf ( perron - frobenius ) eigenvectors @xmath29 and @xmath23 of the time - evolution operator @xmath16 as @xmath30 , where @xmath29 is normalised such that @xmath31 . [ [ population - and - ancestral - means . ] ]
population and ancestral means .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + any function on the type space given , say , by @xmath32 can be averaged with respect to the population or the ancestral distribution distribution .
the population mean of @xmath33 is given by @xmath34 whereas the ancestral mean is @xmath35 note that the time - dependence of the means only comes from the distribution , whereas the function @xmath33 is considered constant in time . in equilibrium , time dependence is again omitted such that the equilibrium population and ancestral means are denoted @xmath36 and @xmath37 , respectively .
an important example of the population mean is the population mean fitness @xmath38 from equation ( [ evolution equation ] ) .
in the previous section , the types are a rather abstract concept . in order to formulate the particular mutation and fitness models ,
they shall now be specified as sequences , mimicking the structure of the dna ( cf .
* ) . for simplicity ,
only two - state sequences are considered , i.e. , sequences that have at each site one out of two possible entries . however , the method used here can immediately be generalised to a more realistic four - state model ( see * ? ? ? * ) .
the types therefore are associated with sequences @xmath39 of fixed length @xmath40 , written in the _ alphabet _ @xmath41 , thus @xmath42 for @xmath43 .
this means that there are @xmath44 different sequences , and thus the type space ( or _ sequence space _ )
@xmath1 has cardinality @xmath45 .
a simple mutation model that neglects any processes changing the length of the sequence , such as deletions or insertions , is used .
mutations are modelled as point processes , where an arbitrary site @xmath46 is switched with rate @xmath47 , such that the mutation rate between sequences that differ only in one particular site is given by @xmath48 .
sequences that differ in more than one site can not mutate into one another within a single mutational step .
this is known as the _ single step mutation model _
, introduced by @xcite .
the mutation model defines a neighbourhood in the sequence space @xcite .
a convenient measure for the distance between sequences is the hamming distance @xmath49 , which counts the number of sites at which the sequences @xmath50 and @xmath51 differ @xcite . with this
, the mutation matrix is explicitly given as @xmath52 the diagonal entry is chosen such that @xmath13 fulfils the markov condition @xmath53 .
a rather simple , though commonly used type of fitness function is the _ permutation - invariant fitness_. there , the fitness of a sequence depends only on the number of mutations it has compared to a reference type , not on their position along the sequence .
thus fitness is a function of the hamming distance to the reference sequence , which is usually chosen as the wild - type .
the hamming distance to the wild - type is also called the _ mutational distance _ @xmath54 . a non - permutation - invariant fitness that contains some ruggedness , but is simple enough to be dealt with in this framework , is the _ hopfield - type fitness _ , a special type of spin - glass model , which has been introduced by @xcite as a model for neural networks . instead of comparing a sequence only to the wild - type , as it is done for the permutation - invariant fitness , the hopfield - type fitness of a sequence
is determined by its hamming distances to @xmath55 reference sequences , the _ patterns _ @xmath56 , @xmath57 .
the hopfield - type fitness shall be defined in terms of the _ specific distances _
@xmath58 , which are the hamming distances with respect to the patterns , @xmath59 and thus the fitness is given as @xmath60 note that in the case of a single pattern ( @xmath61 ) , this yields again a permutation - invariant fitness .
one problem of the sequence space approach is the large number of types , which grows exponentially with the sequence length @xmath40 , @xmath45 .
the time - evolution operator @xmath16 is a matrix of size @xmath62 , and in this set - up one is interested in its leading eigenvalue @xmath20 and the corresponding right and left eigenvectors @xmath23 and @xmath29 .
the relevant sequence length depends on the particular application one has in mind , but it is typically rather long . if one aims to model the whole genome of a virus or a bacterium , @xmath40 has to be in the region of @xmath63 , but even a single gene has of the order of @xmath64 base pairs . these values lead to matrices of a size that makes the eigenvalues and eigenvectors inaccessible . for some types of fitness functions ,
this problem can be reduced by _ lumping _ together types into _ classes _ of types , and considering the new process on a reduced sequence space , which contains the classes rather than the individual types . under certain circumstances ,
mutation is described as a markov process in the emerging lumped process as well , such that this process is accessible to markov process methods , and the framework developed in section [ the mutation selection model in sequence space ] can directly be applied to the lumped system .
the lumping of the mutation process is a standard procedure in the theory of markov chains @xcite , see also @xcite for an application to mutation selection models .
this lumping leads to a meaningful mutation selection model on the reduced type space , if all sequences lumped together into one class have the same fitness .
it is possible to lump the markov chain given by the mutation matrix @xmath13 with state space @xmath1 with respect to a particular partition @xmath65 , if and only if for each pair @xmath66 the cumulative mutation rates @xmath67 from type @xmath68 into @xmath69 , are identical for all @xmath68 , cf . the example shown in figure [ fig lumping ] .
( 258,171 ) ( 15 , 5)(0,0)@xmath70 ( 200 , 30)(0,0)@xmath69 ( 23 , 90)(0,0)@xmath71 ( 25 , 40)(0,0)@xmath72 visualisation of the compatibility with lumping : consider two classes @xmath70 and @xmath69 .
the mutation rates from the types in @xmath70 to the types in @xmath69 ( given next to the arrows ) are compatible with a lumping with respect to @xmath70 and @xmath69 , because the sum of the mutation rates from type @xmath71 to all types in @xmath69 is given by @xmath73 , which is identical with those from @xmath72 , @xmath74 .
, title="fig:",scaledwidth=60.0% ] in this case , the lumped process , with @xmath75 states @xmath76 and mutation rates @xmath77 for any @xmath68 , is again a markov chain ( * ? ? ?
* theorem 6.3.2 ) . whereas in the case of a permutation - invariant fitness function ,
the lumping procedure is fairly simple , collecting all sequences with the same hamming distance to the wild - type into classes , and considering cumulative mutation rates between these classes , the lumping for the hopfield - type fitness is somewhat more complex .
for a two - state model with hopfield fitness , it has been performed for instance by @xcite , and this shall be recollected in the remainder of this section .
first , the quantities with respect to which the lumping shall be performed must be defined . to this end , consider as an example the case of sequence length @xmath78 with three patterns ( @xmath79 ) , and let the patterns @xmath56 be collected in a @xmath80 matrix @xmath81 , such that the @xmath82th row of @xmath81 is pattern @xmath56 . without loss of generality , the pattern @xmath83 can always be chosen as @xmath84 . let the patterns in this example be given as @xmath85 note that there are only @xmath86 different types of sites ( corresponding to the columns of @xmath81 ) .
these are collected in a @xmath87 matrix @xmath88 , the columns of which correspond to the possible types of sites in the matrix of patterns @xmath81 . for the case
@xmath79 , this @xmath89 matrix is given as @xmath90 using the column vectors @xmath91 of the matrix @xmath88 , the patterns given in equation ( [ example pattern ] ) can alternatively be expressed as @xmath92 classifying the sites into @xmath86 classes according to which of the column vectors @xmath91 of @xmath88 coincides with the column vector @xmath93 of the patterns @xmath81 at site @xmath94 .
let @xmath95 be the index set of sites , with a partition into @xmath86 subsets @xmath96 induced by the patterns @xmath81 such that @xmath97 with @xmath98 patterns can be characterised by the number of sites @xmath99 in each subset . the example patterns ( [ example pattern ] ) can therefore be described by @xmath100 considering only subsequences @xmath101 , the _ partial distances _
@xmath102 of a sequence @xmath50 with respect to the pattern @xmath83 are defined as the hamming distance between @xmath50 and @xmath83 , _ restricted to the subsequence @xmath103_. therefore , they can be written as @xmath104 such that the specific distance with respect to pattern @xmath83 is given by @xmath105 . because the differences between each of the patterns @xmath56 within the subsets @xmath96 of @xmath106 are known ( and recorded in the matrix @xmath88 )
, it is sufficient to consider only the partial distances @xmath102 with respect to one pattern , here @xmath83 ; the partial distances @xmath107 with respect to any other pattern can be expressed in terms of the @xmath102 as @xmath108 using the matrix elements @xmath109 of @xmath88 .
the specific distance @xmath58 to any pattern @xmath56 can be expressed as @xmath110 where the index set @xmath111 of classes is partitioned into two subsets , @xmath112 with @xmath113 and @xmath114 .
hence , by specifying the @xmath86 partial distances @xmath102 with respect to pattern @xmath83 , the specific distances @xmath58 with respect to any pattern @xmath56 are determined , which in turn determine the fitness .
this implies that all sequences with the same partial distances @xmath102 have the same fitness .
thus the partial distances @xmath102 to pattern @xmath83 , collected in a _ mutational distance _
vector @xmath115 , shall be the quantities that label the classes in the lumped system . the relevant partition of the sequence space is given by @xmath116 with @xmath117 , and the reduced sequence space , or _
mutational distance space _ @xmath118 , contains the classes @xmath119 .
considering again the subsequences @xmath120 , there are @xmath121 possible different @xmath102 ( as @xmath102 takes values from @xmath122 to @xmath123 ) , and there are @xmath124 different subsequences @xmath103 for each @xmath102 . hence , considering all sites , there are @xmath125 , and @xmath126 sequences @xmath50 that are mapped onto each @xmath127 . for the patterns
@xmath81 chosen as example ( [ example pattern ] ) , we have @xmath128 , while the full sequence space has dimension @xmath129 . in the single step mutation model ,
the only neighbours of a sequence @xmath50 with distance vector @xmath127 lie in the classes @xmath130 , where the @xmath131 are the unit vectors of mutation .
thus the only non - zero cumulative mutation rates are @xmath132 as a sequence with @xmath127 has @xmath133 @xmath122-sites and @xmath102 @xmath134-sites in @xmath96 , and the single - site mutation rates are @xmath48 for all sites , the cumulative mutation rates are given by @xmath135 irrespective of the particular order within the subsequences @xmath103
. therefore the cumulative mutation rates @xmath136 are the same for all sequences @xmath50 with the same @xmath127 , which is the condition for lumping .
the mutation selection model with hopfield - type fitness
is indeed `` lumpable '' with respect to the partition induced by the distance vectors @xmath127 . the mutation
selection process on the mutational distance space @xmath118 is described by the lumped time - evolution operator @xmath137 with lumped reproduction and mutation matrices @xmath138 and @xmath139 of dimension @xmath140 .
whereas the lumped reproduction matrix @xmath138 is still diagonal and contains the same entries as @xmath12 , i.e. @xmath141 , the off - diagonal entries of the lumped mutation matrix @xmath139 are given by the cumulative mutation rates with unchanged diagonal entries compared to @xmath13 , which still fulfil the markov property @xmath142 , and thus @xmath143 with the cumulative mutation rates from equation ( [ cumulative mutation rates ] ) .
for a more general derivation of the lumped reproduction and mutation matrices , see @xcite .
note that the time - evolution operator @xmath144 acting on @xmath118 describes the evolution of a population under mutation and selection determined by the evolution equation ( [ evolution equation ] ) , and thus the theory developed in section [ the mutation selection model in sequence space ] applies .
although the lumping procedure reduces the number of types very efficiently , the evaluation of the eigenvalues and eigenvectors of the time - evolution operator @xmath144 still remains a difficult problem for many applications , due to the size of the eigenvalue problem .
if one is interested solely in the equilibrium behaviour of the system , however , it is possible to determine the population mean fitness ( at least asymptotically for large sequence length @xmath40 ) , given by the leading eigenvalue of @xmath144 .
this can be done by a simple maximum principle that can be derived from rayleigh s general maximum principle , which specifies that the leading eigenvalue @xmath145 of an @xmath146 matrix @xmath144 can be obtained via a maximisation over @xmath147 , @xmath148 the vector @xmath149 for which the supremum is attained is the eigenvector corresponding to the eigenvalue @xmath145 .
the simple maximum principle derived from this guarantees that the population mean fitness @xmath20 can be obtained by maximising a function on the mutational distance space @xmath118 .
it can be shown that the maximiser itself is the ancestral mean mutational distance .
such a maximum principle has first been derived by @xcite for two - state sequences with permutation - invariant fitness .
this has been generalised to apply for four - state sequences with permutation - invariant fitness by @xcite , and subsequently the restriction to permutation - invariant fitness function has been relaxed by @xcite .
the results from @xcite apply directly to the hopfield - type fitness treated here . whereas the original mutation matrix @xmath13 is symmetric
, the lumped mutation matrix @xmath139 is no longer symmetric , as different numbers of sequences are lumped into the different classes , therefore giving rise to unequal cumulative forward and backward mutation rates . to derive the maximum principle , it is necessary to symmetrise the mutation matrix @xmath139 .
@xmath139 is reversible , i.e. , @xmath150 where @xmath151 is the stationary distribution of the pure mutation process , which is given by the equidistribution of types on @xmath1 , and thus given by the number of sequences @xmath152 that are lumped onto the same mutational distance vector @xmath127 .
the reversibility of @xmath139 implies that it can be symmetrised by the means of a diagonal transformation @xmath153 , which yields the symmetrised mutation matrix as @xmath154 with off - diagonal entries @xmath155 using the cumulative mutation rates @xmath156 , this reads @xmath157 as can be seen by using the explicit representation for the cumulative mutation rates from equation ( [ cumulative mutation rates ] ) with the @xmath152 from equation ( [ nd ] ) .
because @xmath158 is diagonal , the diagonal entries of the mutation matrix are unchanged , @xmath159 as @xmath138 is diagonal as well , it is not changed by the transformation @xmath160 , and thus this transformation also symmetrises the time - evolution operator such that @xmath161 is symmetric . before symmetrisation , @xmath144 was expressed as the sum of a markov generator @xmath139 and a diagonal remainder @xmath138 . as the transformation @xmath158 does not preserve the markov property ,
this is not the case for the symmetrised time - evolution operator in ( [ symmetrised time - evolution operator ] ) .
it is however useful to split it up this way . to this end , let @xmath162 where @xmath163 is a ( symmetric ) markov generator and @xmath164 is the ( diagonal ) remainder .
the off - diagonal entries of @xmath163 are given by those of @xmath165 from equation ( [ symmetrised mutation rates ] ) , @xmath166 for @xmath167 , whereas the markov property requires as diagonal entries @xmath168 the remainder @xmath164 is given by @xmath169 to deal with the case of infinite sequence length , it will prove useful to use intensively scaled normalised versions of the extensively scaled variables like the mutational distances .
the pattern in the hopfield model , previously characterised by the _ number _ of sites @xmath123 in each subset @xmath96 , will now be described by the _ fraction _ of sites in @xmath96 , given by @xmath170 .
similarly , we use normalised partial distances @xmath171 where @xmath172 $ ] , with the normalised mutational distance vector @xmath173 .
the permutation - invariant model is obtained in the case @xmath61 . for finite @xmath40
, @xmath174 takes rational values in a normalised version of the mutational distance space @xmath175 , where @xmath176 is a compact domain in @xmath177 . for @xmath178 ,
the vectors @xmath174 become dense in @xmath176 .
assume that the entries of @xmath179 can be approximated by functions @xmath180 and @xmath33 from @xmath181 , i.e. , twice continuously differentiable functions with bounded second derivatives that map @xmath176 onto @xmath182 such that @xmath183 where @xmath184 and the notation @xmath185 is used to emphasise that the normalised mutational distance @xmath174 corresponding to a particular @xmath127 is meant .
in fact , assumption ( [ assumption f ] ) can readily be verified for the cumulative mutation rates from equation ( [ cumulative mutation rates ] ) : let the functions @xmath186 be @xmath187 where the functions @xmath188 are given by @xmath189 with the cumulative mutation rates @xmath190 , which read explicitly @xmath191 using a taylor approximation , it can be shown that the differences between the exact entries of @xmath192 as given in equations ( [ symmetrised mutation rates ] ) and ( [ diagonal entries of f ] ) , and their approximations from equation ( [ function f ] ) , are indeed of @xmath193 .
assuming that also the reproduction rates @xmath194 can be approximated by a @xmath195 function @xmath196 as @xmath197 then from equation ( [ remainder e ] ) , the matrix @xmath164 is approximated by @xmath198 fulfilling equation ( [ assumption e ] ) . with the definition of the _ mutational loss function _ @xmath199 as @xmath200 \,,\end{aligned}\ ] ] this reads explicitly @xmath201 we are now in a position to apply theorems 1 and 2 from @xcite , which read for the mutation selection model with hopfield - type fitness considered here + _ ( i )
assume that for the lumped mutation selection model as set up in section [ lumping for the hopfield - type fitness ] it is possible to approximate the reproduction rates @xmath194 by a @xmath195 function @xmath202 as specified in equation ( [ assumption r ] ) , and that the @xmath195 function @xmath180 assumes its global maximum in the interior _
@xmath203_. then the population mean fitness in equilibrium is given by @xmath204 + { \mathcal{o}\left(\frac{1}{n}\right)}\,.\ ] ] ( ii ) assume furthermore that @xmath180 assumes its maximum at a unique point @xmath205 in _ @xmath203 _ , and that the hessian of @xmath180 at @xmath205 is negative definite . then in the limit of @xmath178 , the maximiser @xmath205 is given by the mean ancestral mutational distance @xmath206 , and in particular @xmath207 _ for a proof the reader is referred to @xcite .
the maximum principle ( [ maximum principle ] ) and ( [ ancestral maximum principle ] ) is a powerful tool to calculate the population mean fitness @xmath20 in equilibrium for arbitrary fitness functions of the permutation - invariant or hopfield type , for any range of mutation rates .
also , the ancestral mean genotype @xmath206 is available .
the general method to identify @xmath20 and @xmath206 is to consider the partial derivatives of @xmath208 with respect to the components @xmath209 of the mutational distance @xmath174 .
a necessary condition for the function @xmath208 to have a maximum at a value @xmath205 is that its derivatives at this @xmath205 vanish , @xmath210_{{\boldsymbol{x}}={\boldsymbol{x}}^ * } = 0 \quad\forall\ , v , k \;.\ ] ] the global maximum of the function @xmath208 must lie on one of the points @xmath205 that fulfil equation ( [ error threshold general method ] ) or on the border of the mutational distance space .
thus by comparing the values of @xmath208 on these possible points , the global maximum can be identified . apart from the general possibility to investigate the dependence of the population mean fitness @xmath20 on the mutation rate @xmath47 , this yields the opportunity to investigate the phenomenon of the _ error threshold _ , which has interested scientists ever since it was first conceived by @xcite .
the phenomenon of the error threshold can be described as the existence of a critical mutation rate , below which the equilibrium population is well localised in sequence space , whereas for mutation rates above the critical mutation rate , the equilibrium population is more delocalised , with a sharp transition between the two phases .
one problem is that there is no generally accepted definition of an error threshold .
the criterion used in the original quasispecies model @xcite is the disappearance of the wild - type from the population , which under the single peaked landscape used there goes in line with the complete delocalisation of the population in sequence space .
however , these two effects do not necessarily coincide for other fitness landscapes .
the definition of the error threshold that shall be used here is equivalent to the definition of a phase transition in physics , differentiating between first and second order transitions as follows : * definition ( first and second order error threshold ) . * + ( i ) a first order error threshold exists at a critical mutation rate @xmath211 , if the ancestral mean mutational distance as a function of the mutation rate @xmath212 shows a discontinuity at this @xmath211 , which is also reflected by a kink in the population mean fitness @xmath213 .
+ ( ii ) a second order error threshold exists at a critical mutation rate @xmath211 , if the ancestral mean mutational distance is continuous , but its derivative with respect to the mutation rate @xmath214_{\mu\rightarrow\mu_c}$ ] is discontinuous at this mutation rate @xmath211 . in the examples shown later in this thesis , the second order error threshold always show an infinite derivative at the critical mutation rates .
note that , like phase transitions in physics , these definitions of the error thresholds apply in the strict sense only to a system with infinite sequence length ( @xmath178 ) , for finite sequence lengths , the thresholds are smoothed out due to the lack of non - analyticities .
@xcite gave a finer classification of different error threshold phenomena .
the first order error threshold they called `` fitness threshold '' . here
, this term shall include also the second order error threshold , making all error thresholds as defined above fitness thresholds .
furthermore , the concept of the `` degradation threshold '' was introduced : * definition ( degradation threshold ) . *
+ a degradation threshold is an error threshold of first or second order , where the population distribution beyond the critical mutation rate @xmath211 is given by the equidistribution in sequence space @xmath1 .
thus here the degradation threshold is a special case of a fitness threshold , going in line with the complete delocalisation of the population in sequence space .
note that in the limit of infinite sequence length ( @xmath178 ) , for which the error threshold definitions apply exactly , this equidistribution is reached immediately above @xmath211 , and beyond the threshold the population is insensitive to any further increase in mutation rates . in the case of finite sequence lengths , where the thresholds are smoothed out ,
the equidistribution is of course only reached asymptotically .
the original error threshold was observed for the single peaked fitness landscape , where a single sequence is attributed a high fitness value , all other sequences are equally disadvantageous ( for a review , see * ? ? ?
this is clearly an oversimplification and should not be regarded as anything but a toy model .
other fitness landscapes that have been investigated comprise , in the permutation - invariant case , linear and quadratic fitness functions , general functions showing epistasis , and as examples lacking permutation - invariance the onsager landscape @xcite , which has nearest neighbour interactions within the sequence , as well as various spin glass landscapes like the hopfield landscape @xcite , the sherrington - kirkpatrick spin glass @xcite , the nk spin glass @xcite , and the random energy model @xcite , assigning random fitness values to each sequence .
one fitness landscape where an analytical solution can be obtained is the linear fitness ( cf . * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
note that this corresponds to a multiplicative landscape in a set - up using discrete time . for a linear fitness function
, there is no error threshold , but the population changes smoothly from localised to delocalised with an increasing mutation rate . for quadratic fitness functions ,
error thresholds only exist for antagonistic epistasis ; they are absent for quadratic fitness functions with synergistic epistasis @xcite .
these results go in line with those for general epistatic fitness functions @xcite .
studies using non - permutation - invariant fitness functions generally report the presence of error thresholds .
of course the discussion of the error threshold phenomenon is academic if the threshold is an artifact of the model rather than a real biological phenomenon .
this issue has been subject to numerous debates , especially because it has first been predicted by a model using the over - simplistic single peaked landscape .
however , over the years biologists have accumulated evidence that particularly rna viruses naturally thrive at very high mutation rates @xcite , of the order of @xmath215 to @xmath216 per base per replication @xcite , corresponding to a genomic mutation rate of about 0.1 to 10 mutations per replication @xcite , and a number of studies have reported that populations of rna viruses only survive a moderate increase of their mutation rate , whereas if the mutation rate is increased further , the populations become extinct @xcite , for reviews see @xcite .
this corresponds to the population being pushed beyond the error threshold .
it has been suggested to use the error threshold for anti - viral therapies @xcite , and in fact , recent experimental results indicate that this is the mechanism via which the broad - spectrum anti - viral drug ribavirin works @xcite .
this clearly warrants some further investigation of the error threshold phenomenon , which shall be done in the remainder of this section .
the original hopfield fitness as introduced by @xcite is a quadratic function of the specific distances and reads @xmath217 using normalised specific distances @xmath218 , similarly to the normalised mutational distances @xmath174 .
the statistical properties of this landscape have been studied in detail @xcite : in the thermodynamic limit @xmath219 , there are @xmath220 global maxima that are associated with the patterns @xmath221 and their complements @xmath222 .
in addition to that , the number of local maxima and saddle points grows exponentially with the number patterns @xmath55 , hence the ruggedness of the fitness landscape can be tuned by the number of patterns .
most works that have studied a hopfield - type fitness used the original hopfield model , a generalisation was however treated by @xcite , using a hopfield - type truncation selection with two patterns .
thus it might be interesting and instructive to investigate the threshold behaviour of different kinds of hopfield - type fitness functions .
applying criteria for the existence of error thresholds that have been obtained by @xcite for permutation - invariant fitness functions to the case of a hopfield - type fitness , it can be shown that for linear hopfield - type fitness functions there are no error thresholds , which is a new result , considering that for all previously investigated hopfield - type fitness functions , the existence of error thresholds was reported .
the next step towards more complex fitness functions is to consider quadratic fitness functions , generalising the original hopfield - fitness , which is a particular example for a quadratic function . here
, the analysis shall be restricted to a symmetry with respect to the normalised specific distances @xmath223 to the patterns @xmath56 , such that @xmath224 the parameter @xmath225 tunes the linear in relation to the quadratic term , and the sign of the quadratic term determines the _ epistasis _ , a measure for the strength of interaction between sites . for a positive quadratic term epistasis
is said to be negative or antagonistic , whereas for a negative quadratic term one speaks of positive or synergistic epistasis .
the case @xmath226 combined with a positive quadratic term ( i.e. , negative epistasis ) yields the original hopfield fitness . in the case of two patterns , @xmath227 ,
the first pattern can be chosen without loss of generality as @xmath228 , such that there is only one pattern to be chosen , usually randomly .
the matrix @xmath88 containing the possible types of sites is given by @xmath229 and thus the index set of sites is partitioned into two subsets , @xmath230 , where @xmath231 contains all sites at which both patterns have entry @xmath122 , whereas @xmath232 corresponds to the sites where the two patterns have entries @xmath122 and @xmath134 , respectively .
the only quantities characterising the patterns are now the fractions of sites in each partition , @xmath233 and @xmath234 .
thus the pattern can be characterised by a single parameter , @xmath235 .
each sequence is characterised with respect to the pattern by the partial hamming distances to pattern @xmath83 ( in normalised form ) , @xmath236 and @xmath237 .
these vary from @xmath122 ( all entries @xmath122 in @xmath96 ) to @xmath134 ( all entries @xmath134 in @xmath96 ) , completely independently from each other . the specific distances with respect to the patterns
are linear combinations of the @xmath209 and given in normalised form by @xmath238 the hopfield - type fitness is defined as an arbitrary function of these patterns , @xmath239 . due to the small number of variables , for the case of two patterns , a lot can be done by analytical treatment . for the quadratic symmetric hopfield - type fitness ( [ quadratic symmetric hopfield - type fitness ] ) with positive epistasis ( i.e. , negative sign of the quadratic term ) and @xmath240 , there are no phase transitions , going in line with the results for permutation - invariant fitness functions , but different from other results for hopfield - type fitness functions . as an example for negative epistasis , consider first the original hopfield fitness ( [ original hopfield fitness ] ) .
since for two patterns there are only two variables , it is possible to visualise the fitness landscape in this case .
figure [ fig 6.14 ] shows the original hopfield fitness ( [ original hopfield fitness ] ) for the cases @xmath241 and @xmath242 . original hopfield fitness in the case of two patterns with different @xmath243.,title="fig:",scaledwidth=45.0% ] original hopfield fitness in the case of two patterns with different @xmath243.,title="fig:",scaledwidth=45.0% ] in the corners of the mutational distance space @xmath118
, one can see the four degenerate maxima .
the ancestral mean partial distances @xmath244 , at which the maxima of @xmath208 are positioned , are obtained by considering the derivatives of @xmath208 .
they are given by @xmath245 for the case of @xmath241 , which corresponds to two completely uncorrelated patterns , the ancestral mean partial distances @xmath244 are shown in figure [ fig 6.15 ] on the left , alongside the ancestral mean specific distances @xmath246 . ancestral mean
partial distances @xmath244 ( top ) and specific distances @xmath247 ( bottom ) depending on mutation rates . the original hopfield fitness ( [ original hopfield fitness ] ) for
two patterns has been used .
results correspond to uncorrelated patterns ( @xmath241 , left ) and correlated patterns with with @xmath242 ( right ) . ] for low mutation rates , there are two possible solutions for each of the ancestral mean partial distances @xmath244 , and as the maxima are degenerate , in equilibrium , the population will be centred equally around all of them
. however , in the approach to equilibrium , the population might well be predominantly concentrated around one of them , depending on initial conditions .
the specific distances @xmath223 that are shown correspond to the combination of @xmath248 and @xmath249 , where both are given by the lower branch .
other combinations yield similar results . for high mutation rates ,
the population is in the mutation equilibrium with @xmath250 , forming a disordered phase . in the limit of low mutation rates , @xmath251 ,
the population is always in the vicinity of one of the patterns ( or its complement ) , such that one of the @xmath252 , which is completely random with respect to the other pattern , and thus the other @xmath253 .
this is the ordered phase . at the critical mutation rate @xmath254
, there is a second order phase transition between these two phases , which is a fitness as well as a degradation threshold , corresponding to the infinite derivative of both @xmath244 at this mutation rate .
as the specific distances @xmath223 are simply superpositions of the partial distances @xmath209 , the phase transitions are also visible in the @xmath223 . in the correlated case @xmath255 ( figure [ fig 6.15 ] , right ) , two second order transitions can be identified . at @xmath256
, @xmath248 has a phase transition , whereas at @xmath257 , @xmath249 has a phase transition .
the threshold occurring at the lower mutation rate is only a fitness threshold , whereas the one happening at the higher mutation rate is both a fitness and degradation threshold , leading to a totally random population . for @xmath258 ,
the population is in an ordered phase , for @xmath259 , it is in a partially ordered phase , which is ordered with respect to one of the variables , but random with respect to the other . finally , for @xmath260 , the population is the equidistribution in sequence space . here again , for low mutation rates the population is close to one of the patterns , but due to the correlation in the chosen patterns , this leads to a non - random overlap with the other pattern . in the uncorrelated case with @xmath241 ,
the two error thresholds coincide , and the partially ordered phase vanishes .
now turn to the question how these phase transitions depend on the particular degeneracy of the fitness functions and consider the quadratic fitness function ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis ( i.e. , positive quadratic term ) for values of @xmath261 .
figure [ fig 6.16 ] shows the fitness landscapes for values of @xmath262 and @xmath263 for uncorrelated patterns @xmath241 and correlated patterns with @xmath242 .
quadratic hopfield - type fitness functions ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis and @xmath262 ( top ) and @xmath263 ( bottom ) for an uncorrelated pattern @xmath241 ( left ) and a correlated pattern with @xmath242 ( right).,title="fig:",scaledwidth=47.0% ] quadratic hopfield - type fitness functions ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis and @xmath262 ( top ) and @xmath263 ( bottom ) for an uncorrelated pattern @xmath241 ( left ) and a correlated pattern with @xmath242 ( right).,title="fig:",scaledwidth=43.0% ] quadratic hopfield - type fitness functions ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis and @xmath262 ( top ) and @xmath263 ( bottom ) for an uncorrelated pattern @xmath241 ( left ) and a correlated pattern with @xmath242 ( right).,title="fig:",scaledwidth=47.0% ] quadratic hopfield - type fitness functions ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis and @xmath262 ( top ) and @xmath263 ( bottom ) for an uncorrelated pattern @xmath241 ( left ) and a correlated pattern with @xmath242 ( right).,title="fig:",scaledwidth=43.0% ] as the pictures indicate , the fitness functions ( and thus the behaviour of the system ) with the same @xmath264 are related by symmetry under @xmath265 ( apart from a constant term , which does not influence the dynamics ) .
note that in @xmath237-direction , the fitness function is independent of @xmath225 .
this is because in the sum of the specific distances , the term with @xmath237 cancels out , which happens generally in the case of an even number of patterns ( i.e. , odd @xmath266 ) for @xmath267 different variables . because in @xmath237-direction the fitness is independent of @xmath225 , the solution for the ancestral mean mutational distance @xmath249 is identical with the solution for the original hopfield fitness as given in equation ( [ two - state two - pattern analytical solution ] ) .
so for all values of @xmath225 , the phase transition with respect to @xmath237 happens at @xmath257 .
for @xmath236 , the solution becomes more complicated , but the inverse function is simpler .
it is given by @xmath268\sqrt{\hat{x}_1(1-\hat{x}_1 ) } } { 2\hat{x}_1 - 1 } \,.\ ] ] the dependence of @xmath248 on the mutation rate is shown in figure [ fig 6.17 ] ( top ) .
ancestral mean partial distance @xmath248 ( top ) and specific distances @xmath269 ( thick lines , bottom ) and @xmath270 ( thin lines , bottom ) depending on mutation rates . the quadratic hopfield - type fitness ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis for two patterns and different values of @xmath225 has been used .
results correspond to uncorrelated patterns @xmath241 ( left ) and correlated patterns with @xmath242 ( right ) .
data are shown for parameter values of @xmath271 ( top to bottom ) .
for clarity , only specific distances @xmath247 corresponding to @xmath272 are shown . ] for @xmath261 , the second order phase transition is smoothed out and thus vanishes . note that the ambiguity in the solutions that exists in the case @xmath226 ( cf .
figure [ fig 6.15 ] ) , does not exist here , due to the lacking degeneracy of the maxima of the fitness function at @xmath273 and @xmath274 ( cf .
figure [ fig 6.16 ] ) . at the bottom ,
figure [ fig 6.17 ] shows the specific distances @xmath247 , using the lower branch of the solution for @xmath249 ( as shown in figure [ fig 6.15 ] ) , which show the second order transition in @xmath249 , a fitness threshold . with this combination of solutions , for low mutation rates
the population is centred around the sequence complementary to pattern @xmath275 .
the general picture for the uncorrelated ( @xmath276 ) and correlated ( @xmath277 ) choice of patterns is very similar , apart from issues like the exact location of the thresholds .
the behaviour of the system with quadratic symmetric hopfield - type fitness has been investigated for three , four and five patterns .
however , due to the complexity of the analysis for a higher number of patterns , the focus is on the case of three patterns . for three patterns ,
the matrix @xmath88 reads @xmath278 thus there are four @xmath279 describing the patterns @xmath81 , fulfilling @xmath280 , and four variables @xmath172 $ ] , describing each sequence .
the specific distances with respect to pattern @xmath56 are given by @xmath281 similarly to the case of two patterns , the original hopfield fitness ( [ original hopfield fitness ] ) shall be considered first , and then variations ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis , but @xmath261 .
the investigation is done by means of numerical calculations . for the hopfield - type fitness , the number of classes in the lumped system is given by @xmath282 . now consider the case where the patterns are chosen randomly with equal probability for each letter at each site .
this results in a multinomial probability distribution for the number of sites @xmath123 in each subset @xcite , @xmath283 where @xmath284 .
the means are given by @xmath285 and the variance @xmath286 such that @xmath287 .
if the patterns @xmath56 for @xmath288 are chosen randomly ( remember @xmath228 ) , the @xmath279 are thus given by @xmath289 .
so to mimic the case of an infinite sequence length , in which the maximum principle is exact , one has to assume uncorrelated patterns with @xmath290 for all @xmath291 .
however , the maximum principle can also be used to investigate the case of finite sequence length by simulating the finite sequence length through choosing pattern distributions that do not follow exactly the infinite distribution @xmath290 , but vary around this mean value with a variance according to the sequence length to be simulated .
practically , patterns corresponding to finite sequence length @xmath40 have been obtained by choosing for the @xmath292 sites entries @xmath122 or @xmath134 with probability 1/2 at each site , and counting the number of sites @xmath123 in each class @xmath96 , similarly to the example pattern given in equation ( [ example pattern ] ) .
thus although the patterns were chosen randomly , they are correlated due to the finite sequence length . as shall be seen in the following , these correlations
do account for some interesting additional features . however , focus first on the case of a genuinely infinite sequence length with @xmath290 for all @xmath293 .
[ [ original - hopfield - fitness - c-1 . ] ] original hopfield fitness ( @xmath226 ) .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + the results for the case of three , four and five patterns with original hopfield fitness ( [ original hopfield fitness ] ) and uncorrelated patterns ( @xmath290 for all @xmath293 , corresponding to random patterns for infinite sequence length ) look exactly like those for two patterns shown in figure [ fig 6.15 ] ( left ) . the solutions for the different @xmath209 all coincide . for small mutation rates
@xmath294 , there are again two degenerate solutions for each @xmath209 , which can be combined in multiple ways for the different @xmath209 , yielding any of the patterns or their complementary sequences . at @xmath295
, there is a second order phase transition , which is a fitness and degradation threshold . for small mutation rates ,
the population is centred around one of the patterns , say @xmath296 , and therefore @xmath297 , whereas it is completely random with respect to the other patterns , yielding @xmath298 for @xmath299 .
[ [ deviations - from - the - original - hopfield - fitness - cneq-1 . ] ] deviations from the original hopfield fitness ( @xmath261 ) .
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + ancestral mean partial distances @xmath244 ( top ) and specific distances @xmath247 ( bottom ) depending on mutation rates . the quadratic hopfield - type fitness ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis for three patterns with @xmath262 ( left ) and @xmath263 ( right ) has been used .
results correspond to uncorrelated patterns for infinite sequence length ( @xmath300 ) . ]
figure [ fig 6.19 ] shows the ancestral mean partial distances @xmath244 and specific distances @xmath247 for the quadratic hopfield - type fitness ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis for different values of @xmath225 in the case of three patterns . data for four patterns look very similar ( not shown ) .
as @xmath225 deviates from @xmath301 , the solutions for the @xmath209 do not coincide , and the phase transition becomes a first order fitness threshold , at which all four partial distances @xmath244 jump , but it is no more a degradation threshold .
so contrary to the case of two patterns , where @xmath249 is independent of @xmath225 and the error threshold in @xmath248 is smoothed out by @xmath225 deviating from @xmath301 , here the threshold concerns all four partial distances @xmath244 and is sharpened to first order by @xmath261 . for @xmath261 , the degeneracy between the patterns and their complements is lifted , and thus for mutation rates below the threshold , there are only @xmath55 different solutions , correlated with the patterns for @xmath302 , and with their complements for @xmath272 . for clarity , only one of the solutions is shown in figure [ fig 6.19 ] . furthermore , the critical mutation rate decreases with increasing @xmath264 .
the dependence of the critical mutation rate @xmath211 on the value of @xmath225 for @xmath303 is shown in figure [ fig 6.20 ] .
the critical mutation rate @xmath211 depending on the value of @xmath225 . the quadratic hopfield - type fitness ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis for three and four patterns has been used .
results correspond to uncorrelated patterns simulating infinite sequence length.,scaledwidth=75.0% ] at @xmath304 and @xmath305 , the critical mutation rate is @xmath306 , and for values of @xmath307 $ ] , there are no first order error thresholds for either @xmath79 or @xmath308 .
this goes in line with a different sequence becoming optimal at @xmath309 for these values of @xmath225 .
however , for @xmath308 , there is an additional line of second order error thresholds , that approaches @xmath310 as @xmath264 grows .
preliminary results for @xmath311 indicate , that in that case the second order line does not occur .
it might thus be conjectured that the existence of the second order error threshold line depends on the number of patterns being even or odd ( remember that for @xmath227 it does exist ) .
this is an interesting result , as for all previously investigated hopfield - type fitness functions ( which are limited to the original hopfield fitness and a hopfield - type truncation selection as far as the author is aware ) , the existence of error thresholds has been reported .
ancestral mean partial and specific distances , @xmath244 and @xmath247 , depending on mutation rates . the original hopfield fitness ( [ original hopfield fitness ] ) for three patterns has been used .
results correspond to two typical examples of random , but correlated patterns chosen for sequences of lengths @xmath312 ( left ) , @xmath313 ( middle ) and @xmath314 ( right ) , specified at the top of each graph as @xmath315 . ]
ancestral mean partial and specific distances , @xmath244 and @xmath247 , depending on mutation rates .
the original hopfield fitness ( [ original hopfield fitness ] ) for three patterns has been used .
results correspond to two typical examples of random , but correlated patterns chosen for sequences of lengths @xmath312 ( left ) , @xmath313 ( middle ) and @xmath314 ( right ) , specified at the top of each graph as @xmath315 . ]
figure [ fig 6.21 ] shows some cases of the ancestral mean partial and specific distances , @xmath244 and @xmath247 , for the original hopfield fitness ( [ original hopfield fitness ] ) with three patterns , which are randomly chosen sequences of finite length .
the correlations between the patterns ( and thus the variations of the @xmath279 ) are characteristic for the sequence length .
the six cases of patterns shown here are typical examples for the sequence lengths considered . in the case of long sequences ( @xmath312 ) , the deviations of the patterns from the infinite sequence limit @xmath316 are small , and grow with decreasing sequence length .
these correlations that are introduced into the system have the same effect as a choice of correlated patterns in the case of two patterns , such that the single critical mutation rate in the case of infinite sequence length is split up into two critical mutation rates , at each of which two of the @xmath209 show threshold behaviour . for short sequence length ( @xmath314 )
, it can be seen that , particularly at the smaller critical mutation rate , the threshold is smoothed out .
ancestral mean partial and specific distances , @xmath244 and @xmath247 , depending on mutation rates .
the hopfield - type fitness ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis for three patterns and @xmath263 has been used .
results correspond to the patterns used in figure [ fig 6.21 ] . ]
ancestral mean partial and specific distances , @xmath244 and @xmath247 , depending on mutation rates .
the hopfield - type fitness ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis for three patterns and @xmath263 has been used .
results correspond to the patterns used in figure [ fig 6.21 ] . ] in figure [ fig 6.22 ] , the ancestral mean partial and specific distances , @xmath244 and @xmath247 , corresponding to the same patterns as in figure [ fig 6.21 ] are shown for the quadratic hopfield - type fitness ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis and @xmath263 . for long sequence length ,
these look very similar to the results for infinite sequence length ( cf .
figure [ fig 6.19 ] ) , showing clearly the single first order phase transition . for shorter sequence lengths ,
they become more and more smoothed out , such that at @xmath313 , roughly only every other pattern that was simulated shows an error threshold , whereas for @xmath314 , in the vast majority of cases , there is no threshold .
note that this effect is present even though the finite sequence length was only simulated by choosing the patterns accordingly , so it is a feature of the model with correlated patterns . in this section ,
the quadratic symmetric hopfield - type fitness given in equation ( [ quadratic symmetric hopfield - type fitness ] ) with negative epistasis ( i.e. , positive quadratic term ) was investigated for a small number of patterns .
the results for the original hopfield fitness ( @xmath226 ) have been compared with those for the generalised quadratic hopfield - type fitness ( @xmath317 ) .
furthermore , both uncorrelated patterns ( @xmath290 for all @xmath293 ) , corresponding to an infinite sequence length , and correlated patterns ( @xmath318 ) , simulating a finite sequence length , were considered . for two patterns , an analytical treatment was possible , making all values of the @xmath279 accessible , whereas the case of three or more patterns was treated numerically due to the larger number of variables .
this means that apart from the uncorrelated choice of pattern ( @xmath290 ) , which was investigated for three , four and five patterns , only some correlated combinations for three patterns with @xmath319 were investigated , some typical examples of which are shown in section [ the case of three patterns ] .
the results are summarised as follows : * * original hopfield fitness ( @xmath226 ) * : * * for _ uncorrelated patterns _ , there is one second order error threshold for all @xmath209 at @xmath254 ( investigated @xmath320 ) . * * for _ correlated patterns _ , there are two second order error thresholds , each for half of the @xmath209 ( investigated @xmath321 ) . * * hopfield - type fitness with @xmath322 * : * * for _ uncorrelated patterns _ , there is a first order error threshold only on a restricted range of @xmath225 ( @xmath323 no first order threshold , @xmath324 $ ] ) .
+ for an even number of patterns ( @xmath325 ) , there is an additional second order error threshold for any @xmath326 ( @xmath327 , @xmath328 ) .
this error threshold does not exist for the investigated cases of an odd number of patterns ( @xmath329 ) . * * for _ correlated patterns _
, at @xmath227 , there is one second order threshold , the other one , which is present for the original hopfield fitness , is smoothed out . at @xmath79 , there is up to one first order threshold , smoothed out for more strongly correlated patterns ( corresponding to shorter sequence length ) .
the evaluation the hopfield - type fitness was limited to the cases of rather small numbers of patterns , simply because an increase in the number of patterns makes the evaluation more complex .
however , the hopfield - type fitness was chosen as a potentially realistic fitness because of its ruggedness that can be tuned by the number of patterns chosen .
the simple cases considered here probably do not show as high a degree of ruggedness as one would expect for realistic fitness functions .
however , the results described here already indicate some features that are common for all numbers of patterns investigated here , and some that depend on whether the number of patterns is odd or even
. it would be very interesting to establish whether these results generalise to an arbitrary number of patterns .
furthermore , the concept of partitioning the set of sites into subsets , which was introduced to analyse the hopfield - type fitness , is very interesting .
one could imagine a different interpretation for this by classifying sites according to the selection strength they evolve under .
some of the behaviour identified for the hopfield system could also occur in such a setting : at intermediate mutation rates , partially ordered phases could exist , such that sites that evolve under weak selection have passed their error threshold and the population is in a phase that is disordered with respect to these sites , whereas at sites that are subject to strong selection the order is still maintained .
the present work has been concerned with the investigation of a deterministic mutation selection model in the sequence space approach , using a time - continuous formulation .
important observables in these mutation
selection models are the population and ancestral distributions @xmath23 and @xmath28 , and means with respect to these distributions , in particular the population mean fitness @xmath20 and the ancestral mean genotype @xmath206 . in equilibrium , @xmath23 is given by the right perron - frobenius eigenvector of the time - evolution operator @xmath144 , whereas the ancestral distribution is given by the product of both right and left pf eigenvectors @xmath23 and @xmath29 , @xmath30 .
types have been modelled as two - state sequences . as mutation model ,
the single step mutation model was used , whereas selection was modelled by hopfield - type fitness functions , using the similarity of a sequence to a number of patterns to determine its fitness .
this allows for a more rugged fitness landscape , and the complexity of the fitness can be tuned by the number of patterns .
the large number of types that arise in the sequence space approach have been lumped into classes of types , labelled by the partial distances @xmath102 introduced in section [ lumping for the hopfield - type fitness ] as a generalisation of the hamming distance . with this , the maximum principle as developed by @xcite can be applied to the case of hopfield - type fitness functions as done in section [ the maximum principle ] , see also @xcite , treating two- and four - state sequences . in section [ error thresholds ] ,
the maximum principle derived in section [ the maximum principle ] was used to investigate the phenomenon of the error threshold .
these error thresholds can be detected with the maximum principle , because the delocalisation of the population distribution manifests itself as a jump ( or at least an infinite derivative with respect to the mutation rate ) of the ancestral mean genotype @xmath206 , the maximiser .
not all fitness functions give rise to error thresholds , and as the error thresholds were first described for a model with highly unrealistic fitness function , it has been argued that they might be an artifact of this rather than a biologically relevant phenomenon .
it is therefore clearly necessary to investigate more complex fitness functions with respect to this phenomenon . here , quadratic
hopfield - type fitness functions with small numbers of patterns have been investigated . for the original hopfield fitness ,
the results for all investigated numbers of patterns are identical . however ,
if the fitness differs from the original hopfield fitness , different behaviours are observed for different numbers of patterns . in the case of uncorrelated patterns , corresponding to random patterns chosen for infinite sequence length ,
the observed features seem to depend on whether the number of patterns is odd or even .
because for correlated patterns , only the cases of two and three patterns were investigated , and found to behave differently , it would be interesting to see how these results generalise to a higher number of patterns . in the original hopfield fitness , error thresholds
were observed for all choices of patterns .
this is not true for a generalised hopfield - type fitness .
for instance , for a hopfield - type fitness with positive epistasis no thresholds were observed , going in line with the results for permutation - invariant fitness . but also for hopfield - type fitness functions with negative epistasis , there are not necessarily any thresholds , if the fitness deviates too much from the original hopfield fitness , challenging the commonly held notion that more complex fitness functions all tend to display error threshold behaviour .
the complexity and ruggedness of the original hopfield fitness have been investigated @xcite and found to be good candidates for realistic fitness functions @xcite
. however , these results do not necessarily transfer to the generalised hopfield - type fitness functions , and therefore it would be very useful to study these properties of a generalised hopfield - type fitness functions to analyse which of these factors are responsible for generating the thresholds .
it is my pleasure to thank uwe grimm , michael baake , ellen baake and robert bialowons for helpful discussions and uwe grimm for comments on the manuscript .
support from the british council and daad under the academic research collaboration programme , project no 1213 , is gratefully acknowledged .
baake , e. , gabriel , w. , 2000 .
biological evolution through mutation , selection , and drift : an introductory review . in : stauffer , d. ( ed . ) , annual reviews of computational physics vii .
world scientific , singapore , pp .
203264 .
holland , j. j. , domingo , e. , de la torre , j. c. , steinhauer , d. a. , 1990 .
mutation frequencies at defined single codon sites in vesicular stromatitis - virus and poliovirus can be increased only slightly by chemical mutagenesis .
journal of virology 64 ( 8) , 39603962 .
loeb , l. a. , essigmann , j. m. , kazazi , f. , zhang , j. , rose , k. d. , mullins , j. i. , 1999 .
lethal mutagenesis of with mutagenic nucleoside analogs .
proceedings of the national academy of sciences of the usa 96 , 14921497 .
sierra , s. , dvila , m. , lowenstein , p. r. , domingo , e. , 2000 .
response of foot - and - mouth disease virus to increased mutagenesis : of viral load and fitness in loss of infectivity .
journal of virology 74 ( 18 ) , 83168323 . | a deterministic mutation selection model in the sequence space approach is investigated .
genotypes are identified with two - letter sequences .
mutation is modelled as a markov process , fitness functions are of hopfield type , where the fitness of a sequence is determined by the hamming distances to a number of predefined patterns . using a maximum principle for the population mean fitness in equilibrium , the error threshold phenomenon is studied for quadratic hopfield - type fitness functions with small numbers of patterns .
different from previous investigations of the hopfield model , the system shows error threshold behaviour not for all fitness functions , but only for certain parameter values .
mutation
selection model , hopfield model , error threshold , maximum principle |
there are ( at least ) two main approaches to solvable lattice models and their associated quantum spin chains : the bethe ansatz @xcite and the corner transfer matrix method @xcite . in his 1931 paper
@xcite , bethe solved the heisenberg spin chain based on the string hypothesis which asserts that the eigenvalues of the hamiltonian form certain strings in the complex plane as the size of the system tends to infinity .
the bethe ansatz has been applied to many models to prove completeness of the bethe vectors .
the eigenvalues and eigenvectors of the hamiltonian are indexed by rigged configurations .
however , numerical studies indicate that the string hypothesis is not always true @xcite .
the corner transfer matrix ( ctm ) method , introduced by baxter @xcite , labels the eigenvectors by one - dimensional lattice paths .
these lattice paths have a natural interpretation in terms of kashiwara s crystal base theory @xcite , namely as highest weight crystal elements in a tensor product of finite - dimensional crystals .
even though neither the bethe ansatz nor the corner transfer matrix method are mathematically rigorous , they suggest that the existence of a bijection between the two index sets , namely rigged configurations on the one hand and highest weight crystal paths on the other ( see figure [ fig : scheme ] ) . for the special case when the spin chain is defined on @xmath3 , where @xmath4 is the irreducible @xmath5 representation indexed by the partition @xmath6 for @xmath7 , a bijection between rigged configurations and semi - standard young tableaux was given by kerov , kirillov and reshetikhin @xcite . this bijection was proven and extended to the case when the @xmath6 are any sequence of rectangles in @xcite .
the bijection has many amazing properties .
for example it takes the cocharge statistics @xmath8 defined on rigged configurations to the coenergy statistics @xmath9 defined on crystals .
rigged configurations and crystal paths also exist for other types . in @xcite the existence of kirillov reshetikhin crystals @xmath10 was conjectured , which can be naturally associated with the dominant weight @xmath11 where @xmath12 is a positive integer and @xmath13 is the @xmath14-th fundamental weight of the underlying algebra of finite type . for a tensor product of kirillov reshetikhin crystals @xmath15 and a dominant weight @xmath16 let @xmath17 be the set of all highest weight elements of weight @xmath16 in @xmath18 . in the same papers
@xcite , fermionic formulas @xmath19 for the one - dimensional configuration sums
@xmath20 were conjectured .
the fermionic formulas admit a combinatorial interpretation in terms of the set of rigged configurations @xmath21 , where @xmath22 is the multiplicity array of @xmath18 ( see section [ sec : def rc ] ) . a statistic preserving bijection @xmath23 has been proven in various cases @xcite which implies the following identity @xmath24 since the sets in are finite , these are polynomials in @xmath2 . when @xmath25 of type @xmath1 , they are none other than the kostka foulkes polynomials .
rigged configurations corresponding to highest weight crystal paths are only the tip of an iceberg . in this paper
we extend the definition of rigged configurations to all crystal elements in types @xmath0 by the explicit construction of a crystal structure on the set of _ unrestricted _ rigged configurations ( see definition [ def : crystal ] ) .
the proof uses stembridge s local characterization of simply - laced crystals @xcite . for nonsimply - laced algebras ,
the local rules provided in @xcite are still necessary , but no longer sufficient conditions to characterize crystals .
crystal operators for rigged configurations associated to nonsimply - laced algebras can be constructed from the ones presented here via `` folding '' of the dynkin diagrams as in the construction of virtual crystals @xcite .
the equivalence of the crystal structures on rigged configurations and crystal paths together with the correspondence for highest weight vectors yields the equality of generating functions in analogy to ( see theorem [ thm : bij new ] and corollary [ cor : x = m ] ) .
denote the unrestricted set of paths and rigged configurations by @xmath26 and @xmath27 , respectively .
the corresponding generating functions @xmath28 are unrestricted generalized kostka polynomials or @xmath2-supernomial coefficients . a direct bijection @xmath29 for type @xmath1 along the lines of @xcite is constructed in @xcite .
rigged configurations are closely tied to fermionic formulas .
fermionic formulas are explicit expressions for the partition function of the underlying physical model which reflect their particle structure . for more details regarding
the background of fermionic formulas see @xcite . for type @xmath1
we obtain an explicit characterization of the unrestricted rigged configurations in terms of lower bounds on quantum numbers ( see definition [ def : extended ] and theorem [ thm : ext = unres ] ) which yields a new fermionic formula for unrestricted kostka polynomials of type @xmath1 ( see equation ) .
surprisingly , this formula is different from the fermionic formulas in @xcite obtained in the special cases of @xmath25 and @xmath30 .
the rigged configurations corresponding to the fermionic formulas of @xcite were related to ribbon tableaux and the cospin generating functions of lascoux , leclerc , thibon @xcite in reference @xcite . to distinguish these rigged configurations from the ones introduced in this paper ,
let us call them ribbon rigged configurations . the lascoux
thibon ( llt ) polynomials @xcite have recently made their debut in the theory of macdonald polynomials in the seminal paper by haiman , haglund , loehr @xcite .
the main obstacle in obtaining a combinatorial formula for the macdonald
kostka polynomials is the schur positivity of certain llt polynomials .
a related problem is the conjecture of kirillov and shimozono @xcite that the cospin generating function of ribbon tableaux equals the generalized kostka polynomial .
a possible avenue to prove this conjecture would be a direct bijection between the unrestricted rigged configurations of this paper and ribbon rigged configurations . for type @xmath1
we can also describe the affine crystal operators @xmath31 and @xmath32 on rigged configurations ( see section [ sec : affine ] ) .
a level-@xmath33 restricted element @xmath34 in a crystal @xmath18 is characterized by @xmath35 .
it is striking that the description of the unrestricted rigged configurations of type @xmath1 ( see definition [ def : extended ] ) is very similar to the characterization of level - restricted rigged configurations as given in @xcite . instead of a modification of the vacancy numbers , which comprise upper bounds for the quantum numbers , an alteration of the lower bounds occurs .
in both cases the modification is governed by a set of tableaux depending on the weight @xmath16 .
it would be interesting to understand the relation between the conditions imposed by classical restriction and level - restriction in a more precise manner .
one of the motivations for considering unrestricted rigged configurations was takagi s work @xcite on the inverse scattering transform , which provides a bijection between states in the @xmath36 box ball system and rigged configurations . in this setting rigged configurations
play the role of action - angle variables .
box ball systems can be produced from crystals of solvable lattice models for algebras other than @xmath36 @xcite .
the inverse scattering transform can be generalized to the @xmath37 case @xcite , which should give a box - ball interpretation of the unrestricted rigged configurations presented here .
another motivation for the study of unrestricted configuration sums , fermionic formulas and associated rigged configurations is their appearance in generalizations of the bailey lemma @xcite .
the andrews
bailey construction @xcite relies on an iterative transformation property of the @xmath2-binomial coefficient , which is one of the simplest unrestricted configuration sums , and can be used to prove infinite families of rogers ramanujan type identities .
the explicit formulas provided in this paper might trigger further progress towards generalizations to higher - rank or other types of the andrews
bailey construction .
the paper is organized as follows . in section [ sec : axiom ] we review basics about crystal bases and stembridge s local characterization of crystals . in section [ sec : rc ] we define rigged configurations and the new crystal structure for types @xmath0 . section [ sec : type a ] is devoted to type @xmath1 , where we give an explicit characterization of the unrestricted rigged configurations in section [ sec : lower bound ] , a new fermionic formula for unrestricted kostka polynomials in section [ sec : fermi ] , and the affine crystal structure in section [ sec : affine ] .
i would like to thank mark haiman , mark shimozono and john stembridge for helpful discussions , peter littelmann for drawing my attention to reference @xcite , and masato okado for his explanations of the box ball system @xcite .
i would also like to thank lipika deka for collaboration on @xcite .
we review the axiomatic definition of crystal graphs in section [ sec : axiom ] and the local characterization of crystals corresponding to representations of simply - laced algebras provided by stembridge @xcite in section [ sec : char ] . in section [ sec : kr ] we review the main properties of kirillov reshetikhin crystals .
kashiwara @xcite introduced a * crystal * as an edge - colored directed graph satisfying a simple set of axioms .
let @xmath38 be a symmetrizable kac moody algebra with associated root , coroot and weight lattices @xmath39 .
let @xmath40 be the index set of the dynkin diagram and denote the simple roots , simple coroots and fundamental weights by @xmath41 , @xmath42 and @xmath43 ( @xmath44 ) , respectively .
there is a natural pairing @xmath45 defined by @xmath46 .
the vertices of the crystal graph are elements of a set @xmath18 .
the edges of the crystal graph are colored by the index set @xmath40 . a @xmath47-weighted @xmath40-crystal satisfies the following properties : 1 .
fix an @xmath44 . if all edges are removed except those colored @xmath48 , the connected components are finite directed linear paths called the * @xmath48-strings * of @xmath18 .
given @xmath49 , define @xmath50 ( resp .
@xmath51 ) to be the vertex following ( resp . preceding ) @xmath34 in its @xmath48-string ; if there is no such vertex , declare @xmath50 ( resp .
@xmath51 ) to be undefined .
define @xmath52 ( resp .
@xmath53 ) to be the number of arrows from @xmath34 to the end ( resp . beginning ) of its @xmath48-string .
2 . there is a function @xmath54 such that @xmath55 kashiwara @xcite constructed crystal graphs for representations of @xmath56 .
crystals of representations form a special subclass of the set of all crystals . in @xcite
, stembridge determines a simple set of local axioms that uniquely characterize the crystals corresponding to representations for simply - laced algebras .
let @xmath57_{i , j\in i}$ ] be the cartan matrix of a simply - laced kac moody algebra @xmath38 .
let @xmath58 be an edge - colored graph .
stembridge @xcite introduces the notion of @xmath1-regularity by requiring the conditions ( p1)-(p6 ) , ( p5 ) , ( p6 ) to hold . 1 .
all monochromatic directed paths in @xmath58 have finite length .
in particular @xmath58 has no monochromatic circuits . 2 . for every vertex @xmath59 and every @xmath44 ,
there is at most one edge @xmath60 , and dually , at most one edge @xmath61 . in the notation of the previous section , the relation @xmath62 or equivalently @xmath63 is graphically depicted by @xmath64 .
set @xmath65 with @xmath66 and @xmath67 as defined in section [ sec : axiom ] .
define @xmath68 whenever @xmath69 is defined , and @xmath70 whenever @xmath71 is defined . for fixed @xmath72 and a distinct pair @xmath73 , assuming that @xmath69 is defined , require 1 .
@xmath74 , and 2 .
@xmath75 , @xmath76 . note that for simply - laced algebras @xmath77 for @xmath73 distinct .
hence ( p3 ) and ( p4 ) allow for only three possibilities : @xmath78 assuming that @xmath79 and @xmath80 both exist , we require 1 .
@xmath81 implies @xmath82 and @xmath83 .
2 . @xmath84 implies @xmath85 and @xmath86 .
dually , assuming that @xmath87 and @xmath88 both exist , we require 1
. @xmath89 implies @xmath90 and @xmath91 .
2 . @xmath92 implies @xmath93 and @xmath94 .
* definition 1.1 ) let @xmath1 be a simply - laced cartan matrix .
an edge - colored directed graph is * @xmath1-regular * if it satisfies ( p1)-(p6 ) and ( p5)-(p6 ) .
stembridge proved ( * ? ? ?
* proposition 1.4 ) that any two @xmath1-regular posets @xmath95 with maximal elements @xmath96 are isomorphic if and only if @xmath97 for all @xmath44 .
moreover this isomorphism is unique .
let @xmath98 be a dominant weight .
denote by @xmath99 the unique @xmath1-regular poset with maximal element @xmath34 such that @xmath100 for all @xmath44 .
* theorem 3.3 ) [ thm : stembridge ] if @xmath1 is a simply - laced cartan matrix , then the crystal graph of the irreducible @xmath101-module of highest weight @xmath16 is @xmath99 .
kirillov reshetikhin crystals are crystals for finite - dimensional irreducible modules over quantum affine algebras .
the irreducible finite - dimensional @xmath102-modules were classified by chari and pressley @xcite in terms of drinfeld polynomials . here
@xmath102 is the derived algebra without the generator @xmath103 , where @xmath104 is degree operator in @xmath38 .
the kirillov reshetikhin modules @xmath105 , labeled by a dynkin node @xmath14 of the underlying algebra of finite type and a positive integer @xmath12 , form a special class of these finite - dimensional modules .
they naturally correspond to the weight @xmath106 , where @xmath107 is the @xmath14-th fundamental weight of @xmath38 .
it was conjectured in @xcite , that there exists a crystal @xmath10 for each @xmath105 . in general ,
the existence of @xmath10 is still an open question . for type @xmath108 the crystal @xmath10 is known to exist @xcite and its combinatorial structure has been studied @xcite . for other types , the existence and combinatorial structure of @xmath10 has been considered in various articles
( see for example @xcite ) . as classical crystals
the kirillov reshetikhin crystals are isomorphic to @xmath109 where @xmath99 is the classically highest weight crystal of highest weight @xmath16 and the sum is over a particular set of weights contained in @xmath11 ( for more details see @xcite ) .
in this section we define a crystal structure on rigged configurations . as alluded to in the introduction rigged configurations form a combinatorial set to index the eigenvalues and eigenvectors of the hamiltonian of an exactly solvable lattice model . the simplest version of rigged configurations appeared in bethe s original paper @xcite and was later generalized by kerov , kirillov and reshetikhin @xcite to models with @xmath5 symmetry .
since the eigenvectors of the hamiltonian can also be viewed as highest weight vectors , one expects a bijection between rigged configurations and semi - standard young tableaux in the @xmath5 case .
such a bijection was given in @xcite .
rigged configurations for other types follow from the fermionic formulas given in @xcite and they correspond to highest weight crystal elements @xcite . here
we extend the notion of rigged configurations to non - highest weight elements called unrestricted rigged configurations and define a crystal structure on this set . in section [ sec : def rc ] we review the definition and known results about the usual rigged configurations . in section
[ sec : def rc crystal ] the set of unrestricted rigged configurations is introduced and the crystal structure is defined for types @xmath0 ( definition [ def : crystal ] ) .
this leads to a bijection between crystal paths and unrestricted rigged configurations ( theorem [ thm : bij new ] ) and the equality of generating functions ( corollary [ cor : x = m ] ) .
let @xmath38 be a simple simply - laced affine kac
moody algebra .
define @xmath110 the index set of the underlying algebra of finite type and set @xmath111 .
the ( highest - weight ) rigged configurations are indexed by a multiplicity array @xmath112 of nonnegative integers and a dominant weight @xmath16 .
the sequence of partitions @xmath113 is a * @xmath114-configuration * if @xmath115 where @xmath116 is the number of parts of length @xmath48 in partition @xmath117 .
denote the set of all @xmath114-configurations by @xmath118 .
the * vacancy number * of a configuration is defined as @xmath119 here @xmath120 is the normalized invariant form on the weight lattice @xmath47 such that @xmath121 is the cartan matrix .
the @xmath114-configuration @xmath122 is * admissible * if @xmath123 for all @xmath124 , and the set of admissible @xmath114-configurations is denoted by @xmath125 .
a rigged configuration is an admissible configuration together with a set of labels of quantum numbers .
a partition can be viewed as a multiset of positive integers .
a rigged partition is by definition a finite multiset of pairs @xmath126 where @xmath48 is a positive integer and @xmath59 is a nonnegative integer .
the pairs @xmath126 are referred to as strings ; @xmath48 is referred to as the length or size of the string and @xmath59 as the * label * or * quantum number * of the string .
a rigged partition is said to be a rigging of the partition @xmath127 if the multiset , consisting of the sizes of the strings , is the partition @xmath127 .
so a rigging of @xmath127 is a labeling of the parts of @xmath127 by nonnegative integers , where one identifies labelings that differ only by permuting labels among equal sized parts of @xmath127 .
a rigging @xmath128 of the @xmath114-configuration @xmath122 is a sequence of riggings of the partitions @xmath117 such that every label @xmath59 of a part of @xmath117 of size @xmath48 satisfies the inequality @xmath129 alternatively , a rigging of a configuration @xmath122 may be viewed as a double - sequence of partitions @xmath130 where @xmath131 is a partition that has at most @xmath116 parts each not exceeding @xmath132 .
the pair @xmath133 is called a * rigged configuration*. the set of riggings of admissible @xmath114-configurations is denoted by @xmath21 .
let @xmath134 be the @xmath135-th rigged partition of @xmath133 .
the * colabel * or * coquantum number * of a string @xmath126 in @xmath134 is defined to be @xmath136 . a string @xmath137 is said to be * singular * if @xmath138 , that is , its label takes on the maximum value . using
, one may easily verify that @xmath139 this implies in particular the * convexity * condition @xmath140 the set of rigged configurations is endowed with a natural statistic @xmath8 called * cocharge*. for a configuration @xmath141 define @xmath142 for a rigged configuration @xmath143 set @xmath144 where @xmath145 is the size of partition @xmath131 . as mentioned in the introduction ,
rigged configurations correspond to highest weight crystal elements .
let @xmath10 be a kirillov
reshetikhin crystal for @xmath146 and @xmath147 .
associate to @xmath18 the multiplicity array @xmath148 where @xmath149 counts the number of tensor factors @xmath10 in @xmath18 .
denote by @xmath150 the set of all highest weight elements of weight @xmath16 in @xmath18 .
there is a natural statistics defined on @xmath18 , called energy function or more precisely tail coenergy function @xmath151 ( see ( * ? ? ?
* eq . ( 5.1 ) ) for a precise definition ) .
the following theorem was proven in @xcite for type @xmath152 and general @xmath15 , in @xcite for type @xmath153 and @xmath154 and in @xcite for type @xmath153 and @xmath25 .
@xcite [ thm : bij ] for @xmath16 a dominant weight , @xmath18 as above and @xmath22 the corresponding multiplicity array , there is a bijection @xmath155 which preserves the statistics , that is , @xmath156 for all @xmath157 . defining the generating functions
@xmath158 we get the immediate corollary of theorem [ thm : bij ] .
@xcite let @xmath16 , @xmath18 and @xmath22 as in theorem [ thm : bij ]
. then @xmath159 . in this section
we introduce the set of unrestricted rigged configurations @xmath160 by defining a crystal structure generated from highest weight vectors given by elements in @xmath161 by the kashiwara operators @xmath162 .
[ def : crystal ] let @xmath22 be a multiplicity array . define the set of * unrestricted rigged configurations * @xmath160 as the set generated from the elements in @xmath163 by the application of the operators @xmath164 for @xmath165 defined as follows : 1 .
define @xmath166 by removing a box from a string of length @xmath167 in @xmath134 leaving all colabels fixed and increasing the new label by one . here
@xmath167 is the length of the string with the smallest negative rigging of smallest length .
if no such string exists , @xmath166 is undefined .
2 . define @xmath168 by adding a box to a string of length @xmath167 in @xmath134 leaving all colabels fixed and decreasing the new label by one . here
@xmath167 is the length of the string with the smallest nonpositive rigging of largest length .
if no such string exists , add a new string of length one and label -1 .
if the result is not a valid unrestricted rigged configuration @xmath168 is undefined . let @xmath169 .
if @xmath170 adds a box to a string of length @xmath167 in @xmath134 , then the vacancy numbers change according to @xmath171 where @xmath172 if the statement @xmath173 is true and @xmath174 if @xmath173 is false .
similarly , if @xmath175 adds a box of length @xmath167 to @xmath134 , then the vacancy numbers change as @xmath176 hence if @xmath177 exists , it is easy to check that @xmath178 and vice versa .
[ rem : lowest label ] note that it follows from that for a string @xmath179 in @xmath180 we have @xmath181 .
hence , @xmath175 only removes a string of length 1 if its label is @xmath182 , which in definition [ def : crystal ] is interpreted as increasing the label by one .
we may define a weight function @xmath183 as @xmath184 for @xmath180 .
it is clear from the definition that @xmath185 .
define @xmath186 let @xmath38 be of type @xmath187 .
let @xmath188 , @xmath189 and all other @xmath190 .
then @xmath191 is in @xmath27 , where the parts of the rigging @xmath131 are written next to the parts of length @xmath48 in partition @xmath117 .
we have @xmath192 before stating our main result , we need some preliminary properties of the crystal operators . [
lem : varphi ] let @xmath169 .
for fixed @xmath165 , let @xmath193 be the vacancy number for large @xmath48 and let @xmath194 be the smallest nonpositive label in @xmath134 ; if no such label exists set @xmath195
. then @xmath196 . by definition
, @xmath170 adds a box to the largest string with @xmath12-rigging in @xmath134 .
let the length of this string be @xmath197 . by the maximality of @xmath197
, we have @xmath198 for all @xmath199 such that @xmath200 exists as a part in @xmath117 . if @xmath201 , then by the convexity of @xmath202 we have @xmath203 for @xmath199 and @xmath204 . under the application of @xmath170 ,
the new label of the string of length @xmath205 would be @xmath206 , but the vacancy number changes by @xmath207 .
hence @xmath208 in this case .
if @xmath209 , by the convexity of @xmath132 we have that @xmath198 for all @xmath199 and @xmath168 is defined .
the new label is @xmath206 and @xmath210 for all @xmath199 . by induction on @xmath211
, @xmath170 adds a box to the largest string with rigging @xmath212 of @xmath213 .
let @xmath214 be the length of this string .
for large @xmath48 , the vacancy number is @xmath215 .
suppose that @xmath216 .
again by the maximality of @xmath217 and the convexity of the vacancy numbers , the vacancy numbers of @xmath218 satisfy @xmath219 for all @xmath220 .
the new label of the added box is @xmath221 and compared to the original vacancy number @xmath222 for all @xmath220 . if @xmath223 , then the new label becomes @xmath221 , but the vacancy number is @xmath224 .
hence @xmath225 is not defined in this case .
this proves @xmath196 .
[ thm : rc crystal ] let @xmath38 be of simply - laced type . for @xmath226 , let @xmath227 be the graph generated by @xmath228 and @xmath162 for @xmath165 .
then @xmath227 is isomorphic to the crystal graph @xmath99 .
let @xmath229 $ ] be a cartan matrix of simply - laced type and @xmath230 .
by theorem [ thm : stembridge ] it suffices to check that the graph @xmath227 generated by the maximal element @xmath228 and operators @xmath162 as defined in definition [ def : crystal ] is @xmath1-regular and that @xmath231 for all @xmath165 .
the claim that @xmath231 for all @xmath165 follows from lemma [ lem : varphi ] . combining and
we find that @xmath232 for large @xmath48 .
note that since @xmath226 , it does not have any negative riggings , so that @xmath195 . hence by lemma [ lem : varphi ] , @xmath231 .
next we check that @xmath227 is @xmath1-regular .
let @xmath233 .
by lemma , @xmath234 is finite .
this proves ( p1 ) .
( p2 ) is clear from definition [ def : crystal ] . to prove ( p3 ) and ( p4 ) we show that one of the following conditions hold @xmath235 it is clear from the definitions that the operators @xmath236=[e_a , e_b]=0 $ ] commute when @xmath237 , so that @xmath238 in this case .
hence assume that @xmath239 .
let @xmath240 be the length of the string in @xmath134 selected by @xmath175 .
let @xmath241 be the length of the string in @xmath242 selected by @xmath243 . under @xmath175 the vacancy number changes according to @xmath244 therefore by lemma we have @xmath245 similarly
, it follows from that @xmath246 ( note that by remark [ rem : lowest label ] the labels @xmath12 of strings of length @xmath48 in @xmath242 satisfy @xmath247 for all @xmath233 .
hence @xmath166 does not exist if this condition does not hold for @xmath166 ) .
this proves ( p3 ) and ( p4 )
. for ( p5 ) assume that @xmath248 .
as before , if @xmath237 , we have @xmath249=[f_a , f_b]=0 $ ] and @xmath250 for any @xmath251 , hence in particular for @xmath252 . therefore assume that @xmath239 . by
, @xmath248 implies @xmath253 .
an explicit calculation yields that @xmath254 .
note that @xmath255 since @xmath256 it follows from that @xmath257 . for ( p6 )
assume that @xmath258 . in this case @xmath259 and @xmath260 , and by definition @xmath261 , @xmath262 .
it can be checked explicitly that @xmath263 in this case . also , @xmath264 it can be shown explicitly that @xmath265 , which implies that @xmath266 and similarly with @xmath135 and @xmath34 interchanged .
( p5 ) and ( p6 ) can be proved analogously .
consider the crystal @xmath267 of type @xmath268 in @xmath269 .
here is the crystal graph in the usual labeling and the rigged configuration labeling : ( 100,200)(0,0 ) ( 45,180)121 ( 10,140)221 ( 10,100)231 ( 10,60)331 ( 45,20)332 ( 80,140)131 ( 80,100)132 ( 80,60)232 ( 50,178)(25,152 ) ( 60,178)(85,152 ) ( 20,138)(20,112 ) ( 20,98)(20,72 ) ( 90,138)(90,112 ) ( 90,98)(90,72 ) ( 25,58)(50,32 ) ( 85,58)(60,32 ) ( 30,165)(0)[b]1 ( 15,120)(0)[b]2 ( 15,80)(0)[b]2 ( 30,40)(0)[b]1 ( 95,120)(0)[b]1 ( 95,80)(0)[b]1 ( 80,40)(0)[b]2 ( 80,165)(0)[b]2 ( 100,200)(0,0 ) ( 40,185)@xmath270 ( 0,145)@xmath271 ( 0,105)@xmath272 ( 0,65)@xmath273 ( 30,17)@xmath274 ( 80,145)@xmath275 ( 80,103)@xmath276 ( 80,63)@xmath277 ( 50,176)(25,152 ) ( 60,176)(85,152 ) ( 20,138)(20,112 ) ( 20,97)(20,72 ) ( 90,138)(90,117 ) ( 90,94)(90,77 ) ( 25,56)(50,32 ) ( 83,53)(60,32 ) ( 30,165)(0)[b]1 ( 15,120)(0)[b]2 ( 15,80)(0)[b]2 ( 30,40)(0)[b]1 ( 95,120)(0)[b]1 ( 95,80)(0)[b]1 ( 80,40)(0)[b]2 ( 80,165)(0)[b]2 [ thm : charge ] let @xmath227 be as in theorem [ thm : rc crystal ] . the cocharge @xmath8 as defined in is constant on @xmath227 .
let @xmath233 such that @xmath168 is defined .
it is easy to check that adding a box to a string of length @xmath167 in @xmath134 changes the cocharge by @xmath278 since @xmath170 changes the label of the new string by @xmath182 and leaves the colabels of all other strings unchanged , it is clear comparing with that @xmath170 does not change the total cocharge , that is @xmath279 . for @xmath15 and @xmath280
let @xmath281 [ thm : bij new ] let @xmath280 , @xmath18 be as in theorem [ thm : bij ] and @xmath22 the corresponding multiplicity array .
then there is a bijection @xmath282 which preserves the statistics , that is , @xmath283 for all @xmath284 . by theorem [ thm : bij ]
there is such a bijection for the maximal elements @xmath285 . by theorems [ thm :
rc crystal ] and [ thm : charge ] this extends to all of @xmath26 . extending the definitions of to @xmath286 we
obtain the corollary : [ cor : x = m ] with all hypotheses of theorem [ thm : bij new ] , we have @xmath28 .
in this section we give an explicit description of the elements in @xmath27 for type @xmath152 .
generally speaking , the elements are rigged configurations where the labels lie between the vacancy number and certain lower bounds defined explicitly ( definition [ def : extended ] and theorem [ thm : ext = unres ] ) .
we use this in section [ sec : fermi ] to write down an explicit fermionic formula for the unrestricted configuration sum @xmath287 .
section [ sec : affine ] is devoted to the affine crystal structure of @xmath27 .
let @xmath288 be a multiplicity array and @xmath280 .
recall that the set of @xmath114-configurations @xmath118 is the set of all sequences of partitions @xmath289 such that holds . as discussed in section [ sec : def rc ] , in the usual setting a rigged configuration @xmath290 consists of a configuration @xmath291 together with a double sequence of partitions @xmath292 such that the partition @xmath131 is contained in a @xmath293 rectangle .
in particular this requires that @xmath294 .
the unrestricted rigged configurations @xmath295 can contain labels that are negative , that is , the lower bound on the parts in @xmath131 can be less than zero . to define the lower bounds we need the following notation .
let @xmath296 be the @xmath297-tuple of nonnegative integers corresponding to @xmath16 , that is @xmath298 .
in this section we use @xmath16 and @xmath299 interchangeably .
let @xmath300 , where @xmath301 is the length of the @xmath167-th column of @xmath302 , and let @xmath303 be the set of tableaux of shape @xmath302 such that the entries are strictly decreasing along columns , and the letters in column @xmath167 are from the set @xmath304 with @xmath305 .
[ rem : row ] denote by @xmath309 the entry of @xmath310 in row @xmath200 and column @xmath167 .
note that @xmath311 since the entries in column @xmath167 are strictly decreasing and lie in the set @xmath312 .
this implies @xmath313 , so that the rows of @xmath314 are weakly decreasing .
let @xmath316 such that @xmath317 .
a @xmath318-quasipartition @xmath319 is a tuple of integers @xmath320 such that @xmath321 .
each @xmath322 is called a part of @xmath319 .
note that for @xmath323 this would be a partition with at most @xmath324 parts each not exceeding @xmath325 .
[ def : extended ] an * extended rigged configuration * @xmath133 is a configuration @xmath326 together with a sequence @xmath292 where @xmath131 is a @xmath327-quasipartition for some @xmath328 .
denote the set of all extended rigged configurations corresponding to @xmath329 by @xmath330 .
let @xmath306 , @xmath331 , @xmath332 and all other @xmath190 .
then @xmath333 is an extended rigged configuration in @xmath330 , where we have written the parts of @xmath131 next to the parts of length @xmath48 in partition @xmath117 . to see that the riggings form quasipartitions , let us write the vacancy numbers @xmath132 next to the parts of length @xmath48 in partition @xmath117 :
@xmath334 this shows that the labels are indeed all weakly below the vacancy numbers .
for @xmath335 we get the lower bounds @xmath336 which are less or equal to the riggings in @xmath133 . 1 .
note that definition [ def : extended ] is similar to the definition of level - restricted rigged configurations ( * ? ? ?
* definition 5.5 ) . whereas for level - restricted rigged configurations the vacancy number had to be modified according to tableaux in a certain set , here
the lower bounds are modified .
2 . for type @xmath337
we have @xmath338 so that @xmath339 contains just the single tableau @xmath340 in this case @xmath341 .
this agrees with the findings of @xcite .
the next theorem shows that the set of unrestricted rigged configurations @xmath342 of type @xmath152 defined in terms of the crystal structure in section [ sec : def rc crystal ] is equal to the set of extended rigged configurations @xmath330 of definition [ def : extended ] .
denote by @xmath227 the graph with maximal element @xmath344 generated by @xmath164 for @xmath165 . by definition @xmath345
we claim that @xmath346 .
the statement @xmath343 then follows since the weight function is defined in the same way on both sets .
let @xmath347 .
then @xmath228 is admissible with respect to @xmath310 where column @xmath167 of @xmath314 is filled with the letters @xmath348 . hence @xmath349 .
now suppose by induction that @xmath350 is admissible with respect to @xmath310 .
we claim that @xmath177 is admissible with respect to some @xmath351 where @xmath352 is obtained from @xmath302 by adding a box to column @xmath135 .
( note that @xmath353 if @xmath354 or equivalently @xmath355 ) .
let @xmath167 be the length of the string in @xmath134 selected by @xmath170 ( see definition [ def : crystal ] ) .
let @xmath356 be minimal such that @xmath357 , where @xmath358 denotes column @xmath135 of @xmath314 .
similarly , let @xmath359 be minimal such that @xmath360
. then @xmath361 is obtained from @xmath314 by adding @xmath14 to column @xmath135 , and by removing @xmath12 from column @xmath362 and adding @xmath363 to column @xmath362 in such a way that the columns are still strictly decreasing .
note that @xmath361 is by construction strictly decreasing in columns and has the property that the elements in column @xmath34 lie in the set @xmath364 where @xmath365 is the length of column @xmath34 in @xmath361 .
hence @xmath351 . to see that @xmath177 is admissible with respect to @xmath361 , note that strings in @xmath366 and @xmath367 change by @xmath368 , and strings in @xmath134 change by @xmath369 .
in addition to this there is a new string @xmath370 in @xmath371 where @xmath372 is the smallest label in @xmath134 . since column @xmath135 of @xmath361 contains an additional entry greater than @xmath167 and in column @xmath362 an entry greater than @xmath167
was increased , @xmath361 certainly provides valid lower bounds for @xmath373 .
note that @xmath374 since by definition of @xmath170 , @xmath167 is largest such that there is a string of this length with label @xmath372 , it is not hard to check that @xmath361 gives proper lower bounds for @xmath251 .
this shows that @xmath375 to prove the reverse inclusion , suppose @xmath376 .
let @xmath328 be such that @xmath133 is admissible with respect to @xmath314 .
if @xmath377 , then certainly @xmath350 .
if @xmath378 , there must be at least one negative rigging .
suppose this occurs in @xmath134
. then @xmath166 exists . to see this note that all colabels remain fixed , so that all labels are still weakly below the vacancy number
the string @xmath379 in @xmath134 selected by @xmath175 becomes @xmath380 .
since by the definition of @xmath167 , @xmath167 is smallest such that its label is @xmath381 , all labels of strings of length less than @xmath167 are strictly bigger than @xmath372 .
hence @xmath382 for all @xmath383 such that @xmath200 appears as a part . by the convexity property of @xmath202
, this is true for all @xmath383
. hence @xmath166 exists .
next we need to show that @xmath384 .
let @xmath385 be maximal such that @xmath386 and let @xmath387 be maximal such that @xmath388 .
note that @xmath14 and @xmath12 must exist , since the rigging of the string of length @xmath167 in @xmath134 is negative so that @xmath389 .
but this implies that @xmath390 .
then define @xmath361 by removing @xmath14 from column @xmath135 of @xmath314 and changing the largest element in column @xmath362 to @xmath12 . by similar arguments as for the previous case
, @xmath166 is admissible with respect to @xmath361 . for type @xmath153 , a simple characterization in terms of lower bounds for
the parts of a configuration @xmath391 does not seem to exist .
for example take @xmath392 of type @xmath393 so that @xmath394 and all other @xmath190 . then the unrestricted rigged configurations @xmath395 which correspond to the crystal elements @xmath396 and @xmath397 respectively , occur in @xmath160 , but @xmath398 on the other hand does not appear .
let @xmath400 be the set of all nonempty subsets of @xmath303 and set @xmath401 by inclusion - exclusion the set of all allowed riggings for a given @xmath326 is @xmath402 the @xmath2-binomial coefficient @xmath403}{0pt}{}{m+p}{m}}$ ] , defined as @xmath404}{0pt}{}{m+p}{m}}=\frac{(q)_{m+p}}{(q)_m(q)_p},\ ] ] where @xmath405 , is the generating function of partitions with at most @xmath324 parts each not exceeding @xmath325 .
hence the polynomial @xmath399 may be rewritten as @xmath406}{0pt}{}{m_i^{(a)}+p_i^{(a)}-m_i^{(a)}(s)}{m_i^{(a)}}}\end{gathered}\ ] ] called * fermionic formula*. by corollary [ cor : x = m ] this is also a formula for the unrestricted configuration sum @xmath407 . as mentioned in the introduction , this formula is different from the fermionic formulas of @xcite which exist in the special case when @xmath22 is the multiplicity array of @xmath25 or @xmath30 .
the kirillov reshetikhin crystals @xmath10 are affine crystals and admit the kashiwara operators @xmath31 and @xmath32 .
it was shown in @xcite that for type @xmath152 they can be defined in terms of the * promotion operator * @xmath408 as @xmath409 the promotion operator is a bijection @xmath410 such that the following diagram commutes for all @xmath411 @xmath412 and such that for every @xmath49 the weight is rotated @xmath413 here subscripts are taken modulo @xmath297 . 1 .
set @xmath416 where @xmath417 acts on @xmath418 .
2 . apply the following algorithm @xmath127 to @xmath251 @xmath419 times : find the smallest singular string in @xmath420 .
let the length be @xmath421 .
repeatedly find the smallest singular string in @xmath422 of length @xmath423 for all @xmath424 .
shorten the selected strings by one and make them singular again .
let @xmath425 , @xmath22 the corresponding multiplicity array and @xmath426 .
then @xmath427 corresponds to the tableau @xmath428 .
after step ( 1 ) of definition [ def : rc pr ] we have @xmath429 which corresponds to the tableau @xmath430 . to prove that @xmath408 is well - defined we need to show that singular strings of length @xmath433 exist in @xmath434 , where @xmath435 .
for a given @xmath436 , set @xmath437 . by definition @xmath438 . hence by lemma [ lem : varphi ] @xmath201 where @xmath193 for large @xmath48 and @xmath12 is the smallest nonpositive label in @xmath439 .
let @xmath33 be the length of the largest part in @xmath439 .
suppose that @xmath440 or @xmath441 has a part of length bigger than @xmath33 . then by the definition of the vacancy number , @xmath442 .
but this contradicts the fact that @xmath443 is the smallest label in @xmath439 .
hence the parts of @xmath444 can not exceed @xmath33 and the string @xmath445 in @xmath439 is singular .
since the application of @xmath446 does not change the colabels in the @xmath135-th rigged partition , the largest string remains singular .
note that the above argument also shows that the longest parts in @xmath447 decrease with @xmath135 .
hence there exist singular strings in @xmath251 such that @xmath423 and @xmath408 is well - defined .
next we show that @xmath408 satisfies .
let @xmath414 so that @xmath448 .
after step ( 1 ) of definition [ def : rc pr ] , we have @xmath449 . hence by
, noting that @xmath450 , we obtain @xmath451 where in the second line we added an @xmath452-th component to the weight . in step ( 2 ) of definition [ def : rc pr ] the size of the @xmath135-th partition changes as @xmath453 .
hence @xmath454 dropping the last component ( which we only added for the intermediate calculation ) we obtain .
it remains to prove .
we treat step ( 1 ) and step ( 2 ) in the definition [ def : rc pr ] of @xmath408 separately .
note that @xmath170 and @xmath243 commute as long as @xmath455 .
hence for step ( 1 ) it suffices to show that for @xmath414 with @xmath456 we have @xmath457 .
note that it is not hard to check that lemma [ lem : varphi ] implies @xmath458 where @xmath459 ( resp .
@xmath460 ) is the length of the string in @xmath134 ( resp .
@xmath367 ) selected by @xmath170 ( resp .
@xmath461 ) . since in our case @xmath462 we must have @xmath463 .
let @xmath464 be the length of the string selected by @xmath170 in @xmath465 .
if @xmath466 , then necessarily @xmath467 since the application of @xmath468 only increases labels in the @xmath135-th rigged partition .
note that in this case @xmath469 .
hence it suffices to repeat the analysis for @xmath470 .
if @xmath471 , it can be checked explicitly that @xmath457 .
the algorithm @xmath127 in step ( 2 ) of definition [ def : rc pr ] commutes with @xmath170 for all @xmath472 ( assuming that both @xmath127 and @xmath170 are defined on the rigged configuration ) @xmath473 to see this , let @xmath180 , let @xmath474 be the string of length @xmath167 and label @xmath12 selected by @xmath170 in @xmath133 and denote by @xmath475 the length of the strings selected by @xmath127 in @xmath133 . if @xmath476 or @xmath477 , then @xmath127 and @xmath170 obviously commute since @xmath170 leaves all colabels of unselected strings fixed and @xmath127 leaves all labels of unselected strings fixed .
hence assume that @xmath478 .
since @xmath12 is the smallest label occurring in @xmath134 , we must have @xmath479 for all @xmath48 such that @xmath480 .
because of the convexity property of the vacancy numbers , it follows that @xmath479 for all @xmath48 .
since @xmath474 is the largest string in @xmath134 with label @xmath12 , all strings @xmath126 with @xmath481 in @xmath134 satisfy @xmath482 .
this together with the convexity condition implies that @xmath483 for all @xmath481 .
also , since there are no singular strings of length @xmath484 in @xmath134 , we must have @xmath485 for @xmath484 and @xmath480 .
let us first assume that @xmath486 .
then either @xmath487 , or @xmath488 and @xmath489 .
first consider the case that @xmath488 and @xmath489 .
for @xmath490 the inequality reads @xmath491 certainly @xmath492 as discussed above .
similarly , @xmath493 since @xmath12 is the smallest label in @xmath134 .
this implies that @xmath494 or @xmath495 . if @xmath496 we must have @xmath495
this is clear if @xmath497 since else @xmath127 would pick the singular string of length @xmath498 rather than @xmath167 . if @xmath499 let @xmath500 be maximal such that @xmath501 . by it
follows that @xmath502 for @xmath503 and @xmath504 for @xmath505 if @xmath506 .
but then @xmath507 and @xmath127 would pick the singular string of length @xmath508 in @xmath134 which is a contradiction to our assumption @xmath509 . hence @xmath495 if @xmath496 .
if @xmath510 we must have @xmath511 so that @xmath494 .
now after the application of @xmath127 there is a new string @xmath512 . since this is the longest string with label @xmath12 , it will be picked by @xmath170 to give the string @xmath513 .
if @xmath170 is applied first , there is a new string @xmath514 . by and the previous findings , we have @xmath515 . under @xmath170 the vacancy number changes to @xmath516 .
hence the string @xmath514 in @xmath517 is singular and will be picked by @xmath127 . note also that by @xmath518 so that @xmath519 .
this shows that @xmath170 and @xmath127 commute in this case .
if @xmath486 and @xmath487 it needs to be shown that the string of length @xmath167 picked by @xmath170 is still picked after the application of @xmath127 .
the only case in which this might not happen is when the new string @xmath520 produced by @xmath127 has label @xmath521 .
note that @xmath522 and @xmath523 under @xmath127 . hence there is only a problem if @xmath524 and @xmath525 .
distinguishing the two cases @xmath497 and @xmath499 , by very similar arguments as above this is not possible . therefore @xmath127 and @xmath170 commute .
hence from now on we assume that @xmath526 .
note that @xmath527 since all strings of length greater than @xmath167 have label greater than @xmath12 .
also @xmath528 if @xmath529 since in this case the string @xmath474 can not be singular .
convexity implies that @xmath530 and @xmath531 unless @xmath532 for all @xmath533 and @xmath534 for all @xmath535 , or @xmath536 .
if @xmath530 and @xmath531 , @xmath170 creates a new string @xmath514 with new vacancy number @xmath537 . hence this string is not singular and @xmath127 still picks the string @xmath538 .
applying @xmath127 first creates a new string @xmath539 with label @xmath540 .
hence @xmath170 picks the same string @xmath474 as before which shows that @xmath170 and @xmath127 commute .
now assume that @xmath532 for @xmath533 , @xmath534 for @xmath541 and @xmath529 . in this case
@xmath170 makes a new string @xmath514 which is singular since @xmath516 .
then @xmath127 picks this string and makes it into @xmath474 since under @xmath542 the vacancy number changes to @xmath543 . on the other hand , first applying @xmath127 picks the string @xmath544 and makes it into @xmath545 .
since this is now the largest string with label @xmath12 , @xmath170 picks it and makes it into @xmath546 which is the same as under @xmath542 since recall that @xmath170 does not change colabels but @xmath547 under @xmath170 .
since @xmath548 for all @xmath533 by , @xmath549 remains unchanged .
hence @xmath170 and @xmath127 commute .
finally assume that @xmath536 and @xmath515 .
this implies that @xmath532 for @xmath533 . by with @xmath550 and @xmath551 we obtain @xmath552 since @xmath553 we have @xmath554 .
also @xmath555 and @xmath556 since @xmath557 and @xmath558 .
hence we must have @xmath489 .
now @xmath170 creates a new string @xmath514 which is singular since @xmath559 .
then @xmath127 picks this string and makes a string @xmath474 .
if @xmath127 is applied first it transforms the string @xmath544 to @xmath545 .
then this becomes the longest string with label @xmath12 , so that @xmath170 picks it and transforms it into @xmath546 .
hence @xmath170 and @xmath127 commute . if @xmath560 it is easy to show that @xmath170 and @xmath127 commute .
lemma 7 of @xcite states that for a single kirillov
reshetikhin crystal @xmath561 the promotion operator @xmath408 is uniquely determined by for @xmath431 and for @xmath432 . hence by lemma [ lem : pr ] @xmath408 on @xmath160 is indeed the correct promotion operator when @xmath22 is the multiplicity array of @xmath561 .
unfortunately , the characterization ( * ? ? ?
* lemma 7 ) does not suffice to define @xmath408 uniquely on tensor products @xmath15 . in @xcite a bijection @xmath563 is defined via a direct algorithm .
it is expected that conjecture [ conj : pr ] can be proven by showing that the following diagram commutes : @xmath564 alternatively , an independent characterization of @xmath408 on tensor factors would give a new , more conceptual way of defining the bijection @xmath565 between paths and ( unrestricted ) rigged configurations . a proof that the crystal operators @xmath170 and @xmath175 commute with @xmath565 for @xmath566 is given in @xcite .
l. deka , a. schilling , _ new explicit expression for @xmath108 supernomials _ , extended abstract , 17th international conference on formal power series and algebraic combinatorics 2005 , university of messina , italy , june 2005 .
m. okado , a. schilling , m. shimozono , _ a crystal to rigged configuration bijection for nonexceptional affine algebras _ , `` algebraic combinatorics and quantum groups '' , edited by n. jing , world scientific ( 2003 ) , 85124 . | rigged configurations are combinatorial objects originating from the bethe ansatz , that label highest weight crystal elements . in this paper a new _ unrestricted _ set of rigged configurations is introduced for types @xmath0 by constructing a crystal structure on the set of rigged configurations . in type @xmath1 an explicit characterization of unrestricted rigged configurations is provided which leads to a new fermionic formula for unrestricted kostka polynomials or @xmath2-supernomial coefficients . the affine crystal structure for type @xmath1 is obtained as well . |
massive elliptical galaxies with similar optical luminosities have hot gas x - ray luminosities that range over two orders of magnitude .
the origin of this scatter , shown in figure 1 , has received much attention but a full understanding remains elusive .
there is evidence that gas loss by ram pressure ( and tidal ) stripping has reduced @xmath0 in elliptical galaxies or groups orbiting within rich clusters of galaxies ( biller et al .
2004 ; machacek et al .
2005 ; sun et al . 2005 ) .
however , an enormous range in @xmath0 also prevails among non - interacting ellipticals that are isolated or at the centers of isolated galaxy groups .
the correlation between the spatial extent of the x - ray emission and @xmath0 suggests that the driver for this scatter is a variation in the virial mass @xmath7 of the halo that surrounds otherwise similar elliptical galaxies ( mathews & brighenti 1998 ) .
the virial mass @xmath7 and radius @xmath8 are found by fitting dark nfw halos to the total mass distribution derived from x - ray observations of the hot gas density and temperature in @xmath9 , assuming hydrostatic equilibrium . to gain further insight into the broad range of x - ray emission from optically similar galaxies ,
we draw attention here to those ellipticals with the largest x - ray luminosities .
these isolated galaxy / groups have been variously referred to as `` [ x - ray ] overluminous elliptical galaxies ( olegs ) '' ( vikhlinin et al .
1999 ) or `` fossil groups '' ( ponman et al . 1994 ) .
the concept of fossilized groups is meant to imply that they are relics of merging among galaxies in a group environment , although all elliptical galaxies may qualify for this designation .
jones et al . ( 2003 ) provide an operational definition for fossil groups in terms of the magnitude difference between the first and second brightest group galaxies . for our purposes here we simply consider those elliptical galaxies with the largest @xmath0 in the @xmath10 plot , many of which have been previously regarded as fossils or olegs .
we then note that several of the best studied of these galaxies have nearly the same baryon mass fraction as the most massive galaxy clusters and the _ wmap _ value , @xmath3 ( spergel et al . 2003 ) , i.e. they appear to be baryonically closed .
most baryons are in the hot intragroup gas .
the data in figure 1 are mostly taken from osullivan et al .
( 2001 ) ( open squares ) , but we have added additional x - ray luminous ellipticals assembled from more recent observations ( filled symbols ) with properties listed in table 1 .
these x - ray luminous systems define the upper envelope of the luminosity distribution in the @xmath10 plane .
while all estimates of the baryon mass fraction @xmath11 require uncertain extrapolations beyond the observations to the virial radius @xmath8 , @xmath11 for several x - ray luminous groups in table 1 indicate near or complete baryon closure .
all data have been scaled to @xmath12 km s@xmath13 mpc@xmath13 .
galaxy groups and poor clusters with @xmath14 @xmath15 and @xmath16 kev are known to deviate systematically below the @xmath4 relation established by more massive clusters , suggesting additional non - gravitational energy by cosmic preheating or agn activity ( as reviewed by voit 2005 ) .
consequently , it is remarkable that groups in table 1 with @xmath17 kev have survived with most or all of their baryonic component intact .
ngc 5044 is a good example of such a group . for the purpose of this discussion ,
we have made a preliminary mass model of ngc 5044 based on gas density and temperature profiles observed to @xmath18 kpc from buote et al .
( 2003 , 2004 , & 2006 in prep . ) . in the central regions
the azimuthally averaged gas density @xmath19 was replaced with @xmath20 where @xmath21 ( buote et al . 2003 ) is the filling factor of the denser gas component at each radius responsible for most of the observed emission .
the model was constructed by first setting the stellar parameters a de vaucouleurs profile with luminosity @xmath22 @xmath23 , effective radiius @xmath24 kpc and stellar mass to light ratio @xmath25 that establish the total stellar mass @xmath26 @xmath15 and potential .
the dark halo is assumed to have an nfw mass profile with an adjustable virial mass @xmath7 and concentration @xmath27 expected for this mass ( bullock et al .
the equation of hydrostatic equilibrium is integrated for @xmath28 , fixing the gas temperature @xmath29 to fit observations and extrapolating to larger radii in a @xmath30 plot . @xmath7 and
the innermost gas density are varied until an excellent fit is achieved to the @xmath31 profile throughout the observed region .
the resulting virial mass , @xmath32 @xmath15 , is similar to our previous estimate ( buote et al .
2004 ) and the virial radius @xmath33 kpc with @xmath34 and @xmath35 gm @xmath36 . when the observed gas density profile in ngc 5044 is extrapolated to @xmath8 ( buote et al .
2004 ; 2006 in prep . ) , maintaining the same power law @xmath37 @xmath36 observed in the region @xmath38 kpc , we find that the total gas mass is @xmath39 @xmath15 , in agreement with the mass model . the mass fraction in gas is @xmath40 .
this corresponds to a baryon ratio @xmath41 , assuming a ( conservative ) star formation efficiency of @xmath42% ( lin & mohr 2004 ) .
at least @xmath43% of the initial baryons in ngc 5044 is still bound to the group . evidently , the non - gravitational heating received by the gas is @xmath44% of the gas binding energy , @xmath45 ergs . where @xmath46 and @xmath47 are the gas density and gravitational acceleration from our mass model for ngc 5044 . ] for simplicity we assume that the percentage difference between the observed @xmath48 and the value @xmath49 expected from wmap is proportional to the amount of non - gravitational energy that the gas received as a percentage of @xmath50
. the gas heating efficiency associated with accretion onto the central black hole in ngc 5044 must be consistent with gas retention .
the central galaxy with mass @xmath51 @xmath15 is expected to contain a black hole of mass @xmath52 @xmath15 ( haering & rix 2004 ) . during the accretion history of the central black hole ,
suppose that a fraction @xmath53 of the rest energy @xmath54 heats the intragroup gas , then gas retention at the 80% level suggests that @xmath55 , although some of this energy will be radiated away .
the energy radiated by ngc 5044 during several gyrs is @xmath56 ergs and , since no gas is observed to cool below @xmath57 in ngc 5044 ( buote et al .
2003 ) , the minimum accretion heating efficiency is @xmath58 .
evidently only a tiny fraction of the accretion energy released by the central black hole ( @xmath59 ) can have heated the intragroup gas in ngc 5044 .
nevertheless , substantial ongoing agn - related heating is currently observed near the center of the ngc 5044 group ( buote et al .
2003 ; mathews , brighenti & buote 2004 ) .
the non - gravitational energy received by the hot intracluster gas from supernovae can be estimated from the total mass of iron observed in the ngc 5044 group , @xmath60 where the estimated total gas mass is @xmath61 @xmath15 , @xmath62 is the ratio of hydrogen mass to total mass including helium , and the total stellar mass in the group is @xmath63 times larger than @xmath64 .
the mass - weighted gas iron abundance in the ngc 5044 group is @xmath65 and we adopt a mean stellar abundance @xmath66 and @xmath67 ( grevesse & sauval 1998 ) . the total iron mass can be expressed in terms of supernova yields ( @xmath68 @xmath15 and @xmath69 @xmath15 ) and @xmath70 , the number of supernovae per @xmath15 of initial stars formed , @xmath71 we assume that none of the supernova iron either cooled ( brighenti & mathews 2005 ) or was buoyantly expelled to @xmath72 . if all stars formed at high redshift with a salpeter imf between 0.08 and 100 @xmath15 the number of stars with mass above @xmath73 that become type ii supernovae is @xmath74 per @xmath15 . if @xmath75 is the current stellar mass in the group , the initial mass is @xmath76 where @xmath77 is the fraction of the original stellar mass ejected from stars ( brighenti & mathews 1999 ) . if the iron mass @xmath78 is eliminated between the two equations above , we find that the number of type ia supernova per @xmath15 of initial stars is @xmath79 . the total energy released by both types of supernovae is @xmath80 where we assume @xmath81 ergs per supernova and @xmath82 . since @xmath83 it is possible that supernovae energy could eject @xmath84% of the baryons from the group .
but our assumption that all of the snii energy is communicated to the hot gas may not be plausible since it makes no allowance for radiation losses in snii remnants .
of course a flatter imf ( e.g. brighenti & mathews 1999 ; nagashima et al .
2005 ) could generate enough snii energy to eject 20% of the gas ( e.g. brighenti & mathews 2001 ) .
clearly , the large baryon fraction @xmath11 observed in ngc 5044 and other x - ray luminous groups imposes a significant constraint on non - gravitational heating .
the x - ray luminosity of the groups with filled symbols in figure 1 along the upper envelope of the @xmath10 distribution correlates with @xmath86 as @xmath87 erg s@xmath13 ( also see jones et al .
2003 ) . if these groups are essentially baryonically closed , as we propose here , then @xmath86 for the group - centered e galaxy should also increase with the virial mass .
this correlation is shown in figure 2a where we plot those groups from table 1 having known estimated @xmath7 .
similar @xmath88 correlations have been found from the 2mass survey ( lin & mohr 2004 ) and the galaxy - galaxy lensing data of cooray & milosavljevic ( 2005 ) who find @xmath5 , which agrees well with the group data in figure 2a . for ngc 6482 ( the leftmost point in figure 2a )
may be significantly underestimated since the x - ray observations extend only to @xmath89 kpc , close to the expected transition between the influence of the dark halo mass and the much smaller stellar mass . ]
the @xmath88 correlation arises because group - centered elliptical galaxies grow by mergers as massive satellite group galaxies undergo dynamical friction in the dark halos .
progressively more massive halos contain more mergeable galaxies so the final @xmath86 of the group - centered elliptical increases with @xmath7 until @xmath90 @xmath15 where dynamical friction is reduced by the small galaxy / halo mass ratio .
all baryonically closed groups have been dynamically processed in this way .
the ratio of the total halo mass to that of the central galaxy @xmath91 is of particular interest .
the correlation @xmath92 @xmath15 shown in figure 2a ( dashed line ) , when combined with the stellar mass to light ratio @xmath93 ( trujillo et al .
2004 ) , results in @xmath94 .
this ratio is larger than the dynamical masses determined from the sloan digital sky survey : @xmath95 ( based on @xmath96 from prada et al .
2003 ) and @xmath97 ( padmanabhan et al . 2004 ) for e galaxies comparable to those in table 1 .
this suggests that optically similar elliptical galaxies with larger @xmath98 may have somewhat more massive dark halos .
the mean gas temperature for these groups plotted in figure 2b also correlates with optical luminosity of the group - centered galaxy as indicated by the dotted line with slope @xmath99 . since @xmath100 from the table 1 data in figure 1 ,
both @xmath7 and @xmath101 also correlate with @xmath98 , @xmath102 and @xmath103 .
the x - ray luminosity of baryonically closed groups ( @xmath104 ) scales approximately as @xmath105 [ for the mean gas temperatures of the groups in table 1 , @xmath106 kev , the bolometric x - ray emissivity is insensitive to temperature ( sutherland & dopita 1993 ) ] .
however , the @xmath107 relation also depends on the nfw concentration @xmath108 ( bullock et al .
we estimate @xmath109 for closed groups ( @xmath110 ) by filling nfw potentials with isothermal gas with @xmath111 k taken from shimizu et al .
the bolometric x - ray luminosity within @xmath8 is @xmath112 erg s@xmath13 for @xmath113 , where we assume @xmath5 from figure 2a ( dashed line ) .
this locus of maximum x - ray luminosity , @xmath1 erg s@xmath13 , is shown with a dotted line in figure 1 . if the gas temperature @xmath114 then @xmath115 suggests that @xmath116 similar to kaiser ( 1986 ) .
we propose that galaxy groups lying near the upper envelope of the @xmath10 distribution in figure 1 are nearly or completely baryonically closed boxes similar to more massive clusters .
this conclusion is supported by the baryon fraction estimates listed in table 1 and the proximity of the observations in figure 1 to the approximate dotted line locus for the maximum @xmath98 expected from baryonically closed groups observed to @xmath8 .
the projected x - ray luminosity of our mass model for ngc 5044 beyond @xmath117 varies as @xmath118 so we expect that @xmath98 for galaxies with filled symbol points in figure 1 will creep upward toward the dotted line when observed with larger apertures and more sensitive detectors .
nevertheless , we do not expect to find galaxy / groups in the future that lie significantly above the filled circles and squares in figure 1 .
baryonically closed groups provide interesting constraints on the amount of non - gravitational heating acquired by the intragroup gas . to retain most or all of the gas in these groups , the gas heating by the central black hole ( agn ) must be @xmath119 or @xmath120 kev per particle , consistent with typical values @xmath121 kev per particle required to account for deviations from self - similarity in the @xmath4 plot for clusters ( e.g. tornatore et al .
2003 ; voit 2005 ) .
we also find that the combined energy of all past supernovae is insufficient to remove significant amounts of intragroup gas unless the imf is flatter than salpeter ( brighenti & mathews 1999 , 2001 ) .
another necessary attribute of baryonically closed groups is that they are spatially isolated , i.e. they have not lost mass by ram - stripping during mergers with comparable or larger systems .
spatially isolated e galaxies and groups are of particular interest because of the strong limits they impose on non - gravitational heating .
it is therefore remarkable that some isolated e galaxies have much lower @xmath98 than ellipticals in baryonically closed groups .
for example , in figure 1 we mark with + symbols two isolated e galaxies found recently by reda et al .
( 2004 ) that are near the bottom of the distribution .
this large variation of @xmath98 for isolated es of similar @xmath86 in figure 1 may result from normal cosmic variance .
it would be interesting to determine if these and other isolated galaxies have unusually undermassive dark halos ( allowing winds ) , if they contain more energetically active ( and massive ) black holes or if their star formation efficiencies are unusually large , since such variations could help explain why these isolated galaxy / groups are not baryonically closed . by this means it will eventually be possible to determine if the non - gravitational heating arises primarily from the central black hole .
it would also be worthwhile to assemble @xmath7 , @xmath122 and optical luminosities for all isolated e galaxies and groups throughout the @xmath10 plane .
finally , we have shown that baryonically closed groups can inform us about the important relationship between the optical luminosity and mass of the group - centered galaxy and the mass of the surrounding ( group ) dark halo .
the preliminary data currently available suggest that the dark halos are about @xmath123 times more massive than the central ( non - cd ) elliptical galaxy .
these conclusions can be explored further in the @xmath10 plane by considering all elliptical galaxies for which x - ray observations provide accurate @xmath7 .
studies of the evolution of hot gas in elliptical galaxies at uc santa cruz are supported by nasa grants nag 5 - 8409 & atp02 - 0122 - 0079 and nsf grants ast-9802994 & ast-0098351 for which we are very grateful .
gastaldello , f. et al . 2005 , _ cluster substructure and its evolution with redshift _ eds .
t. e. jeltema , c. r. canizares , m. w. bautz & d. a. buote , 35th cospar scientific assembly , july 25 , 2004 , paris , france p. 3139
lccclcl ngc5044 & 10.76 & 42.80 & 1.2 & 13.60 & @xmath124 & 1rxj1159 & 11.09 & 43.05 & 2.2 & 14.15 & @xmath125 & 2wj943.7 & 10.94 & 43.07 & 1.7 & 13.74 & @xmath126 & 3rxj0419 & 10.64 & 42.75 & 1.4 & @xmath12713.6 & @xmath128 & 4ngc6482 & 10.47 & 42.04 & 0.7 & 12.60 & @xmath129 & 5eso3060170 & 11.30 & 43.82 & 2.7 & 14.25 & na & 6rxj1340 & 11.02 & 43.11 & 2.3 & 14.33 & na & 2,10rxj2114 & 10.96 & 43.01 & 2.1 & 14.27 & na & 2rxj2247 & 11.13 & 43.32 & 2.8 & 14.45 & na & 2ngc1132 & 10.73 & 42.71 & 1.0 & 13.52 & na & 7,8rxj1416 & 11.24 & 44.05 & 1.5 & na & na & 9rxj1119 & 10.36 & 41.94 & na & na & na & 9rxj1256 & 11.16 & 43.49 & na & na & na & 9rxj1331 & 10.68 & 42.48 & na & na & na & 9rxj1552 & 11.12 & 43.51 & na & na & na & 9rxj0116 & 10.76 & 42.94 & na & na & na & 9 | elliptical galaxies and their groups having the largest @xmath0 lie close to the locus @xmath1 expected for closed systems having baryon fractions equal to the cosmic mean value , @xmath2 .
the estimated baryon fractions for several of these galaxies / groups are also close to @xmath3 when the gas density is extrapolated to the virial radius .
evidently they are the least massive baryonically closed systems .
gas retention in these groups implies that non - gravitational heating can not exceed about 1 kev per particle , consistent with the heating required to produce the deviation of groups from the @xmath4 correlation for more massive clusters .
isolated galaxies / groups with x - ray luminosities significantly lower than baryonically closed groups may have undermassive dark halos , overactive central agns , or higher star formation efficiencies .
the virial mass and hot gas temperatures of nearly or completely closed groups correlate with the group x - ray luminosities and the optical luminosities of the group - centered elliptical galaxy , i.e. @xmath5 , an expected consequence of their merging history .
the ratio of halo mass to the mass of the central galaxy for x - ray luminous galaxy / groups is @xmath6 . |
the mystery of the ultra - high energy cosmic rays ( uhecrs ) with energies exceeding @xmath0 ev the greisen - zatsepin - kuzmin ( gzk ) ` cutoff ' @xcite continues to deepen .
this energy sets the threshold for photomeson production on the cosmic microwave background so the observation of such uhecrs ( assumed to be protons or heavier nuclei ) would indicate that the sources are relatively nearby , within the local supercluster of galaxies @xcite .
recent observations by the hires air fluorescence detector @xcite are however inconsistent with previously published data from the akeno giant air shower array ( agasa ) which ruled out such a cutoff with a significance @xmath1 @xcite .
hires has reported only 1 event above @xmath2 ev , whereas about 20 would have been expected on the basis of the agasa spectrum .
the two spectra can be made to agree _ below _ this energy , if the energies of the agasa events are systematically lowered by 20% ( within the quoted uncertainty ) , however 5 of them still remain above this energy @xcite .
subsequently the agasa collaboration have carefully assessed their energy measurement uncertainties and reaffirmed that their observed spectrum does extend well beyond the gzk energy @xcite .
to resolve this situation requires making simultaneous measurements using both the air shower and air fluorescence methods ; such measurements are underway at the pierre auger observatory being constructed in argentina @xcite .
another development has been the agasa observation that the uhecr arrival directions , although consistent with isotropy overall , exhibit clustering on small angular scales @xcite . among the 59 agasa events above @xmath3 ev , there are 5 ` doublets ' and 1 ` triplet ' with separation angle less than the estimated angular resolution of @xmath4 @xcite .
ev , 68% have a reconstructed arrival direction within @xmath5 of the true direction and 90% within @xmath6 ; the corresponding angles for all events above @xmath7 ev are @xmath8 and @xmath9 , keeping in mind that the energy resolution is @xmath10 @xcite . ] the probability for this to arise by chance from an isotropic distribution is less than 0.1% .
however this probability is very sensitive to the assumed angular resolution @xcite , e.g. increasing to @xmath11 if the angular resolution is @xmath6 @xcite .
moreover adding data from three other air shower experiments ( volcano ranch , haverah park , and yakutsk ) _ dilutes _ the significance . in an earlier such analysis @xcite , 8 doublets and 2 triplets
were found in a dataset of 47 agasa plus 45 other events with @xmath12 ev , taking the effective angular resolution of the dataset to be @xmath13 .
the chance probability for this to arise from an uniform distribution is @xmath14 , thus statistically not significant .
nevertheless , the existence of such clusters has been linked to the possibility of ( repeating ) point sources of uhecr @xcite , specifically cosmologically distant bl lacertae @xcite a sub - class of active galactic nuclei ( agn ) which have been long discussed as possible accelerators of uhecrs @xcite .
however the expected deflections of uhecrs ( assumed to be charged particles ) by galactic and intergalactic magnetic fields ought to smear out such tight source correlations @xcite .
contrary to these results , it has been claimed recently that the correlations with bl lacs are preserved , even improved , if the uhecrs are protons , after allowing for deflections by the galactic magnetic field @xcite .
little is known about the intergalactic magnetic field @xcite ; requiring rectilinear propagation of protons over the attenuation length of @xmath15 mpc at @xmath12 ev ( decreasing to @xmath16 mpc at @xmath17 ev @xcite ) would imply that its homogeneous component on such scales is extremely weak : @xmath18 g @xcite .
it has also been claimed @xcite that such clustering is predicted in a model where the uhecr arise from the decay of superheavy relic particles accumulated in the galactic halo @xcite , due to the expected clumping of halo dark matter . in this paper
we examine both these claims in detail , using as our basic statistical tool the two - point correlation function .
our intention is to determine whether the claimed correlations are meaningful , given the present limited event statistics .
it is natural to look for correlations between the observed uhecr arrival directions and plausible astrophysical sources , however it is essential to take care not to generate spurious correlations by introducing biases .
for example it has been claimed that the 5 highest energy events with @xmath17 ev are all aligned with compact radio - loud quasars ( crqsos ) having redshifts between 0.3 and 2.2 , and the chance probability for this coincidence was estimated to be 0.5% @xcite .
however this rises to 3% when the event used to formulate the hypothesis itself ( the previously noted @xcite alignment of the quasar 3c147 with the @xmath19 ev fly s eye event @xcite ) is excluded from the sample @xcite .
a careful recent analysis @xcite based on an updated event list ( 5 agasa @xcite , 4 haverah park @xcite and 1 fly s eye @xcite ) demonstrates that there are _ no _ significant correlations between uhecrs and crqsos .
these authors show also that another recent claim @xcite of significant correlations with crqsos is based on inadequate data , and , in addition , that there are _ no _ significant correlations with an interesting sub - group of these sources , viz .
@xmath20-ray blazars @xcite .
a correlation between events with @xmath12 ev and nearby galaxies likely to host quasar remnants ( qrs ) has also been found at the @xmath21 level , although this disappears if attention is restricted to events above @xmath2 ev @xcite . what has revived
interest in the possibility of such correlations is the claimed clustering in the arrival directions of uhecrs @xcite .
this may arise for example if the sources are ` compact ' ( i.e. smaller than the experimental angular resolution ) with the clusters corresponding to more than one uhecr being received from the _ same _ source . since the number of events in such clusters is much smaller than the total number of events , the majority of such sources have clearly not been seen at all .
however it is possible to estimate their number density using poisson statistics . allowing for the attenuation of uhecrs from distant sources due to gzk energy losses , the observed occurences of `
triplets ' and ` doublets ' relative to the number of single events was used to estimate the spatial density of such sources to be @xmath22 mpc@xmath23 @xcite .
this would obviously place stringent constraints on candidate astrophysical sources , e.g. @xmath20-ray bursts ( grbs ) have a spatial density of only @xmath24 mpc@xmath23 .
however a more careful analysis @xcite shows that the uncertainties in this estimate are very large .
the true number is @xmath25 mpc@xmath23 at the 68% ( 95% ) c.l . ;
moreover relaxing the assumptions made , viz .
that the sources all have the same luminosity and a spectrum @xmath26 , increases the allowed range even further , e.g. to @xmath27 mpc@xmath23 for a schechter luminosity function and a spectrum @xmath28 @xcite .
clearly the present limited event numbers do _ not _ permit any candidate class of sources to be definitively excluded .
note that the observed clustering may also arise because of a higher density of local sources in certain directions , e.g. due to the clumpiness of halo dark matter in the decaying dark matter model .
the next step taken was the construction of the angular autocorrelation function of uhecrs @xcite .
for the agasa data this displays a clear peak at separation angles less than @xmath4 , consistent with the point spread function @xcite . moreover the chance probability estimated by monte carlo to match or exceed the observed count in the first angular bin , when plotted versus energy , is seen to have a minimum at @xmath29 ; the peak in the autocorrelation function for @xmath12 ev is stated to have a significance of 4.6@xmath30 @xcite .
an equally significant autocorrelation was claimed using a different method of analysing the data in which a ` triplet ' was taken to correspond to three or two ` doublets ' depending on whether the events are bunched together or linearly aligned @xcite .
these authors found the probability for chance coincidences to be minimum for events above @xmath31 ev in the agasa data @xcite , and above @xmath32 ev in the yakutsk data @xcite . restricting attention to events above these energies , the chance probability for the observed clustering in the first angular bin was quoted as @xmath33 for agasa and @xmath34 for yakutsk , taking @xmath4 and @xmath13 respectively for the size of the first bin , corresponding to the respective experimental angular resolutions @xcite . motivated by the results quoted above which implicate compact sources for uhecrs , tinyakov & tkachev @xcite
have proposed that the sources are in fact bl lacs .
the physical motivation they provided for this is that only agns in which the central jet points towards us ` blazars ' are likely to be uhecr sources ( since particles accelerated in a relativistic jet are strongly beamed ) , and among all blazars , bl lacs in particular have few emission lines in their spectra , indicating low density of ambient matter and radiation , thus presumably more favorable conditions for particle acceleration .
tinyakov & tkachev @xcite used a catalogue of agns and quasars containing 306 confirmed ( out of 462 listed ) bl lacs @xcite .
they asserted that since the ability of bl lacs to accelerate uhecrs may be correlated with optical and radio emissions , it would be appropriate to select the most _ powerful _ bl lacs by imposing cuts on redshift , apparent magnitude and 6 cm radio flux .
in fact the redshift is unknown for over half of all confirmed bl lacs but they assumed that all such bl lacs are at @xmath35 and included them in the sample anyway . by imposing the cuts @xmath36 ( or unknown ) , @xmath37 , and @xmath38 jy , they selected a sample of 22 bl lacs .
they considered 39 agasa events with @xmath39 ev and 26 yakutsk events with @xmath40 ev , the energy cuts being motivated by their earlier autocorrelation analysis @xcite which had indicated that the small - angle clustering of uhecrs is most pronounced above these energies in the respective datasets . assuming that the event energies are not important for small angle correlations , they combined these into one set of 65 events .
then they computed the correlation between the arrival directions of these uhecrs and the selected 22 bl lacs , finding a significant number of coincidences .
eight uhecrs were found to be within @xmath4 of 5 bl lacs , the chance probability of which is only @xmath41 .
note that these events are all _ below _ the gzk cutoff .
moreover the uncertainty in the arrival direction of yakutsk events is @xmath13 at @xmath3 ev , and even higher at lower energies , so these close alignments are unlikely to be physically significant . ]
the authors acknowledged that the imposition of the arbitrary cuts made on the bl lac catalogue can affect the significance of this result and estimated the ` penalty factor ' to be about 15 ; however the significance of the coincidences taking this into account was then quoted as @xmath42 ( implying a penalty factor of only 3 ) @xcite .
this was the basis for their claim that bl lacs are the probable sources of uhecrs .
since these are cosmologically distant sources , it is pertinent to ask how the uhecrs get to the earth .
initially tinyakov & tkachev @xcite inferred that the primaries have to be neutral , i.e. photons or neutrinos ( unless the gzk effect is inoperative because of violation of lorentz invariance ) .
however in subsequent work @xcite they found that the correlations are _ improved _ if the primaries are assumed to be protons , whose trajectories are modified by the galactic magnetic field ( gmf ) . in this work
they used the full set of 57 agasa events with @xmath12 ev ( but no yakutsk events ) and allowed for deflections by the regular component of the gmf ( but ignored the fluctuating component which is in fact of comparable strength @xcite ) , while assuming that deflections by intergalactic magnetic fields ( igmf ) are _ negligible_. the same bl lac catalogue @xcite was used but this time no cuts were made on redshift or the 6 cm radio flux , only on the apparent magnitude ( @xmath37 ) since this maximised the correlations .
it was found that 18 bl lacs then lie within @xmath4 of the reconstructed arrival directions of 22 uhecrs , if these mainly have charge + 1 ( however 8 might alternatively be neutral and 4 must be neutral ) .
of these 18 bl lacs , only 6 have measured redshifts and these authors now proposed @xcite , contrary to their previous supposition , that the rest must in fact have redshifts _ less than _ 0.1 in order that the protons they emit can overcome the gzk losses and reach the earth . )
is near enough to be a possible source of the @xmath43 ev event ( assumed to be charge + 1 ) it is associated with ; the other object ( txs 0806 + 315 , @xmath44 ) which is also closely aligned with this event is probably too far on the basis of uhecr propagation calculations @xcite .
the remaining 4 pairings are also implausible these are rx j1058.6 + 5628 ( @xmath45 ) aligned with a ` doublet ' ( charge 0 ) having energies @xmath46 ev , tex 1428 + 370 ( @xmath47 ) aligned with an event ( charge + 1 ) of energy @xmath48 ev , 1es 1853 + 671 ( @xmath49 ) aligned with an event ( charge + 1 ) of energy @xmath50 ev , and exo 1118.0 + 4228 ( @xmath51 ) aligned with an event ( charge 0 or + 1 ) of energy @xmath52 ev .
we have estimated distances from the redshifts using the mattig formula , @xmath53 $ ] , indicated by measurements of cosmological parameters .
] they asserted further @xcite that their success at finding significant correlations between bl lacs and uhecrs in this manner confirmed that bl lacs are the sources , _ as well _ as validating their adopted model of the gmf , _ and _ their assumption that there are no significant deflections due to the igmf .
moreover tinyakov & tkachev @xcite noted that many of these 22 bl lacs are x - ray sources . in subsequent work
@xcite , an updated catalogue of qsos containing 350 confirmed bl lacs @xcite was examined for correlations with the third egret catalogue of @xmath20-ray sources @xcite and 14 were identified as strong @xmath20-ray emitters ( of which 8 were known to be so already ) .
correlations between these 14 bl lacs and the set of 39 agasa plus 26 yakutsk events selected earlier @xcite were then studied , again allowing for deflections by the gmf modelled as in earlier work @xcite .
it was found that there are 13 possible coincidences within @xmath54 ( for charge 0 or + 1 ) with a chance probability of @xmath55 @xcite . leaving out the 2 bl lacs that are invisible to the northern hemisphere cosmic ray experiments ,
8 of the remaining 12 are found to be along the ( reconstructed ) arrival directions of uhecrs .
it was concluded @xcite that @xmath20-ray emission is the physical criterion for a bl lac to be a uhecr source .
however uhecrs are known _ not _ to correlate with @xmath20-ray blazars @xcite .
it was stated that there is no contradiction because the bl lacs considered display a low degree of polarisation whereas @xmath20-ray blazars are highly polarised @xcite .
given this set of interesting claims we wish to ascertain to what extent the strong correlations found depend on the selection criteria used .
to do so we calculate the correlation function in the same manner as tinyakov & tkachev @xcite and use monte carlo simulations to determine the probability of chance coincidences .
we consider four cases using the agasa data @xcite ; we do not consider the yakutsk data @xcite because the events which contribute dominantly to the correlations found earlier @xcite have energies _ below _ the gzk cutoff , where the uncertainty in the arrival directions exceeds @xmath13 , so the correlations found at smaller angles can not be meaningful .
our 4 models correspond to considering : \(1 ) the 39 uhecr with @xmath39 ev @xcite and 22 bl lacs selected by the tinyakov & tkachev criteria @xcite , \(2 ) the full set of 57 uhecr with @xmath12 ev @xcite , but retaining the cuts on the bl lacs @xcite , \(3 ) the full set of 57 uhecr with @xmath12 ev @xcite and the full set of 306 bl lacs with no cuts @xcite , \(4 ) the full set of 57 uhecr with @xmath12 ev @xcite and 915 grbs @xcite .
the last case is a control to determine whether there are equally significant correlations with other suggested sources of uhecrs at cosmological distances , which are _ not _ expected to contribute events beyond the gzk cutoff @xcite . in figure
[ fig : skybl ] we plot the positions on the sky ( hammer - aitoff projections in equatorial coordinates ) of both the uhecrs and the selected objects in order to give a visual impression of how the coincidences arise , particularly for the ` doublets ' and ` triplet ' in the agasa data .
the two - point correlation function for the four cases are shown in figure [ fig:2pbl ] , calculated according to the tinyakov & tkachev prescription @xcite , adopting the same angular bin size of @xmath4 . to determine the significance of these correlations we run @xmath56 monte carlo simulations , as they did , to calculate the probability of chance coincidences .
it is evident that while there is indeed a @xmath57 correlation between uhecrs and bl lacs if suitable cuts are employed @xcite , the significance weakens to @xmath58 if the energy cut is relaxed , and _ disappears _ if the cuts on bl lacs are also relaxed .
thus there is _ no _ basis for the claim that bl lacs are the sources of uhecrs ; indeed cosmologically distant grbs correlate just as well with post - gzk uhecrs as do bl lacs
! c [ cols="^,^ " , ] of the correlations.,width=283,height=245 ] let us start by presenting different properties of the clump populations generated according to the recipes of blasi & sheth and ourselves . for a galaxy halo , we have typically the same numbers of clumps ( @xmath59 ) in both models . with our prescription ,
the clumps tend to be smaller in extent which makes their detection more difficult . however , the main difference is in the distribution of clumps with galactocentric radius , as shown in figure [ fig : histgc ] . in blasi &
sheth s model , the peak in the radial distribution of the clumps occurs at @xmath60 kpc , which is close to the solar circle .
this is of course optimum for causing visible consequences . in our model ,
the peak occurs at 220 kpc , very much in the outer parts of the galaxy halo .
let use remark that the numerical simulations clearly show that most of the surviving substructure _ is _ in the outer parts .
both nfw and sis profiles are singular at the origin , so a small regularizing core radius must be included in the computations .
for the nfw profiles , we take our cue from our earlier paper @xcite where the regularizing core @xmath61 was chosen as @xmath62 kpc ( the resolution limit ) for a halo of @xmath63 kpc extent . for the nfw clumps ,
@xmath61 is scaled to be the same fraction of the extent . for the sis profiles ,
we adopt @xmath64 kpc , as suggested by the limit imposed by particle dark matter self - annihilation @xcite .
figure [ fig : skyag ] shows the incoming uhecr flux in a hamer - aitoff projection in equatorial coordinates folded with the response of the agasa detector .
the flux from the underlying smooth model is calculated by integrating the emissivity density along the line of sight ( see e.g. , ref .
@xcite ) . however , the angular size of clumps can be smaller than the angular resolution of the instrument , in which case the flux is computed by the volume integral of the emissivity over the clump , divided by the square of the distance of the clump from earth . there are four panels showing random arrival directions , a smooth nfw galactic halo , a smooth nfw halo plus nfw clumps and a smooth halo plus sis clumps .
the grey region corresponds to no detection , as agasa is a northern hemisphere experiment .
the lower right panel shows irregularities , caused by the sis clumps which are brighter than any contribution from the underlying smooth halo .
the lower left panel corresponding to nfw clumps shows much less evidence for irregularities .
the flux is much more uniform with less clustering .
we can not verify the claim made by blasi & sheth that a smooth nfw halo alone is able to provide almost half the observed clustering @xcite .
in fact , it is hard to see how a smooth halo can be responsible for small scale flux variations . to quantify this
, we use the 58 events in the agasa experiment above @xmath3ev as our dataset .
the two point autocorrelation function for the four cases is shown in figure [ fig : tp25 ]
. samples of 58 uhecrs , with the agasa response function folded in , are generated , and the average autocorrelation is compared to the one found for the experimental data .
the clustering in the experimental dataset is not well reproduced . even when sis clumps are present , the disagreement is at the @xmath65 level and it is beyond @xmath21 for the other models .
we also note from panels ( a ) , ( b ) and ( c ) that the smooth nfw profile ( with or without clumps ) is almost indistinguishable from the isotropic model .
table [ table : probs ] gives the explicit probabilities deduced from @xmath56 monte carlo calculations .
we see that in the first three models , we fall short of obtaining the 8 clusters required as less than 2 are expected on average .
the probability of obtaining 8 or more clumps is of order @xmath66 .
when sis clumps are introduced in the halo , however , the probability increases to 0.056 .
note that , in every case , the discrepancy between model predictions and experiment is always less than @xmath67 .
the angular width of each bin was taken to be @xmath4 in the calculations reported above .
if this is enlarged to @xmath68 , then the discrepancies between model predictions and data are smaller .
this is a reasonable angular size on which to look for correlations as it corresponds to the typical deflection of a @xmath3 ev proton in the galactic magnetic field @xcite .
figure [ fig : tp5 ] shows the autocorrelation functions for the cases of isotropic arrival directions and for a smooth nfw halo with nfw clumps .
we expect typically 6 and 7 clusters for these two cases .
the probability of obtaining as many clusters as the experimental data ( 11 ) rises to 0.06 and 0.07 respectively .
note that , even for the isotropic model , the discrepancy is now at only the @xmath69 level .
the dependence of the probability on the bin width is shown in figure [ fig : bina ] for the model with nfw clumps in an nfw halo .
it reaches a minimum around @xmath4 .
this is where the discrepancy with the data is at a maximum .
either decreasing or increasing @xmath70 gives model predictions in better agreement with the data .
so , this figure shows how sensitive the clustering results are to the bin width .
for example , changing @xmath4 to @xmath6 causes almost an order of magnitude change in the probability . using the 92 events from the combined datasets of agasa , volcano ranch , haverah park and yakutsk
@xcite , blasi & sheth estimated that the probability of attaining more doublets than the data from sis clumps is 12% ( for @xmath6 bin widths ) and and 47% ( for @xmath13 ) .
these high numbers appear to be primarily a consequence of placing the sis clumps nearby . using our model for the mass and spatial distribution of the clumps
, we obtain corresponding probabilities of 7% ( for @xmath6 ) and 29% ( for @xmath13 ) .
if the clumps have an nfw profile as is more likely , these numbers are reduced further to 3.5% and 15% respectively .
we conclude that clustering is _ not _ a generic prediction of the decaying dark matter model .
even though clumps are indeed expected in the dark matter distributions in galaxy haloes , any clustering of uhecrs depends sensitively on the density profile of the clumps .
in particular , nfw clumps do not give rise to much clustering , but sis clumps may do so .
however , sis profiles are not very natural , and almost all the signal comes from the @xmath71 singularity .
in fact , the self - similarity of the structure formation process suggests that nfw clumps are more natural .
we note that the 10% of mass that we placed in clumps is already generous , and so it is not clear whether any real clustering of uhecrs can be expected from dark matter substructure .
however , it is also not clear that there is any real clustering of the uhecrs at all , as the signal is less than @xmath72 at the most natural angular scale of @xmath73 .
it has been a general expectation that it should be possible to identify the long - sought sources of cosmic rays at energies exceeding @xmath74 ev , when their arrival directions can no longer be randomised by galactic magnetic fields .
however the sky distribution has remained consistent with isotropy up to the highest energies observed .
it is clear that at such energies the sources can not be in the disk of the galaxy and all experiments indicate that the energy spectrum of such sources is significantly flatter than the component at lower energies .
however it is not clear whether the isotropy of arrival directions implicates a relatively _ local _ population of sources in the galactic halo ( e.g. decaying supermassive dark matter ) or a cosmologically _ distant _ population of astrophysical accelerators ( e.g. active galaxies or @xmath20-ray bursts ) .
support for the latter possibility has come from the new hires data which shows a cutoff in the spectrum beyond @xmath75 ev , as has long been expected for extragalactic sources .
however the agasa collaboration has reaffirmed that there is no gzk cutoff in their data , which strongly favours the former possibility .
given this confusing situation , the indication of small - angle clustering in the agasa data has naturally been seized upon as a possible further clue as to the nature of the sources .
it has been argued both that the clusters coincide with a specific class of extragalactic objects , viz .
bl lacs @xcite , and , alternatively , that such clustering might arise due to the expected clumping of halo dark matter @xcite .
the bl lac hypthesis is in fact _ inconsistent _ with the absence of the gzk cutoff in the same agasa data since many of the identified objects are at very large distances .
hence it appeared more plausible that the sources are clumps of dark matter in the galactic halo ( composed in part of supermassive decaying particles ) which would explain the absence of the gzk cutoff .
we have shown that the correlations claimed between bl lacs and the observed clusters of uhecrs are spurious , being entirely due to selection effects .
this is not the first time that such correlations with a particular class of astrophysical accelerators has been claimed ; the moral is clearly that care must be taken to not become intrigued by weak accidental correlations and then make arbitrary cuts on the dataset to emphasise them further .
we have also found that the extent to which dark matter may be clumped in the halo is not sufficient to generate the observed small - angle clustering , if the uhecrs indeed arise from decaying dark matter . here
the proponents have been misled due to the use of an unphysical density profile for the clumps , as well as a radial distribution in the galaxy which is inconsistent with the general expectations for hierarchical structure formation .
the net result of our investigations is thus rather negative .
the claimed small - angle clustering in the arrival directions of post - gzk uhecrs does not definitively implicate either extragalactic compact sources such as bl lacs , or decaying clumps of dark matter in the galactic halo . on the positive side , the forthcoming increase in statistics from the pierre auger observatory will enable us to identify the expected signal from dark matter decays if this is indeed the source of uhecrs .
moreover auger will also definitively resolve the current contradiction between the air shower and atmospheric fluorescence methods for energy measurement , thus clarifying whether the spectrum does have a gzk cutoff .
if so , searches for coincidences with cosmologically distant candidate sources such as active galaxies or @xmath20-ray bursts would be of interest . again
the increased statistics provided by auger would enable the significance of such coincidences to be meaningfully assessed .
we will soon know whether the mystery of uhecrs implicates astrophysical sources or new physics beyond the standard model .
we are grateful to motohiko nagano and alan watson for discussions and for providing uhecr data sets , and to ben moore for clarifications of the results obtained in numerical simulations of halo substructure .
we also thank pascuale blasi , ravi sheth and igor tkachev for helpful correspondance .
1truecm note added in proof : while this paper was being refereed , a preprint ( astro - ph/0301336 ) by tinyakov and tkachev appeared claiming that our critique of their work is not justified .
they calculate that the ` penalty ' for making a - posteriori cuts on the bl lac and uhecr catalogues reduces the significance of the coincidences they find by only a factor of 15 .
however as our monte carlo simulations demonstrate directly , this is a vast underestimate and in fact the claimed coincidences are likely to have arisen just by chance .
k. greisen , phys .
* 16 * , 748 ( 1966 ) . g. t. zatsepin and v. a. kuzmin , jetp lett .
* 4 * , 78 ( 1966 ) [ pisma zh .
fiz . * 4 * , 114 ( 1966 ) ] .
f. w. stecker , phys .
* 21 * , 1016 ( 1968 ) .
t. abu - zayyad _ et al .
_ [ high resolution fly s eye collaboration ] , arxiv : astro - ph/0208301 . m. takeda _ et al . _ , phys . rev .
* 81 * , 1163 ( 1998 ) [ arxiv : astro - ph/9807193 ] .
n. hayashida _ et al .
_ , astrophys .
j. * 522 * , 225 ( 1999 ) [ arxiv : astro - ph/0008102 ] .
d. r. bergman , arxiv : hep - ex/0208024 .
m. takeda _ et al .
_ , arxiv : astro - ph/0209422 .
a. a. watson , arxiv : astro - ph/0112474 .
h. goldberg and t. j. weiler , phys .
d * 64 * , 056008 ( 2001 ) [ arxiv : astro - ph/0009378 ] .
l. a. anchordoqui , h. goldberg , s. reucroft , g. e. romero , j. swain and d. f. torres , mod . phys .
a * 16 * , 2033 ( 2001 ) [ arxiv : astro - ph/0106501 ] .
y. uchihori , m. nagano , m. takeda , m. teshima , j. lloyd - evans and a. a. watson , astropart .
* 13 * , 151 ( 2000 ) [ arxiv : astro - ph/9908193 ] .
s. l. dubovsky , p. g. tinyakov and i. i. tkachev , phys .
lett . * 85 * , 1154 ( 2000 ) [ arxiv : astro - ph/0001317 ] .
p. g. tinyakov and i. i. tkachev , jetp lett .
* 74 * , 1 ( 2001 ) [ pisma zh .
* 74 * , 3 ( 2001 ) ] [ arxiv : astro - ph/0102101 ] .
p. g. tinyakov and i. i. tkachev , jetp lett .
* 74 * , 445 ( 2001 ) [ pisma zh .
* 74 * , 499 ( 2001 ) ] [ arxiv : astro - ph/0102476 ] .
p. l. biermann and p. a. strittmatter , astrophys .
j. * 322 * , 643 ( 1987 ) .
t. stanev , astrophys .
j. * 479 * , 290 ( 1997 ) .
g. a. medina tanco , e. m. de gouveia dal pino and j. e. horvath , astrophys .
j. * 492 * , 200 ( 1998 ) .
[ arxiv : astro - ph/9707041 ] .
j. alvarez - muniz , r. engel and t. stanev , astrophys .
j. * 572 * , 185 ( 2001 ) .
g. a. medina tanco , astrophys .
j. * 495 * , l71 ( 1998 ) [ arxiv : astro - ph/9801060 ] .
d. harari , s. mollerach and e. roulet , jhep * 0002 * , 035 ( 2000 ) [ arxiv : astro - ph/0001084 ] .
d. harari , s. mollerach , e. roulet and f. sanchez , jhep * 0203 * , 045 ( 2002 ) [ arxiv : astro - ph/0202362 ] .
p. g. tinyakov and i. i. tkachev , astropart .
phys . * 18 * , 165 ( 2002 ) [ arxiv : astro - ph/0111305 ] .
l. m. widrow , arxiv : astro - ph/0207240 .
t. stanev , r. engel , a. mucke , r. j. protheroe and j. p. rachen , phys .
d * 62 * , 093005 ( 2000 ) [ arxiv : astro - ph/0003484 ] .
v. berezinsky , a. z. gazizov and s. i. grigorieva , arxiv : astro - ph/0210095 .
p. blasi and r. k. sheth , phys .
b * 486 * , 233 ( 2000 ) [ arxiv : astro - ph/0006316 ] .
p. blasi and r. k. sheth , arxiv : astro - ph/0108288 . v. berezinsky , m. kachelriess and a. vilenkin , phys .
lett . * 79 * , 4302 ( 1997 ) [ arxiv : astro - ph/9708217 ] .
m. birkel and s. sarkar , astropart .
* 9 * , 297 ( 1998 ) [ arxiv : hep - ph/9804285 ] .
g. r. farrar and p. l. biermann , phys .
lett . * 81 * , 3579 ( 1998 ) [ arxiv : astro - ph/9806242 ] .
g. r. farrar and p. l. biermann , phys .
lett . * 83 * , 2472 ( 1999 ) [ arxiv : astro - ph/9901315 ] . j. w. elbert and p. sommers , astrophys .
j. * 441 * , 151 ( 1995 ) [ arxiv : astro - ph/9410069 ] .
d. j. bird _
et al . _ ,
astrophys . j. * 441 * , 144 ( 1995 ) . c. m. hoffman , phys .
lett . * 83 * , 2471 ( 1999 ) [ arxiv : astro - ph/9901026 ] .
g. sigl , d. f. torres , l. a. anchordoqui and g. e. romero , phys .
d * 63 * , 081302 ( 2001 ) [ arxiv : astro - ph/0008363 ] . m. a. lawrence , r. j. reid and a. a. watson , j. phys .
g * 17 * , 733 ( 1991 ) .
m. ave , j. a. hinton , r. a. vazquez , a. a. watson and e. zas , phys .
lett . * 85 * , 2244 ( 2000 ) [ arxiv : astro - ph/0007386 ] .
a. virmani , s. bhattacharya , p. jain , s. razzaque , j. p. ralston and d. w. mckay , astropart . phys .
* 17 * , 489 ( 2002 ) [ arxiv : astro - ph/0010235 ] .
d. f. torres , e. boldt , t. hamilton and m. loewenstein , phys .
d * 66 * , 023001 ( 2002 ) [ arxiv : astro - ph/0204419 ] .
z. fodor and s. d. katz , phys .
d * 63 * , 023002 ( 2001 ) [ arxiv : hep - ph/0007158 ] .
a. inoue , e. sakamoto ( eds ) , _ catalogue of highest energy cosmic rays , no.3 _ ( world data center for cosmic rays , institute of physical and chemical research , saitama , 1988 ) .
r. beck , space sci .
* 99 * , 243 ( 2001 ) [ arxiv : astro - ph/0012402 ] .
d. s. gorbunov , p. g. tinyakov , i. i. tkachev and s. v. troitsky , astrophys .
j. * 577 * , l93 ( 2002 ) [ arxiv : astro - ph/0204360 ] . m. p. veron - cetty and p. veron , astron .
astrophys . * 374 * , 92 ( 2001 ) .
j. n. bahcall and e. waxman , arxiv : hep - ph/0206217 .
b. moore , s. ghigna , f. governato , g. lake , t. quinn , j. stadel and p. tozzi , astrophys .
j. * 524 * , l19 ( 1999 ) [ arxiv : astro - ph/9907411 ] .
s. ghigna , b. moore , f. governato , g. lake , t. quinn and j. stadel , astrophys .
j. * 544 * , 616 ( 2000 ) [ arxiv : astro - ph/9910166 ] .
r. arnold and g. gilmore , mnras * 257 * , 225 ( 1990 ) . c. s. kochanek and n. dalal , [ arxiv : astro - ph/0111401 ] .
j. * 572 * , 25 ( 2002 ) r. b. metcalf and h. zhao , astrophys .
j. * 567 * , l5 ( 2002 ) [ arxiv : astro - ph/0111427 ] .
evans and h.j .
witt , arxiv : astro - ph/0212013 .
navarro , c.s .
frenk and s.d.m .
white , astrophys .
j. * 462 * , 563 ( 1996 ) [ arxiv : astro - ph/9508025 ] .
evans , in _ idm 2000 : third international workshop on the identification of dark matter _ , edited by n. spooner and v. kudryavtsev ( world - scientific , singapore , 2000 ) p.85 [ arxiv : astro - ph0102082 ] . n. w. evans , f. ferrer and s. sarkar , astropart .
* 17 * , 319 ( 2002 ) [ arxiv : astro - ph/0103085 ] . c. tyler , phys .
d * 66 * , 023509 ( 2002 ) [ arxiv : astro - ph/0203242 ] .
p. blasi , a. v. olinto and c. tyler , arxiv : astro - ph/0202049 . | the sky distribution of cosmic rays with energies above the ` gzk cutoff ' holds important clues to their origin . the agasa data ,
although consistent with isotropy overall , shows evidence for small - angle clustering , and it has been argued that such clusters are aligned with bl lacertae objects , implicating these as the sources .
it has also been suggested that such clusters can arise if the cosmic rays come from the decays of very massive relic particles in the galactic halo , due to the expected clumping of cold dark matter .
we examine these claims and show that both are , in fact , unjustified . |
the question whether the morphology of galaxies is imprinted by the initial conditions of their formation or rather determined by secular evolution remains a subject of intense debate .
the existence of the hubble sequence has for many years provided important constraints on this issue . in very simple terms
the hubble sequence tells us that galaxies are made up of two components : a bulge and a disk .
the canonical view of these components has long been that bulges have @xmath1 surface brightness profiles @xcite while disks have exponential surface brightness profiles .
as one goes from early - type to late - type galaxies one goes from galaxies that are bulge - dominated to galaxies that are disk - dominated .
while this simplistic interpretation of the hubble sequence has definite value , reality is considerably more complicated . in recent years
, our views of the hubble sequence have evolved and gained more nuance . for elliptical galaxies
it has become clear that they are not necessarily pure bulge systems : many elliptical galaxies contain embedded disks .
there is evidence from other information ( e.g. , kinematics ) that elliptical galaxies form a heterogeneous class of galaxies that may have formed in different ways ( e.g. * ? ? ?
* ) . for spiral galaxies
a clearer understanding has developed of their bulge properties .
high - resolution imaging - both from the ground ( e.g. * ? ? ?
* ) , and with the _ hubble space telescope _
( hst , * ? ? ?
* ) - has shown that the central surface brightness profile ( sbp ) of many late - type spirals can not be fit by the classical @xmath1 law that is well suited to describe the bulge profiles of early - type spirals . instead , the sbps of many late - type spirals rise above the extrapolation of the exponential disk in a way that can be well described by a second exponential @xcite .
this has led to the now popular view that spiral bulges come in two flavors : on the one hand , the classical @xmath1 bulges which are mostly observed in early - type spirals , and on the other the `` pseudo - bulges '' @xcite or `` exponential bulges '' @xcite which are prevalent in later hubble types . in reality
there is probably a continuum of properties , instead of a dichotomy .
when @xmath0 profiles @xcite are fit to available sbps , the profile shape parameter spans the full range of values @xmath2 ; the profile shape parameter correlates with both hubble type and bulge - to - disk ratio of the galaxy , in the sense that spiral galaxies with earlier hubble type have bulges with higher @xmath3 values @xcite .
the existence of different types of bulges in disk galaxies can be plausibly explained in the context of popular scenarios for the formation and secular evolution of galaxies .
the classical massive @xmath1 law bulges fit in with the `` primordial collapse '' formation scenario first suggested by @xcite , in which the bulge forms during the initial collapse of a galaxy - sized density perturbation , and later `` acquires '' a disk through accretion processes .
by contrast , the pseudo - bulges may have formed by secular evolution of a pre - existing disk , so that they formed after the disk , out of disk material .
some support for this scenario comes from the fact that pseudo - bulges are dynamically similar to their host disks @xcite .
plausible secular evolution scenarios include the accretion of satellite galaxies @xcite , buckling instabilities in a stellar bar @xcite , and the disruption of a stellar bar through the accumulation of a central mass concentration @xcite .
many discussions of these and related topics can be found in the review by @xcite and in the proceedings of the recent workshop on ` the formation of galactic bulges ' @xcite . in the present paper
we study the presence and properties of bulges in the very latest - type spiral galaxies ( scd or later ) .
this is an important topic for several reasons .
first , these galaxies are generally classified as very late type spirals because they do not have a very prominent bulge . as a result , many observational studies of bulges have avoided these galaxies .
second , it has become clear from recent work with hst that the majority of spiral galaxies contain a central star cluster . in the very latest - type spiral galaxies
we find that @xmath4% of the galaxies contain such a star cluster ( * ? ? ?
* hereafter paper i ) . in late hubble types , these clusters are easily mistaken for a small bulge when observed from the ground , even in good seeing conditions .
so the bright , compact `` bulges '' in late - type spirals which were used as a classification criterion in the original work of @xcite may in fact be dense star clusters occupying the photocenter of the galaxy .
the purpose of this paper is to shed some light on these issues . in particular
, we investigate whether the very latest - type spirals are completely bulgeless , whether they show excess light above the constant scale - length disk , and if so , whether this in fact implies the presence of a separate entity which could rightfully be called a bulge .
hst resolution is needed to separate the luminous nuclear star cluster from a putative bulge .
our i - band snapshot survey of late - type spiral galaxies conducted with the _ wide field and planetary camera 2 _ ( wfpc2 ) and discussed in paper i therefore forms the basis of our analysis .
we complement the hst observations with ground - based data that extends to larger radii .
the paper is organized as follows : in [ sec : data ] , we describe the data and the analysis methods that form the basis of our work .
the results of our analysis are summarized in [ sec : results ] .
we discuss the implications of our findings , and present our conclusions in
[ sec : disc ] .
the target list and selection criteria for the hst sample are described in paper i. in brief , we selected 113 nearby ( @xmath5 ) galaxies with low inclination ( @xmath6 ) and hubble type between scd and sm ( @xmath7 ) .
77 galaxies have been observed to date with wfpc2 on board hst in the f814w filter which is similar to johnson i - band . in all cases ,
the integration time was @xmath8 .
after standard data reduction to correct for instrumental effects and cosmic rays , we performed an isophotal analysis using the ellipse task in iraf .
for all galaxies studied in the present analysis , the ellipse center , ellipticity , and position angle were allowed to vary freely from isophote to isophote .
the wfpc2 sbps have been published in figures 1 and 2 of paper i. in this study , we discuss the 19 galaxies for which we were able to find ground - based , wide - field surface brightness data for comparison . our focus lies on the structural properties of the disks rather than the photometry of the nuclear clusters ( which was the subject of paper i ) . due to the short integration time ( @xmath9 ) of the wfpc2 snapshot observations ,
the signal - to - noise ratio is generally insufficient to make use of the wf chips in our analysis .
it is therefore prudent to ask whether the rather limited field - of - view of the pc chip yields a representative view of the galaxy disk . in order to address this question , we searched the literature for published wide - field observations of our sample galaxies
. in general , these are scarce , most likely due to the less than spectacular star formation activity and rather low surface brightness of the latest hubble type spirals . in this context
, it is worth pointing out that all - sky surveys such as 2mass do not reach the required sensitivity to map the outer disk of most of our sample galaxies , and hence are not useful for our purposes .
we found published wide - field sbps for 19 of our sample galaxies which are listed in table [ tab : sample ] .
all but one ( ngc4904 ) of these 19 galaxies harbor a nuclear cluster as described in paper i. the wide - field surface brightness data for 18 galaxies are based on images in the _ photometric atlas of northern bright galaxies _
( panbg , * ? ? ? * ) .
the digitized profiles , which have been studied in detail by @xcite , were kindly provided to us in machine readable form by m. hamabe .
the panbg images were taken in the v - band , with sbps derived from cuts along the major and minor semi - axes of the galaxy , typically extending out to radii between @xmath10 and @xmath11 .
for comparison to the high - resolution wfpc2 data , we use ( in most cases ) the average of the two semi - major axis cuts .
for some galaxies , visual inspection of the images and profiles in the panbg showed that one or both semi - major axis cuts appeared affected by large - amplitude , small - scale variations . in these cases , we either used the remaining semi - major axis cut , or the average of the _ semi - minor _ axes , scaled by the ellipticity of the envelope ellipse ( the elliptical fit to the isophote contour at @xmath12 ) listed in column13 of the panbg .
the remaining galaxy ( ngc2805 ) was part of the sample of @xcite , who used a red passband centered at 6500 .
figure [ fig : baghst ] shows the available data for the 19 galaxies . for each galaxy ,
the left panel contains the wfpc2 i - band sbps ; the right panel contains the panbg data for each galaxy .
the ordering of the galaxies in the figure and the solid lines in the panels are discussed in [ subsec : seq ] below . in order to normalize both datasets , we shifted the ( v - band ) panbg profiles by the amount needed to match the average i - band surface brightness in the region between @xmath13 and @xmath14 .
typical offsets were in the range @xmath15 .
this is not unreasonable , given that the typical integrated color of galaxies of these hubble types is @xmath16 ( e.g. * ? ? ?
the colors that we infer appear to be somewhat too red , probably indicating a systematic offset between the two datasets in their photometric calibration .
however , in this study , we are interested mainly in the structural properties of the light distribution and this problem therefore does not affect our analysis .
there are a few limitations to the combined data set thus obtained .
for one , the panbg profiles were derived from simple cuts along the galaxy major or minor axis , rather than from azimuthally averaged isophote fitting methods ( which is what we used for the hst data ) .
the panbg profiles are therefore affected by spiral arms , dust lanes , star forming regions , or other asymmetries , even though we made every attempt to minimize these by selecting the least - affected semi - axes .
also , possible mass - to - light variations or color gradients could complicate the direct comparison of the two sbps .
however , the results of @xcite and @xcite have shown that color gradients in late - type disks are generally small , with variations in ( b - i ) of @xmath17mag over a few scalelengths .
it thus appears that - except for the very center - the stellar populations of late - type disks are rather uniform , and our approach of profile matching seems justified . in summary
, the panbg profiles should yield a fairly robust impression of the sbp of the outer disk .
they are certainly useful for comparison to our wfpc2 images , especially to test whether the central @xmath18 are representative for the profile of the stellar disk .
although the galaxies in our sample were selected as a fairly homogeneous group with respect to hubble type , distance , and inclination , the sbps show a wide range of shapes . to address this quantitatively , we started by fitting for each galaxy an exponential profile to the i - band sbps inside the pc field
. we will refer to this as the ` inner exponential fit ' . in this fit we always excluded the nuclear cluster , as well as any ` central excess emission ' ( the definition and nature of which is discussed below ) , if present .
the fits are shown as solid lines in the left panel of each plot .
we show the same exponential fit also in the right panel of each plot ( so the solid line in the right panel is _ not _ a fit to the panbg data , but only the continuation of the fit in the left panel ) .
the extent to which the inner exponential fit describes the data at the different radii differs considerably for the different galaxies . to illustrate this point ,
we have ordered the galaxies in figure [ fig : baghst ] in a rough sequence , according to the following scheme .
the sequence starts with galaxies for which there is no evidence for a bulge of other inner component in the sbp , except for the nuclear star cluster ( which has a typical hwhm radius of @xmath19 ; paper i ) .
this group itself is somewhat heterogeneous .
for some galaxies the inner exponential fit yields a good description of the large - scale stellar disk as measured by the panbg profiles ( e.g. , ngc428 ) .
these galaxies are fairly well described by a single exponential . for other galaxies ( e.g. , ngc4299 )
, the outer disk falls below the inner exponential fit , which implies that the sbp at small radii falls _
below _ the inward extrapolation of the outer disk .
this is exactly opposite to what one would expect if a bulge were present ( see [ subsec : really ] below ) .
we interpret both types of galaxy as pure disk systems .
interestingly , these pure disks have sbps that can differ from a single exponential .
progressing along the sequence , the profiles show an increasing amount of light in excess of the inner exponential fit on radial scales of a few hundred pc ( the affected radii were excluded from the inner exponential fit ) .
we refer to this light as ` central excess emission ' .
this emission has a distinct profile shape : while the overall galaxy profile steepens towards the center , the excess component flattens towards the center ( see , e.g. , ngc275 , ngc2139 , or ngc3346 ) .
this is different from what one would expect for a bulge .
bulges tend to have @xmath0 profiles with @xmath20 @xcite .
such profiles steepen towards the center , when plotted as a function of ( linear ) @xmath21 .
consequently , we interpret the central excess emission as a phenomenon that is different from a central bulge .
we discuss this component further in
[ subsec : exc ] .
approximately halfway along the sequence the outward extrapolation of the inner exponential fit begins to underpredict considerably the true brightness of the outer disk ( e.g. , ngc5068 ) .
conversely , this implies that there is excess light in the central few kpc over the inward extrapolation of the outer disk .
this is the component that is traditionally called a bulge .
we discuss the nature of this component further in
[ subsec : really ] .
given the narrow range of hubble types of our sample ( between scd and sm ) , and the uniformity in inclination ( less than @xmath22 from face - on ) , the variety of profile shapes is somewhat surprising .
in addition , there seems to be no clear trend of position along our sequence with either hubble type or galaxy luminosity .
for example , the profiles of ngc275 and ngc1042 are very different , even though both galaxies have nearly identical hubble type ( sb(rs)cd vs. sab(rs)cd ) and total blue magnitude ( @xmath23 ) .
clearly , late - type spirals are not simple systems , their morphologies ( and possibly evolutionary states ) can differ drastically .
a considerable source of confusion in discussions of bulges is that different definitions are often used when referring to a bulge .
modern theorists tend to think of a bulge as a kinematically hot component with an extended three - dimensional structure .
however , the observational definition of a bulge is often a different one , and relies on the assumption that bulges have different sbps than disks .
the standard wisdom is that a disk can be well fit by an exponential profile , @xmath24 .
by contrast , bulges can be described empirically with either a second exponential with different scale length ( e.g. * ? ? ? * ) , a de vaucouleurs @xmath1 profile ( e.g. * ? ? ? * ; * ? ? ? * ) , or most generally , a srsic @xmath0 profile ( e.g. * ? ? ? * ) .
if one defines a bulge as ` the central concentration of mass [ or light ] in excess of the inward extrapolation of the outer , constant scale - length , exponential disk ' @xcite , one must make sure that exponential profiles indeed provide a highly accurate description of galaxy disks .
here one is on shaky ground .
the exponential model has been widely used to describe galaxy disks because it provides a fairly good fit to many observed galaxy profiles ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however , the fits are generally restricted to the region outside the central few kpc , so that it is actually very hard to know whether disks remain exponential all the way into the center .
the disks of many late - type spirals in fact show sharp , well - defined `` breaks '' in their brightness profiles , which separate sections of the disk that follow exponential profiles with different scale lengths @xcite .
so observationally , it is not at all clear how photometry alone can distinguish between an exponential bulge and a somewhat steeper disk section in the inner part of the galaxy .
theoretical considerations also do not clinch the argument .
there is no theory that unequivocally predicts that disks must be exponential . as an ` a posteriori ' explanation ,
( semi-)analytical models that invoke angular momentum redistribution via viscosity - driven radial gas flows can provide a plausible mechanism for building profile shapes close ( but not identical ) to an exponential .
this has been demonstrated in the initial work by @xcite and @xcite , and recently confirmed by the refined study of @xcite .
an alternative scenario which invokes self - propagating stochastic star formation in a disk of constant density atomic hydrogen gas @xcite also results in brightness profiles that approximate the exponential shape
. however , all these models are dependent on a number of parameters , which can be fine - tuned to reproduce a range of profile shapes .
in fact , most model profiles presented by @xcite show significant curvature at all radii . it is therefore not obvious that a simple exponential is a good analytic model for disks to begin with . in order to demonstrate the ambiguity of profile decompositions , and the ensuing interpretation of the derived `` bulge '' properties ,
we have compared two different fits to the combined ( wfpc2 plus panbg ) datasets , excluding again the nuclear cluster .
the two models are : ( a ) a single srsic profile ; and ( b ) the sum of two exponentials of different scale lengths .
interestingly , we find that most galaxies in the second half of our sequence are equally well fit with either model , thus questioning the need for a `` bulge '' component .
this is demonstrated in figure [ fig : comps ] which compares these two fits for five galaxies with apparent disk - bulge transitions . in most cases
the two profiles provide equally adequate fits , and for one galaxy , ngc5668 , the srsic profile provides an obviously better fit . only for one galaxy ,
ngc1042 , does a sum of two exponentials provide the better fit .
overall , we find that roughly half of our sample galaxies show excess light ( over the inward extrapolation of the outer exponential disk ) that can be fit by a second exponential .
this is similar to the ` pseudo bulges ' or ` exponential bulges ' that have been reported by other authors , mostly in earlier type spirals .
however , the extent to which this excess light is associated with a bona - fide bulge component remains an open question .
the sbps of these galaxies are generally well fit with srsic profiles with @xmath25@xmath26 ( as listed in the panels of figure [ fig : comps ] ) . in the absence of information on kinematics or three - dimensional
structure it is difficult to rule out that we are dealing merely with non - exponential disks , or possibly small - scale stellar bars which are known to also have exponential - like sbps ( e.g. * ? ? ?
as described in [ subsec : seq ] , several galaxies of our sample show evidence for central excess emission on scales of a few hundred pc which does not show the steepening characteristic of bulges . in order to investigate the nature of this emission ,
we have measured the amount of excess light for four example cases ( ngc275 , ngc2139 , ngc3346 and ngc5584 ) .
we first determined by eye the radius @xmath27 at which the sbp starts to deviate from the inward extrapolation of the exponential fit .
we then analyzed the wfpc2 i - band images presented in paper i with the _ iraf _ photometry package _ apphot _ in order to measure the total flux within a circular aperture of radius @xmath27 , centered on the nuclear star cluster . to give a better visual impression of the four case studies , figure [ fig : four ] reproduces their images presented in paper i. here , we have overlaid circles with radius @xmath27 to indicate the aperture over which we measure the excess emission .
also , we have slightly modified the grey scale stretch to emphasize the disk structure rather than the nuclear cluster . from the total flux inside @xmath27 ,
we subtract the contribution of the galaxy disk which is calculated by integrating the exponential fit ( the solid lines in figure [ fig : baghst ] ) inside @xmath27 .
we also subtract the light from the nuclear cluster as listed in paper i , which usually constitutes only a small correction .
table [ tab : excess ] summarizes the results of this analysis . for the four galaxies of figure [ fig : four ] ,
the apparent magnitude of the excess emission is in the range @xmath28 which corresponds to luminosities between @xmath29 and @xmath30 .
this is between 10 and 100 times brighter than the median luminosity of the nuclear star cluster in our sample ( @xmath31 or @xmath32 , paper i ) , but constitutes only between 1% and 5% of the total galaxy luminosity .
it is interesting that with absolute luminosities of @xmath33 ( using the distances in table [ tab : excess ] ) , this excess emission appears to be a faint end continuation of the distribution of spiral bulge luminosities shown in figure14 of @xcite , if one assumes a mean color of @xmath34 . the images in figure
[ fig : four ] reveal a variety of morphologies for the excess emission . for ngc275 , ngc5584 , and ngc3346 , the emission appears fairly smooth and follows the larger scale disk structure .
however , this is clearly not the case for ngc2139 , for which the observed excess emission ( which is also by far the brightest in our sample ) is due to a prominent , bright star forming region which is elongated in the e - w direction .
inspection of the wf chips shows that this structure is the inward continuation of at least one spiral arm ( the other direction is off the wfpc2 field of view ) .
this underlines the danger of relying solely on one - dimensional surface brightness data for evaluating galaxy morphology .
we have presented an investigation into the structural properties of 19 spiral galaxies with hubble type between scd and sm . from a combination of our high - resolution hst data and wide - field ground - based images ,
we obtain surface brightness profiles ( sbps ) for the sample galaxies that cover a large dynamic range in galactic radius .
we use these profiles to study quantitatively the presence and properties of bulges in spiral galaxies of the very latest hubble types .
of course , by the very definition of the hubble sequence , one does not expect very prominent bulges in these galaxies .
however , it has been somewhat of an open question whether bulges are present at all in these galaxies .
it has been realized only recently that space - based resolution is required to properly address this question .
the nuclear morphologies of most spiral galaxies are complex , and a large fraction of spiral galaxies has a nuclear star cluster .
these clusters are easily mistaken for `` compact bulges '' when observed with ground - based resolution .
previous studies with the hst have focused in majority on earlier hubble types , and the present study is the first to focus exclusively on the very latest hubble types .
approximately 30% of the sample galaxies seem to be more - or - less `` pure '' exponential disks , without any type of a central bulge .
our sample was selected to focus on face - on galaxies , but studies of edge - on galaxies support the view that disk galaxies can indeed be completely bulgeless . @xcite and @xcite have studied edge - on `` super - thin '' galaxies such as ugc7321 , and have demonstrated that in these galaxies there is no evidence for a spheroidal component . despite being disk - dominated systems ,
most galaxies in our sample have sbps that can not be well fit by a single exponential , in the sense that the surface brightness in the central few kpc exceeds the inward extrapolation of the outer exponential disk .
this has generally been found for other samples of spirals as well , and has generally prompted the addition of a bulge component to analytic models of the sbp .
in particular , numerous studies have shown that the sbps of intermediate- to late - type spirals can be well fit by a sum of two exponentials .
the inner exponential has traditionally been interpreted as an ` exponential ' bulge or ` pseudo ' bulge .
such bulges can be ( qualitatively ) explained theoretically as a result of secular evolution of the disk . while this may be correct , we point out that this is not an unambiguous interpretation . in the absence of information on three - dimensional structure or dynamics
there is no guarantee that one is dealing with a bona - fide bulge component .
we explicitly illustrate this point for the galaxies in our sample that are not well fit by a single exponential .
indeed , the sbps of these galaxies are generally well fit by a sum of two exponentials .
however , most profiles can be described at least equally well with a srsic - type @xmath0 model over the entire radial range ( outside the nuclear star cluster ) .
the shape parameter @xmath3 is in the range @xmath35 , which is not unrealistically large . as we have discussed in
[ subsec : really ] , there is no a priori theoretical reason for the sbps of disk galaxies to be pure exponentials ; srsic - type profiles are in some sense equally arbitrary as the model of choice .
so it may well be that we are simply dealing with non - exponential disks in most of our sample galaxies , and that bona - fide bulges are rare at these hubble types .
we have found that a number of late - type galaxies show central excess emission on spatial scales of a few hundred parsec that can be attributed neither to the nuclear cluster , nor to the srsic - type description of the stellar disk , nor to what one would generally consider to be a bulge component .
the origin of this light component remains unclear .
one of the interesting findings from our work is that , despite the narrow range in hubble type , the sbps of the sample galaxies are far from uniform .
our study finds no systematic trends in the structural properties with morphological type .
the exact hubble type of spirals between scd and sm appears somewhat arbitrary , in the sense that it provides little information about the presence and relative importance of galaxy bulges .
it is quite possible that this may be due to the presence of nuclear clusters , which may have played an important role in the morphological classification in photographic catalogs .
in general our results fit in well with the picture that emerged from a ground - based imaging study of 49 late - type spirals by @xcite .
they found that late - type spirals exhibit a diverse array of structural properties and morphologies , even in galaxies with otherwise similar parameters , and they concluded that bulges are often very weak or non - existent . in studies of this kind
one does need to be concerned about selection effects and observational bias .
the @xcite sample was selected for low galaxy luminosity ( @xmath36 ) , and one might worry that it did not provide a representative view of the family of very late - type spirals . our galaxy sample , on the other hand , has been selected only for hubble type , distance , and inclination .
it thus includes more luminous ( and presumably more massive ) galaxies than the @xcite sample .
the blue absolute luminosity of the galaxies in our sample ranges from @xmath37 ( ngc2552 , ngc4701 ) to @xmath38 ( ngc2805 ) .
however , we find no systematic correlation between absolute luminosity and position along the sequence of sbp shapes discussed in [ subsec : seq ] .
this suggests that more luminous late - type spirals are not systematically different from their faint - end relatives , and that presumably the @xcite results are valid for most late - type spirals .
late type spiral galaxies are in most ways `` normal '' spiral galaxies .
their angular momenta and rotation velocities are not atypical , as demonstrated by a recent survey of optical rotation curves @xcite .
they are in some sense the dynamically simplest type of disk galaxies .
they are ( mostly ) disk - dominated and often have only faint spiral arm structure or even no detectable density perturbations at all . yet , we are clearly a long way from understanding their formation and evolution .
the latest - type disk galaxies provide the most stringent observational constraints on the well - known angular momentum problem in cold dark matter ( cdm ) galaxy formation models ( e.g. * ? ? ?
* ) . because of their unevolved disks and apparent history of only modest star formation , they are a challenge for proposed solutions for the angular momentum problem which rely on energy feedback from supernova explosions @xcite
. to gain a better understanding of these issues it will be important to continue to improve our knowledge of the structure of late - type spiral galaxies .
observational studies such as the one presented here will continue to be essential .
we are grateful to m. hamabe for providing us with the digital surface brightness data from the panbg , to roelof de jong for helpful discussions , and to the anonymous referee for useful comments .
r. s. acknowledges a grant from the stsci summer student program .
support for proposal # 8599 was provided by nasa through a grant from the space telescope science institute , which is operated by the association for research in astronomy , inc .
, under nasa contract no .
nas 5 - 26555 .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , california institute of technology , under contract with nasa .
it has also benefited greatly from use of the lyon - meudon extragalactic database ( leda , http://leda.univ-lyon1.fr ) .
aguerri , j. a. l. , balcells , m. , & peletier 2001 , , 367 , 428 andredakis , y. c. , & sanders , r. h. 1994 , , 267 , 283 andredakis , y. c. , peletier , r. f. , & balcells , m. 1995 , mnras , 275 , 874 baggett , w. e. , baggett , s. m. , & anderson , k. s. j. 1998 , aj , 116 , 1626 bker , t. , laine , s. , van der marel , r. p. , sarzi , m. , rix , h .- w . , ho , l. c. , & shields , j. c. 2002 , aj , 123 , 1389 ( paper i ) boroson , t. a. 1981 , apjs , 46 , 177 bosma , a. , & freeman , k. c. 1993 , aj , 106 , 1394 carollo , c. m. , stiavelli , m. , & mack , j. 1998 , aj , 116 , 68 carollo , c. m. , ferguson , h. c. , & wyse , r. f. g. 1999 , eds .
, the formation of galactic bulges .
cambridge : cambridge university press de jong , r. s. 1996a , a&as , 118 , 557 de jong , r. s. 1996b , a&a , 313 , 377 de vaucouleurs , g. 1948 , ann .
dastrophysique , 11 , 247 eggen , o. , lynden - bell , d. , & sandage , a. 1962 , apj , 136 , 748 elmegreen , b. g. , & elmegreen , d. m. 1985 , apj , 288 , 438 ferguson , a. m. n. , & clarke , c. j. 2001 , , 325 , 781 freeman , k. c. 1970 , apj , 160 , 811 graham , a. w. 2001 , aj , 121 , 820 hubble , e. 1926 , apj , 64 , 321 hubble , e. 1936 , the realm of the nebulae ( new haven : yale univ . press ) kent , m. s. 1984 , apjs , 56 , 105 kodaira , k. , okamura , s. , & ichikawa , s. 1990 , _ photometric atlas of northern bright galaxies _
( tokyo : univ . of tokyo press ) ( panbg ) kormendy j. 1993 , in _ galactic bulges _ , iau symp .
153 , ( dordrecht : kluwer ) , p. 209 kormendy j. , & bender , r. 1996 , apj , 464 , 119 lin , d. n. c. , & pringle , j. e. 1987 , , 320 , l87 matthews , l. d. , & gallagher , j. s. 1997 , aj , 114 , 1899 matthews , l. d. , gallagher , j. s. & van driel , w. 1999 , aj , 118 , 2751 matthews , l. d. 2000 , aj , 120 , 1764 matthews , l. d. & gallagher , j. s. 2002 , apjs , in press [ astro - ph/0203188 ] navarro , j. f. , & steinmetz , m. 2000 , apj , 538 , 477 norman , c. a. , sellwood , j. a. , & hasan , h. 1996 , apj , 462 , 114 raha , n. , sellwood , j. a. , james , r. a. , & kahn , f. d. 1991 , nature , 352 , 411 sandage a. , & tammann , g. 1990 , apj , 365 , 1 seiden , p. e. , schulman , l. s. , & elmegreen , b. g. 1984 , apj , 282 , 95 srsic , j .- l . 1968 , atlas de galaxias australes ( cordoba : obs .
astron . ) silk , j. 2001 , mnras , 324 , 313 wyse , r. f. g. , gilmore , g. , & franx , m. 1997 , , 35 , 637 yoshii , y. , & sommer - larsen , j. 1989 , , 236 , 779 lcccccc ngc275 & 00 51 04.20 & -07 04 00.0 & 1681 & sb(rs)cd pec & 13.16 & 1.5 + ngc428 & 01 12 55.60 & -00 58 54.4 & 1130 & sab(s)m & 11.91 & 4.1 + ngc1042 & 02 40 23.63 & -08 25 59.8 & 1271 & sab(rs)cd & 12.50 & 4.7 + ngc2139 & 06 01 07.90 & -23 40 21.3 & 1649 & sab(rs)cd & 11.99 & 2.6 + ngc2552 & 08 19 20.14 & + 50 00 25.2 & 695 & sa(s)m ? & 12.56 & 3.5 + ngc2805 & 09 20 24.56 & + 64 05 55.2 & 1968 & sab(rs)d & 11.52 & 6.3 + ngc3346 & 10 43 38.90 & + 14 52 18.0 & 1315 & sb(rs)cd & 12.41 & 2.9 + ngc3423 & 10 51 14.30 & + 05 50 24.0 & 1025 & sa(s)cd & 11.59 & 3.8 + ngc3445 & 10 54 35.87 & + 56 59 24.4 & 2245 & sab(s)m & 12.90 & 1.6 + ngc4027 & 11 59 30.50 & -19 15 44.0 & 1588 & sb(s)dm & 11.66 & 3.2 + ngc4299 & 12 21 40.90 & + 11 30 03.0 & 306 & sab(s)dm : & 12.88 & 1.7 + ngc4540 & 12 34 50.90 & + 15 33 06.9 & 1383 & sab(rs)cd & 12.44 & 1.9 + ngc4701 & 12 49 11.71 & + 03 23 21.8 & 768 & sa(s)cd & 12.80 & 2.8 + ngc4775 & 12 53 45.79 & -06 37 20.1 & 1565 & sa(s)d & 12.24 & 2.1 + ngc4904 & 13 00 56.97 & -00 01 31.9 & 1204 & sb(s)cd & 12.60 & 2.2 + ngc5068 & 13 18 54.60 & -21 02 19.7 & 607 & sb(s)d & 10.52 & 7.2 + ngc5585 & 14 19 48.08 & + 56 43 43.8 & 571 & sab(s)d & 11.20 & 5.8 + ngc5584 & 14 22 23.65 & -00 23 09.2 & 1695 & sab(rs)cd & 12.63 & 3.4 + ngc5668 & 14 33 24.30 & + 04 27 02.0 & 1665 & sa(s)d & 12.2 & 3.3 + lcccccccc ngc275 & 24.0 & 2.5 & 290 & 15.33 & 15.54 & 19.47 & 17.38 & 2.9 + ngc5584 & 24.2 & 3.5 & 410 & 15.11 & 15.36 & 22.53 & 16.83 & 5.2 + ngc3346 & 18.8 & 5.0 & 460 & 14.61 & 14.73 & 19.64 & 17.21 &
2.0 + ngc2139 & 23.6 & 7.0 & 800 & 12.98 & 13.13 & 19.28 & 15.24 & 19.2 + | we investigate the stellar disk properties of a sample of 19 nearby spiral galaxies with low inclination and late hubble type ( scd or later ) .
we combine our high - resolution _ hubble space telescope _
i - band observations with existing ground - based optical images to obtain surface brightness profiles that cover a high dynamic range of galactic radius .
most of these galaxies contain a nuclear star cluster , as discussed in a separate paper .
the main goal of the present work is to constrain the properties of stellar bulges at these extremely late hubble types .
we find that the surface brightness profiles of the latest - type spirals are complex , with a wide range in shapes .
we have sorted our sample in a sequence , starting with `` pure '' disk galaxies ( approximately 30% of the sample ) .
these galaxies have exponential stellar disks that extend inwards to within a few tens of pc from the nucleus , where the light from the nuclear cluster starts to dominate .
they appear to be truly bulge - less systems . progressing along the sequence ,
the galaxies show increasingly prominent deviations from a simple exponential disk model on kpc scales .
traditionally , such deviations have prompted `` bulge - disk '' decompositions .
indeed , the surface brightness profiles of these galaxies are generally well fit by adding a second ( exponential ) bulge component .
however , we find that most surface brightness profiles can be fit equally well ( or better ) with a single srsic - type @xmath0 profile over the entire radial range of the galaxy , without requiring a separate `` bulge '' component .
we warn in a general sense against identification of bulges solely on the basis of single - band surface brightness profiles . despite the narrow range of hubble types in our sample , the surface brightness profiles are far from uniform .
the differences between the various galaxies appear unrelated to their hubble - types , thus questioning the usefulness of the hubble - sequence for the sub - categorization of the latest - type spirals .
a number of galaxies show central excess emission on spatial scales of a few hundred parsec that can be attributed neither to the nuclear cluster , nor to the srsic - type description of the stellar disk , nor to what one would generally consider to be a bulge component .
the origin of this light component remains unclear . |
index coding ( introduced by birk and kol @xcite in 1998 ) , a sender broadcasts messages through a noiseless shared channel to multiple receivers , each knowing some messages a priori , which are known as side information .
side information occurs frequently in many communication networks , e.g. , in a web browsers cache .
knowing the side information of the receivers , the sender can send coded symbols , known as an index code , in such a way that all of the receivers can decode their requested messages using their side information and the received coded symbols .
the aim is to find the shortest ( optimal ) index code .
how to optimally design an index code for an arbitrary index - coding instance is an open problem to date . in the literature
, various approaches have been adopted to solve the index - coding problem .
we broadly classify these approaches into four categories : ( i ) numerical , ( ii ) shannon s random coding , ( iii ) interference alignment , and ( iv ) graph - based .
numerical approaches include rank minimization over finite fields @xcite ( which is np - hard to compute in general @xcite ) , and mathematical optimization programming ( semi - definite programming @xcite , linear programming @xcite , and integer - linear programming @xcite ) .
these approaches do not provide much intuition on the interaction between the side - information configuration and the index codes .
shannon s random coding approaches @xcite require infinitely long message packets .
interference - alignment approaches treat index coding as an interference - alignment problem @xcite , and construct index codes via two alignment techniques , namely one - to - one alignment and subspace alignment .
these alignment techniques have no well - defined algorithms to construct index codes for arbitrary index - coding instances .
graph - based approaches @xcite provide intuition on the side - information configurations and index codes .
these approaches represent index - coding instances by graphs , and construct index codes as functions of the graphs .
these graph - based schemes provide linear ( scalar and vector ) index codes .
although linear index codes are not always optimal @xcite , they have simpler encoding and decoding processes .
we classify graph - based approaches into two sub - categories : ( i ) maximum distance separable ( mds ) code based interference alignment approaches , and ( ii ) graph structure based approaches .
the mds code based interference alignment approaches construct index codes by treating messages not known to a receiver as interference , and aligning all interference with the help of mds codes .
these approaches include the partial - clique - cover scheme @xcite and its fractional version @xcite , the local - chromatic - number scheme and its fractional version @xcite , and the partitioned - local - chromatic - number scheme and its fractional version @xcite .
graph structure based approaches exploit special graph structures , based on messages known to the receivers that can provide savings on index - coding instances .
it has been shown that no structure in an acyclic graph can provide any savings @xcite .
furthermore , if an arc does not belong to any cycle , then removing it does not change the optimal index code @xcite . these observations point to the importance of cycles on index coding
. in the literature ,
only disjoint cycles and cliques , a specific combination of overlapping cycles , have been exploited so far .
more precisely , disjoint cycles in digraphs are exploited by the cycle - cover scheme @xcite and its fractional version @xcite , and disjoint cliques in digraphs are exploited by the clique - cover scheme @xcite and its fractional version @xcite .
overlapping cycles can provide more savings than disjoint cycles .
we take a clique as an example . in a clique ,
every vertex forms a cycle with any other vertex , and we see overlapping of cycles at every vertex . if we consider only disjoint cycles in the clique , we get an index code strictly longer than that by considering the clique . however , not all forms of overlapping cycles are useful , in the sense that they provide more savings than considering only disjoint cycles and cliques . in this work ,
we consider a graph structure based approach , and propose structures of overlapping cycles that can be exploited in graphs to provide potentially more savings than the cycle - cover scheme , the clique - cover scheme , and other existing schemes .
the proposed structures are called interlinked - cycle ( @xmath2 ) structures , and they generalize cycles and cliques .
furthermore , we define a scheme , called the interlinked - cycle cover ( @xmath3 ) scheme , that constructs index codes based on @xmath2 structures . 1 .
we propose a new index - coding scheme ( called the @xmath1 scheme ) that generalizes the clique - cover scheme and the cycle - cover scheme .
the new scheme constructs scalar linear index codes .
we characterize a class of digraphs ( with infinitely many members ) for which the @xmath1 scheme is optimal ( over all linear and non - linear index codes ) .
this means scalar linear index codes are optimal for this class of digraphs .
3 . for a class of digraphs
, we prove that the @xmath1 scheme performs at least as well as the partial - clique - cover scheme .
we conjecture that the result is valid in general .
furthermore , we present a class of digraphs where the additive gap between these two schemes grows linearly with the number of vertices in the digraph .
4 . for a class of digraphs
, we prove that the @xmath1 scheme performs at least as well as the fractional - local - chromatic - number scheme .
moreover , we present a class of digraphs where the additive gap between these two schemes grows linearly with the number of vertices in the digraph . 5 .
we show that the @xmath1 scheme can outperform all of the existing graph - based schemes and the composite - coding scheme in some examples .
we extend the @xmath1 scheme to the fractional-@xmath1 scheme .
this modified scheme time - shares multiple @xmath0 structures , and constructs vector linear index codes that can be , for certain digraphs , shorter than the scalar linear index codes obtained from the @xmath1 scheme .
consider a transmitter that wants to transmit @xmath4 messages @xmath5 to @xmath4 receivers @xmath6 in a _
unicast _ message setting , meaning that each message is requested by only one receiver , and each receiver requests only one message . without loss of generality
, let each receiver @xmath7 request message @xmath8 , and possess side information @xmath9 .
this problem can be described by a digraph @xmath10 , where @xmath11 , the set of vertices in @xmath12 , represents the @xmath13 receivers .
an arc @xmath14 exists from vertex @xmath7 to vertex @xmath15 if and only if receiver @xmath7 has packet @xmath16 ( requested by receiver @xmath17 ) as its side information .
the set of the side information of a vertex @xmath7 is @xmath18 , where @xmath19 is the out - neighborhood of @xmath7 in @xmath12 .
let @xmath20 and all @xmath21 be ordered sets , where the ordering can be arbitrary but fixed . [ index code ] suppose @xmath22 for all @xmath7 and some integer @xmath23 , i.e. , each message consists of @xmath24 bits .
given an index - coding problem modeled by @xmath25 , an index code ( @xmath26,@xmath27 ) is defined as follows : 1 .
an encoding function for the source , @xmath28 , which maps @xmath29 to a @xmath30-bit index for some positive integer @xmath30 .
2 . a decoding function @xmath31 for every receiver @xmath7 , @xmath32 , that maps the received index code @xmath33 and its side information @xmath34 to the requested message @xmath8 .
[ broadcast rate or index codelength ] the broadcast rate of an index code ( @xmath35 ) is the number of transmitted bits per received message bits at every user , or equivalently the number of transmitted coded symbols ( each of @xmath24 bits ) .
this is denoted by @xmath36 , and also referred as the normalized length of the index code .
[ optimal broadcast rate ] the optimal broadcast rate for a given index coding problem @xmath25 with @xmath37-bit messages is @xmath38 , and the optimal broadcast rate over all @xmath37 is defined as @xmath39 .
[ maximum acyclic induced sub - digraph ] for a digraph @xmath12 , an induced acyclic sub - digraph formed by removing the minimum number of vertices in @xmath25 , is called a maximum acyclic induced sub - digraph ( mais ) .
the order of the mais is denoted as @xmath40 .
it has been shown that for any digraph @xmath25 and any message length of @xmath24-bits , @xmath40 lower bounds the optimal broadcast rate @xcite , @xmath41 in this sub - section , we describe the clique - cover , the cycle - cover and the partial - clique - cover schemes in detail .
these schemes provide some basic intuitions about our proposed @xmath1 scheme .
[ clique ] a clique is a sub - digraph where each vertex has an out - going arc to every other vertex in that sub - digraph .
[ clique - covering number , @xmath42 the clique - covering number is the minimum number of cliques partitioning a digraph @xmath12 ( over all partitions ) such that if each clique is denoted as @xmath43 for @xmath44 , then @xmath45 , @xmath46 and @xmath47 ( here a vertex is a clique of size one ) .
the clique - covering number is equal to the chromatic number of the underlying undirected graph of the complement digraph , i.e. , @xmath48 , where @xmath49 is the underlying undirected graph of @xmath50 , @xmath50 is the complement digraph of @xmath12 , and @xmath51 denotes the chromatic number of @xmath12 .
[ clique - cover scheme ] the clique - cover scheme finds a set of disjoint cliques that provides the clique - covering number , and constructs an index code in which the coded symbol for each of the disjoint cliques is the bit - wise xor of messages requested by all of the vertices in that clique .
the clique - cover scheme achieves the following rate : the optimal broadcast rate of an index coding instance is upper bounded by the clique cover number , i.e. , @xmath52 .
[ path and cycle ] a _ path _
consists a sequence of distinct ( except possibly the first and last ) vertices , say @xmath53 , and an arc @xmath54 for each consecutive pair of vertices @xmath55 for all @xmath56 .
we represent a path from the vertex @xmath57 to the vertex @xmath58 in a digraph @xmath12 as @xmath59 . here
@xmath60 is the _ first vertex _ and @xmath58 is the _
last vertex_. a path with the same first and last vertices is a _ cycle_. [ cycle - covering number , @xmath61 the difference between the total number of vertices in @xmath12 and the maximum number of disjoint cycles in @xmath12 is the cycle - covering number . [ cycle - cover scheme ]
the cycle - cover scheme finds a set of disjoint cycles in @xmath12 that provides the cycle - covering number , and constructs an index code that has ( i ) coded symbols for each disjoint cycle ( for a cycle @xmath62 , a set of coded symbols are @xmath63 ) , and ( ii ) uncoded messages which are requested by those vertices not included in any of the disjoint cycles in @xmath12 .
the cycle - cover scheme achieves the following rate : the optimal broadcast rate of an index - coding problem is upper bounded by the cycle - covering number , i.e. , @xmath64 .
a sub - digraph @xmath65 is a @xmath66-partial clique , where @xmath67 and @xmath68 is the minimum out - degree of @xmath69.[partialclique1 ] if @xmath70 , for some positive integer @xmath71 , partition a digraph @xmath12 such that @xmath72 and @xmath73 , @xmath74 and @xmath47 , then @xmath75= \sum_{i=1}^{m } ( \kappa(d_i)+1),\ ] ] and the partial - clique number of the digraph is @xmath76,\ ] ] where the minimum is taken over all partitions .
the partial - clique - cover scheme finds a set of disjoint partial cliques in @xmath12 that provides the partial - clique number , and constructs an index code that has ( i ) @xmath77 coded symbols for each disjoint @xmath66-partial clique with @xmath78 ( a partial clique uses mds codes to generate coded symbols ) , and ( ii ) an uncoded message for each disjoint @xmath66-partial clique with @xmath79 .
the partial - clique - cover scheme achieves the following rate : [ prepbrik ] the optimal broadcast rate of an index coding instance is upper bounded by the partial - clique number , i.e. , @xmath80 the partial - clique - cover scheme performs at least as well as the cycle - cover and the clique - cover schemes , i.e. , @xmath81 this is because the partial - clique - cover scheme includes the cycle - cover scheme or the clique - cover scheme as a special case . by definition ,
a clique is a @xmath82-partial clique , and a cycle with @xmath83 vertices is a @xmath84-partial clique . despite the fact that the partial - clique - cover scheme uses mds codes , which require sufficiently large message length @xmath37 in general , to construct index codes
, one can find mds codes for any cycle and any clique for any @xmath37 .
the clique - cover , the cycle - cover and the partial - clique - cover schemes provide scalar linear index codes .
we can also construct vector linear index codes by time - sharing all possible cliques , cycles , partial cliques in their respective schemes , and these are called the fractional versions of those schemes .
the fractional version can strictly decrease the broadcast rates ( over the non - fractional version ) for some digraphs , e.g. , a 5-cycle @xcite .
we present an example that illustrates the importance of overlapping cycles on index coding .
consider the digraph @xmath85 in fig .
[ figmotivationa ] . in @xmath85 , the cycles @xmath86 and @xmath87 overlap at vertex @xmath57 , and some cycles similarly overlap at vertices @xmath88 and @xmath89 . note that @xmath90 .
index codelengths for @xmath85 by existing graph - based schemes ( some schemes require a sufficiently large @xmath37 ) are depicted in table [ table:1 ] .
.index codelengths for the digraph @xmath85 in fig .
[ figmotivationa ] from existing schemes [ cols="^,^",options="header " , ] [ table:2 ] the index codelength provided by the aforementioned existing schemes are strictly greater than @xmath91 except the composite - coding scheme .
there exist digraphs where the @xmath3 scheme outperforms the composite - coding scheme . for an example , the digraph shown in fig .
[ 4a ] is a @xmath92-@xmath93 structure , and it is denoted @xmath94 .
an index code from the @xmath95 scheme is @xmath96 , which is of length @xmath97 . from theorem [ theorem3 ] , @xmath98 .
for @xmath99 , the index codelength provided by the composite - coding scheme is 3.5 , which is greater than @xmath100 .
in this section , firstly , we extend the @xmath1 scheme using time - sharing to code over overlapping @xmath0 structures in a digraph , and to obtain vector linear index codes . secondly , we extend the definition of the @xmath0 structure in such a way that we can extend the @xmath1 scheme to code on extended @xmath0 structures .
let @xmath101 denote a sub - digraph induced by a subset @xmath102 of the vertices in @xmath103 .
we define a function @xmath104 the index codelength from the fractional @xmath1 scheme is represented as @xmath105 , and given by the following linear program : @xmath106,\ s\in s. \nonumber \end{aligned}\ ] ] here @xmath107 is the power set of @xmath108 . in the fractional @xmath1 scheme ,
each sub - digraph induced by the subset @xmath102 in @xmath109 is assigned a weight @xmath110 $ ] such that the total weight of each message over all of the subsets it belongs to is at least one . in this scheme , @xmath105 is the minimum sum of weights .
the @xmath1 scheme is a special case of the fractional @xmath1 scheme where @xmath111 , so @xmath112 .
we start with an example that provides an insight to the extension of the @xmath0 structure .
consider a digraph @xmath12 that has three cliques @xmath113 , @xmath114 and @xmath115 each of size two .
let @xmath116 and @xmath117 be vertex sets of those three cliques respectively .
furthermore , for the clique pair ( @xmath118 ) , all vertices in @xmath119 have out - going arcs to all vertices in @xmath120 , and the result follows similarly for clique pairs ( @xmath121 ) and ( @xmath122 ) .
this digraph is depicted in fig .
[ fig2a ] .
one can verify that @xmath12 is a @xmath92-@xmath0 structure with an inner vertex set @xmath123 .
we can not get a @xmath124-@xmath0 structure in @xmath12 .
suppose that we pick @xmath125 , there is no path from vertices @xmath57 and @xmath88 to vertex @xmath124 without passing through the inner vertex @xmath89 or @xmath92 . by symmetry ,
choosing any 5 vertices as an inner vertex set will have the same issue .
now the @xmath1 scheme gives an index code @xmath126 of length @xmath127 .
however , from the coding point of view of an @xmath0 structure , vertices @xmath124 and @xmath128 need not be separated during encoding because they have the same arc sets @xmath129 and @xmath130 , and have arcs to each other .
this means we can get another index code by removing @xmath131 , and replacing @xmath132 with @xmath133 , i.e. , @xmath134 of length two . here due to the special connectivity of clique @xmath135 , we have treated it as a single vertex , and used @xmath133 in the code construction . in light of this
, we will now extend the definition of an @xmath0 structure to capture cliques with such special configurations . to achieve this
we define a term called a super - vertex .
[ super - vertex ] in a digraph @xmath12 , let @xmath136 be a vertex set where ( i ) all vertices in @xmath136 have arcs to each other , i.e. , they form a clique , and ( ii ) every vertex @xmath137 has the same @xmath138 and the same @xmath139 .
such a group of vertices ( all @xmath137 ) is called a super - vertex and denoted as @xmath140 .
now we define extended @xmath0 structures and an index - coding scheme for them .
[ extended @xmath0 structure ] the extended @xmath141 ( @xmath142 ) structure is defined as an @xmath0 structure that allows super - vertices in its non - inner vertex set .
-15pt [ extended @xmath1 scheme ] for any digraph @xmath12 , the extended @xmath1 ( @xmath143 ) scheme finds a set of disjoint @xmath142 structures covering @xmath12 .
it then codes each of these @xmath142 structures using the code construction described in the following : * each super - vertex ( non - inner vertices ) is treated as a single vertex during the construction and the encoding process of the @xmath142 structure .
* we consider the message requested by the super - vertex to be the xor of all messages requested by the vertices forming the super - vertex .
* each of these @xmath142 structures are treated as an @xmath0 structure , and an index code is constructed using the @xmath144 scheme .
along with super - vertices , and taking their definition into account , one can prove the validity of the code constructed by the @xmath143 scheme similar to the proof of proposition [ propositionvalidic ] .
denote the length of the index code produced by the @xmath143 scheme by @xmath145 . for a digraph @xmath12 ,
the index codelength obtained from the @xmath143 scheme is a better upper bound to the optimal broadcast rate than the codelength obtained from the @xmath1 scheme , i.e. , @xmath146 .
it follows from the definition of @xmath142 structures that include @xmath0 structures as special cases .
graph - based approaches have been shown to be useful for index coding , in which cycles play an important role . prior to this work ,
disjoint cycles and disjoint cliques ( including the timeshared version ) were used to construct index codes . in this work
, we attempted to extend the role of cycles on index coding .
we took a step further and showed the benefits of coding on interlinked - cycle structures ( one form of overlapping cycles ) .
our proposed scheme generalizes coding on disjoint cycles and disjoint cliques . by identifying a useful interlinked - cycle structure
, we were able to characterize a class of infinitely many graphs where scalar linear index codes are optimal .
for some classes of digraphs , we proved that the @xmath1 scheme performs at least as well as some existing schemes such as the partial - clique - cover scheme and the fractional - local - chromatic - number scheme .
furthermore , for a class of digraphs , we proved that the partial - clique - cover scheme and the @xmath1 scheme have linearly - growing additive gap in index codelength with the number of vertices in the digraphs .
we proved a similar result for the fractional - local - chromatic - number scheme and the @xmath1 scheme for another class of digraphs .
we extended the @xmath1 scheme , to allow time - sharing over all possible @xmath0 structures in digraphs .
we also extended the @xmath0 structure to allow super vertices as its non - inner vertices .
however , it remains an open problem to identify cycles overlapping in other useful ways .
a @xmath147-@xmath0 structure @xmath148 has some properties captured in the following lemmas , which will be used to prove proposition [ propositionvalidic ] . here
we consider @xmath149 and @xmath150 as any two distinct directed rooted trees present in @xmath148 with the root vertices @xmath7 and @xmath17 respectively .
[ lemma1 ] for any vertex @xmath151 , and @xmath152 , the set of leaf vertices that fan out from the common vertex @xmath153 in each tree is a subset of @xmath154 . in a tree @xmath149 ( see fig . [ fig3 ] ) , for any vertex @xmath155 and @xmath152 , let @xmath156 be a set of leaf vertices that fan out from vertex @xmath153 .
if vertex @xmath157 , then there exists a path from @xmath153 to @xmath17 in @xmath149 .
however , in @xmath150 , there is a path from @xmath17 to @xmath153 .
thus in the sub - digraph , a path present in any @xmath149 also present in @xmath148 . ] @xmath148 , we obtain a path from @xmath153 to @xmath17 ( via @xmath149 ) and vice versa ( via @xmath150 ) . as a result , a cycle including non - inner vertices and only one inner vertex ( i.e. , @xmath17 ) exists .
this cycle is an i - cycle , and condition 1 ( i.e. , no i - cycle ) for @xmath148 is violated .
hence , @xmath158 . in other words , @xmath159
similarly , @xmath160 . and @xmath150 with root vertices @xmath7 and @xmath17 respectively , and a non inner vertex @xmath153 in common .
here we have used solid arrow to indicate an arc , and dashed arrow to indicate a path.,height=196 ] [ lemma2 ] for any vertex @xmath161 , and @xmath152 , the out - neighborhood of vertex @xmath153 is same in both trees , i.e. , @xmath162 . here the proof is done by contradiction .
let us suppose that @xmath163 .
for this proof we refer to fig .
this proof has two parts . in the first part
, we prove that @xmath164 , and then prove that @xmath165 in the second part .
( part 1 ) suppose that @xmath166 . from lemma [ lemma1 ] , @xmath167 is a subset of @xmath168 .
now pick a vertex @xmath169 that belongs to @xmath168 such that @xmath170 but @xmath171 ( such @xmath169 exists since we suppose that @xmath166 , and we swap the indices @xmath7 and @xmath17 if @xmath172 ) . in tree @xmath149 , there exists a directed path from vertex @xmath7 , which includes @xmath153 to the leaf vertex @xmath169 .
let this path be @xmath173 .
similarly , in tree @xmath150 , there exists a directed path from vertex @xmath17 , which does nt include @xmath153 ( since @xmath171 ) , and ends at the leaf vertex @xmath169 .
let this path be @xmath174 . in the digraph @xmath148
, we can also obtain a directed path from @xmath17 which passes through @xmath153 ( via @xmath175 ) , and ends at the leaf vertex @xmath169 ( via @xmath149 ) .
let this path be @xmath176 .
the paths @xmath174 and @xmath177 are different , which indicates the existence of multiple i - paths from @xmath17 to @xmath169 in @xmath148 , this violates the condition 2 for @xmath148 .
consequently , @xmath164 .
( part 2 ) now we pick a vertex @xmath178 such that , without loss of generality , @xmath179 but @xmath180 ( such @xmath178 exists since we assumed that @xmath163 , and we swap the indices @xmath7 and @xmath17 if @xmath181 ) . furthermore , we have two cases for @xmath178 , which are ( case 1 ) @xmath182 , and ( case 2 ) @xmath183 .
case 1 is addressed in the first part of this proof . on the other hand , for case 2 ,
we pick a leaf vertex @xmath184 such that there exists a path ( see fig . [ fig3 ] ) that starts from @xmath153 followed by @xmath178 , and ends at @xmath185 , i.e. , @xmath186 exists in @xmath149 .
a path @xmath187 exists in @xmath150 .
thus a path @xmath188 exists in @xmath148 . from the first part of the proof
, we have @xmath164 , so @xmath189 . now in @xmath150 , there exists a path from @xmath17 to @xmath185 , which includes @xmath153 followed by a vertex @xmath190 such that @xmath191 and @xmath192 ( as @xmath193 ) , and the path ends at @xmath185 , i.e. , @xmath194 which is different from @xmath188 .
note that in trees @xmath149 and @xmath150 , only the root and the leaf vertices are from @xmath195 , so multiple i - paths are observed at @xmath185 from @xmath17 .
this violates condition 2 for @xmath148 .
consequently , @xmath196 .
[ lemma3 ] if a vertex @xmath197 such that @xmath198 , then its out - neighborhood is the same in @xmath149 and in @xmath148 , i.e. , @xmath199 . for any @xmath197 from lemma [ lemma2 ] , @xmath165 for all @xmath200 . since @xmath201
, vertex @xmath153 must have the same out - neighborhood in @xmath148 as well .
[ proof of proposition [ propositionvalidic ] ] from , all @xmath202 which are non - inner vertices , can decode their requested messages .
this is because the coded symbol @xmath203 is the bitwise xor of the messages requested by @xmath17 and its all out - neighborhood vertices , and any @xmath17 knows messages requested by all of its out - neighborhood vertices as side information . for an inner vertex @xmath7 , rather than analyzing the sub - digraph @xmath148
, we will analyze its tree @xmath204 , and show that it can decode its message from the relevant symbols in @xmath205 .
we are able to consider only the tree @xmath149 due to the lemma [ lemma3 ] .
now let us take any tree @xmath149 .
assume that it has a height @xmath206 where @xmath207 .
the vertices in @xmath149 are at various depths , i.e. , @xmath208 from the root vertex @xmath7 .
the root vertex @xmath7 has depth zero , and any vertex at depth equal to the height of the tree is a leaf vertex .
firstly , in @xmath149 , we compute the bitwise xor among coded symbols of all non - leaf vertices at depth greater than zero , i.e. , @xmath209 .
however , in @xmath149 , the message requested by a non - leaf vertex , say @xmath210 , at a depth strictly greater than one , appears exactly twice in @xmath211 ; 1 .
once in @xmath212 , where @xmath213 is parent of @xmath30 in tree @xmath149 , and 2 .
once in @xmath214 . refer to for mathematical details .
thus they cancel out each other while computing @xmath215 in the tree @xmath149 .
hence , in the tree @xmath149 , the resultant expression is the bitwise xor of 1 .
messages requested by all non - leaf vertices at depth one , and 2 .
messages requested by all leaf vertices at depth strictly greater than one .
refer to for mathematical details .
secondly , in @xmath149 , we compute @xmath216 ( refer to for mathematical details ) which yields the bitwise xor of 1 .
the messages requested by all non - leaf vertices at depth one , which are out - neighbors of @xmath7 , 2 .
the messages requested by all leaf vertices at depth one , which are also out - neighbors of @xmath7 , and 3 .
the message requested by @xmath7 , i.e. , @xmath8 .
this is because the message requested by each leaf vertex at depth strictly greater than one in the tree @xmath149 is present in both the resultant terms of @xmath215 and in @xmath217 , thereby canceling out itself in @xmath216 .
hence , @xmath216 yields the bitwise xor of @xmath8 and @xmath218 . as @xmath7 knows all @xmath218 as side - information
, any inner vertex @xmath7 can decode its required message from @xmath216 .
the mathematical computations of @xmath215 and @xmath219 in the tree @xmath149 are as follows : @xmath220 + where , @xmath221 , and @xmath222 here is obtained because each @xmath223 has only one parent in @xmath149 , and we exclude all @xmath223 whose parent is @xmath7 , and @xmath224 is bitwise xor of messages requested by all of the leaf vertices not in the out - neighborhood of @xmath7 .
if we expand @xmath225 as per the group of vertices according to their depth , we get @xmath226 \hskip-2pt \right ] \hskip-3pt \dotsc \hskip-2pt \right ] \hskip-2pt \right ] \nonumber \\ & = \bigoplus \limits_{j\in n^+_{t_i}(i)\setminus v_{\mathrm{i } } } x_{j}. \end{aligned}\ ] ] note that the intermediate terms in @xmath225 cancel out ( we have used the same color to indicate the terms that cancel out each other ) .
now substituting @xmath225 of and @xmath227 of in , we get @xmath228 @xmath229 let @xmath256 be a @xmath89-@xmath0 structure having an inner - vertex set @xmath257 .
now we prove that any non - inner vertices of @xmath258 belonging to an i - path could not contribute to form a cycle including only non - inner vertices of @xmath256 . to show this , we start by picking two inner vertices @xmath259 and @xmath178 , and the i - path @xmath260 .
we assume that the i - path includes @xmath261 non - inner vertices , i.e. , @xmath262 .
further , let @xmath263 be denoted by @xmath264 .
if @xmath265 , then @xmath266 for @xmath267 . now for @xmath268 , we have the following in @xmath256 : * @xmath269 , @xmath270 and @xmath271 can not contain any vertex in @xmath264 of @xmath260 because of the following : 1 .
if @xmath269 or @xmath270 contains any vertex @xmath272 , then @xmath273 ( part of @xmath260 ) and @xmath274 ( part of @xmath269 or @xmath270 ) form an i - cycle at @xmath259 .
the existence of the i - cycle in @xmath256 contradicts the definition of an @xmath0 structure .
if @xmath271 contains any vertex @xmath272 , then @xmath275 ( part of @xmath271 ) and @xmath276 ( part of @xmath260 ) form an i - cycle at @xmath178 .
the existence of the i - cycle in @xmath256 contradicts the definition of an @xmath0 structure .
* one can verify that only the remaining i - paths @xmath277 and @xmath278 can contain vertices in @xmath264 without forming an i - cycle in @xmath256 .
now for @xmath260 and @xmath277 , these two i - paths must form a directed rooted tree @xmath279 with the root vertex @xmath259 ( by the definition of an @xmath0 structure ) .
thus these two i - paths alone could not form a cycle including only non - inner vertices . furthermore ,
if @xmath260 and @xmath277 contain some common non - inner vertices , then let a vertex @xmath280 , which is in @xmath279 , be the vertex from where @xmath279 branches to @xmath178 and @xmath169 ( refer to fig . [ abc1 ] ) .
* if they contain some common non - inner vertices , then let @xmath281 be the first vertex where @xmath260 and @xmath278 meet each other ( refer to fig .
[ abc2 ] ) . considering only these two i - paths ,
a cycle including only non - inner vertices can form only if a part of @xmath278 contributes to form a path from @xmath282 to @xmath280 for some @xmath283 .
this path is not possible because multiple i - paths would be created from @xmath259 or @xmath169 to @xmath178 in @xmath256 ( contradiction of the definition of an @xmath0 structure ) .
in fact , both the i - paths have the same destination , i.e. , @xmath178 , so there must be only one path from @xmath284 to @xmath178 common in both of them ( to avoid any multiple i - paths ) . for @xmath277 and @xmath278 ,
if these i - paths contain a non - inner vertex @xmath153 in common , then @xmath285 ( part of @xmath278 ) and @xmath286 ( part of @xmath277 ) forms an i - cycle at @xmath169 ( contradiction to the definition of an @xmath0 structure ) .
thus @xmath277 and @xmath278 can not contain any vertex in common except @xmath169 .
consequently , considering @xmath260 , @xmath277 and @xmath278 ( along with assumptions that there are some common non - inner vertices in ( i ) @xmath260 and @xmath277 , and ( ii ) @xmath260 and @xmath278 ) , we end up with the structure as shown in fig .
[ abc3 ] , where @xmath287 and @xmath288 .
this structure contradicts one of the necessary conditions ( let this vertex be @xmath289 ) and an out - going path from a vertex in @xmath264 ( let this vertex be @xmath290 ) , for some @xmath291 , where @xmath292 .
one can easily verify this necessary condition .
] for any vertex in @xmath264 to contribute to form a cycle including only non - inner vertices in @xmath256 .
thus there is no vertex in @xmath264 to contribute to form a cycle including only non - inner vertices in @xmath256 . due to symmetry ,
the result ( any non - inner vertices of @xmath258 belonging to @xmath277 could not contribute to form a cycle including only non - inner vertices of @xmath256 ) implies similarly for non - inner vertices belonging to any of the i - paths in @xmath256 .
therefore , there is no cycle among the non - inner vertices in @xmath256 .
we first prove one lemma that will help to prove the optimality of the @xmath1 scheme .
[ proof of theorem [ theorem3 ] ] we will show that the mais lower bound is tight for all @xmath24 .
we denote the digraph which is also an @xmath0 structure by @xmath230 , and consider that it has @xmath4 vertices . for @xmath231 , the digraph contains only one vertex , and @xmath232 . for @xmath233
, we have the following : ( case 1 ) from lemma [ lemma6 ] , any cycle must include at least two inner vertices , or no inner vertex , thus if we remove @xmath234 inner vertices , then the digraph @xmath148 becomes acyclic .
thus @xmath235 from theorem [ theorem1 ] , we get @xmath236 it follows from , and that @xmath237 . thus @xmath238 .
( case 2 ) a @xmath239 can be viewed in two ways .
the first way is considering the whole @xmath148 as a @xmath147-@xmath0 structure . the second way is considering induced sub - digraphs of @xmath239 which consist of 1 . @xmath58 disjoint cycles together consisting of a total of @xmath240 ( @xmath241 ) non - inner vertices ( if @xmath242 or @xmath57 , then @xmath243 , which is case 1 ) , 2 .
@xmath244 disjoint @xmath0 structures each with @xmath245 vertices and @xmath246 inner vertices in such a way that @xmath247 , we consider that each @xmath0 structure is also disjoint from all @xmath58 cycles among non - inner vertices , and 3 .
total remaining of @xmath248 non - inner vertices ( which are not included in @xmath58 cycles , or the @xmath244 @xmath0 structures ) .
now we will show that both ways of looking at @xmath148 are equivalent in the sense of the index codelength generated from our proposed scheme , and both equal to @xmath249 .
we prefer the second way of viewing @xmath148 for our proof since it is easier to find the mais lower bound . for the partitioned @xmath148 ( looking at in the second way ) ,
the total number of coded symbols is the summation of the coded symbols for ( i ) each of the @xmath58 disjoint cycles ( each cycle has saving equal to one ) , ( ii ) each of the @xmath244 disjoint @xmath0 structures ( each of the @xmath0 structures has savings equal to @xmath234 ) , and ( iii ) @xmath250 uncoded symbols for the remaining non - inner vertices , i.e. , @xmath251 from and , @xmath252 , thus from both perspectives the code length is the same . now for @xmath148 ( looking at in our second way ) , if we remove one vertex from each of the @xmath58 cycles among non - inner vertices ( @xmath58 removal in total ) , and remove @xmath253 vertices from each of the @xmath244 @xmath0 structures ( @xmath254 ) , i.e. , total removal of @xmath234 , then the digraph becomes acyclic .
thus @xmath255 it follows from , , and that @xmath237 .
thus @xmath238 . in this section , firstly , for every minimal partial clique @xmath294 with @xmath295 , we prove that there exists an @xmath0 structure within it such that both of the schemes ( partial - clique - cover and @xmath144 ) provide the same savings .
secondly , we conjecture that the result is valid in general ( this is the main reason that results the conjecture [ conj1 ] ) . the summary is depicted in table [ table1 ] . finally , we prove the theorem .
[ lemmaa1 ] a minimal partial clique @xmath294 with @xmath296 vertices and the minimum out - degree @xmath297 is a cycle , and both the partial - clique - cover scheme and the @xmath144 scheme achieve savings of one .
the properties of @xmath294 with @xmath296 and @xmath297 are as follows : in @xmath294 , ( i ) as @xmath297 , the partial - clique - cover scheme provides the savings of one by proposition [ lemm1 ] , ( ii ) further partitioning could not provide any savings because it is a minimal - partial - clique , and ( iii ) there exists at least a cycle because an acyclic digraph should have a sink vertex ( a vertex with out - degree zero ) , and that does not exist since @xmath297 .
thus @xmath294 , which is a minimal partial clique , having all of the above properties is a cycle . from theorem
[ theorem2 ] , a cycle is a @xmath88-@xmath0 structure , and the @xmath144 scheme provides savings of one by . note that such @xmath294 with @xmath298 is also a clique of order two .
[ lemmaa2 ] a minimal partial clique @xmath294 with @xmath303 vertices and the minimum out - degree @xmath304 , is a clique of order @xmath305 , and @xmath294 has the savings of @xmath306 from both the partial - clique - cover scheme and the @xmath144 scheme . in @xmath294 ,
if @xmath304 , then each vertex has an out - going arc to every other vertex , which is a clique by definition . thus the partial - clique - cover scheme provides savings of @xmath306 by proposition [ lemm1 ] . from theorem [ theorem2 ] ,
a clique is a @xmath305-@xmath0 structure , and the @xmath144 scheme provides savings of @xmath306 by .
[ lemmaa3 ] in any minimal partial clique @xmath294 with @xmath307 vertices and @xmath308 , there exists a @xmath89-@xmath0 structure , and @xmath294 has the savings of two from both the partial - clique - cover scheme and the @xmath144 scheme .
[ def1 ] a path in the forward direction indicates a path from a vertex @xmath7 to any vertex @xmath17 such that @xmath287 , and a path in the reverse direction indicates a path from a vertex @xmath17 to any vertex @xmath7 such that @xmath309 . for simplicity , we refer a _ forward path _ to a path in the forward direction , and a _ reverse path _ to a path in the reverse direction .
[ farthest path ] [ def2 ] consider a sequence of vertices labeled in an increasing order such that there are multiple paths from a vertex @xmath7 in the sequence to other vertices in the sequence . among those paths from @xmath7 , the path to the vertex with the largest label
is called the farthest path from vertex @xmath7 . for an example , let @xmath310 be a sequence of vertices in an increasing order , and @xmath57 has paths to vertices @xmath311 .
the path from from @xmath57 to @xmath312 is the farthest path from @xmath57 . *
a minimal partial clique @xmath294 with @xmath307 number of vertices and @xmath308 has at least one cycle by lemma [ lemmaa4 ] . without loss of generality ,
denote the cycle by @xmath313 , and its vertices by @xmath314 . for simplicity ,
vertices of @xmath313 are labeled in an increasing order as shown in fig .
* for @xmath315 , @xmath316 denotes a path from @xmath259 to @xmath178 including only vertices and arcs of @xmath313 , and @xmath317 denotes a path from @xmath259 to @xmath178 consisting of arcs and vertices outside @xmath313 except vertices @xmath259 and @xmath178 .
* for the vertices in @xmath318 ( see fig . [ 2 ] ) , vertex @xmath57 ( the first vertex of @xmath313 ) has only forward paths to other vertices of @xmath319 ( definition [ def1 ] ) , and vertex @xmath30 ( the last vertex of @xmath320 ) has only reverse paths to other vertices of @xmath320 ( definition [ def1 ] ) .
let @xmath321 be the first vertex in the vertex sequence that has a reverse path to any vertex in @xmath318 , i.e. , there exist a @xmath322 such that @xmath323 and all vertices in @xmath324 have no reverse path . 1 .
there exists at least one cycle , @xmath313 , and 2 . for a vertex @xmath326 ,
one of its out - going arcs ( beside the arc in @xmath313 ) contribute to form a path that always returns to some vertex in @xmath313 .
in @xmath294 with @xmath307 vertices and @xmath308 , there exists at least one cycle , denoted @xmath313 . this is because if any @xmath327 is acyclic , then it must contain at least one sink vertex , and there can not be sink vertices in @xmath294 since @xmath308 .
assume that vertices @xmath328 such that there is an arc from @xmath301 to @xmath329 in @xmath313 .
now in @xmath294 , @xmath301 has out - degree @xmath330 , and the next out - going arc of @xmath301 ( other than the arc in @xmath313 ) consider the terminal vertex of the path @xmath332 to be @xmath333 such that @xmath334 .
vertex @xmath333 can not be a sink vertex because @xmath335 , and it must contribute to form a path . since there are no disjoint cycles ( all cycles are connected , otherwise any two disjoint cycles in @xmath294 provides savings of two and such @xmath327 is not a minimal partial clique with @xmath308 ) , @xmath333 must have a return path to a vertex in @xmath318 .
we had considered that vertex @xmath7 of @xmath313 is the first vertex having a reverse path .
now if any path @xmath336 meets path @xmath322 at some vertices , then there exists a reverse path from vertex @xmath213 to @xmath17 .
this is not possible because such @xmath213 would have been the first vertex in the vertex sequence that has a reverse path ( contradiction ) .
if the farthest path from @xmath17 , @xmath340 , meets any path @xmath342 at some vertices , then there exists @xmath345 .
this is not possible ; otherwise , path @xmath345 would have been the farthest path ( contradiction ) . if @xmath348 or @xmath349 , then we get a figure - of - eight structure at @xmath17 .
two closed paths at @xmath17 will be ( i ) @xmath350 , and ( ii ) @xmath320 . if @xmath351 , then let @xmath352 be the common vertex which is nearest to @xmath17 in these two paths @xmath340 and @xmath346 .
now we get a figure - of - eight structure at @xmath17 .
two closed paths at @xmath17 will be ( i ) @xmath353 ( part of @xmath340 ) , @xmath354 ( part of @xmath346 ) , and ( ii ) @xmath320 . these closed paths are vertex - disjoint except @xmath17 . [ proof of lemma [ lemmaa3 ] ] the proof is done by the detailed structural analysis of the minimal partial clique @xmath294 .
we divide the proof into two parts : in part - i , we prove that @xmath294 has a figure - of - eight structure ( see definition [ fig8 ] ) at a vertex , and in part - ii , we prove there exist a @xmath89-@xmath0 structure within @xmath294 having a figure - of - eight structure .
( part - i ) consider the cycle @xmath313 in @xmath294 has @xmath355 vertices .
now based on @xmath313 , there are two cases in @xmath294 , and those are ( i ) @xmath313 with @xmath356 , i.e. , @xmath313 is a clique of size two , and ( ii ) @xmath313 with @xmath357 .
we assume @xmath358 and @xmath359 are the two distinct vertices belonging to @xmath318 in such a way that there exist a directed arc from @xmath301 to @xmath153 . in @xmath294 , @xmath330 and
the next out - going arc of @xmath301 goes out of @xmath313 , and contributes to form a path , say @xmath332 that returns to some vertex in @xmath318 by lemma [ lemmaa4 ] .
( case ( i ) ) the cycle @xmath313 is a clique of size two . therefore , the path @xmath332 returns to @xmath313 either at @xmath301 or @xmath153 .
now we get a figure - of - eight structure at @xmath301 if the path returns to @xmath301 , otherwise it returns to @xmath153 . for the latter case
, we get a new cycle that includes the path @xmath332 ( which includes @xmath301 , some vertices other than @xmath153 ) , vertex @xmath153 and the arc from @xmath153 to @xmath301 ( i.e. , arc of the cycle @xmath313 ) .
the new cycle has more than two vertices , thus this ends up with case ( ii ) .
on the basis where path @xmath332 returns , we have the following sub - cases : @xmath332 returns to ( ii - a ) @xmath301 ( see in fig . [ 1a ] ) providing a figure - of - eight structure , ( ii - b ) @xmath153 ( see in fig .
[ 1c ] and [ 1d ] ) , and ( ii - c ) some vertex @xmath360 ( see in fig . [ 1b ] ) . for sub - case ( ii - b )
, we have a vertex @xmath361 such that there is a direct arc from @xmath352 to @xmath153 .
since there are no disjoint cycles , the next out - going arc of @xmath352 ( besides the arc from @xmath352 to @xmath153 ) contributes to form a path to a vertex in @xmath362 .
if @xmath352 has path to @xmath301 , then we get a figure - of - eight at @xmath301 ( shown in fig .
[ 1c ] ) , otherwise we have the following : 1 .
@xmath352 has a path to some @xmath360 , so this ends up with sub - case ( ii - c ) ( shown in fig . [ 1b ] ) , or 2 .
@xmath352 has a path to @xmath153 , but for this case , we have another path ( shown in fig . [ 1d ] ) from @xmath301 to @xmath153 ( beside the direct arc from @xmath301 to @xmath153 , and path including the direct arc from @xmath352 to @xmath153 ) , and in this path , we can repeat sub - case ( ii - b ) by considering the predecessor of @xmath153 in place of @xmath363 . for sub - case
( ii - c ) , note that we have a path @xmath332 , which is vertex - disjoint from @xmath313 except the first and the last vertices , which starts from any vertex @xmath364 and returns to some vertex @xmath365 .
we start analyzing the sub - case ( ii - c ) in @xmath294 considering vertices in @xmath318 . for the vertices in @xmath318 ( see fig . [ 2 ] ) , in a sequential order starting from vertex @xmath57 , we track their out - going paths ( which may include vertices not in @xmath366 ) to other vertices in @xmath318 . for this sub - case , we know that there exist a @xmath322 such that @xmath323 and all vertices in @xmath324 have only forward paths .
now we consider out - going paths from @xmath17 ( the paths are always forward paths ) , and get the following subsub - cases : 1 .
if the farthest path from @xmath17 is @xmath367 for some @xmath368 , then there exists a figure - of - eight at @xmath17 .
two closed paths at @xmath17 will be ( i ) @xmath369 , @xmath322 , and ( ii ) @xmath367 , @xmath370 ( refer fig .
if @xmath367 and @xmath322 are vertex - disjoint except @xmath17 , then by recalling the definition of @xmath371 and @xmath317 for any @xmath315 , one can show that the two closed paths are vertex - disjoint except @xmath17 .
otherwise , @xmath367 and @xmath322 are not vertex - disjoint except @xmath17 , and by lemma [ prop4 ] , one can find a figure - of - eight at @xmath17 .
if the farthest path from @xmath17 is @xmath367 for @xmath372 , then there exists a figure - of - eight at @xmath17 .
two closed paths at @xmath17 will be ( i ) @xmath373 , @xmath322 , and ( ii ) @xmath374 , @xmath375 for @xmath376 ( one of the forward paths of @xmath377 ) , @xmath378 ( refer fig . [ 2b ] ) .
using lemmas [ prop2 ] and [ prop3 ] , and recalling the definition of @xmath371 and @xmath317 for any @xmath315 , one can show that these two closed paths are vertex - disjoint except @xmath17 .
3 . otherwise , the farthest path from @xmath17 is @xmath340 for some @xmath379 . starting from @xmath17 , we have at least two forward paths to some vertices in @xmath318 such that these paths are vertex - disjoint except @xmath17 .
we assume these paths are path - a ( the path from @xmath17 to @xmath380 ) and path - b ( the path from @xmath17 to @xmath213 ) .
paths @xmath381 and @xmath367 will be path - a and path - b respectively ( refer fig . [ 3 ] ) . now considering the out - going paths from @xmath380 into account ( the paths are always forward paths ) , we get the following subsubsub - cases : 1 .
if the farthest path from @xmath380 is @xmath382 for some @xmath376 , then there exists a figure - of - eight at @xmath17 .
two closed paths at @xmath17 will be ( i ) path - a , @xmath382 , @xmath378 , and ( ii ) path - b , @xmath383 , @xmath322 ( refer fig . [ 3a ] ) .
using lemmas [ prop2 ] and [ prop3 ] , and recalling the definition of @xmath371 and @xmath317 for any @xmath315 , one can show that these two closed paths are vertex - disjoint except @xmath17 .
if the farthest path from @xmath380 is @xmath382 for @xmath384 , then there exists a figure - of - eight at @xmath17 .
two closed paths at @xmath17 are ( i ) path - a , @xmath385 , @xmath322 , and ( ii ) path - b , @xmath386 , @xmath387 for @xmath388 ( one of the forward paths of @xmath377 ) , @xmath389 ( refer fig . [ 3b ] ) .
if @xmath367 and @xmath322 are vertex - disjoint except @xmath17 , then by using lemmas [ prop2 ] and [ prop3 ] , and recalling the definition of @xmath371 and @xmath317 for any @xmath315 , one can show that the two closed paths are vertex - disjoint except @xmath17 .
otherwise , @xmath367 and @xmath322 are not vertex - disjoint except @xmath17 , and by lemma [ prop4 ] , one can find a figure - of - eight at @xmath17 .
3 . otherwise , the farthest path from @xmath380 is @xmath382 for some @xmath390 .
the union of path - b and @xmath391 give new path - a and we update @xmath392 .
similarly , the union of path - a and @xmath382 give new path - b and we update @xmath393 ( refer fig .
[ 3c ] ) . considering new path - a , new path - b , and updating @xmath394 , we repeat the subsubsub - cases of subsub - case 3 . during this iteration ,
if we get subsubsub - case ( a ) or ( b ) , then there is a figure - of - eight . otherwise , we have subsubsub - case ( c ) , where @xmath213 strictly increase to a value up to @xmath377 .
however , when @xmath213 reaches @xmath377 , we must have either subsubsub - case ( a ) or ( b ) .
now for the figure - of - eight structure at @xmath301 in @xmath294 ( see fig .
[ fig4 ] ) , consider a vertex @xmath153 in a cycle @xmath299 ( indicated in blue color ) , and a vertex @xmath395 in another cycle @xmath300 ( indicated in red color ) in such a way that both @xmath153 and @xmath395 have direct arcs going to @xmath301 .
one of the out - going arcs of @xmath153 must contribute to form a path , say @xmath332 , returning to a vertex in @xmath396 .
this is because all other cases are not possible ; * @xmath332 can not return to any vertex in @xmath397 , otherwise a disjoint cycle to @xmath300 will be created .
* @xmath332 can return to @xmath301 , but for this case , we have another path from @xmath153 to @xmath301 ( beside path including direct arc from @xmath153 to @xmath301 ) , and in this path , we can repeat the case by considering the predecessor of @xmath301 in place of @xmath153 .
thus this case ends up with the same consideration as of the cycle @xmath299 with the vertex @xmath153 having direct arc to the vertex @xmath301 .
* @xmath332 must return to some vertex in @xmath398 and @xmath399 due to lemma [ lemmaa4 ] .
now in a similarly way , one of the out - going arcs of @xmath395 must also contribute a path to a vertex in @xmath398 other than @xmath301 .
rearrange @xmath294 to get the structure in fig .
now consider @xmath400 .
we can see that any vertex in @xmath401 has only one i - path each to other two vertices in @xmath401 with no i - cycles .
thus a @xmath89-@xmath0 exists in @xmath294 ( for an example , see fig . [ 123 ] ) .
consecutively , the @xmath144 scheme provides savings of two by .
again , for @xmath294 , the partial - clique - cover scheme provides savings of two by proposition [ lemm1 ] . for all minimal partial cliques that we have analyzed , there exists an ( @xmath402)-@xmath141 structure within each minimal partial clique having the minimum out - degree @xmath403 .
we conjecture that this holds in general ( the following conjecture is not a part of the proof of the theorem [ theorem4 ] , but this provides an intuition about the conjecture [ conj1 ] ) . [ proof of theorem [ theorem4 ] ] given a digraph @xmath12 , if minimal partial cliques , partitioning the digraph to provide partial - clique number , have @xmath404 , then using the @xmath1 scheme on each of the minimal partial cliques , we achieve @xmath405 ( by using lemmas [ lemmaa1 ] , [ lemmaa2 ] and [ lemmaa3 ] , and considering zero savings for @xmath406 from any schemes ) .
the @xmath1 scheme may produce a shorter index code by considering a different partitioning of the digraph ( for an example , see fig .
[ ex1b ] and [ ex1a ] ) .
[ eg3 ] this example illustrates that for the class of digraphs stated in theorem [ theorem4 ] , the @xmath1 scheme performs as least as well as the partial - clique - cover scheme .
consider two digraphs that are depicted in fig .
the digraph in fig .
[ ex1b ] has more savings from the @xmath1 scheme than that obtained from the partial - clique - cover scheme , and the digraph in fig .
[ ex1a ] has equal savings from both schemes .
[ prop6 ] for any digraph @xmath12 of the class a , the index codelength obtained from the partial - clique - cover scheme and the @xmath1 scheme are @xmath147 and @xmath411 respectively , i.e. , @xmath412 and @xmath413 .
if any minimal partial clique @xmath415 in @xmath12 of the class a includes a vertex @xmath416 for any @xmath417 , then @xmath418 .
this is because the out - degree of any vertex @xmath416 in @xmath12 is two , i.e. , @xmath419 by construction .
now any minimal partial clique without any vertex in @xmath420 has @xmath421 .
this is because for any vertex in @xmath422 , only one out - neighbor is in this set , and the rest are in @xmath420 .
so , any minimal partial clique in the digraph can have @xmath423
. now by proposition [ lemm1 ] , we know that @xmath424 . in @xmath12 , we try to construct a minimal partial clique @xmath294 that has @xmath425 ( we do not need to consider @xmath294 having @xmath426 because of lemma [ lemma10 ] ) by starting from @xmath294 with only one vertex and then adding vertices into its vertex set in such a way that we can obtain @xmath425 .
we start from any vertex @xmath416 for an @xmath417 ( we will show a similar result if we start from some vertex @xmath7 ) .
let @xmath427 be the vertex set of @xmath327 .
now we include both of the two out - neighbor vertices of @xmath416 , i.e. , vertices in @xmath428 in @xmath429 . if we include only one out - neighbor vertex of @xmath416 in @xmath429 , then @xmath430 resulting @xmath431 .
the new vertex - induced sub - digraph @xmath294 has @xmath432 , and @xmath431 because @xmath433 . to get a minimum out - degree of two
, we must include another vertex in @xmath434 from @xmath435 . by symmetry
, it does not make any difference which vertex to add .
we arbitrarily choose @xmath436 , for some @xmath437 .
now the new vertex - induced sub - digraph @xmath294 has @xmath438 for a @xmath17 .
here @xmath439 because @xmath440 .
we further include all vertices in @xmath441 in @xmath442 because @xmath443 , and if we include only one of its out - neighbor vertices , then @xmath431 . the new @xmath294 has @xmath444 and @xmath445 .
further , including any vertex set in @xmath446 could not increase @xmath447 because any @xmath294 must have @xmath448 ( lemma [ lemma10 ] ) .
if we start building @xmath294 with some @xmath449 , we will end up with a sub - digraph ( i ) that includes @xmath450 for some @xmath451 where @xmath452 , or ( ii ) that includes @xmath453 for some @xmath451 where @xmath454 and @xmath455 .
altogether , any minimal partial clique with @xmath308 must contain @xmath456 for some @xmath451 where @xmath454 .
however , @xmath294 is not the minimal because by simply considering two partial cliques among the vertices in @xmath294 , i.e. , partial cliques with vertex sets @xmath457 and @xmath458 , we get savings of two ( one in each ) . for a minimal partial clique @xmath294 in @xmath12 of the class
a , lemma [ lemma10 ] provides @xmath423 , and lemma [ lemma11 ] proves that there exists no minimal partial clique with @xmath425 . thus the minimal partial cliques in a digraph @xmath12 of the class a are only cycles . for this case ,
the cycle - cover scheme and the partial - clique - cover scheme for @xmath12 are the same . since any cycle must include two vertices from @xmath422 , @xmath459 . the minimal partial cliques with vertex sets @xmath460
provide @xmath461 . consider a vertex set @xmath462 with @xmath147 number of vertices . in the digraph @xmath12
, any vertex @xmath463 has only one path each to all other vertices in @xmath464 such that only the first and the last vertices of each of these paths belong to @xmath465 ( i.e. , i - path ) .
moreover , there is no i - cycle at any vertex in @xmath465 .
thus this forms an @xmath0 structure with inner - vertex set @xmath465 .
now from , @xmath466 . thus @xmath467 .
a. blasiak , r. d. kleinberg , and e. lubetzky , `` broadcasting with side information : bounding and approximating the broadcast rate , '' _ ieee trans .
inf . theory _ ,
59 , no . 9 , pp . 58115823 , sept .
2013 .
s. e. rouayheb , a. sprintson , and c. georghiades , `` on the index coding problem and its relation to network coding and matroid theory , '' _ ieee trans .
inf . theory _ ,
56 , no .
3187 3195 , jul . | we consider a graphical approach to index coding .
while cycles have been shown to provide coding gain , only disjoint cycles and cliques ( a specific type of overlapping cycles ) have been exploited in existing literature . in this paper
, we define a more general form of overlapping cycles , called the interlinked - cycle ( @xmath0 ) structure , that generalizes cycles and cliques .
we propose a scheme , called the interlinked - cycle - cover ( @xmath1 ) scheme , that leverages @xmath0 structures in digraphs to construct scalar linear index codes .
we characterize a class of infinitely many digraphs where our proposed scheme is optimal over all linear and non - linear index codes .
consequently , for this class of digraphs , we indirectly prove that scalar linear index codes are optimal .
furthermore , we show that the @xmath1 scheme can outperform all existing graph - based schemes ( including partial - clique - cover and fractional - local - chromatic number schemes ) , and a random - coding scheme ( namely , composite coding ) for certain graphs .
shell : bare demo of ieeetran.cls for journals index coding problem , unicast , linear index codes , interlinked - cycle cover , optimal broadcast rate . |
this paper is devoted to studying the stability of dynamical objects which are called by very different terms such as one - mode solutions ( omss ) @xcite , simple periodic orbits ( spos ) @xcite , low - dimensional solutions @xcite , one - dimensional bushes @xcite etc .
below we refer to them as _ nonlinear normal modes _ ( nnms ) .
let us comment on this terminology .
the concept of similar nonlinear normal modes was developed by rosenberg many years ago @xcite .
each nnm represents a periodic vibrational regime in the conservative @xmath1-particle mechanical system for which the displacement @xmath2 of every particle is proportional to the displacement of an arbitrary chosen particle , say , the first particle [ @xmath3 at any instant @xmath4 : @xmath5 where @xmath6 are constant coefficients .
note that convenient _ linear normal modes _ ( lnms ) also satisfy eq .
( [ eqch1 ] ) since , for any such mode , one can write @xmath7 where @xmath8 are constant amplitudes of individual particles , while @xmath9 and @xmath10 are the frequency and initial phase of the considered mode . as a rule
, nnms can exist in the mechanical systems with rather specific interparticle interactions , for example , in systems whose potential energy represents a _ homogeneous _ function with respect to all its arguments .
however , in some cases , the existence of nnms is caused by certain symmetry - related reasons .
we refer to such dynamical objects as _ symmetry - determined _ nnms . in @xcite
, we have found all symmetry - determined nnms in all @xmath1-particle mechanical systems with any of 230 space groups .
this proved to be possible due to the group - theoretical methods developed in @xcite for constructing _ bushes _ of vibrational modes . at this point
, it is worth to comment on the concept of bushes of modes introduced in @xcite ( the theory of these dynamical objects can be found in @xcite ) . in rigorous mathematical sense
, they represent symmetry - determined _ invariant manifolds _ decomposed into the basis vectors of _ irreducible representations _ of the symmetry group characterizing the considered mechanical system ( `` parent '' group ) . because of the specific subject of the present paper , it is sufficient to consider only bushes of vibrational modes in nonlinear monoatomic chains .
such bushes have been discussed in @xcite .
let us reproduce here some ideas and results from these papers .
every bush b@xmath11 $ ] describes a certain vibrational regime , and some specific _ pattern _ of instantaneous displacements of all the particles of the mechanical system corresponds to it . in turn , this pattern is characterized by a symmetry group @xmath12 ( in particular , such group can be trivial ) which is a _ subgroup _ of the symmetry group @xmath13 of the mechanical system in its equilibrium state .
for example , let us consider the _ two - dimensional _ bush b@xmath14 $ ] in the monoatomic chain with periodic boundary conditions whose displacement pattern @xmath15 can be written as follows @xmath16 this pattern is determined by two time - dependent functions @xmath17 , @xmath18 , and the corresponding _ vibrational state _ of the @xmath1-particle chain is fully described by displacements inside the _ primitive cell _ , which is four time larger than that of the equilibrium state .
we will refer to the ratio of the primitive cell size of the vibrational state to that of the equilibrium state as _ multiplication number _
( @xmath19 ) and , therefore , for the pattern ( [ eqch3 ] ) , one can write @xmath20 .
the symmetry group @xmath21}$ ] of the bush b@xmath14 $ ] is determined by two _ generators _ : the translation ( @xmath22 ) by four lattice spacing ( @xmath23 ) and the inversion ( @xmath24 ) with respect to the center of the chain ( note that the condition @xmath25 must hold for existence of such bush ) .
if we decompose the displacement pattern ( [ eqch3 ] ) into the linear normal coordinates @xmath26\right|n=1 .. n\right\}\nonumber\\ & ( j=0 .. n-1),\label{eqch10}\end{aligned}\ ] ] we get the following form of the bush b@xmath14 $ ] in the _ modal space _ : @xmath27 where @xmath28 @xmath29 while @xmath30 and @xmath31 are time - dependent coefficients in front of the normal coordinates @xmath32 and @xmath33 .
thus , only @xmath34 normal coordinates from the full set ( [ eqch10 ] ) contribute to the `` configuration vector '' @xmath15 corresponding to the given bush and we will refer to @xmath35 as the _ bush dimension_. in @xcite , we developed a simple crystallographic method for obtaining the displacement pattern @xmath15 for any subgroup @xmath12 of the parent group @xmath13 . using this method one can obtain bushes of different dimensions for an arbitrary nonlinear chain .
the _ one - dimensional bushes _
( @xmath36 ) represent symmetry - determined nonlinear normal modes .
the displacement pattern @xmath15 corresponding to a given nnm depends on only one ( time - periodic ) function @xmath31 : @xmath37 where @xmath38 is a constant vector , which is formed by the coefficients @xmath39 ( @xmath40 ) from eq .
( [ eqch1 ] ) , while the function @xmath31 satisfies a certain differential equation .
this so - called `` governing '' equation can be obtained by substitution of the ansatz ( [ eqch20 ] ) into the dynamical equations of the considered chain . in some sense
, the concept of bushes of vibrational modes can be considered as a certain _ generalization _ of the notion of nnms by rosenberg . indeed , if we substitute the ansatz ( [ eqch12 ] ) into dynamical equations of the chain , we obviously get two `` governing '' equations for the functions @xmath31 and @xmath30 , that determines the above - discussed two - dimensional bush ( note that , in contrast to a nnm , such dynamical object describes , in general , a _
motion ) . finally , one can conclude that @xmath35-dimensional bush is determined by @xmath35 time - dependent functions for which @xmath35 governing differential equations can be obtained from the dynamical equations of the considered mechanical system .
let us emphasize that bushes of modes represent a new type of _ exact _ excitations in nonlinear systems with discrete symmetries and the excitation energy proves to be trapped in a given bush for the case of hamiltonian systems .
it is very important to emphasize that there exist only a _ finite number _ of vibrational bushes of any fixed dimension in every @xmath1-particle mechanical system . as a consequence
, there is sufficiently small number of nnms ( one - dimensional bushes ) in the fpu chains ( three nnms for the fpu-@xmath41 model and six for the fpu-@xmath0 model ) .
all possible one - dimensional bushes are explicitly listed in our papers @xcite ( see also @xcite ) .
the stability of some nnms in the fpu chains has been studied in @xcite by numerical and analytical methods .
let us comment explicitly on the recent papers @xcite .
in @xcite , t. bountis and coworkers have investigated the destabilization thresholds ( @xmath42 and @xmath43 ) of two nonlinear normal modes which they call spo-1 and spo-2 ( simple periodic orbits ) by numerical methods .
the authors of the above papers try to reveal some relations between the destabilization thresholds @xmath42 , @xmath43 and the origin of the weak chaos in connection with the famous fermi - pasta - ulam problem of the energy equipartition between different modes . in particular , they conclude that the main role in the weak chaos appearance in the thermodynamic limit ( @xmath44 ) plays spo-2 , because @xmath45 , @xmath46 , and , therefore , @xmath47 . however , there are some other spos in the fpu-@xmath0 chain and one can be interested in their role in the origin of the weak chaos in the thermodynamic limit .
some comments are appropriate at this point . according to lyapunov @xcite , some strictly periodic orbits for nonlinear systems
can be obtained from the linear normal modes ( which are introduced in the harmonic approximation ) by continuation with respect to a parameter characterizing the strength of nonlinearity . from this point of view , there exist @xmath1 different spos for longitudinal vibrations of an @xmath1 particle monoatomic chain .
however , only few of the modes , constructed in such a way , possess an _
identical _ time dependence of the displacements of all the particles .
more exactly , only few of the lyapunov modes can be written in the form ( [ eqch20 ] ) implying a separation of time and space variables that is typical for the rosenberg nonlinear normal modes . indeed , in general case , @xmath48 where @xmath49 ( @xmath50 ) are _ different _ functions of time with identical periods .
note that in the present paper we consider only _ extended _ spos , but the same problem there exists for _ localized _ periodic modes ( discrete breathers ) and we have discussed it in detail in @xcite .
as far as we aware , all periodic solutions in monoatomic chains that have been studied up to now ( see the above cited papers @xcite and references therein ) belong namely to the class of the rosenberg nonlinear normal modes determined by eq .
( [ eqch20 ] ) .
moreover , the spatial profiles @xmath51 of these modes possess certain symmetry properties . in particular
, every such mode can be characterized by a multiplication number ( @xmath19 ) determining the enlargement of the primitive cell of the vibrational state in comparison with that of the equilibrium state .
as was already noted , we refer to these modes as symmetry - determined nnms and there exist only finite number of such modes ( even for the case @xmath44 ! ) for each nonlinear chain @xcite . above considered
spo-1 , spo-2 and the well - known @xmath52-mode ( zone boundary mode ) represent nnms with multiplication numbers 4 , 3 , and 2 , respectively . however , among _ six _ symmetry - determined nnms in the fpu-@xmath0 chain @xcite there exist another three nnms with @xmath53 , @xmath20 and @xmath54 .
the stability properties of the second nnm with @xmath20 were studied by m. leo and r.a .
leo in @xcite .
the stability of this mode was investigated in the thermodynamic limit by both numerical and analytical methods .
the stability diagrams for all the nonlinear normal modes in the fpu-@xmath0 chain , as well as for the fpu-@xmath41 chain , can be found in our paper @xcite . with the aid of these diagrams
, one may reveal many stability properties of nnms for an arbitrary @xmath1 , in particular , for the thermodynamic limit ( @xmath44 ) .
note that these diagrams were obtained numerically . in this paper
, we present some _ analytical _ results for the stability properties of all nnms in the fpu-@xmath0 chain in the _ thermodynamic _ limit ( @xmath44 ) .
we also compare our results with those by different authors when it is possible . in sec . 2
, we consider all the possible symmetry - determined nonlinear normal modes in the fpu-@xmath0 chain . in sec .
3 , the stability diagrams for these nnms are discussed . in sec . 4 , the analytical method for studying the stability of nnms in the thermodynamic limit is presented . in sec . 5 , we list our results on the stability properties for every nnm .
as was already mentioned , there exists only _
finite _ number of symmetry - determined nnms in any monoatomic chain .
every nnm corresponds to a certain _ subgroup _ of the symmetry group of the chain dynamical equations .
the difference in the number of nonlinear normal modes for the fpu-@xmath41 chain ( three nnms ) and the fpu-@xmath0 chain ( six nnms ) is associated with the fact that the symmetry group of the fpu-@xmath0 chain dynamical equations is higher than that of the fpu-@xmath41 chain @xcite . in @xcite , we have investigated the stability of all nnms both in the fpu-@xmath41 and fpu-@xmath0 chains ( for the case @xmath55 ) by numerical methods .
let us comment on the main idea of this investigation . following the standard method of the linear stability analysis , we linearize the fpu-@xmath0 dynamical equations near a given nnm and get the linearized system in the form @xmath56 , where @xmath57 represents the infinitesimal perturbation vector , while @xmath58 is the jacobian matrix of the original system of nonlinear differential equations .
thus , we obtain @xmath1 linear differential equations with time - periodic coefficients depending on the considered nnm
. then the floquet method can be applied for studying the stability of the zero solution of the system @xmath56 .
however , such straightforward way for the stability analyzing becomes practically impossible for @xmath59 . in @xcite
, the general group - theoretical method has been developed for splitting ( decomposition ) of the original system @xmath56 of @xmath1 linear differential equations into certain subsystems of sufficiently small dimensions @xmath60 .
for the fpu-@xmath0 chain , these dimensions do not exceed three ( see below ) .
then we have applied the floquet method for such subsystems of small dimensions . moreover , proceeding in this manner , one can reveal those subsets of the vibrational modes , which are responsible for the loss of stability of the considered nnm . as a consequence of this approach
, it proves to be possible to construct very transparent diagrams , which demonstrate explicitly stability properties of each fpu nonlinear normal mode @xcite .
the fpu-@xmath0 model represents a chain of unit masses coupled with each other by the appropriate nonlinear springs .
the dynamical equations describing longitudinal vibrations of the fpu-@xmath0 chain can be written in the form @xmath61 where @xmath2 is the displacement of the @xmath24th particle from its equilibrium state at the instant @xmath4 , while the force @xmath62 depends on the spring deformation @xmath63 as @xmath64 the periodic boundary condition is assumed to hold : @xmath65 let us mention some results of the paper @xcite , which are necessary for our further discussions .
every nnm in the fpu-@xmath0 chain can be written as follows [ see eq .
( [ eqch20 ] ) ] : @xmath66 where @xmath31 satisfies the duffing equation @xmath67 with different values @xmath9 and @xmath68 for different nnms .
the function @xmath31 describes the time - evolution of a given nnm , while the @xmath1-dimensional vector @xmath38 determines the pattern of the displacements of all particles of the chain .
below , we list all possible nnms in the fpu-@xmath0 chain . 1
. b@xmath69 $ ] : @xmath70 this is a boundary zone mode or @xmath52-mode .
2 . b@xmath71 $ ] : @xmath72 there exist three `` dynamical domains '' of this nnm ( see below ) .
3 . b@xmath73 $ ] : @xmath74 there exist two dynamical domains of this nnm .
4 . b@xmath75 $ ] : @xmath76 there exist three dynamical domains of this nnm .
b@xmath77 $ ] : @xmath78 there exist two dynamical domains of this nnm
b@xmath79 $ ] : @xmath80 there exist three dynamical domains of this nnm .
let us comment on the above listed nnms in the fpu-@xmath0 chain .
every nnm , denoted by the symbol b@xmath11 $ ] , is characterized by the corresponding symmetry group @xmath12 , that represents a certain subgroup of the symmetry group @xmath81}$ ] of the fpu-@xmath0 dynamical equations ( [ eqch35],[eqch36 ] ) .
we determine every such group by the set of its generators using the following notations : @xmath23 : : the translation of the chain by one lattice spacing , @xmath24 : : the inversion with respect to the center of the chain , @xmath82 : : the operator , that changes signs of the displacements of all particles without any their transposition .
the symmetry group @xmath81}$ ] of the fpu-@xmath0 dynamical equations is described by _ three _ generators ( @xmath23 , @xmath24 , and @xmath82 ) .
the corresponding transformations @xmath23 , @xmath24 and @xmath82 of @xmath1-dimensional vectors @xmath83 _ do not _ change the dynamical equations ( [ eqch35],[eqch36 ] ) of the fpu-@xmath0 chain .
all the above listed groups of nnms are fully described by only _ two _ generators , but these generators can be written as some _ products _ of the generators @xmath23 , @xmath24 , and @xmath82 of the group @xmath13 . for example , @xmath84 , @xmath85 , @xmath22 are translations of the chain by two , three and four lattice spacings , respectively .
the transformation @xmath86 means that we must perform the inversion of the displacement pattern with respect to the chain center and then translate it by one lattice spacing .
note that transformations @xmath23 and @xmath24 do not commute : @xmath87 [ the relation @xmath88 holds because of the periodic boundary condition ( [ eqch37 ] ) ] . on the other hand
, the transformation @xmath82 does commute with both @xmath23 and @xmath24 transformations : @xmath89 the transformation @xmath90 means that we must change signs of all displacements and then translate the displacement pattern @xmath91 by two lattice spacings .
some simple examples are worth mentioning at this point . for the chain with @xmath92 particles ,
we can write the following relations : @xmath93 the displacement pattern corresponding to a given nnm can be obtained as _ invariant vector _ of its symmetry group @xmath94 .
for example , let us obtain the displacement pattern for the nnm with @xmath95 $ ] [ see eq .
( [ eqch_62 ] ) ] . for simplicity ,
we demonstrate the method for obtaining displacement patterns with the case @xmath96 .
let @xmath97 where @xmath98 ( @xmath99 ) are arbitrary displacements of eight particles of the chain .
the vector @xmath91 must be invariant with respect to the action of our two generators @xmath22 and @xmath86 of the symmetry group of the considered nnm : @xmath100 the former equation is reduced to the following form : @xmath101 from which we conclude that @xmath102 this displacement pattern is formed by two primitive cells whose size four times larger than that of the fpu-@xmath0 chain in its equilibrium state .
the sets of the displacements in both cells are identical : @xmath103 now let us take into account the second equation ( [ eqch65 ] ) .
acting on the vector ( [ eqch66 ] ) by @xmath86 , we obtain @xmath104 then using the equation @xmath105 , we get @xmath106 thus , the invariant ( under the action of the group @xmath107}$ ] ) vector @xmath91 depends on _ only one _ arbitrary parameter , which we denote by @xmath108 : @xmath109 ( note that this vector being invariant with respect to generators of the group @xmath107}$ ] will automatically be invariant relative to all its other elements ) . then the nnm corresponding to the invariant vector ( [ eqch60 ] ) can be written as follows @xmath110 to find all nnms , we can try _ all subgroups _ of the symmetry group @xmath81}$ ] to choose those displacement patterns , which depend on _ only one _ arbitrary parameter .
the patterns depending on @xmath35 arbitrary parameters with @xmath111 form the @xmath35-dimensional bushes of vibrational modes .
namely in this sense nonlinear normal modes may be called one - dimensional bushes . in @xcite , three different group - theoretical methods for constructing the bushes of vibrational modes in _ arbitrary _ @xmath1-particle nonlinear mechanical systems were developed .
the most efficient of these methods uses the concept of irreducible representations of the symmetry groups . taking into account the above method that was used for constructing eq .
( [ eqch70 ] ) , we conclude that every nnm can be written in the form @xmath112 where @xmath38 is a certain time - independent vector . substituting ansatz ( [ eqch71 ] ) into the dynamical equations ( [ eqch35][eqch36 ] ) of the fpu-@xmath0 chain , with explicit forms of the vectors @xmath38 from eqs .
( [ eqch_60][eqch_65 ] ) , one can find that fpu-@xmath0 equations are reduced to only one differential equation ( governing equation of the corresponding nnm ) of the form : @xmath113 this is the duffing equation with different values @xmath9 and @xmath68 for different nnms which are listed in eqs .
( [ eqch_60][eqch_65 ] ) .
above , we have mentioned the existence of so - called `` dynamical domains '' of all nonlinear normal modes presented in eqs .
( [ eqch_60][eqch_65 ] ) .
let us comment on this notion borrowed from the theory of phase transitions .
we have already emphasized that a certain symmetry group @xmath12 corresponds to every nnm .
this group is a subgroup of the symmetry group of the considered mechanical system in its equilibrium state ( @xmath114 ) . if we act on the vector @xmath15 corresponding to a given nnm by operator @xmath115 , that _ does not _
belong to subgroup @xmath12 , we get the _ equivalent _ configuration vector @xmath116 .
the equivalent vector @xmath117 corresponds to a new nnm , which is described by the _
same _ dynamical equations as that of the nnm associated with the vector @xmath15 .
for example , three dynamical domains are associated with the nnm from eq .
( [ eqch_61 ] ) : @xmath118:&\vec c=\frac{3}{\sqrt{6n}}\{1,0,-1~|~1,0,-1~|~1,0,-1~|~\ldots\},\label{eqch90}\\ \text{b}[a^3,ai]:&\vec c=\frac{3}{\sqrt{6n}}\{0,1,-1~|~0,1,-1~|~0,1,-1~|~\ldots\},\label{eqch91}\\ \text{b}[a^3,a^2i]:&\vec c=\frac{3}{\sqrt{6n}}\{1,-1,0~|~1,-1,0~|~1,-1,0~|~\ldots\}.\label{eqch92}\end{aligned}\ ] ] all the displacement patterns ( [ eqch90][eqch92 ] ) differ from each other by a cyclic transposition of the displacements inside each primitive cell of the chain equilibrium state . let us note that the symmetry groups @xmath119 ( @xmath120 ) of nnms from eqs .
( [ eqch90][eqch92 ] ) prove to be _ conjugate _ subgroups in the parent group @xmath13 , for example , @xmath121 ( @xmath115 ) .
since the above - discussed `` domains '' possess equivalent dynamical properties , we study below the stability of only one copy of the full set of dynamical domains for every nnm in the fpu-@xmath0 chain .
all symmetry - determined nnms that can exist in the fpu-@xmath0 chain with an appropriate number of particles are listed in table [ table10 ] . [
cols="<,^,^,^ " , ] the time - depending function @xmath31 entering eq .
( [ eqch100 ] ) is a periodic solution to the governing eq .
( [ eqch38 ] ) . for every nnm in the fpu-@xmath0 chain , this governing equation represents duffing equation @xmath122 where @xmath23 is the squared frequency of the harmonic approximation , while @xmath123 is a nonlinearity coefficient .
( [ eqch39 ] ) is called the _ hard _
( _ soft _ ) duffing equation if @xmath124 ( @xmath125 ) . for initial conditions
@xmath126 , @xmath127 , the solution of the hard duffing equation can be written in the form @xmath128 here @xmath129 while modulus @xmath130 of the jacobi elliptic cosine is determined by the relation @xmath131 the solution ( [ eqch40a ] ) represents periodic function with the period @xmath132 where @xmath133 is the complete elliptic integral of the first kind .
for the same initial conditions , the solution of the soft duffing equation ( @xmath125 ) can be written in the form @xmath134 with @xmath135 @xmath136 @xmath137 it is convenient to introduce the time scaling @xmath138 which transforms eq .
( [ eqch100 ] ) into the equation with @xmath52-periodic coefficients . as a result of this scaling the form of eq .
( [ eqch100 ] ) does not change , but the constant @xmath23 and @xmath139 , entering this equation , must be multiplied by @xmath140 . we do not change notations in eq .
( [ eqch100 ] ) , however , we imply below that the above transformations are already fulfilled .
our further stability analysis of nnms reduces to investigating the stability of the zero solution of eq .
( [ eqch100 ] ) .
the analysis consists of the following steps : * step 1 . *
simplification of eq .
( [ eqch100 ] ) in the thermodynamic limit ( @xmath44 ) . in this case
@xmath141 and we can decompose the coefficients of eq .
( [ eqch100 ] ) into power series with respect to the small dimensionless parameter @xmath142 * step 2 . * search for the general solution of the approximate equation that was obtained as a result of step 1 . * step 3 .
* construction of the monodromy matrix with the aid of the above solution . * step 4 .
* construction of the characteristic polynomial of the monodromy matrix .
* step 5 . * analyzing discriminant of the characteristic polynomial in the limit @xmath143 .
let us consider these steps in turn .
the function @xmath31 in the form ( [ eqch40a ] ) for the case @xmath55 and in the form ( [ eqch50 ] ) for @xmath144 must be substituted into eq .
( [ eqch100 ] ) taking into account that modulus @xmath130 goes to zero when @xmath44 .
to simplify @xmath31 , we use the following formulas from the theory of elliptic functions @xcite : @xmath145 @xmath146 where @xmath147 ^ 2\kappa^{2n}+\ldots\right\}\label{eqch122}\end{aligned}\ ] ] is the complete elliptic integral of the first kind , while @xmath148 ( @xmath149 is the complimentary modulus of the elliptic functions ) .
note that the modulus @xmath130 depends on @xmath150 in a different manner for the cases @xmath55 and @xmath144 [ see eqs .
( [ eqch42 ] ) and ( [ eqch52 ] ) , respectively ] .
now , we have to decompose the left - hand - side of eq .
( [ eqch100 ] ) into the power series with respect to the small parameter @xmath151 .
this very cumbersome decomposition has been performed with the aid of the maple^^ mathematical package .
the corresponding result can be written as follows : @xmath152{\varepsilon}+\right.\nonumber\\ & \quad\left[\frac{75}{128}\hat\omega^2+\left(-\frac{13}{32}-\frac{3}{8}\cos 2\tau+\frac{1}{32}\cos 4\tau\right)\hat m\right]{\varepsilon}^2+\label{eqmy40}\\ & \quad\left.\left[-\frac{243}{512}\hat\omega^2+\left(\frac{87}{256}+\frac{597}{2048}\cos 2\tau-\frac{3}{64}\cos 4\tau+\frac{3}{2048}\cos 6\tau\right)\hat m\right]{\varepsilon}^3\right\}\vec\mu+\nonumber\\ & o\left({\varepsilon}^4\right)=0,\nonumber\end{aligned}\ ] ] eq .
( [ eqmy40 ] ) represents a system of differential equations with time - periodic coefficients and to construct the corresponding monodromy matrix we must obtain its solution for @xmath153 . on the other hand , for small time intervals , the solution of eq .
( [ eqmy40 ] ) can be found by a simple perturbation theory . to that end
we decompose @xmath154 into a formal series @xmath155 substitute it into eq .
( [ eqmy40 ] ) and equate to zero the terms with every fixed power of the small parameter @xmath156 . as a result , we get the following set of differential equations [ eqmy41 ] @xmath157\vec\mu_0,\label{eqmy41b}\\ \ddot{\vec\mu}_2+\hat\omega^2\vec\mu_2=&-\left[-\frac{3}{4}\hat\omega^2+\left(\frac{1}{2}+\frac{1}{2}\cos 2\tau\right)\hat m\right]\vec\mu_1\label{eqmy41c}\\ & -\left[\frac{75}{128}\hat\omega^2+\left(-\frac{13}{32}-\frac{3}{8}\cos 2\tau+\frac{1}{32}\cos 4\tau\right)\hat m\right]\vec\mu_0,\nonumber\\ \ldots\nonumber\end{aligned}\ ] ] because of the diagonal form of the matrix @xmath158 ( see table [ table20 ] ) , these equations determine certain sets of harmonic oscillators with different time - periodic external forces .
each of these oscillators is described by equation @xmath159 the general solution to this equation , obtained by the method of variation of arbitrary constants , can be written in the form @xmath160dt,\ ] ] where @xmath161 , in our case , represents a _ sum _ of time - periodic functions whose frequencies are _
incommensurable_. indeed , for the most nnms from table [ table20 ] , @xmath158 are matrices with different diagonal elements and , therefore , the components of the vector @xmath162 from ( [ eqmy41a ] ) vibrate with different frequencies .
substituting @xmath162 into ( [ eqmy41b ] ) leads to mixing its time - depended components because of multiplying by the matrix @xmath163 , and such a mixing produces more and more complicated terms in r.h.s . of eqs .
( [ eqmy41 ] ) when we take into account higher orders in the decomposition ( [ eqmy40 ] ) the usual way to study stability of a given periodic dynamical regime is the floquet method . in this method , we linearize nonlinear equations of motion in the vicinity of the periodic solution and calculate the _ monodromy matrix _ @xmath164 by integrating @xmath165 times the linearized equations with time - periodic coefficients over one period @xmath166 using specific initial conditions [ @xmath167 is the number of equations in ( [ eqmy40 ] ) ] .
these conditions are determined by the successive columns of @xmath168 identity matrix .
solving eqs .
( [ eqmy41 ] ) in step - by - step manner , we can construct the approximate analytical solution to eq .
( [ eqmy40 ] ) up to a fixed order of the small parameter @xmath156 . with the aid of this solution , we are able to obtain the monodromy matrix @xmath169 for eq .
( [ eqmy40 ] ) , where @xmath52 is the period of its coefficients .
the stability of the considered periodic solution is determined by floquet multipliers representing eigenvalues of the monodromy matrix .
if all these multipliers lie on the unit circle , the solution is _
linear stable_. in other case , the solution _ linear unstable_. we obtain eigenvalues of the monodromic matrix @xmath169 as the roots of its characteristic polynomial .
let us remind that according to the newton formulas , the coefficients of the characteristic polynomial @xmath170 of any @xmath171 matrix @xmath172 can be expressed via the sums @xmath173 with the aid of the recurrence relation @xmath174 thus , we have @xmath175 on the other hand , all sums @xmath176 in eq .
( [ eqch500 ] ) can be found directly by means of traces of the matrix @xmath172 : @xmath177 it is well known , that in the case of any hamiltonian system with @xmath167 degree of freedom floquet multipliers @xmath178 form pairs @xmath179 ( @xmath50 ) and , as a consequence , the characteristic polynomial @xmath180 of the monodromic matrix @xmath169 proves to be _ palindromic _ :
@xmath181 with the following coefficients @xmath8 : @xmath182,\\ & \cdots\end{aligned}\ ] ] now we have to obtain formulas for discriminants @xmath183 , @xmath184 and @xmath185 for the corresponding palindromic characteristic polynomials of @xmath186 degrees via traces of monodromy matrices . from the above newton formulas applied to the polynomial ( [ eqcha1 ] )
, we obtain the following formulas for its coefficients : @xmath187,\label{eqcha2}\\ a_3=&\frac{1}{6}\left[\operatorname{tr}^3\hat x(\pi)-3\operatorname{tr}\hat x(\pi)\operatorname{tr}\hat x^2(\pi)+2\operatorname{tr}\hat x^3(\pi)\right].\nonumber\end{aligned}\ ] ] on the other hand , one can express discriminant @xmath188 explicitly via these coefficients of the characteristic polynomial . with the aid of maple^^ , we finally find @xmath189 where @xmath190 , @xmath191 , @xmath192 are given by eqs .
( [ eqcha2 ] ) .
now we present some illustrations of the above discussed technique .
_ example 1 .
_ nonlinear normal mode b@xmath69 $ ] ( @xmath52-mode ) : @xmath193 .
firstly , let us discuss the case @xmath55 .
one can see that there exist modes , corresponding to the left and right sides of the black region in fig .
[ figstabdiag]a , which are not excited by parametric interactions with the @xmath52-mode .
this fact was revealed analytically in @xcite with the aid of rotating wave approximation ( rwa ) . in @xcite , in the framework of the same approximation the following relation between the amplitude threshold value @xmath194 and the wavenumber @xmath195 ( i.e. the boundary curve of the black region in fig .
[ figstabdiag]a ) was obtained : @xmath196 this analytical formula is in good agreement with the numerical results .
let us now consider the stability threshold of the @xmath52-mode in the thermodynamic limit @xmath143 using the above discussed method .
we have to consider the vicinity of the point @xmath197 on the @xmath198 plane ( @xmath199 ) .
the one - dimensional constant matrices of the decoupled variational system can be found in table [ table20 ] : @xmath200 the monodromy matrix also depends on the wavenumber @xmath195 and we can decompose its trace @xmath201 into the taylor series in two small parameters @xmath202 and @xmath151 ( in our case , @xmath203 , @xmath204 and , therefore , @xmath205 ) .
this decomposition read @xmath206 in the considered case , @xmath207 and the corresponding discriminant is @xmath208 where @xmath209 .
the condition @xmath210 leads to the equations @xmath211 from the first of these equation , we find @xmath212 and then we obtain @xmath213 substitution of @xmath214 and @xmath215 leads us to the following result @xmath216 the corresponding energy per one particle is @xmath217 the second equation ( [ eqch261 ] ) leads to a contradiction with the condition of smallness of the parameter @xmath156 and , therefore , this equation does nt produce instability of the @xmath52-mode in the limit @xmath143 . note that the analytical dependence @xmath218 was revealed in @xcite , and later was recovered in @xcite . in the case
@xmath144 , the stability properties of the @xmath52-mode are utterly different .
indeed , one can see in fig .
[ figstabdiag]a ( right column ) that this mode turns out to be stable up to the finite value of its amplitude @xmath194 . using a numerical method , we found that in the thermodynamic limit @xmath143 @xmath219 ( @xmath220 ) .
_ example 2 .
_ nonlinear normal mode b@xmath77 $ ] : @xmath221 .
the variational system for this nnm is decoupled into @xmath222 subsystems with time - periodic coefficients . as a result , for studying the stability loss of the mode
b@xmath77 $ ] we have to vanish the discriminant @xmath184 from ( [ eqch301 ] ) . the first factor of @xmath184 near the resonance wavenumber @xmath199 is equal to @xmath223 being positive
, this factor can not lead to the condition @xmath224 .
the second factor of @xmath184 from eq .
( [ eqch301 ] ) reads @xmath225 also does nt vanish in the thermodynamic limit @xmath143 .
only the last factor of the discriminant @xmath184 @xmath226 can lead to fulfilment of the condition @xmath224 .
this yields @xmath227 and , therefore , @xmath228 note that this nnm has been investigated quite enough in @xcite . above
, we simply reproduced the main result of this paper by our method .
_ example 3 .
_ nonlinear normal modes b@xmath71 $ ] : @xmath229 and b@xmath75 $ ] : @xmath230 .
the variational systems for these nnms can be decoupled into @xmath231 independent subsystems whose matrices are presented in table [ table20 ] .
now one has to vanish the discriminant @xmath185 from eq .
( [ eqch302 ] ) . with the aid of our method
, we get the following results . for both modes
b@xmath71 $ ] and b@xmath75 $ ] , we have obtained identical scaling for the case @xmath143 : @xmath232 some numerical results on stability of the nnm b@xmath71 $ ] have been found in @xcite , but we do nt know any results on stability properties of the nnm b@xmath75 $ ] . in conclusion ,
let us consider another scenario of the stability loss of nnms .
indeed , up to this point , we have discussed only the loss of stability associated with parametric interactions of a given nnm with other ( linear ) normal modes of the fpu-@xmath0 chain .
some nnms in the fpu-@xmath0 chain , when @xmath233 , transform not into _ one _ linear normal mode ( lnm ) , but into a certain _ superposition _ of such modes .
for example , nnm b@xmath73 $ ] @xmath234 transforms , in the case @xmath235 , into the linear combination @xmath236 of two linear normal modes @xmath237 where @xmath238 one can also say that nnm b@xmath73 $ ] is the result of the continuation of the superposition ( [ eqch701 ] ) with respect to the nonlinearity parameter @xmath156 of the considered fpu-@xmath0 chain .
we have to emphasize that the continuation of an _ arbitrary _ linear combination of two above discussed lnms does nt represent an exact solution to nonlinear dynamical equations of the fpu-@xmath0 chain , while the superposition with @xmath239 produces an exact solution .
then , the following question arises : if we will increase the parameter @xmath156 from zero , is it possible that the discussed nnm ( [ eqch700 ] ) loses stability because of transformation into a _ two - dimensional bush _ @xmath240 with two _ different _ functions of time @xmath31 , @xmath30 ? in general , such a bush describes not periodic , but a _ quasiperiodic _ dynamical regime in the fpu-@xmath0 chain .
in contrast to the previous case , this stability loss scenario does nt imply any extension of the vibrational modes set , but it means breaking the correlation ( [ eqch701 ] ) between two lnms , @xmath241 and @xmath242 .
we do nt discuss this scenario of the stability loss in the present paper .
however , our analysis allows us to assert that such scenario is inessential for all six symmetry - determined nnms in the fpu-@xmath0 chain .
with the aid of the above discussed method , we investigated scaling of stability thresholds in the thermodynamic limit @xmath143 for all possible in the fpu-@xmath0 chain nonlinear normal modes for both cases @xmath55 and @xmath144 .
the summary results are presented in tables [ table3 ] , [ table4 ] .
let us comment on these results .
@xmath243 & \frac{\pi}{\sqrt{6\beta n } } & \frac{\pi^2}{3\beta n^2 } & \text{\cite{budinsky1983,bermankolovskij1984,sanduskypage1994,flach1996,dauxois1997,poggiruffo1997,bountis2006a , yoshimura2004,dauxois2005}}\\ b[a^3,i ] & \frac{2\pi}{3\sqrt{\beta n } } & \frac{2\pi^2}{3\beta n^2 } & \text{\cite{bountis2006}}\\ b[a^4,ai ] & \sqrt{\frac{2\pi}{3\beta } } & \frac{2\pi}{3\beta n } & \text{\cite{bountis2006,bountis2006a}}\\ b[a^3,iu ] & \frac{2\pi}{3\sqrt{\beta n } } & \frac{2\pi^2}{3\beta n^2 } & \\ b[a^4,iu ] &
\frac{\sqrt{2}\pi}{\sqrt{3\beta n } } & \frac{2\pi^2}{3\beta n^2 } & \text{\cite{leoleo2007}}\\ b[a^6,ai , a^3u ] & \frac{2\pi}{\sqrt{3\beta n } } & \frac{2\pi^2}{3\beta n^2 } & \\ \hline \end{array}\ ] ] @xmath244 & \sqrt{\frac{2\pi}{3|\beta| } } & \frac{2\pi}{3|\beta|n } & \\ b[a^4,iu ] & \frac{\sqrt{2}\pi}{\sqrt{3|\beta|n } } & \frac{2\pi^2}{3|\beta|n^2}&\text{\cite{leoleo2007}}\\ b[a^6,ai , a^3u ] & 0 & 0 & \\ \hline \end{array}\ ] ] in the first column of tables [ table3 ] , [ table4 ] , the symbols of nnms are given .
in two next columns , we present the scaling laws for @xmath143 of stability thresholds for the amplitude of each nnm , @xmath245 , and for its specific energy , @xmath246 ( the energy per one particle of the chain ) . in the last column of the above tables , we give references to the papers in which stability thresholds of the corresponding nnms are also discussed .
firstly , let us pay attention to the following interesting fact : four of six nnms in the fpu-@xmath0 chain , namely , , obeys different duffing equations [ see eqs .
( [ eqch38])-([eqch_65 ] ) ] . ] @xmath247 & = & \nu(t)\cdot\frac{3}{\sqrt{6n}}\{1,0,-1\,|\,1,0,-1\,|\,\ldots\},\\ \text{b}[a^3,iu ] & = & \nu(t)\cdot\frac{1}{\sqrt{2n}}\{1,-2,1\,|\,1,-2,1\,|\,\ldots\},\\ \text{b}[a^4,iu ] & = & \nu(t)\cdot\frac{1}{\sqrt{n}}\{1,-1,-1,1\,|\,1,-1,-1,1\,|\,\ldots\},\\ \text{b}[a^3u , ai ] & = & \nu(t)\cdot\frac{3}{\sqrt{6n}}\{{\scriptstyle 0,1,1,0,-1,-1\,|\,0,1,1,0,-1,-1\,|\,\ldots}\},\\ \end{array}\ ] ] possess _ identical _ scaling of the stability threshold in the limit @xmath143 : @xmath248 the scaling of @xmath246 for the @xmath52-mode @xmath249 = \nu(t)\cdot\frac{1}{\sqrt{n}}\{1,-1\,|\,1,-1\,|\,\ldots\}\ ] ] is _ exactly _ twice less than that determined by eq .
( [ eqch_500 ] ) . only for the mode @xmath250 = \nu(t)\cdot\frac{2}{\sqrt{2n}}\{0,1,0,-1\,|\,0,1,0,-1\,|\,\ldots\}\\\ ] ]
the scaling law of the stability threshold turns out to be cardinally different : @xmath251 a qualitative difference in scaling between this mode and all other nnms for the fpu-@xmath0 chain with @xmath55 can be seen in the left column of fig .
[ figstabdiag ] .
the stability properties of the same nnms in the fpu-@xmath0 chain with @xmath144 prove to be essentially different , as one can see in the right column of fig .
[ figstabdiag ] .
firstly , for three nnms , namely , b@xmath69 $ ] , b@xmath71 $ ] and b@xmath75 $ ] the stability thresholds @xmath246 do nt tend to zero when @xmath143 .
these modes lose their stability for a certain _ finite _ value @xmath194 of the nnm s amplitude . for the above listed nonlinear normal modes , these values are equal respectively to @xmath252 , @xmath253 , and @xmath254 . the fundamental difference between scaling of @xmath246 for the modes b@xmath73 $ ] and b@xmath77 $ ] takes place in the case @xmath144 , as well as for the above discussed case @xmath55 : @xmath255 for b@xmath73 $ ] and @xmath256 for b@xmath77 $ ] ( see table [ table4 ] ) . studying of the stability threshold for the nnm b@xmath79 $ ] proves to be more difficult .
in this case , the loss of stability is determined by the second scenario discussed at the beginning of the present section .
normalizing the variational equations , described dynamics of the vibrational modes with basis vectors @xmath32 , @xmath257 , simultaneously with the duffing equation , corresponding to nnm b@xmath79 $ ] , we infer that for @xmath144 there exists an exponential _ detuning _ between the above modes for an arbitrary small amplitude of the investigated nnm .
this means that @xmath246 turns out to be equal to zero as indicated in table [ table4 ] .
in the present paper , a certain asymptotic technique for studying the stability loss of nonlinear normal modes in the fpu-@xmath0 chain in the thermodynamic limit @xmath143 is developed . using this technique
we were able to obtain the scaling laws of the stability threshold @xmath246 for all six symmetry - determined nnms that are possible in the fpu-@xmath0 chain for both cases @xmath55 and @xmath144 .
the general method @xcite for splitting the variational system for a given dynamical regime in a physical system with discrete symmetry into independent subsystems of small dimensions was applied for investigation of stability of nnms in the fpu-@xmath0 chain .
the above dimensions for the considered case turn out to be equal to 1 , 2 and 3 .
this splitting allows us to construct numerically the stability diagrams ( fig .
[ figstabdiag ] ) that can help to reveal many interesting properties of nonlinear normal modes , such as qualitative behaviour of stability thresholds @xmath246 in the thermodynamic limit @xmath143 , the existence of the `` second stability threshold '' for some nnms , existence of finite limits of @xmath246 for certain modes , etc .
chechin , d.s .
ryabov , and v.p .
sakhnenko , _ bushes of normal modes as exact excitations in nonlinear dynamical systems with discrete symmetry _ ( in : `` nonlinear phenomena research perspectives '' , pp . 225327 , ed . c. w. wang , nova science publishers , ny , 2007 ) . | all possible symmetry - determined nonlinear normal modes ( also called by simple periodic orbits , one - mode solutions etc . ) in both hard and soft fermi - pasta - ulam-@xmath0 chains are discussed . a general method for studying their stability in the thermodynamic limit , as well as its application for each of the above nonlinear normal modes
are presented . |
in this paper we shall show that imposition of the correct lorentz structure in the parton model allows us to relate two issues which , _ a priori _ , do not seem to be connected with each other : the derivation of a consistent expression for the polarized deep inelastic scattering ( dis ) structure function @xmath0 in the parton model , and the derivation of the higher twist target - mass corrections i.e. corrections of the form @xmath2 , where @xmath3 is the nucleon mass , to @xmath4 and @xmath5 . as will be discussed below
, the target - mass corrections have previously been derived in a very complicated way from the operator product expansion ( ope ) , and the derivation of the wandzura - wilczek ( ww ) expression for @xmath0 has involved a questionable analytic continuation in the ope moments @xcite .
it is thus particularly interesting that these results can be derived in a field theoretic context without use of the ope .
target - mass corrections for _ unpolarized _ dis were first derived by nachtmann @xcite employing a very elegant mathematical approach in which the power series expansion used in the ope was replaced by an expansion into a series of hyperspherical functions ( representation functions of the homogeneous lorentz group ) .
later , also within the context of the ope , georgi and politzer @xcite re - derived nachtmann s results using what they called an alternative analysis for simple - minded souls like ourselves " i.e. based on a straightforward power series expansion but , in fact , requiring a very clever handling of the combinatoric aspects of the problem .
the derivation of target - mass corrections for _ polarized _ dis turned out to be much more difficult .
several papers @xcite succeeded in expressing the reduced matrix elements @xmath6 , @xmath7 of the relevant operators in terms of combinations of moments of the structure functions , but did not manage to derive closed expressions for the structure functions @xmath8 themselves .
the latter was finally achieved in 1997 by piccione and ridolfi @xcite and later generalized to weak interaction , charged current reactions , by blmlein and tkabladze @xcite .
these calculations , based on the ope , are extremely complicated , and we shall see presently how the same results can be obtained in a much simpler field - theoretic approach .
the clue to this entire approach is contained in the classic paper of ellis , furmanski and petronzio ( efp ) @xcite , which gave the first derivation of the dynamic higher twist corrections to unpolarized dis in terms of amplitudes involving not just the handbag " diagram of fig .
[ hand ] , whose soft part is the quark - quark @xmath9 correlator @xmath10 , but the higher order diagrams in fig . [ qqg ] , whose soft parts are the @xmath11 and @xmath12 correlators respectively .
the dis handbag " diagram involving the @xmath13-correlator.,scaledwidth=30.0% ] dis diagrams involving the @xmath11 and @xmath12-correlators.,scaledwidth=40.0% ] efp begin with a brief discussion of a parton model , which they refer to as a reference model " , in which the active quark , momentum @xmath14 , emitted from the nucleon is on mass - shell . handling the kinematics
exactly they arrive at expressions for the unpolarized structure functions @xmath15 in terms of the quark densities @xmath16 which are identical to those of nachtmann @xcite . at first sight
it seems surprising that such a naive model should give the exact results of the theory .
but the point is and this is something not stressed in the literature that any off - shellness of the quark is a direct consequence of qcd , i.e. if the strong interaction coupling @xmath17 then it follows that ( taking the quark mass @xmath18 ) @xmath19 [ see eq .
( [ e ] ) ] .
target - mass corrections are , by definition , kinematic in origin , therefore are independent of the value of @xmath20 .
hence it is not miraculous that a model with @xmath19 , i.e. equivalent to putting @xmath17 , should yield exact results for the target - mass corrections .
however , it is crucial that the lorentz structure is built into the model , as is done by efp @xcite .
the implications of this are that the exact target - mass corrections must be derivable from the handbag " diagram , fig .
[ hand ] , alone , since the diagrams of fig .
[ qqg ] vanish when @xmath17 .
we shall carry out the derivation for polarized dis and also show that when the lorentz structure is respected the @xmath19 model " yields an unambiguous result for @xmath0 , namely the wandzura - wilczek ( ww ) result @xcite @xmath21 where @xmath22 is the well - known bjorken variable for dis .
this suggests that the analytic continuation necessary in the ope derivation of the ww result is in fact correct , and also implies that in a correctly formulated parton model , with on - shell quarks , @xmath0 is exactly given by the ww expression .
in this paper we shall also show that careful attention to the lorentz structure in the parton model imposes strong constraints on the possible @xmath1 dependence of the so - called transverse momentum dependent ( tmd ) parton densities , that in recent years have received much emphasis ( for an up - to - date review see e.g. ref .
@xcite and references therein ) . indeed , taking into account
the transverse momentum of partons is important for an understanding of the large transverse single spin asymmetries observed in many reactions .
it is also essential in order to generate the parton orbital angular momentum which appears necessary as a consequence of the small contribution to the nucleon angular momentum provided by the parton spins .
as we are going to show , imposing lorentz covariance we are able to _ derive _ , for the unpolarized densities @xmath23 , the longitudinal densities @xmath24 and the transversity densities @xmath25 or @xmath26 , their dependence on @xmath1 from the functional form of the usual collinear , @xmath1-integrated , parton densities @xmath27 , @xmath28 and
@xmath29 [ these relations are spelled out in detail in eqs .
( [ vaa ] ) , ( [ wa2 ] ) and ( [ wa ] ) , and in eqs .
( [ gb ] ) , ( [ q2 ] ) ] .
moreover , if perturbative evolution with @xmath30 is taken into account , then the evolution of the @xmath1-dependent densities is entirely controlled via the known evolution of the collinear densities .
it should be noted that the functional form found in the parton model is quite different from the often used factorized form , typically @xmath31 , though the latter may be a reasonable starting point for analyzing the presently available data .
further , we obtain positivity and soffer - type bounds @xcite , based on the @xmath1-dependent densities , which are stronger , in general , than those usually imposed on the purely @xmath32-dependent , @xmath1-integrated , collinear densities , and we suggest that it would be interesting to impose these bounds on the collinear densities when extracting parton densities from deep inelastic and semi - inclusive deep inelastic scattering data .
therefore , as will become even more clear in the following , careful attention to the lorentz structure in the parton model with on - shell partons , i.e. with @xmath19 , has dramatic consequences .
given the standard belief that the _ soft _ functions appearing in qcd have support only in a very narrow range of values of @xmath33 around @xmath19 , it is tempting to suppose that the results derived in this paper may not be too different from those holding in a full qcd treatment .
covariant parton models have been discussed in the literature for many decades , beginning with the work of landshoff , polkinghorne and short in 1971 @xcite , and franklin in 1977 @xcite .
these papers dealt only with the unpolarized structure functions @xmath34 .
the collinear spin - dependent structure functions @xmath35 were studied , in response to the spin crisis in the parton model " , by jackson , roberts and ross @xcite , who also commented upon some aspects of quark transverse momentum .
more recently , a very detailed study of covariant models was initiated by zavada in 1997 @xcite , and later developed by zavada @xcite and efremov , teryaev and zavada @xcite , to include the polarized parton densities as well as parton transverse momentum .
while our paper was in its last phase of preparation we became aware of a further paper by efremov , schweitzer , teryaev and zavada @xcite ( which we shall refer to as estz ) and a very latest work by zavada @xcite in which tmd parton distributions are discussed in detail . in these papers
the analysis is based upon reasonable assumptions about the structure of the quark wave function in the nucleon _ rest _ frame . because the treatment is covariant the results
can then be translated to any other lorentz frame .
it is remarkable that many of the results of this kind of analysis are in complete agreement with our more general , frame independent , treatment .
the principal differences are that : _ i _ ) in our treatment the three leading twist densities , the unpolarized density @xmath23 , the longitudinal density @xmath36 and the transversity density @xmath37 or @xmath38 are _ independent _ , whereas in the zavada et al .
papers only two of these are independent ; _ ii _ ) we take into account target mass corrections .
because our approaches are so different , we believe it will be instructive to present our derivation in full .
the plan of the paper is the following : in section [ corr ] we discuss the hadronic correlator for polarized dis in the parton model , with @xmath17 , and give the general expression of the structure functions @xmath39 and @xmath0 ; in section [ coeff ] we evaluate explicitly @xmath39 and @xmath0 including target - mass corrections and derive the ww relation ; in section [ tmd ] we extend our analysis to the tmd quark distributions and discuss the consequences of imposing lorentz invariance ; in section [ bounds ] we examine the positivity and soffer bounds for the @xmath1-dependent unpolarized , longitudinally and transversely polarized quark distributions , discuss how their correct implementation leads to new , more stringent , bounds on the @xmath1-integrated collinear distributions and show some phenomenological implications of this approach ; finally , in section [ concl ] we give some additional comments and conclusions .
the correlator @xmath41 , where @xmath42 and @xmath43 are the momentum and spin - polarization 4-vectors for the nucleon ( @xmath44 ) , @xmath14 is the quark 4-momentum and @xmath45 are dirac indices , is defined by @xmath46 to preserve colour gauge invariance a path - ordered wilson line ( or gauge - link ) should be inserted between the quark fields .
since we are mostly interested in the @xmath17 case , we will take this operator as the identity .
we ignore for the moment flavour and the quark charge
they are trivially reinstated at the end and , as already stated , work with @xmath18 . by partial integration
we see that , when @xmath17 , @xmath47 \ > = \
> i \,\int\,\frac{{\rm d}^4z}{(2\pi)^4}\ , e^{ik\cdot z}\ , \langle p , s|\bar{\psi}(0 ) { \not\!\partial } \ , \psi(z)|p , s\rangle \>=\ >
0\ > , \label{c}\ ] ] and that for any of the 16 dirac matrices @xmath48 ( @xmath49 , @xmath50 , @xmath51 , @xmath52 , @xmath53 ) , @xmath54 = 0\>.\ ] ] finally , again by partial integration , we get for all @xmath33 @xmath55 implying that @xmath56 @xmath10 is a @xmath57 matrix in dirac spin space .
the first attempt to write down its most general form was made by ralston and soper @xcite who , however , missed one term @xcite . the full dynamical twist - two structure , including the
three so - called
naively @xmath58-odd " amplitudes , contains twelve terms @xcite .
eight of these , including all the naively @xmath58-odd " amplitudes , are eliminated by the requirements of eqs .
( [ c])-([e ] ) and two become related to each other . we are left with @xmath59\>. \label{h}\ ] ] the scalar functions @xmath60 are , in general , functions of @xmath61 and @xmath33 . in our case , bearing in mind eq .
( [ f ] ) , we shall later use @xmath62\ ] ] and @xmath63\>,\ ] ] where the factors are for later convenience and , following ref .
@xcite , the @xmath64-function ensures that the nucleon remnant has positive energy .
notice again the factor @xmath65 in these expressions . for the moment we shall only deal with the part relevant to @xmath66 i.e. @xmath67 .
polarized dis is controlled by the antisymmetric part @xmath68 of the hadronic tensor , related to the structure functions @xmath66 via @xcite @xmath69\ , g_2 \bigg\ } \>,\ ] ] where @xmath70 is the photon 4-momentum and our convention is @xmath71 .
the contribution from the handbag diagram , fig .
[ hand ] , to @xmath68 is then @xmath72 \ , { \rm tr}(\gamma_\sigma\,\gamma_5\,\phi ) \nonumber \\ & = & -\,\epsilon^{\mu \nu \rho \sigma } \int { \rm d}^4k\,(k_\rho + q_\rho)\,\delta [ ( k + q)^2 ] \,{\rm tr}(\gamma_\sigma\,{\not\ ! k})\ , \frac{a_8}{m}\,(k\cdot s ) \nonumber \\ & = & - \,\frac{4}{m}\,\epsilon^{\mu \nu \rho \sigma } q_\rho\,s^\beta i_{\beta \sigma}\>,\end{aligned}\ ] ] where @xmath73 \,a_8\>.\ ] ] note that the hadronic tensor @xmath68 in eq .
( [ i ] ) is electromagnetic gauge invariant .
we introduce the standard two auxiliary null vectors @xmath74 and @xmath75 , @xmath76 , @xmath77 , in a slightly unusual way via @xmath78 where @xmath79 is the nachtmann variable @xmath80 and @xmath81 note that @xmath82 switching off the target - mass corrections thus corresponds to putting @xmath83 . since by definition
, see eq .
( [ j ] ) , @xmath84 is symmetric under @xmath85 , its most general expression can be written as @xmath86 on the other hand , from eq .
( [ j ] ) , bearing in mind the factor @xmath65 in @xmath67 , eq .
( [ p ] ) , and introducing the shorthand @xmath87 \,a_8\>,\ ] ] one finds that @xmath88 by comparing with eq .
( [ igen ] ) and expressing the factors @xmath89 in terms of the @xmath90 s one can then write @xmath91 + c_{n}\ , p_\beta p_\sigma + m^4 c_{p}\ , n_\beta n_\sigma\>.\ ] ] moreover , from eq .
( [ k ] ) , because of the factor @xmath92 in eq .
( [ i ] ) and the fact that @xmath93 , we can replace @xmath94 in eq .
( [ q ] ) .
thus @xmath95\ , q_\beta p_\sigma\>.\ ] ] putting this into eq .
( [ i ] ) and comparing with eq .
( [ h2 ] ) , we can read off directly expressions for @xmath96 and @xmath97 and thereby obtain @xmath98 and @xmath99\,\big\}\>,\ ] ] where we have used eq .
( [ n1 ] ) .
in order to evaluate the coefficients @xmath100 , @xmath101 , @xmath102 in eqs .
( [ s ] ) , ( [ t ] ) we parameterize @xmath14 in the usual way , bearing in mind that @xmath19 , @xmath103 where @xmath104 .
then @xmath105 and in terms of these variables @xmath106 ^ 2\ , \rangle\>.\ ] ] bearing in mind eqs .
( [ n1 ] ) and ( [ w ] ) one has @xmath107 = \frac{x}{\xi}\ , \delta \big [ \bm{k}_t^2 - q^2 \,\frac{x}{\xi}(\frac{x}{\xi } - 1)\big]\>.\ ] ] we therefore integrate over @xmath108 , introduce @xmath109\>,\ ] ] via eq .
( [ z ] ) , and have @xmath110 where the upper integration limit comes from the @xmath111-function in eqs .
( [ o ] ) , ( [ p ] ) , and the lower limit from @xmath112 . in terms of @xmath113
the variables appearing in eq .
( [ y ] ) are given by @xmath114 we substitute the expressions for the coefficients , eq .
( [ y ] ) , into eqs .
( [ s ] ) , ( [ t ] ) using eq .
( [ p ] ) , restore the quark charge , separate the leading twist terms , i.e. those that do not vanish when @xmath115 , and , inside the integral , order the rest according to the powers of @xmath116 and @xmath113 involved , to obtain , for each flavour , @xmath117 @xmath118 from here we also get a separate parton - model expression for @xmath97 : @xmath119 the expression for @xmath120 , eq .
( [ da ] ) , must agree with the standard collinear parton model result , with no target - mass corrections ( ntm ) , in terms of the ( longitudinal ) polarized quark density , @xmath121 when target masses are neglected , i.e. when @xmath122 and @xmath123 .
thus @xmath124 this relation between @xmath125 and @xmath126 allows us to express the integrals in eq .
( [ da ] ) in terms of integrals over @xmath127 , but also has powerful consequences for the structure of transverse momentum dependent parton densities , as will be discussed shortly . turning to @xmath128 , we have from eq .
( [ ea ] ) : @xmath129 consider now [ see eq .
( [ ga ] ) ] @xmath130 changing the order of integration , we obtain ( @xmath131 ) @xmath132 putting this relation in eq .
( [ ha ] ) and using eq . ( [ fa ] )
yields @xmath133 which is the wandzura - wilczek relation in the absence of target - mass corrections .
returning to eqs .
( [ da ] ) , ( [ ea ] ) we note first that from eqs .
( [ ja ] ) , ( [ ga ] ) @xmath134 then , using again eq .
( [ ga ] ) in the second step , @xmath135 \nonumber \\ & = & -\frac{1}{\xi}\,\big\{\,\delta q(\xi ) + 2 \int _
\xi^1 \ , \frac{{\rm d}\eta}{\eta } \,\delta q(\eta)\,\big\}\>.\end{aligned}\ ] ] further , integrating by parts , @xmath136 substituting eqs .
( [ la])-([na ] ) into eqs .
( [ da ] ) , ( [ ea ] ) we obtain expressions which , for reasons to be explained in the next subsection , we shall call the _ bare parton model _ ( bpm ) results @xmath137 @xmath138 @xmath139 which , as can be shown after some algebra , agree exactly with the dynamical twist - two ( i.e. taking @xmath140 ) ope results of ref .
furthermore we find that @xmath141 = -\frac{d } { n}\ >
g_1^{\rm bpm}(x_{\rm bj})\>,\ ] ] so that , bearing in mind eqs .
( [ n ] ) and ( [ n1 ] ) , @xmath142 and the wandzura - wilczek relation holds also when target - mass corrections are included @xcite .
there is a long - standing problem about the behaviour of the structure functions , both unpolarized and polarized , at @xmath143 . to see this
most simply consider the first term in the expansion of eq .
( [ oa ] ) in powers of @xmath2 , where we show the kinematic dependence in explicit detail @xmath144\>.\ ] ] it is usually stated that one must have @xmath145 , implying the bizarre result that @xmath146 vanishes at @xmath147 i.e. at a continuous infinity of @xmath30 dependent points .
of course without target mass corrections the vanishing of @xmath148 simply implies that @xmath149 .
there are no satisfactory prescriptions for avoiding this issue in the literature .
georgi and politzer @xcite and piccione and ridolfi @xcite argue that higher twist terms must be taken into account in the region of large @xmath22 , whereas accardi and melnitchouk @xcite impose some constraints on the virtuality of the struck quark .
we would like to propose a very simple resolution to this problem . in the standard parton model treatment
there is the underlying assumption that the struck quark and target fragments materialize into physical hadrons with probability one .
this is tantamount to the assumption of _ equivalent completeness _ of partonic and hadronic states i.e. @xmath150 while this might be reasonable for massless hadrons , it can not be true when hadrons possess their physical masses .
thus what is needed is a _ non - perturbative _ function to express the failure of equivalent completeness .
it has to express the fact that there is zero probability for the final partonic state ( the struck parton and the target fragments ) to hadronize if its total energy is too small to produce an inelastic hadronic event , and that the probability for hadronization is one when the energy is large enough .
the simplest possibility is @xmath151 , where @xmath152 is the maximum kinematically allowed value of @xmath22 for _ inelastic _ scattering .
we note that , strictly speaking , and excluding photon production which is higher order in the fine structure constant , this is _ not _
@xmath153 , but @xmath154 where @xmath155 is the pion mass .
thus we propose that in a _ physical parton model _
( ppm ) the bpm target mass dependent expressions , eqs .
( [ oa ] ) , ( [ pa ] ) , for @xmath156 and @xmath157 should be modified to @xmath158 @xmath159\ , \theta ( x_{\rm th } - x_{\rm bj})\>.\ ] ]
in the standard collinear parton model the transverse momentum @xmath160 is integrated over up to the large scale @xmath161 .
but in the last few years many polarized reactions have been studied in a more general framework in which the fundamental parton densities @xmath162 and @xmath163 are allowed to depend on the _ intrinsic _ @xmath1 ( see e.g. refs .
@xcite and references therein ) . in all these approaches
it is often assumed that the @xmath32 and @xmath164 dependence can be factorized .
this , as we shall see , is in contradiction with the lorentz structure of the amplitudes , _ at least in the @xmath17 case_. we shall comment later on the more general situation . for unpolarized dis the analogue of eq .
( [ ga ] ) is @xcite @xmath165 moreover , on changing integration variables from @xmath166 to @xmath167 via eq .
( [ aa ] ) , we see that @xmath168 for which @xmath169 must be given by @xmath170\>.\ ] ] thus the @xmath32 and @xmath171 dependence are intimately linked . moreover , the maximum size of @xmath172 is bounded and @xmath32-dependent , and in general @xmath173 gev@xmath174 .
most importantly , if the functional form of @xmath175 is known ( for consistency it must , as usual , satisfy @xmath176 ) then from eq .
( [ sa ] ) @xmath177\>.\ ] ] therefore we have the remarkable result that , in the @xmath17 case , the @xmath171 dependence of @xmath178 is completely determined by the @xmath22 dependence of the standard collinear quark density @xmath175 : @xmath179_{x=\eta}\ , \theta [ x(1-x)\,m^2 -\bm{k}_t ^2 ] \>,\ ] ] where , we remind the reader , @xmath180 notice that eqs .
( [ va])-([eta ] ) are in agreement with eqs . ( 47 ) , ( 48 ) of ref .
@xcite . as a consequence ,
the average transverse momentum squared , @xmath181 , can be easily obtained : @xmath182 for the case of polarized dis the result is slightly more complicated . given the functional form of @xmath126 we can obtain @xmath183 via eq .
( [ la ] ) , in agreement with eq .
( 42 ) of ref .
@xcite : @xmath184 the transverse momentum dependent polarized parton density is then given , via eq .
( [ ga ] ) and the relation between @xmath166 and @xmath171 in eq . ( [ eta ] ) , by @xmath185 \,\varphi _ 8\big ( x + \frac{\bm{k}_t ^2}{x m^2}\big)\ , \theta [ x(1-x)\,m^2 -\bm{k}_t ^2 ] \>,\ ] ] with @xmath186 just as for the unpolarized density , we see , via eq .
( [ wa ] ) , that the @xmath187 dependence of @xmath188 is completely determined by the @xmath22 dependence of the collinear polarized quark density @xmath126 . of great interest at present is the transversely polarized quark density @xcite ( also referred to as the transversity , see e.g. ref .
@xcite for a review ) which is concerned with quarks transversely polarized along the @xmath189-direction inside a nucleon transversely polarized along @xmath190 : @xmath191\ , \cos\phi .
\end{aligned}\ ] ] here we specify the parton distributions using the physically motivated probabilistic notation of ref .
@xcite , referring to the polarization of a parton , whose spin direction ( in its helicity frame whose axes are labelled @xmath192 , @xmath193 , @xmath194 @xcite ) is indicated by @xmath195 for transverse polarization , inside a nucleon moving along @xmath196 in some fixed frame with axes labelled @xmath197 , @xmath190 , @xmath196 ( it could be the laboratory frame or the c.m .
frame of the @xmath198-nucleon collision ) , whose spin direction is indicated by @xmath199 for transverse polarization , namely @xmath200 the connection with the more formal notation of ref .
@xcite will be given in section [ tmd - others ] ( see also appendix c of ref .
@xcite ) . in eq .
( [ db ] ) @xmath201 is the azimuthal angle of @xmath164 in the fixed frame .
note that @xmath202 this is because @xmath203 corresponds to having the quark polarized along @xmath193 perpendicular to _ its _ momentum and not along @xmath190 , perpendicular to the nucleon s momentum .
the @xmath1-dependent density which does integrate to the usual collinear transversely polarized or transversity density is one in which the quark is polarized along the direction @xmath190 in the fixed frame : @xmath204 so that @xmath205 one finds @xmath206\>.\ ] ] it follows that @xmath207 moreover , from eq .
( [ gb1 ] ) one can show that @xmath208\>.\ ] ] thus , via eqs .
( [ gb ] ) , ( [ q1 ] ) , the @xmath1 dependence of both @xmath209 and @xmath203 is completely determined by the functional form of the standard collinear transversity density @xmath29 . it is an intriguing question whether , in the real case of @xmath210 , it is possible to have a factorized form @xmath211 for the parton densities . aside from the more complex tensorial structure when @xmath210 , the functional dependence will be controlled by expressions of the form @xmath212 where the range of @xmath33 is limited to a small region around @xmath19 .
while it seems unlikely that this could lead to a factorized form , we have been unable to demonstrate this mathematically . in this subsection
we shall give expressions for several other tmd quark densities , which , while playing no role in the @xmath1-integrated hadronic correlator for dis , are important in other processes , like semi - inclusive dis or drell - yan dilepton production , and have been widely used in the literature @xcite .
related to @xmath67 we consider quarks polarized longitudinally ( i.e. with helicity @xmath213 ) inside a transversely polarized nucleon ( here it is irrelevant which transverse direction is involved ) .
then ( see also ref .
@xcite ) @xmath214 with @xmath215\>.\ ] ] for the quark polarized in the @xmath32-direction , we have @xmath216 \,\varphi _ { 11}\big ( x + \frac{\bm{k}_t ^2}{x m^2}\big)\ , \theta [ x(1-x)\,m^2 -\bm{k}_t ^2 ] \,\sin\phi \>.\end{aligned}\ ] ]
one also finds that @xmath217\>.\ ] ] for the function @xmath218 defined via @xmath219 we find @xmath220 \,\varphi _ { 11}\big ( x + \frac{\bm{k}_t ^2}{x m^2}\big)\ , \theta [ x(1-x)\,m^2 -\bm{k}_t ^2 ] \>.\ ] ] finally , for a quark polarized along the @xmath32-direction inside a longitudinally polarized nucleon , @xmath221 with @xmath222\>.\ ] ] note that for consistency in the parton model , with @xmath17 , we have , in agreement with the results of ezts , @xmath223
in attempting to extract information on the collinear polarized density @xmath224 and the transversity density @xmath225 by fitting data using parameterized forms for the densities , it is usual to impose the positivity bound @xmath226 .
however , it turns out that while these bounds are _ necessary _ they are not _ sufficient_. + the reason why eq .
( [ oba ] ) is not sufficient is that from a probabilistic point of view one should expect @xmath227 ) implies eq .
( [ oba ] ) , the reverse is not true . actually , as we shall now show , it turns out that it is sufficient to require just @xmath228\>,\ ] ] and from eq .
( [ ua ] ) @xmath229\>.\ ] ] since eq .
( [ obb ] ) must hold for all @xmath32 we find that ( of course @xmath230 ) @xmath231 and this is sufficient to guarantee that eq .
( [ ob ] ) holds .
thus , we suggest that in parameterizing the collinear polarized density @xmath224 one should perhaps impose the stronger bound @xmath232 which follows from eqs .
( [ pb2 ] ) , ( [ wa ] ) and ( [ va ] ) and is in agreement with eq . ( 48 ) of ref .
@xcite ; of course , one must have @xmath233 .
notice that there is a further bound , originally derived in ref .
@xcite , eq .
( 60 ) , @xmath234 which can be easily obtained from our eqs .
( [ ka ] ) , ( [ fa ] ) , ( [ sa ] ) and ( [ pb2 ] ) .
soffer derived his bound on the collinear transversity density @xcite @xmath235-channel absorptive part of _ forward _ elastic quark - nucleon scattering .
but collinear parton densities include all partons with transverse momentum @xmath236 and thus do not consist of strictly forward moving partons .
hence , in the light of the existence of intrinsic @xmath237 , soffer s analogy could be said to have been misinterpreted , and the correct bound should read @xmath238\>.\ ] ] now , from eqs .
( [ q1 ] ) , ( [ gb ] ) we have @xmath239\>,\ ] ] and substituting this and eqs .
( [ pb ] ) , ( [ qb ] ) into eq .
( [ nb ] ) we obtain @xmath240\>.\ ] ] we suggest that in parameterizing models for the collinear transversity density @xmath163 , the scalar functions @xmath241 should be calculated in terms of @xmath16 , @xmath28 and @xmath163 , via eqs .
( [ va ] ) , ( [ wa ] ) , ( [ q2 ] ) , and , perhaps , should be forced to satisfy the bound in eq .
( [ rb ] ) .
finally , we have also explicitly verified that the bounds in eqs .
( [ pb2 ] ) , ( [ rb ] ) are equivalent to the bounds on the @xmath1-dependent functions derived in ref .
@xcite , in the @xmath17 limit . as a simple and illustrative example of our results , we consider here some of the consequences of imposing lorentz invariance on the @xmath1-dependence of tmd parton distributions .
we will start from a well - known set of unpolarized and longitudinally polarized distributions extracted within the standard collinear approach , namely the grv98 @xcite and grsv2000 @xcite sets respectively , and show the crucial role played by the new bounds .
since we are working in the @xmath17 approximation , we will consistently consider only @xmath242 and @xmath243 quark distributions , neglecting sea quarks .
moreover , since we are here mainly interested in the intrinsic @xmath1 dependence , we choose as reference scale a relatively low one , @xmath244 gev@xmath174 , in order to exclude large effects due to perturbative evolution . from these parameterizations for @xmath245 and @xmath246
we therefore generate , via eq .
( [ vaa ] ) and eqs .
( [ wa ] ) , ( [ wa2 ] ) the @xmath1-dependent distributions @xmath247 and @xmath248
. concerning transversity , we have to recall that at present no parameterization in the standard collinear approach is available .
double transverse spin asymmetries , @xmath249 , in the drell - yan processes @xmath250 , or @xmath251 , are considered the best tool to this end , but are unfortunately still out of reach for present experimental setups . in the meantime , the study of transverse single spin asymmetries in polarized sidis processes in the tmd approach has been found to be the most promising way to get information on the unintegrated , @xmath1-dependent transversity distribution @xcite .
we will therefore utilize in our discussion the first ever parameterization available for the transversity of @xmath242 and @xmath243 quarks obtained by fitting , in the tmd approach , the so - called collins azimuthal asymmetries measured in sidis pion and kaon production by the hermes and compass experiments .
we have to stress , however , that this parameterization has been obtained by assuming a factorized @xmath32 and @xmath1-dependence for the tmd distributions , which is at variance with our findings in the simplified , @xmath17 , limit .
given the still poor knowledge of the transversity distribution , we believe however that , despite this inconsistency , it is of interest to investigate the consequences of our approach .
starting from the most updated parameterization of the transversity distribution for valence quarks , ref .
@xcite , and using eqs .
( [ q2 ] ) , ( [ gb ] ) , we therefore generate the @xmath17 , @xmath1-dependent distribution @xmath252 .
our results are summarized in figs .
[ u - k0-x]-[kt2-x ] . in fig .
[ u - k0-x ] , left panel , we show the unpolarized and longitudinally polarized @xmath242 quark distributions , and their ratio , as a function of @xmath32 at fixed @xmath253 .
although the collinear parton distributions adopted here fulfill the usual positivity bound , eq .
( [ oba ] ) , we see that the stronger positivity bound of eq .
( [ obb ] ) is indeed violated for some @xmath32 values .
this is most clearly seen by looking at the ratio @xmath254 ( dotted line ) , where the violation reachs about 15% . in fig .
[ u - k0-x ] , right panel , we also show , again for @xmath242 quarks , the transversity distribution @xmath255 and the positive - helicity distribution @xmath256 and their ratio ( related to the soffer bound ) , as a function of @xmath32 at fixed @xmath253 . in this case
also the stronger , soffer - type , bound , eq .
( [ nb ] ) , is fulfilled .
similar results are shown in fig .
[ d - k0-x ] for @xmath243 quarks . this time , while the stronger positivity bound derived in this section is always fulfilled , the soffer - type bound , eq .
( [ nb ] ) , is instead violated over a large range of @xmath32 values . in fig .
[ u - x02-k ] we show , again for @xmath257 in the left panel and for @xmath258 in the right panel , the @xmath172 dependence of the @xmath242-quark distributions at a fixed value of @xmath32 in the valence region , @xmath259 . for the longitudinal distribution
, we see that the violation of the stronger positivity bound at @xmath253 persists up to some relatively large values of @xmath172 .
a similar behaviour should result for any value of @xmath32 in the region of violation of the positivity bound at @xmath253 , shown in the left panel of fig .
[ u - k0-x ] .
notice also the node in @xmath260 , which is a general feature of @xmath261 coming from eq .
( [ wa2 ] ) .
finally , notice from the right panel that , as expected , the validity of the soffer - type bound at @xmath253 guarantees its validity at any allowed value of @xmath172 .
analogous results for @xmath243 quarks are presented in fig .
[ d - x02-k ] , and similar comments hold on the violation , this time , of the soffer - type bound . for completeness , in fig .
[ kt2-x ] we show the resulting mean transverse momentum squared , @xmath181 , eq . ( [ ave - k2 ] ) , as a function of @xmath32 , for @xmath242 and @xmath243 quarks , at the same reference scale , @xmath244 gev@xmath174 .
we briefly note that the shape of @xmath181 is quite reasonable and approximately flavour - independent , although its size is considerably smaller than expected from available phenomenological analyses within the tmd approach @xcite .
this is perhaps a hint that the larger values of @xmath262 needed in the phenomenological analyses may be due to hard gluon emission and may therefore not be a measure of the size of the genuinely intrinsic transverse momentum .
it would be interesting to construct models based upon such a picture .
we have shown that the simple parton model has a very rich structure when due care is given to its lorentz properties .
on the basis that _ kinematic _ relations should not depend on the value of the strong coupling @xmath20 , and thus should be valid when @xmath17 , we have been able to give a simple derivation of the exact target mass corrections for polarized dis , without recourse to the operator product expansion , and have verified that the wandzura - wilczek relation holds when target mass corrections are included .
we have also demonstrated that that there is a surprising and intimate connection between the intrinsic @xmath1 dependence of parton densities and the functional form of the @xmath1-integrated collinear densities .
indeed the @xmath1 dependence can actually be _ derived _ from the collinear densities .
using some published versions of the collinear densities we have studied the implications of this relationship and found that our @xmath1-dependent distributions , thus derived , are much narrower than those used in phenomenological work i.e. our @xmath262 is much smaller .
this suggests that some of the @xmath1 dependence usually described as intrinsic , may , in fact , be due to gluon radiation .
one outcome of our work is the clear demonstration that the factorized form @xmath263 often used in phenomenological studies , is untenable , _ at least when @xmath17_. while not proved , it seems likely that this result is quite general , though , as commented on , a factorized form does seem a reasonable starting point for the analysis of the presently available data .
in addition , because of this interrelationship between @xmath1 and @xmath32 dependence , we have been able to derive new positivity and soffer - type bounds on the collinear densities , which are stronger than the those usually utilized in dis and sidis analyses .
we have shown numerically that some of the published collinear parton densities , which satisfy the traditional bounds , do not always satisfy these stronger bounds .
the particular case of the down - quark transversity may be of some interest .
is grateful to dr .
ekaterina christova for many discussions in the early stages of this work , and to the department of physics of the university of cagliari for the hospitality extended to him during several visits .
10 s. wandzura and f. wilczek , phys
b72 * , 195 ( 1977 ) .
o. nachtmann , nucl . phys . *
b63 * , 237 ( 1973 ) .
h. georgi and h. d. politzer , phys . rev . * d14 * , 1829 ( 1976 ) . s. matsuda and t. uematsu , nucl .
b168 * , 181 ( 1980 ) .
s. wandzura , nucl .
b122 * , 412 ( 1977 ) .
a. piccione and g. ridolfi , nucl
b513 * , 301 ( 1998 ) , hep - ph/9707478 .
j. blmlein and a. tkabladze , nucl . phys .
* b553 * , 427 ( 1999 ) , hep - ph/9812478 .
r. k. ellis , w. furmanski , and r. petronzio , nucl .
b212 * , 29 ( 1983 ) .
u. dalesio and f. murgia , prog .
61 * , 394 ( 2008 ) , 0712.4328[hep - ph ] .
j. soffer , phys .
* 74 * , 1292 ( 1995 ) , hep - ph/9409254 .
v. landshoff , j. c. polkinghorne , and r. d. short , nucl . phys . *
b28 * , 225 ( 1971 ) .
j. franklin , phys . rev . *
d16 * , 21 ( 1977 ) .
j. d. jackson , g. g. ross , and r. g. roberts , phys .
b226 * , 159 ( 1989 ) .
p. zavada , phys . rev . *
d55 * , 4290 ( 1997 ) , hep - ph/9609372 .
p. zavada , phys . rev . * d67 * , 014019 ( 2003 ) , hep - ph/0210141 .
p. zavada , eur .
j. * c52 * , 121 ( 2007 ) , 0706.2988[hep - ph ] . a. v. efremov , o. v. teryaev , and p. zavada , phys .
rev . * d70 * , 054018 ( 2004 ) , hep - ph/0405225 .
a. v. efremov , p. schweitzer , o. v. teryaev , and p. zavada , phys .
* d80 * , 014021 ( 2009 ) , 0903.3490[hep - ph ] .
p. zavada , 0908.2316[hep - ph ] .
ralston and d. e. soper , nucl .
b152 * , 109 ( 1979 ) .
e. leader , private communication to ralston and soper ( 1979 ) .
p. j. mulders and r. d. tangerman , nucl .
b461 * , 197 ( 1996 ) , hep - ph/9510301 .
m. anselmino , a. efremov , and e. leader , phys .
261 * , 1 ( 1995 ) , hep - ph/9501369 .
a. accardi and w. melnitchouk , phys
* b670 * , 114 ( 2008 ) , 0808.2397[hep - ph ] .
m. anselmino et al . , phys . rev . * d73 * , 014020 ( 2006 ) , hep - ph/0509035 .
v. barone , a. drago , and p. g. ratcliffe , phys . rep . * 359 * , 1 ( 2002 ) , hep - ph/0104283 .
e. leader , spin in particle physics , cambridge university press ( 2001 ) .
a. bacchetta , m. boglione , a. henneman , and p. j. mulders , phys .
* 85 * , 712 ( 2000 ) , hep - ph/9912490 .
m. glck , e. reya , and a. vogt , eur .
j. * c5 * , 461 ( 1998 ) , hep - ph/9806404 .
m. glck , e. reya , m. stratmann , and w. vogelsang , phys . rev . *
d63 * , 094005 ( 2001 ) , hep - ph/0011215 . m. anselmino et al . ,
054032 ( 2007 ) , hep - ph/0701006 . m. anselmino et al .
* 191 * , 98 ( 2009 ) , 0812.4366[hep - ph ] .
u. dalesio and f. murgia , phys . rev .
* d70 * , 074009 ( 2004 ) , hep - ph/0408092 .
m. anselmino et al .
, phys . rev . * d71 * , 074006 ( 2005 ) , hep - ph/0501196 . | we show that respecting the underlying lorentz structure in the parton model has very strong consequences .
failure to insist on the correct lorentz covariance is responsible for the existence of contradictory results in the literature for the polarized structure function @xmath0 , whereas with the correct imposition we are able to derive the wandzura - wilczek relation for @xmath0 and the target - mass corrections for polarized deep inelastic scattering _ without _ recourse to the operator product expansion .
we comment briefly on the problem of threshold behaviour in the presence of target - mass corrections .
careful attention to the lorentz structure has also profound implications for the structure of the transverse momentum dependent parton densities often used in parton model treatments of hadron production , allowing the @xmath1 dependence to be _
derived explicitly_. it also leads to stronger positivity and soffer - type bounds than usually utilized for the collinear densities . |
the low energy structure of qcd may be investigated in the context of chiral perturbation theory @xcite . in refs .
@xcite gasser and leutwyler constructed the generating functional of @xmath2 for @xmath3 to order @xmath4 .
an explicit representation of this functional , including graphs with up to two propagators and at most four external pion fields , was also given .
for the first two flavours @xmath5 and @xmath6 , this construction was extended recently to include graphs with three propagators @xcite . in the case of qcd including electromagnetic interactions , the initial theory depends on the strong coupling constant @xmath7 , the fine structure constant @xmath8 and the light quark masses .
the corresponding effective theory was formulated in refs . .
it is based on a systematic expansion , which combines the chiral power counting scheme with the expansion in powers of the electromagnetic coupling @xmath9 . within this framework ,
virtual photon effects were calculated for a number of processes . in the two flavour case , electromagnetic corrections to the @xmath10 scattering amplitude as well as to the vector and scalar form factors @xcite have been evaluated at next - to - leading order .
further , pionic beta decay and radiative @xmath11 decay have been analyzed @xcite in a generalized framework including leptons and virtual photons @xcite .
virtual photons have also been included in the three
flavour case @xcite .
the purpose of the present article is to include virtual photons in the explicit generating functional at one
loop @xcite .
the advantage of having the explicit functional at hand is evident : the calculation of s - matrix elements boils down to performing traces over flavour matrices - the combinatorics has already been carried out and all quantities are expressed in terms of ultraviolet finite integrals . in the following ,
we stick to the two flavour case .
the extension to three flavours will be considered elsewhere .
the article is organized as follows : in the first part ( sects .
[ slagr ] , [ szoneloop ] ) , we construct the generating functional for @xmath12 to @xmath13 in the low energy expansion .
this allows us to calculate green functions to next - to - leading order in a simultaneous expansion in powers of the external momenta , of the quark masses and of the electromagnetic coupling . in order to extract form factors or scattering amplitudes from our explicit representation of the generating functional
, one simply has to perform traces over flavour matrices .
the extraction procedure is demonstrated in sects .
[ sextract ] and [ spipi ] by means of the @xmath0 scattering amplitude .
the variables of the effective theory are the pion field @xmath14 and the photon field @xmath15 . as shown in ref .
@xcite , the electric charge must be - for consistency - treated as a quantity of order @xmath16 in the chiral expansion . at @xmath17 , the leading order lagrangian
which is consistent with chiral symmetry becomes @xcite , @xmath18 with @xmath19 and @xmath20 the symbol @xmath21 denotes the trace in flavour space .
the external fields @xmath22 , @xmath23 , @xmath16 and @xmath24 are given by @xmath25 where @xmath26 denote the pauli matrices .
we restrict ourselves to isovector axial fields , i.e. we take @xmath27 . for the transformation properties of the external fields , we refer to @xcite .
the mass matrix of the two light quarks is contained in @xmath24 , @xmath28 to ensure the chiral symmetry of the effective lagrangian ( [ lagr2 ] ) , the local right and left handed spurions @xmath29 and @xmath30 transform under @xmath31 , according to @xmath32 in the following , we work with a constant charge matrix @xmath33 the quantity @xmath34 denotes the gauge fixing parameter and the parameters @xmath35 , @xmath36 and @xmath37 are the three low energy coupling constants occurring at leading order . in the chiral limit and in absence of electromagnetic interactions , @xmath35 coincides with the pion decay constant , normalized such that @xmath38 @xcite , while @xmath36 is related to the quark condensate .
the coupling constant @xmath37 occurs in the low energy expansion of the charged physical pion mass only .
finally , the matrix @xmath39 is determined by the equation of motion @xmath40 while the field equations for the photon field @xmath41 read @xmath42\bar{a}^\nu + \frac{i f^2}{2}{\langle d_\mu \bar{u}[\bar{u}^+ , q ] \rangle } = 0 .
\label{cleqa}\ ] ] the next - to - leading order lagrangian reads @xmath43 the lagrangian at order @xmath4 was constructed in refs . , . ]
@xmath44 with right and left handed fields strengths defined as @xmath45 , \quad \hat{r}_{\mu\nu } = r_{\mu\nu}-\frac{1}{2}{\langle r_{\mu\nu } \rangle},{\nonumber}\\ & & l_{\mu\nu } = { \partial}_\mu l_\nu-{\partial}_\nu l_\mu - i\left[l_\mu , l_\nu\right ] , \quad \hat{l}_{\mu\nu } = l_{\mu\nu}-\frac{1}{2}{\langle l_{\mu\nu } \rangle}.\end{aligned}\ ] ] the most general list of counterterms occurring at order @xmath46 was given in refs .
, @xmath47u + d_{\mu}u[c_{\rm l}^{\mu}q , q]u^+\rangle { \nonumber}\\ & & + k_{10}\langle c_{\rm
r}^{\mu}q u c_{\rm l \mu}q u^+\rangle + k_{11}\langle c_{\rm r}^\mu q c_{\rm r \mu } q + c_{\rm l}^\mu q c_{\rm l\mu}q \rangle\big\},\end{aligned}\ ] ] where @xmath48 , \quad i = r , l.\ ] ] in the following we consider the next - to - leading order contributions where at most one virtual photon is running in a loop .
therefore , we drop the term @xmath49 in eq .
( [ lagr4 ] ) , ( terms with @xmath50 and @xmath51 in ref . ) .
the renormalized couplings are defined by @xmath52 with @xmath53.\ ] ] the coupling constant @xmath54 is finite at @xmath55 @xcite , because the singlet fields @xmath56 and @xmath57 do not occur in the leading order lagrangian ( [ lagr2 ] ) .
the coefficients @xmath58 and @xmath59 are specified in refs . @xcite and .
to evaluate the one loop contribution to the generating functional , we expand @xmath60 and @xmath61 around their solutions @xmath62 and @xmath41 to the classical equations of motion , @xmath63 collecting the fluctuations in @xmath64 , the euclidean action can be written as a quadratic form , @xmath65 here , @xmath66 is evaluated at the solution to the classical equations of motion .
the one loop functional @xmath67 then takes the form : @xmath68 in the feynman gauge @xmath69 , the euclidean differential operator is given by @xmath70 with @xmath71 the pion field indices @xmath72 run from @xmath73 to @xmath74 , while the photon field components are labeled by greek letters @xmath75 . to renormalize the determinant we work in @xmath76 dimensions , @xmath77[\delta_\mu,\tau^k ] \rangle}+\frac{1}{2}\delta^{ik}{\langle \sigma \rangle}-\frac{d f^2}{16}{\langle { h_{-}}\tau^i \rangle}{\langle { h_{-}}\tau^k \rangle}{\nonumber}\\ & & -\frac{c}{8f^2}{\langle [ { h_{+}}\!+\!{h_{-}},\tau^i][{h_{+}}\!-\!{h_{-}},\tau^k]+i\leftrightarrow k \rangle},{\nonumber}\\ \gamma^{ik}_\mu & = & -\frac{1}{2}{\langle [ \tau^i,\tau^k]\gamma_\mu \rangle},{\nonumber}\\ \gamma^i_\mu & = & f{\langle ( [ { h_{+}},\delta_\mu]+\frac{1}{2}d_\mu{h_{-}})\tau^i \rangle},{\nonumber}\\ x^{i\rho}_\mu & = & -x^{\rho i}_\mu = -\frac{f}{4}{\langle { h_{-}}\tau^i \rangle}\delta^\rho_\mu,{\nonumber}\\ \rho & = & \frac{3f^2}{8}{\langle { h_{-}}^2 \rangle},\end{aligned}\ ] ] where @xmath78,{\nonumber}\\ \gamma_\mu & = & \frac{1}{2}\left[u^+,{\partial}_\mu
u \right]-\frac{i}{2}u^+(v_\mu+\bar{a}_\mu { q_{\rm r}}+a_\mu ) u{\nonumber}\\ & & -\frac{i}{2}u(v_\mu+\bar{a}_\mu { q_{\rm l}}-a_\mu)u^+,\end{aligned}\ ] ] and @xmath79 in the absence of external fields @xmath80 reduces to @xmath81 the parameters @xmath82 denote the pion masses at leading order @xmath83 with @xmath84 .
the full operator is given by @xmath85 where @xmath86 to work out explicitly the one loop contributions to a given green function , we have to expand the determinant of the differential operator in powers of the external fields .
since @xmath87 vanishes if the external fields are switched off , we may expand @xmath88 in powers of @xmath87 , @xmath89 following the counting scheme used in refs .
@xcite , the external fields @xmath23 and @xmath16 count as @xmath90 , whereas @xmath91 and @xmath92 are of @xmath93 . in the presence of electromagnetic interactions
@xmath87 is of @xmath94 rather than @xmath93 , since the quantities @xmath95 and @xmath96 count as @xmath94 . to achieve an accuracy of @xmath97 ,
it is sufficient to stop the expansion in ( [ expdet ] ) at @xmath98 .
this is due to the fact that both , @xmath95 and @xmath96 , are of order @xmath9 , and their contributions at @xmath97 are beyond the precision of our calculation . in the following , we work in minkowski spacetime . if we add @xmath99 to @xmath100 and renormalize the low energy couplings according to , the result for the generating functional to @xmath101 is ultraviolet finite as @xmath102 . at order @xmath103
, it takes the form , ( see figure [ fig : diagrams ] ) @xmath104 explicitly , @xmath105 consists of @xmath106 where the first term contains tadpole contributions , the next three terms are generated by expanding @xmath107 around @xmath108 , while the last two terms are tree graphs .
repeated isospin indices are summed over .
the quantity @xmath109 now corresponds to in @xmath110 dimensions . however , eq . ( [ zt ] ) is both , ultraviolet and infrared finite as @xmath111 . ]
@xmath112\left[\delta^\mu,\tau_k\right ] \rangle}+\frac{1}{2}\delta_{ik}{\langle \sigma \rangle}-\frac{d f^2}{16}{\langle { h_{-}}\tau_i \rangle}{\langle { h_{-}}\tau_k \rangle}{\nonumber}\\ & & -\frac{c}{8f^2}{\langle \left[{h_{+}}\!+\!{h_{-}},\tau_i\right]\left[{h_{+}}\!-\!{h_{-}},\tau_k\right]+i\leftrightarrow k \rangle}- m_i^2\delta_{ik},\end{aligned}\ ] ] and @xmath113 the unitarity correction @xmath114 contains one loop graphs with two vertices , @xmath115 { \nonumber}\\ & & -k_\mu(x\!-\!y , m_i^2,0)\left[y_i(x)\gamma_i^\mu(y)+2y_i(x)\gamma^{\mu}_{ik}(y)y_k(y)\right ] { \nonumber}\\ & & + \tfrac{1}{4}j^r(x\!-\!y , m_i^2,m_k^2)\left[\bar{\sigma}_{ik}(x)\bar{\sigma}_{ki}(y)+2d \bar{\sigma}_{ik}(x)y_k(y)y_i(y)\right ] { \nonumber}\\ & & -\tfrac{1}{8}j^r(x\!-\!y , m_i^2,0)\left[\gamma^\mu_i(x)\gamma_{i\mu}(y)+4\gamma_{i\mu}(x)\gamma^\mu_{ik}(y)y_k(y)\right]\big\ } , \label{zu2}\end{aligned}\ ] ] as well as the triangle vertex diagram with two pion and one photon propagator , @xmath116 for the definition of the various kernels , we refer the reader to appendix [ kernels ] . for non - exceptional momenta , the generating functional ( [ zt ] ) , ( [ zu2 ] ) and ( [ zu3 ] ) is finite as @xmath111 .
however , we want to extract on shell matrix elements and this leads in general to infrared singularities .
we treat these infrared pole terms using dimensional regularization .
both , the kernels as well as the flavour matrices are thus written in @xmath117 dimensions . in the isospin symmetry limit @xmath118 and @xmath119
, our expression agrees with the explicit one loop functional of @xmath120 in ref .
@xcite .
the representation for the generating functional , specified in ( [ zt ] ) , ( [ zu2 ] ) and ( [ zu3 ] ) , contains the one loop contributions to green functions in the even intrinsic parity sector , formed with the quark currents , @xmath121 and @xmath122 the accuracy @xmath97 suffices to calculate all two point functions , the three point functions containing one vector or scalar current , as well as the pseudoscalar and axial four point functions .
the calculation of form factors or scattering amplitudes from the explicit representation of the generating functional ( [ zt ] ) , ( [ zu2 ] ) and ( [ zu3 ] ) amounts to perform the traces over the flavour matrices . in sect .
[ sextract ] , we demonstrate the extraction procedure by means of the @xmath10 scattering amplitude .
in the following , it is convenient to work in @xmath123-parameterization , i.e. to represent the matrix @xmath124 in the form @xmath125 however , all the steps listed are parameterization independent .
we expand the generating functional in powers of @xmath126 , @xmath127 where the pion field @xmath128 counts as @xmath90 and the photon field @xmath41 is of order @xmath129 .
this can be seen by solving the equations of motion ( [ cleq ] ) and ( [ cleqa ] ) to first and second order in @xmath126 , respectively .
the field equations for the pion field are discussed in subsection [ eom ] . in the presence of electromagnetic interactions ,
the quadratic term @xmath130 contains non - local contributions coming from @xmath131 .
if we expand their fourier transform around the physical pion masses @xmath132 , we arrive at @xmath133 the ellipsis stands for terms which do not contribute to the residue of the pole in the axial or pseudoscalar two point functions .
the scaling factors @xmath134 and @xmath135 are given by @xmath136+\frac{m_+^2}{8\pi^2f^2}{\rm ln } \frac{m_+^2}{\mu^2},{\nonumber}\\ z_+ & = & 1 + 4e^2{\lambda_{i\!r}}-\frac{20e^2}{9}\left(k^r_1+k^r_2\right){\nonumber}\\ & & + \frac{1}{16\pi^2 f^2}\left[m_+^2{\rm ln}\frac{m_+^2}{\mu^2}+m_0 ^ 2{\rm ln}\frac{m_0 ^ 2}{\mu^2}\right ] .
\label{zfactor}\end{aligned}\ ] ] we used dimensional regularization to treat the infrared singularity , @xmath137.\ ] ] note that the scaling factors @xmath138 and @xmath139 are parameterization independent , since unitarity determines the expansion of the field @xmath60 to @xmath140 .
we choose the charged pion mass as normalization point @xcite . in the isospin - symmetry limit @xmath118 and @xmath141 ,
the position of the pole in the correlator of two axial currents is identified with the charged pion mass .
the physical masses then read at next - to - leading order differs by a factor @xmath142 from the corresponding term obtained in ref . . ] , @xmath143+\frac{4e^2}{9}k_7{\nonumber}\\ & & -\frac{{m_{\pi^{\scriptscriptstyle{+}}}}^2 - 4e^2f^2z}{32\pi^2f^2}{\rm ln } \frac{{m_{\pi^{\scriptscriptstyle{+}}}}^2}{{m_{\pi^{\scriptscriptstyle{0}}}}^2}\bigg\}-\frac{4e^2b({m_d - m_u})}{3}k_7+{\mathcal{o}}{(e^4 , p^6 ) } , \label{mcphys}\end{aligned}\ ] ] and @xmath144{\nonumber}\\ & & -\frac{2}{f^2}({m_{\pi^{\scriptscriptstyle{+}}}}^2 - 4e^2f^2z){\rm ln } \frac{{m_{\pi^{\scriptscriptstyle{+}}}}^2}{{m_{\pi^{\scriptscriptstyle{0}}}}^2}\bigg\}+\frac{2b^2({m_d - m_u})^2}{f^2}l_7+{\mathcal{o}}{(e^4 , p^6 ) } , \label{deltapi}\end{aligned}\ ] ] with @xmath145 .
the scale independent low energy constants @xmath146 and @xmath147 are defined by @xmath148 the quantities @xmath149 and @xmath150 are the coupling constants of the isovector axial and pseudoscalar currents to the pion , @xmath151 while @xmath152 stands for the one - particle matrix element of the isoscalar density , @xmath153 the coupling constants @xmath149 , @xmath150 and @xmath152 are given explicitly in appendix [ decayconstants ] .
for a constant charge matrix @xmath154 the generating functional is invariant under the subgroup of @xmath155 transformations @xmath156 this implies in particular that the relation , @xmath157 holds at @xmath158 .
we checked that the explicit expressions for @xmath159 , @xmath160 and @xmath152 satisfy this relation to order @xmath161 .
to extract the scattering amplitude from the generating functional , we may concentrate on the pseudoscalar four point function .
the external fields @xmath23 , @xmath22 and @xmath162 are switched off and the scalar field @xmath24 is set to @xmath163 .
we may even switch @xmath41 off , because there are no one - photon exchange contributions to @xmath0 at tree level .
now @xmath164 is a functional of @xmath165 only , depending on @xmath165 both explicitly and implicitly through the equation of motion ( [ cleq ] ) .
the pseudoscalar four point function is given by the coefficient of @xmath166 in the expansion of @xmath167 in powers of @xmath126 .
the solution of the equation of motion ( [ cleq ] ) represents a power series in @xmath126 , @xmath168 to order @xmath103 , only the leading order term gives a contribution to the action .
explicitly , @xmath169 reads @xmath170 where @xmath171 stands for the feynman propagator of a pion with mass @xmath82 .
since @xmath128 represents an extremum of the action @xmath172 , the classical solution may be replaced with the extremum of ( [ zphi2 ] ) , @xmath173 where @xmath174 denotes the feynman propagator with physical mass @xmath132 .
the shift @xmath175 affects the classical action at order @xmath176 only , @xmath177 which is beyond one loop accuracy .
we may determine the scattering matrix element from the on shell residue of the fourfold pole in the fourier transform of the pseudoscalar four point function , @xmath178 the amplitude is related to the connected part of the matrix element through @xmath179 subsequently , we present a simple recipe to extract the @xmath180 amplitude from the explicit representation of the generating functional ( [ zt ] ) , ( [ zu2 ] ) and ( [ zu3 ] ) : * switch all explicit external fields off , including the field @xmath165 . now the generating functional @xmath181 depends on @xmath182 only . *
perform the traces over the flavour matrices @xmath123 , @xmath183 , @xmath184 * the pion fields are rescaled , according to @xmath185 .
this step brings the quadratic terms ( [ zphi2 ] ) to normal form . *
the scattering matrix element may be read off from the terms of order @xmath186 .
these contributions generate the on - shell fourfold pole in eq .
( [ fourpoint ] ) .
as a check , we calculated the @xmath0 amplitude to @xmath101 from ( [ zt ] ) , ( [ zu2 ] ) and ( [ zu3 ] ) .
we compared the thus obtained expression with the result of knecht and urech , who used a photon mass as an infrared regulator . upon identifying the infrared pole term in dimensional regularization with @xmath187
the two results are in agreement .
the on shell matrix element contains infrared singularities generated by virtual photon contributions . since we used dimensional regularization to treat the infrared pole terms ,
the amplitude is well defined in @xmath188 only .
the essential loop function , which develops an infrared singularity as @xmath111 , is the scalar triangle diagram @xmath189 with one photon and two charged on shell pion propagators .
a discussion of the analytic properties of @xmath189 , as well as an explicit representation of this function , is given in appendix [ triangle ] .
expanding the amplitude around threshold @xmath190 leads to @xmath191 in the condon - shortley phase conventions . here , @xmath192 denotes the relative 3-momentum of the charged pions and the coulomb pole stems from the infrared region of the scalar triangle diagram ( [ gdcoulomb ] ) .
the quantity @xmath193 is free of infrared singularities .
we may now expand @xmath193 in powers of the isospin breaking parameter @xmath194 , @xmath195 according to @xmath196 the combination @xmath197 denotes the s - wave scattering length difference in the isospin symmetry limit @xmath118 , @xmath119 @xcite , @xmath198+{\mathcal{o}}{(\hat{m}^3)}.\ ] ] the coefficients @xmath199 and @xmath200 of the isospin breaking contributions are given by @xcite , @xmath201{\nonumber}\\ & & + \frac{3{m_{\pi^{\scriptscriptstyle{+}}}}^2}{256\pi^2f^2}p(k_i)+{\mathcal{o}}{(\hat{m}^2)},\end{aligned}\ ] ] where @xmath202 denotes the physical mass difference at @xmath118 , @xmath203 and @xmath204 stands for the following combination of low energy constants @xmath205
in the present paper we constructed an explicit representation of the generating functional for @xmath206 with virtual photons at first non - leading order in the low energy expansion [ @xmath101 ] .
the explicit form includes graphs , where up to two pions and at most one photon are running in the loop .
it has been shown in that the ultraviolet divergences of the one loop functional are absorbed by the corresponding counterterms of @xmath208 . for non - exceptional
4-momenta the generating functional is thus finite as @xmath102 . nevertheless , if we want to extract on shell matrix elements , we encounter in general infrared singularities , generated by soft photon contributions .
we use dimensional regularization to treat these infrared pole terms and therefore specify the explicit form of the generating functional in @xmath117 dimensions . as a check of the one loop functional
, we evaluated the @xmath209 scattering amplitude at order @xmath210 .
the thus obtained result agrees with the one in ref . .
furthermore , we specify the electromagnetic corrections to the decay constants @xmath211 and @xmath212 , which measure the one - particle matrix elements of the isovector axial and pseudoscalar currents .
the extension to @xmath213 is in progress .
it will provide a comparison with the @xmath214 amplitudes already available in the literature @xcite .
i am grateful to j. gasser for many illuminating discussions as well as for helpful comments on the manuscript .
further , i thank r. kaiser for useful remarks .
this work was supported in part by the swiss national science foundation , and by rtn , bbw - contract n0 .
01.0357 and ec - contract hprn - ct-2002 - 00311 ( euridice ) .
here , we explain the notation used in subsection [ sexplz ] .
we consider the fourier transforms ( @xmath215 ) @xmath216 the kernel @xmath217 remains ultraviolet finite as @xmath111 , @xmath218 where @xmath219 and @xmath220 the renormalized kernel @xmath221 reads : @xmath222 with @xmath223\bar{j}(s , m_i^2 , m_k^2){\nonumber}\\ & & + \frac{d}{s}(m_i^2-m_k^2)^2\bar{{\bar{j}}}(s , m_i^2 , m_k^2)\big\}-\frac{1}{6}k(m_i^2 , m_k^2)+\frac{1}{288\pi^2},{\nonumber}\\ l(s , m_i^2 , m_k^2 ) & = & \frac{1}{4s}(m_i^2-m_k^2)^2\bar{j}(s , m_i^2 , m_k^2),\end{aligned}\ ] ] and @xmath224.\end{aligned}\ ] ] we list the ultraviolet finite kernels @xmath217 and @xmath221 in @xmath225 dimensions , because we encounter an infrared singularity , if we expand the charged pion self - energy around its physical mass .
this pole term , which comes from @xmath226 , leads to an infrared divergent scaling factor @xmath139 in eq .
( [ zfactor ] ) .
furthermore , we introduce the fourier transform @xmath227 and similarly for the other kernels @xmath228 and @xmath229 in eq .
( [ zu3 ] ) .
then we have @xmath230 with @xmath231 the integrals @xmath232 and @xmath233 are ultraviolet finite , but @xmath232 develops an infrared singularity for on shell 4-momenta ( see appendix [ triangle ] ) .
the functions @xmath234 and @xmath235 have ultraviolet poles at @xmath108 : @xmath236 for non - exceptional 4-momenta the renormalized quantities @xmath237 and @xmath238 remain finite as @xmath111 .
the scalar vertex function @xmath189 is given by @xmath239 where @xmath240 , @xmath241 denote the on shell 4-momenta of the incoming @xmath242 , @xmath243 mesons and @xmath244 .
the vertex function is ultraviolet finite , but develops an infrared singularity at @xmath55 .
( at @xmath245 , the singularity would be absent ) .
( 40,40 ) ( 0,0)(1,0)40 ( 0,37)(0,0)@xmath6 ( 0,0)(0,1)35 ( 42,0)(0,0)@xmath24 ( 0,20)(1,0)40 ( -5,20)(0,0)@xmath108 ( 20,0)(0,1)35 ( 23.5,-3)(0,0)@xmath246 ( 0,0)(1,1)19.8 ( 0,5)(1,1)14.8 ( 0,10)(1,1)9.8 ( 0,15)(1,1)4.8 ( 5,0)(1,1)19.8 ( 10,0)(1,1)19.8 ( 15,0)(1,1)19.8 ( 20,0)(1,1)19.8 ( 25,0)(1,1)14.8 ( 30,0)(1,1)9.8 ( 35,0)(1,1)4.8 ( 30,30)(-1,0)7 ( 23,30)(-1,0)3 ( 30,30 ) ( 30,30)(0,-1)7 ( 30,23)(0,-1)3 ( 30,20)(-1,0)10 ( 20,30)(0,-1)10 ( 17.5,25)(0,0)@xmath247 ( 32.5,25)(0,0)@xmath34 for @xmath248 , the function has a cut along the positive real axis . expanding the vertex function in @xmath249 for arbitrary @xmath248 yields @xmath250 with @xmath251 and @xmath252 if we expand ( [ gds ] ) near threshold @xmath253 , which corresponds to path @xmath34 in fig . [
fig : gd ] , we find @xmath254 where @xmath192 denotes the relative momentum of the incoming pions in the cm frame .
the singular contributions @xmath255 are generated by small values of the variable of integration @xmath256 in ( [ gd ] ) . on the contrary
, if we first evaluate the vertex function at threshold in @xmath117 dimensions @xmath257 and then expand in @xmath249 , i.e we choose path @xmath247 in fig .
[ fig : gd ] @xmath258+{\mathcal{o}}{(d-4)},\ ] ] we are left with the constant term in eq .
( [ gdcoulomb ] ) .
in presence of electromagnetic interactions the coupling constants of the isovector axial and pseudoscalar currents to the pion are given by in @xmath259 differs by a factor @xmath260 from the corresponding expression in ref . .
] , @xmath261-\frac{e^2z}{8\pi^2}\bar{l}_4{\nonumber}\\ & & -\frac{e^2}{64\pi^2}\left [ \left(3+\tfrac{4z}{9}\right)\bar{k}_1 -\tfrac{40z}{9}\bar{k}_2 - \bar{k}_9\right]{\nonumber}\\ & & + \frac{{m_{\pi^{\scriptscriptstyle{+}}}}^2 - 2z e^2 f^2}{32\pi^2 f^2}{\rm ln}\frac{{m_{\pi^{\scriptscriptstyle{+}}}}^2}{{m_{\pi^{\scriptscriptstyle{0}}}}^2}+{\mathcal{o}}{(e^2 m_q)}\bigg\ } , { \nonumber}\\ f_{{\pi^{\scriptscriptstyle{0 } } } } & = & f_\pi - f\bigg\ { \frac{e^2 z}{8\pi^2}\bar{l}_4 + \frac{e^2}{64\pi^2}\left[\left(3+\tfrac{4z}{9}\right)\bar{k}_1-\tfrac{40z}{9}\bar{k}_2 { \nonumber}\right.\\ & & \left . -3 \bar{k}_3 - 4z \bar{k}_4\right]+{\mathcal{o}}{(e^2 m_q)}\bigg\},\end{aligned}\ ] ] and @xmath262 + \frac{4e^2}{9}k_7+{\mathcal{o}}{(e^2 m_q)}\bigg\ } , { \nonumber}\\ g_{{\pi^{\scriptscriptstyle{0 } } } } & = & g_\pi+2bf\bigg\{\frac{e^2z}{16\pi^2}\left(\bar{l}_3 - 2\bar{l}_4\right)+\frac{e^2}{64\pi^2}\left[\left(3+\tfrac{4z}{9}\right)\bar{k}_1-\tfrac{40z}{9}\bar{k}_2\right.{\nonumber}\\ & & \left.-3\bar{k}_3 - 4z\bar{k}_4-\tfrac{2}{9}\left(5 + 4z\right)\bar{k}_5 + \tfrac{10}{9}\left(1 + 8z\right)\bar{k}_6\right]+\frac{4e^2}{9}k_7{\nonumber}\\ & & + \frac{{m_{\pi^{\scriptscriptstyle{+}}}}^2 - 2e^2f^2z}{32\pi^2f^2}{\rm ln}\frac{{m_{\pi^{\scriptscriptstyle{+}}}}^2}{{m_{\pi^{\scriptscriptstyle{0}}}}^2 } + { \mathcal{o}}{(e^2 m_q)}\bigg\}. \ ] ] the quantities @xmath211 and @xmath212 denote the coupling constants at @xmath263 and @xmath118 @xcite , @xmath264,{\nonumber}\\ g_\pi & = & 2bf\big[1-\frac{{m_{\pi^{\scriptscriptstyle{+}}}}^2}{32\pi^2f^2}\left(\bar{l}_3 - 2\bar{l}_4\right)+{\mathcal{o}}{(m_q^2)}\big].\end{aligned}\ ] ] due to the infrared pole term , @xmath265 and @xmath266 are well defined in @xmath117 dimensions only .
30 s. weinberg , physica * a96 * ( 1979 ) 327 j. gasser and h. leutwyler , ann .
* 158 * ( 1984 ) 142 j. gasser and h. leutwyler , nucl .
b250 * ( 1985 ) 465 r. unterdorfer , jhep * 0207 * ( 2002 ) 053 [ arxiv : hep - ph/0205162 ] g. ecker , j. gasser , a. pich and e. de rafael , nucl . phys .
* b321 * ( 1989 ) 311 r. urech , nucl .
* b433 * ( 1995 ) 234 [ arxiv : hep - ph/9405341 ] m. knecht and r. urech , nucl . phys .
* b519 * ( 1998 ) 329 [ arxiv : hep - ph/9709348 ] ulf - g .
meissner , g. mller and s. steininger , phys .
b406 * ( 1997 ) 154 [ arxiv : hep - ph/9704377 ] m. knecht and a. nehme , phys . lett .
* b532 * ( 2002 ) 55 [ arxiv : hep - ph/0201033 ] b. kubis and ulf - g .
meissner , nucl .
a671 * ( 2000 ) 332 [ arxiv : hep - ph/9908261 ] v. cirigliano , m. knecht , h. neufeld and h. pichl , [ arxiv : hep - ph/0209226 ] v. cirigliano , g. ecker and h. neufeld , jhep * 0208 * ( 2002 ) 002 [ arxiv : hep - ph/0207310 ] v. cirigliano , g. ecker and h. neufeld , phys . lett . *
b513 * ( 2001 ) 361 [ arxiv : hep - ph/0104267 ] m. knecht , h. neufeld , h. rupertsberger and p. talavera , eur .
j. * c12 * ( 2000 ) 469 [ arxiv : hep - ph/9909284 ] h. neufeld and h. rupertsberger , z. phys .
* c68 * ( 1995 ) 91 r. baur , j. kambor and d. wyler , nucl . phys .
* b460 * ( 1996 ) 127 [ arxiv : hep - ph/9510396 ] v. cirigliano , j. f. donoghue and e. golowich , phys . rev .
* d61 * ( 2000 ) 093001 [ arxiv : hep - ph/9907341 ] g. ecker , g. isidori , g. mller , h. neufeld and a. pich , nucl . phys . * b591 * ( 2000 ) 419 [ arxiv : hep - ph/0006172 ] v. cirigliano , m. knecht , h. neufeld , h. rupertsberger and p. talavera , eur .
j. * c23 * ( 2002 ) 121 [ arxiv : hep - ph/0110153 ] b. kubis and ulf - g .
meissner , nucl .
* a699 * ( 2002 ) 709 [ arxiv : hep - ph/0107199 ] b. kubis and ulf - g .
meissner , phys .
b529 * ( 2002 ) 69 [ arxiv : hep - ph/0112154 ] j. schweizer , work in progress review of particle physics , k. hagiwara _ et al_. , phys . rev .
* d66 * ( 2002 ) 010001 r. kaiser , phys . rev .
* d63 * ( 2001 ) 076010 [ arxiv : hep - ph/0011377 ] j. gasser , v. e. lyubovitskij , a. rusetsky and a. gall , phys . rev .
* d64 * ( 2001 ) 016008 [ arxiv : hep - ph/0103157 ] j. gasser , v. e. lyubovitskij and a. rusetsky , phys . lett . * b471 * ( 1999 ) 244 [ arxiv : hep - ph/9910438 ] | we construct the explicit one loop functional of chiral perturbation theory for two light flavours , including virtual photons .
we stick to contributions where 1 or 2 mesons and at most one photon are running in the loops . with the explicit functional at hand ,
the evaluation of the relevant green functions boils down to performing traces over the flavour matrices . for illustration ,
we work out the @xmath0 scattering amplitude at threshold at order @xmath1 . butp-02/13 * explicit generating functional for pions and virtual photons * j. schweizer + institute for theoretical physics , + university of bern , + sidlerstr . 5 , ch3012 bern , switzerland + e - mail : [email protected] march 31 , 2003 |
observations are currently being undertaken for the detection of microlensing events , with survey observations being carried out by the ogle @xcite and moa @xcite groups , and with follow - up observations being performed by the planet @xcite and microfun @xcite groups .
the follow - up observations are focusing on high magnification events toward the galactic bulge , so that they are efficient for detecting events with the characteristic features of binary lens systems , such as the caustic - crossing appearing in high magnification events .
moreover , if future space missions such as _ microlensing planet finder _ ( _ mpf _
@xcite ) with high photometric precision and monitoring cadence are carried out , the efficiency will significantly increase , and thus many more binary lensing events with the characteristic features will be found .
detection of binary microlensing events is important for providing various information about the lensing parameters .
first , the binary mass ratio and projected separation are obtained .
if many binary lensing events are detected , these parameters can help to infer statistical properties of binaries in the lens population @xcite .
second , the special case of caustic - crossing events provides the caustic - crossing time , @xmath2 , which is given by @xmath3 where @xmath4 is the source radius normalized to the einstein ring radius , corresponding to the total mass of the binary , @xmath5 is the einstein timescale of the binary lens , @xmath6 is the angular source radius , @xmath7 is the caustic - crossing angle , and @xmath8 is the lens - source proper motion . the angular source radius can be measured from its color and apparent magnitude . the einstein timescale and caustic - crossing angle can be determined from fitting of the observed light curve .
third , we can obtain the information of the lens - source proper motion and the normalized source radius from equation ( 1 ) .
fourth , because @xmath9 , if the proper motion is determined , the angular einstein ring radius , @xmath10 , can be measured .
the einstein ring radius is related to the total mass of the binary lens , @xmath11 , which is expressed as @xmath12 where @xmath13 and @xmath14 are the distances to the lens and the source from the observer , respectively .
therefore , measuring the proper motion is very important to determine the lens mass and distance to the lens .
@xcite first reported that there exist multiple solutions to a photometrically observed binary lensing event .
@xcite investigated the pattern of the centroid shift of the source images selecting four of the multiple solutions in the paper of @xcite . as a result , all four cases showed different centroid shift patterns , and thus they demonstrated that the photometric degeneracy could be astrometrically resolved .
@xcite , however , showed that for the macho 98-smc-1 event @xcite , both of the light curves and centroid shift patterns of the close and wide binary solutions are extremely similar .
the degeneracy of the close and wide binaries is rooted in the similarity of the lens equations of the two solutions , and thus it is very difficult to find a unique solution @xcite .
nevertheless they found that the centroid shift curves of the two solutions have an offset at the point of the caustic - crossing time , and if the offset is observable through a space mission such as _ space interferometry mission planetquest _ ( _ sim planetquest _
@xcite ) with high astrometric precision , a severe degeneracy could be astrometrically resolved .
in addition to the method of @xcite , there is a further method for breaking severe close / wide degeneracies of binary microlensing events . for a wide binary lensing event ,
if a source star is lensed by one of the binary components ( the first event ) , after this lensing event finishes , the event induced by the other binary star ( the second event ) can also be detected . in this paper
, we investigate whether the close / wide degeneracies in binary microlensing events can be resolved by detecting the centroid shift of the second event in wide binary lensing events .
we determine the probability of detecting the additional centroid shift in high magnification binary lensing events .
we also estimate the waiting time for the detection of the centroid shift of the second event after the caustic - crossing of the first event in wide binary lensing events .
the paper is organized as follows . in 2 , we describe the astrometric properties of microlensing . in 3 , we investigate a new astrometric method for breaking the degeneracy of binary lensing events and determine the probability of detecting the additional centroid shift in binary lensing events with high magnification .
in addition , we estimate the waiting time for the detection of the additional centroid shift in wide binary lensing events . we conclude with a summary of our results and discussion in 4 .
when a source star is lensed by a single lens , it is split into two images with individual magnifications and the centroid of the images is defined by @xmath15}{\theta_{\rm e}}\ ; \ \ \ { \mbox{\boldmath$u$}}= { \left ( t - t_0 \over { t_{\rm e}}\right ) { \hat{\it x}}+ { u}_{0}}\ { \hat{\it y}}\ , \ ] ] where @xmath16 represents the projected lens - source separation in units of @xmath10 .
the lensing parameters @xmath17 and @xmath18 are the lens - source impact parameter in units of @xmath10 and the time of the maximum magnification , respectively .
the @xmath19 and @xmath20 axes represent the directions that are parallel with and normal to the lens - source transverse motion .
the centroid shift of the images with respect to the source star is determined by @xmath21 for a widely separated binary lensing event , each of the binary lens components works as the individual single lenses @xcite .
the centroid shift of this event can be described as the superposition of events induced by the individual lens components , and thus is expressed as @xmath22 where @xmath23 and @xmath24 are the centroid shifts of the individual lens components , respectively .
the condition for wide binary lensing events to become photometric double lensing events is that the source should pass in the einstein ring of the lens causing the second event .
however , astrometric double lensing events can occur even though the source passes away from the einstein ring of the lens .
therefore , the range of detecting the astrometric double lensing events is wider than for the photometric double lensing events and is represented by @xmath25 @xmath26 @xmath27 for @xmath28 , @xmath29 for @xmath30 , and @xmath31 for @xmath32 . here
@xmath33 is the angle between the source trajectory and the binary axis , @xmath34 is the projected binary separation normalized to the einstein radius of the primary , @xmath35 and @xmath36 are the threshold impact parameters to the primary and the companion normalized to the einstein ring radii of the individual lens components @xcite .
the indices @xmath37 and @xmath38 represent the cases when the impact parameters are located on the opposite and same sides to the binary axis , respectively .
for resolving the close / wide degeneracies in binary lensing events using a new astrometric method , we use two solutions of the macho 98-smc-1 event that are very similar in terms of the light curve and centroid shift ( afonso et al .
2000 ) , but which have opposite source trajectories .
figure 1 shows the geometries of the close and wide solutions of a severely degenerate binary lensing event . in figure 1
, @xmath39 is the projected binary separation and @xmath40 is the mass ratio of the binary .
the coordinates @xmath41 represent the axes that are parallel with and normal to the binary axis .
the coordinates for the close and wide binaries are centered at the center of mass and the midpoint of the binary lens , respectively . here
all lengths in the close and wide binaries are normalized to the individual einstein ring radii of the two cases .
the straight line with an arrow represents the source trajectory and the curve that varies with the source trajectory represents the trajectory of the image centroid .
the solid curve represents the image centroid dominated by the companion , @xmath42 , while the dot curve represents the image centroid dominated by the primary , @xmath43 .
the dashed circles for the wide binary represent the einstein radii of the individual lens components , while for the close binary the dashed circle represents the combined einstein radius of the binary .
two dots are the locations of the individual lens components .
the concave curves with cusps located around or away from the lenses are caustics that represent the set of source positions at which the jacobian determinant of the lens equation becomes zero , and thus the magnification of a point source becomes infinite .
a plus sign in the caustic represents the center of mass of the binary for the close binary and represents the effective position for the wide binary .
the criterion for which lens of the binary lens components has a greater effect on the binary lensing event is determined by @xmath44 where @xmath45 , @xmath46 , and @xmath47 represent the effective lens - source separation , the source position , and the effective positions of the individual lens components , respectively , which are normalized to the combined einstein radius of the binary lens system .
therefore , the solid curve in figure 1 is for @xmath48 , while the dotted curve is for @xmath49 .
the light curves and the centroid shift trajectories resulting from the source trajectories represented in figure 1 are shown in figure 2 . in figure 2 ,
@xmath50 and @xmath51 represent the einstein radii corresponding to the total masses of the close and wide binaries , respectively .
an arrow in the lower panel represents the orientation of the centroid shift trajectory .
the dot curve in the left side of the lower panel is caused by the primary .
since the first observed event is caused by the companion in the wide binary , the additional deviation in the centroid shift trajectory is caused by the primary .
the additional deviation can be induced by the primary or the companion , depending on which lens of the binary lens components causes the first observed event . assuming that the first event is a typical galactic lensing event caused by @xmath52 and with @xmath53 and @xmath54 , the einstein timescale is @xmath55 days , the angular einstein radius is @xmath56 , and the centroid shift is usually several tens of @xmath57 . under this assumption ,
the crosses marked in figure 2 show the progress of the event in the interval of 30 days after the caustic - crossing .
as shown in the figure , the additional deviation appears about 30 days after the caustic - crossing and the centroid shift at the position of the next cross changes by @xmath58 from the first cross position .
if this remarkable additional deviation is detected by using future astrometric follow - up observations , such as _ sim planetquest _ with the position accuracy of @xmath59 , the degeneracy in binary microlensing events can be easily resolved . however , if the source passes too far away from the einstein ring of the lens causing the second event , it will be difficult to detect the additional deviation .
the source should cross right on the caustic or pass in the vicinity of the caustic to be securely identified as binary lensing events , so that most of the observed binary lensing events are high magnification events . to determine the probability of detecting the remarkable additional centroid shift of the second event in wide binary lensing events with high magnification , we set @xmath60 and @xmath61 in equations @xmath62 . because of the long range astrometric effect of the lens causing the second event , the additional deviation presents well in the centroid shift trajectory , despite the fact that @xmath36 is 2.5 times larger than the einstein ring of the lens .
we also set that the lens causing the first event is the primary and the lens causing the second event is the companion . in this case , since the mass of the companion causing the second event can be heavier than that of the primary , the binary mass ratio can be larger than @xmath63 .
the probability is determined by @xmath64 and the result is presented in figure 3 .
as shown in figure 3 , the probability increases as the binary mass ratio increases and the binary separation decreases , because the efficient astrometric range increases as the mass ratio increases and the separation decreases .
the separation distribution of three of the four binary lensing events with known close and wide solutions is @xmath65 .
we thus present the probability under the two different binary separation distributions in table 1 .
the physical separation is determined by assuming that the physical einstein radius of the typical galactic lensing event is @xmath66 . from the determination
, we find that in the case where the mass of the companion is 1.5 times heavier than that of the primary , i.e. , @xmath67 , the degeneracy of binary lensing events with a separation of @xmath68 can be resolved with an efficiency @xmath69 .
the issue of how long we should wait for the detection of the centroid shift of the second event after the caustic - crossing of the first event in wide binary lensing events depends on both the binary separation and the einstein timescale of the first event for a given binary mass ratio .
the @xmath70 for the detection of the additional centroid shift in wide binary lensing events decreases as the binary separation and the einstein timescale of the first event decrease .
when the caustic - crossing events are detected , various lensing parameters can be estimated from the fit of those light curves .
the separation and the einstein timescale of them would be helpful to foresee the waiting time . for predicting the waiting time , we estimate it by using the distributions of the separation and the einstein timescale of observed galactic lensing events .
figure 4 represents the waiting time for wide binary lensing events with @xmath67 . from this
, we find that for typical galactic events with @xmath71 days at a separation of @xmath72 , one can detect the additional centroid shift within @xmath1 days .
we have investigated whether the close / wide degeneracies in binary microlensing events can be resolved by detecting the additional centroid shift induced by the lens causing the second event in wide binary microlensing events . from this investigation
, we found that in the case where the source passes close to the einstein ring of the lens causing the second event , the degeneracy could be easily resolved by using future astrometric follow - up observations .
we have determined the probability of detecting the additional centroid shift in high magnification binary lensing events . from this
, we found that the degeneracy of binary lensing events with the separation of @xmath73 could be resolved with a significant efficiency .
we have also estimated the waiting time for the detection of the additional centroid shift in wide binary lensing events . from this
, we found that for typical galactic binary lensing events , the degeneracy could be sufficiently resolved within a year .
the targets of _ sim planetquest _ are usually limited to rather bright source stars and thus the degeneracy for very faint sources with @xmath74 could not be broken by this mission . the probability presented in sec .
3 has not considered the source brightness ; if the source brightness were to be considered , then the probability would actually more decrease . considering that for the galactic lensing events with @xmath71 days the additional deviation would appear about a month after the caustic - crossing and the degeneracy could be broken within 100 days , the astrometric observations once a month would be enough for breaking the close / wide degeneracy .
unfortunately , the mission of _ sim planetquest _ has been deferred and there are also no other astrometric missions with the capabilities of _ sim _ , while in the near future the ground - based photometric microlensing survey would be carried out for 24-hour by the global telescope network with three 1.6 m class telescopes located at good seeing sites ( _ korea microlensing telescope network _
( _ kmtnet _ ) project @xcite ) .
recently , the newly redesigned _
sim - lite _ , which costs less than the original mission , and which has improved , has been proposed by @xcite . however , even though the mission is due to restart in the near future , it is very difficult to foresee its launching date . | when a source star is microlensed by one stellar component of a widely separated binary stellar components , after finishing the lensing event , the event induced by the other binary star can be additionally detected . in this paper
, we investigate whether the close / wide degeneracies in binary lensing events can be resolved by detecting the additional centroid shift of the source images induced by the secondary binary star in wide binary lensing events . from this investigation
, we find that if the source star passes close to the einstein ring of the secondary companion , the degeneracy can be easily resolved by using future astrometric follow - up observations with high astrometric precision .
we determine the probability of detecting the additional centroid shift in binary lensing events with high magnification . from this , we find that the degeneracy of binary lensing events with a separation of @xmath0 au can be resolved with a significant efficiency .
we also estimate the waiting time for the detection of the additional centroid shift in wide binary lensing events . we find that for typical galactic lensing events with a separation of @xmath0 au , the additional centroid shift can be detected within @xmath1 days , and thus the degeneracy of those events can be sufficiently broken within a year . |
since abrikosov s @xcite prediction of the flux - line lattice in type - ii superconductors from ginzburg - landau ( gl ) theory , most theoretical work on this vortex lattice in bulk and thin film superconductors considered the situation when the applied magnetic field and the average induction @xmath5 are close to the upper critical field @xmath6 , since analytical solutions may be obtained for this particular case . in the opposite limit of low induction @xmath7 , the properties of an isolated vortex and the interaction between vortices are obtained to good approximation from the london theory when the gl parameter @xmath0 is not too small , @xmath8 .
@xcite the problem of an isolated vortex in thin films was solved analytically within london theory by pearl @xcite ; the interaction energy of such pearl vortices ( or pancake vortices @xcite ) is easily calculated by noting that within london theory the currents and magnetic fields of the vortices superimpose linearly and that the force on a vortex equals the thickness - integrated super - current density at the vortex core times the quantum of flux @xmath9 . in thin films with thickness @xmath10 smaller than the london magnetic penetration depth @xmath11 the range of the vortex
vortex interaction is increased to the effective penetration depth @xmath12 since the interaction now occurs mainly via the magnetic stray field outside the film .
@xcite vortices in superconducting films of finite thickness ( @xmath13 and @xmath14 ) and in the superconducting half space ( @xmath15 ) were calculated from gl theory @xcite and london theory .
@xcite at larger reduced induction @xmath16 when the london theory does not apply , the properties of the gl vortex lattice have to be computed numerically .
a very efficient method @xcite uses fourier series as trial functions for the gl function @xmath17 and magnetic field @xmath18 and minimizes the gl free energy with respect to a finite number of fourier coefficients . this numerical method
was recently improved @xcite by solving the gl equations iteratively with high precision .
the present paper extends this two - dimensional ( 2d ) method to the 3d problem of a film of arbitrary thickness containing a periodic lattice of gl vortices oriented perpendicular to the film plane . due to the fourier ansatz
, the magnetic stray field energy is easily accounted for in this method .
moreover , it turns out that the extension from 2d to a 3d problem only slightly increases the required computation time and computer memory , so that high precision can be achieved easily on a personal computer . like in refs .
, we consider here vortex lattices with arbitrary shape of the unit cell containing one vortex , i.e. , our method computes triangular , rectangular , square lattices , etc . , and yields also the shear moduli @xcite of the equilibrium lattices .
the approximate shear modulus @xmath2 of the triangular vortex lattice in thin films was computed from gl theory for @xmath19 and @xmath20 in ref . .
for early work on films with perpendicular vortex lattice see refs . .
though we consider here isotropic superconductors , the corresponding results for anisotropic superconductors with principal symmetry axes along @xmath21 may be obtained from this isotropic method by scaling the coordinates and introducing an effective gl parameter @xmath22 .
@xcite the magnetic field of a vortex inside a uniaxially anisotropic superconductor with surface parallel to the @xmath23 symmetry plane and perpendicular to the vortex line was calculated from anisotropic london theory @xcite and compared with experiments in ref . .
the main effect of the flat surface of a superconductor film or half space is the widening of the magnetic field lines of the vortices as they approach the surface .
this widening minimizes the sum of the bulk free energy plus the energy of the magnetic stray field outside the superconductor .
the resulting magnetic field lines cross the superconductor surface smoothly , see fig . 1 for the vortex lattice and figs .
1,2 of ref . for isolated vortices .
one can see that for the _ vortex lattice _
the field lines at the boundary of the wigner - seitz cell are exactly parallel to @xmath24 , inside and outside the superconductor , and at some distance outside from the surface ( @xmath25 half the vortex spacing ) the magnetic field becomes uniform and thus the field lines are parallel and equidistant . for the _ isolated vortex _ , the field lines away from the surface approach radial lines as if they would originate from a point source , a magnetic monopole with strength @xmath26 situated on the vortex core at a distance @xmath27 below the surface .
@xcite in ref .
the widening of the field lines inside the superconductor was missed , but some modification of the superconductor order parameter near the surface was calculated from gl theory . below we obtain that the correct modification of @xmath28 is very small : the vortex core , visualized as contour lines of @xmath29 , widens near the surface by only a few percent .
the outline of this paper is as follows . in sct .
2 the solution method is outlined .
section 3 presents a selection of results for thin and thick films : magnetic field lines , profiles of the order parameter and magnetic field , the variances of the periodic order parameter and magnetic field inside and outside the film , surface energy and stray - field energy , and shear modulus of the triangular vortex lattice in the film .
a summary is given in sct .
the properties of the vortex lattice within gl theory are calculated by minimizing the gl free energy of the superconductor with respect to the complex gl function @xmath30 and to the vector potential @xmath31 of the local magnetic induction @xmath32 . in the usual reduced units @xcite ( length @xmath11 , magnetic induction @xmath33 , energy density @xmath34 , where @xmath35 is the thermodynamic critical field ) the spatially averaged free energy density @xmath36 of the gl theory referred to the meissner state ( @xmath37 , @xmath38 ) within the superconductor reads @xmath39 here @xmath40 means spatial averaging over the superconductor with volume @xmath41 . introducing the supervelocity @xmath42 and the magnitude @xmath43 of @xmath44 $ ] one
may write @xmath36 as a functional of the real and gauge - invariant functions @xmath45 and @xmath46 , @xmath47 in the presence of vortices @xmath48 has to be chosen such that @xmath49 has the appropriate singularities along the vortex cores , where @xmath45 vanishes . by minimizing this @xmath36 with respect to @xmath50 , @xmath51 or @xmath45 , @xmath46 ,
one obtains the gl equations together with the appropriate boundary conditions . for the superconducting film considered here
, one has to add the energy of the magnetic stray field outside the film , which makes @xmath52 continuous at the film surface , see below .
the 3d solution of the gl equations for an infinitely large , thick or thin film with periodic lattice of vortices perpendicular to the film plane , can be obtained numerically by a modification of the 2d method developed in refs . .
this is possible since in any plane @xmath24 = const parallel to the film the solutions for the ideal vortex lattice are still periodic .
actually this periodicity applies even to tilted and arbitrarily curved vortex lines , and to anisotropic superconductors , which may be computed by a similar method .
these more complex problems will be considered in future work .
for the present problem of straight vortices along @xmath24 one may choose a general ansatz for the magnitude of the gl function @xmath53 in form of the following 3d fourier series for the smooth function @xmath54 : @xmath55 here @xmath56 , @xmath57 , @xmath58 , @xmath59 .
in all sums here and below the term @xmath60 is excluded . for vortex positions
@xmath61 the reciprocal lattice vectors are @xmath62 with @xmath63 the unit cell area and @xmath64 , @xmath65 .
the @xmath24-component of @xmath66 is chosen as @xmath67 with @xmath68 , 1 , 2 , @xmath69 and @xmath10 the film thickness .
this ansatz guarantees that @xmath70 at the ( straight ) vortex cores and that at the two surfaces of the film @xmath71 one has @xmath72 , as it follows from the variation of the gl free energy functional ( 2 ) .
if only the term @xmath73 is kept , the ansatz ( 3 ) reduces to that for the 2d vortex lattice in ref . .
formally , the 3d fourier series ( 3 ) may also be expressed as a 2d fourier series with @xmath24 dependent coefficients @xmath74 . for the supervelocity @xmath46 and magnetic induction @xmath75 inside the film we chose the general ansatz @xmath76 here @xmath77 is the supervelocity of the abrikosov @xmath78 solution , which satisfies @xmath79 { \bf\hat z}\ , , \end{aligned}\ ] ] where @xmath80 is the 2d delta function and @xmath9 the quantum of flux , @xmath81 in reduced units .
formula ( 5 ) shows that @xmath82 is the velocity field of a lattice of ideal vortex lines but with zero average rotation . near each vortex center one
has @xmath83 and @xmath84 with @xmath85 .
@xmath86 may be expressed as a slowly converging fourier series by integrating ( 5 ) using @xmath87 .
it is , however , more convenient to take @xmath82 from the exact relation @xmath88 where @xmath89 is the abrikosov @xmath78 solution given by a rapidly converging series of type ( 3 ) with @xmath24-independent coefficients @xmath90 \end{aligned}\ ] ] for general vortex - lattice symmetry , and @xmath91 ( @xmath92 ) for the triangular lattice .
the @xmath93 from ( 7 ) is normalized to @xmath94 ; this yields the strange relation @xmath95 for any lattice symmetry .
another strange property of the abrikosov solution ( 7 ) is that @xmath96 , although both terms diverge at the vortex positions ; this relation follows from ( 5 ) and ( 6 ) using @xmath97 .
the useful formula ( 6 ) may be proven via the complex @xmath78 solution @xmath98 ; it means that near @xmath78 the second and third terms in the @xmath36 , eq .
( 2 ) , are equal .
the general ansatz for @xmath99 is a fourier series for all three components , satisfying @xmath100 . for simplicity here
i shall assume @xmath101 , which means planar supercurrents .
in the considered case of vortices perpendicular to the film plane this is an excellent approximation , which is exact in the limit of small induction and probably also at large inductions @xmath102 , and it is exact for thin films .
i further assume @xmath103 , which is exact in several special cases ( e.g. for @xmath7 and @xmath102 ) and is possibly exact even in the general case , though i did not find a proof for this .
note also that within the circular cell approximation @xcite both assumptions are satisifed . with these two assumptions @xmath104
is fully determined by the @xmath24-component of @xmath105 : @xmath106 with @xmath107 .
formally , these 3d fourier series ( 8) may also be expressed as 2d fourier series with @xmath24 dependent coefficients @xmath108 and their derivatives @xmath109 .
the solution is thus completely determined by the two infinite sets of scalar fourier coefficients @xmath110 and @xmath111 , which are obtained by minimizing the total free energy with respect to these coefficients for given parameters @xmath0 and @xmath5 and film thickness @xmath10 . for
the computation i shall use a large but finite number of @xmath110 and @xmath111 in the sense of a ritz variational method .
the total free energy @xmath112 per unit volume of the infinite film is the sum of the gl free energy , eq .
( 2 ) , and the stray - field energy @xmath113 . in reduced units and referred to the state where @xmath114 and @xmath115 one has with @xmath116 : @xmath117 the factor of 2 comes from the two half spaces above and below the film , which contribute equally to @xmath113 .
the stray field @xmath118 with constant planar average @xmath119 is determined by the laplace equation @xmath120 ( since @xmath121 and @xmath122 in vacuum ) and by its perpendicular component at the film surface @xmath123 , since @xmath124 has to be continuous across this surface .
this yields with ( 8) the stray field : @xmath125 + \bar b\ , , \nonumber \\
\nonumber { \bf b}_\perp ( x , y , z \ge d/2 ) = \nonumber \\ \sum_{{\bf ~k}_\perp } b_{{\bf k}_\perp}^s \ !
{ { \bf k}_\perp \over k_\perp } \sin ( { \bf k}_\perp { \bf r}_\perp ) \exp [ -k_\perp ( z - d/2 ) ] \ , , \nonumber \\ b_{{\bf k}_\perp}^s \ ! = b_{{\bf k}_\perp}\ ! ( z = d/2 ) = \sum_l b_{\bf k } \cos(\pi l ) \ , .
\end{aligned}\ ] ] ( @xmath126 ) . for spatial averaging
we shall need the orthonormality relations valid for @xmath127 : @xmath128 @xmath129 averaging the squared stray field over @xmath130 and @xmath131 and using ( 11 ) , ( 12 ) one obtains terms @xmath132 $ ] , and thus @xmath113 in eq .
( 9 ) becomes @xmath133 the fourier coefficients @xmath110 and @xmath111 may be computed by iterating appropriate gl equations as shown in ref . .
minimizing @xmath36 , eq .
( 2 ) , with respect to @xmath45 and @xmath46 yields the two gl equations for bulk superconductors @xmath134 the first gl equation ( 14 ) applies also to our film ; the second gl equation ( 15 ) , too , but when it is written as an equation for the @xmath111 it has to be supplemented by a stray - field term @xmath135 on its r.h.s . , which originates from the boundary condition for @xmath52 .
a possible iteration equation for the @xmath110 is obtained from ( 14 ) using the relation @xmath136 to give @xmath137 with @xmath138 as above .
note that @xmath139 here means the 3d nabla operator , while the similar eq .
( 9 ) of ref . is 2d . to obtain better convergence of the iteration
i subtract a term @xmath140 on both sides of ( 16 ) , such that @xmath141 is replaced by @xmath142 ; this choice yields fastest convergence . using the ansatz ( 3 ) and the orthonormalities ( 11 ) , ( 12 ) we then obtain an iteration equation for the @xmath110 : @xmath143 where @xmath144 averages over @xmath21 .
in particular , if @xmath145 and @xmath46 do not depend on @xmath24 , eq .
( 17 ) reduces to eq . ( 11 ) of ref . and yields @xmath146 for all @xmath147 .
other forms of iteration equations for the @xmath110 are possible , e.g. one containing in the denominator @xmath148 instead of @xmath141 , but one should choose that which yields fastest convergence of the iteration .
in general , the iteration of any equation for some parameter @xmath149 given in the original form @xmath150 may be made more stable or faster converging by rewriting it in the form @xmath151 with some constant @xmath152 ( or even @xmath153 in some cases ) .
the convergence is accelerated by alternating the iteration step ( 17 ) with an iteration step that changes only the amplitude of @xmath145 but not its shape .
namely , from @xmath154 we obtain + @xmath155 similarly , an iteration equation for the @xmath111 is obtained from the equation @xmath156 by reordering the terms appropriately . from eq .
( 8)-(13 ) one has @xmath157 with @xmath158 from eq .
( 10 ) and @xmath159 + equating the sum of the terms ( 19)(21 ) to zero and adding and subtracting an appropriate term @xmath160 that improves the convergence ( with some constant @xmath161 or larger ) , one obtains an iteration equation for the @xmath111 : @xmath162 the solutions @xmath163 , @xmath164 , and @xmath165 are obtained iteratively by first finding the 2d solution as in ref . , keeping only the terms with @xmath73 and starting , e.g. , with @xmath166 and @xmath167 and then iterating the three equations ( 17 ) , ( 18 ) , ( 23 ) by turns a few times ; after this , the 3d solution is obtained by continuing this iteration with the terms for all @xmath168 included until the coefficients @xmath110 and @xmath111 do not change any more . with the empirical choice @xmath169 this iteration is stable for all @xmath170 , @xmath0 , and @xmath10 and the free energy decreases smoothly until it becomes stationary ( with accuracy @xmath171 ) after @xmath172 iteration steps .
= .981.0 = .981.0 = .981.0 = .981.0 = .981.0 = .981.0 = .981.0
figure 1 shows one example for the resulting magnetic field lines and some cross sections of @xmath173 and @xmath174 along @xmath130 in the plane @xmath175 at @xmath176 ( center plane of the film ) and @xmath177 ( film surface ) , for a film of finite thickness @xmath178 at reduced induction @xmath179 and gl parameter @xmath180 , yielding for the triangular vortex lattice a vortex spacing of @xmath181 . the left half of fig . 1 shows the field lines that result if the unchanged 2d bulk solutions for @xmath18 and @xmath182 are assumed inside the film .
the right half shows the correct solution , exhibiting smooth field lines across the film surface , and a very weak widening of the vortex core near the surface .
figure 2 shows the magnetic field lines for a film with thickness @xmath183 at @xmath184 for @xmath185 ( @xmath186 , left ) and @xmath187 ( @xmath188 , right ) , triangular lattice .
the solid lines are the stream lines of @xmath189 ; they have the correct slope of @xmath52 and start at equidistant points far away from the film surface , where @xmath190 constant ( in infinitely extended films the average induction @xmath191 equals the applied induction @xmath192 outside and inside the film ) , but their 1d density is not proportional to the 2d flux density @xmath193 . the dashed lines in fig .
2 are field lines that have approximately the slope of @xmath194 and have a density proportional to @xmath195 , see appendix a. this type of field lines is depicted also in figs . 1 and 3 . in fig .
3 the magnetic field lines are shown for films of various thicknesses @xmath196 , 0.2 , 0.1 , 0.05 for @xmath184 and @xmath197 as in fig . 1 , where @xmath198 . these thicknesses correspond to @xmath199 , 2 , 1 , 0.5 ( and 0.25 in fig . 1 ) . at low inductions @xmath200 and not too small @xmath201 , these field patterns may also be obtained by linear superposition of the fields of isolated london vortices given by eqs .
( 5)-(9 ) of ref . ,
with appropriately cut - off vortex core introduced to consider the finite coherence length @xmath202 .
this superposition also applies to nonperiodic vortex arrangements .
figures 4 and 5 show the two relative variances @xmath203 and @xmath204 of the magnetic induction defined here as @xmath205 ^ 2 \,\rangle_{x , y}^{1/2 } / \bar b \\ \sigma_\perp(z)&= & \langle\,b_x(x , y , z)^2\,+b_y(x , y , z)^2 \,\rangle_{x , y}^{1/2 } / \bar b \ , .
\end{aligned}\ ] ] these measures of the relative variation of the periodic induction depend on @xmath24 : deep inside thick films , @xmath206 reduces to the variance of the 2d vortex lattice in the bulk , @xmath207 , computed , e.g. , in ref . as function of @xmath170 and @xmath0 , and one
has @xmath208 since @xmath209 for the considered case .
as one approaches the surface from inside the film , @xmath206 decreases and @xmath210 increases until they coincide at the surface @xmath177 . outside the superconductor one
has exactly @xmath211 \ , .
\end{aligned}\ ] ] this follows from eqs . ( 10 ) for the magnetic stray field . at large @xmath212
, the variance decreases exponentially with @xmath24 , @xmath213 , where @xmath214 is the shortest reciprocal lattice vector of the vortex lattice defined below eq .
thus , @xmath206 decreases monotonically from its bulk value @xmath207 inside a thick film to zero far away from the film , reaching at the surface approximately half its bulk value ( for thick films ) .
in contrast to this , the transverse variance @xmath210 increases from zero at @xmath176 and reaches a sharp cusp - shaped maximum at the surface , where it joins @xmath206 and then decreases again to zero away from the film . for large @xmath215 and @xmath216 these curves are approximately symmetric , @xmath217
this is so since for @xmath218 the outer and inner magnetic stray fields are symmetric . shown in figs . ( 4 ) and ( 5 ) are both variances for various values of @xmath170 and @xmath0 for 8 film thicknesses @xmath219 = 0.1 , 0.2 , 0.4 , 0.6 , 0.8 , 1 , 1.2 , and 1.4 for the triangular vortex lattice ( solid lines ) and for @xmath220 , 0.4 , 0.8 , and 1.2 for the square vortex lattice ( dashed lines )
. the variances for these two lattice symmetries are almost identical .
figure 6 shows the variance of the order parameter @xmath221 inside films with periodic vortex lattice , @xmath222 at large reduced inductions @xmath223 in the middle of thick films @xmath224 coincides with the abrikosov parameter @xmath225 for the triangular vortex lattice and @xmath226 for the square lattice . at low inductions @xmath227
one has @xmath228 since the order parameter @xmath145 is constant except in the small vortex cores .
figure 6 shows that @xmath224 , and thus the shape of @xmath173 , changes very little with @xmath24 . for films of thickness @xmath229 ,
one has constant bulk @xmath230 in a central region around @xmath176 , and as @xmath24 approaches the surface of the film , @xmath224 increases by at most a factor of 1.0033 within a layer of thickness @xmath231 .
here @xmath232 is the vortex distance at @xmath233 .
one has @xmath234 for the triangular and @xmath235 for the square vortex lattice .
the maximum change occurs at @xmath236 and @xmath237 . for thinner films and larger and smaller @xmath170 ,
the variation of @xmath224 is even smaller .
thus , to a very good approximation , one may assume that the order parameter @xmath173 of the vortex lattice inside films in perpendicular magnetic field is independent of @xmath24 and for not too thin films has the same form as the 2d order parameter @xmath182 of the bulk vortex lattice .
for very thin films with @xmath238 at @xmath19 the vortex cores are slightly wider than in the bulk .
for example , at @xmath184 , @xmath197 , the core width increases by about 25 % when @xmath239 decreases from 0.5 to 0.005 , but then saturates and does not increase further in thinner films .
this is just the interval of @xmath10 in which the modulation @xmath240 of the periodic magnetic field @xmath241 decreases from @xmath242 to @xmath243 since the effective penetration depth @xmath244 becomes larger .
close to @xmath78 the constancy along @xmath24 of the gl function @xmath245 applies to thicker and thicker films .
this numerical result is consistent with the finding in ref . of a correlation length @xmath246 that diverges for @xmath247 and describes the extension along the vortex lines of perturbations in @xmath245 caused by small material inhomogeneities ( pins ) .
interestingly , a similar diverging length @xmath248 describes the long axis ( along @xmath24 ) of a cigar - shaped superconducting region ( nucleus ) that nucleates at applied fields above @xmath78 at a small defect with transition temperature @xmath249 higher than the bulk @xmath250 .
@xcite in superconducting films of thickness @xmath251 , or at applied fields satisfying @xmath252 , small inclusions or precipitates are thus expected to cause a virtually 2d perturbation that has no @xmath24-dependence .
next i consider the additional free energy caused by the presence of the two surfaces of the film .
this energy per unit area of the film is composed of two terms : \(a ) @xmath113 , the magnetic energy of the stray field outside the film , defined by eq .
( 9 ) and expressed in eq .
( 13 ) in terms of the fourier coefficients @xmath253 of the field component @xmath254 at the surface , \(b ) @xmath255 , the actual surface energy defined as the difference of the free energy of the film per unit area , @xmath256 , minus the 2d bulk free energy density of the infinite vortex lattice , @xmath257 , times @xmath10 , thus @xmath258 the total surface energy , originating from both surfaces of the film , is the sum of these two terms , @xmath259 .
both terms tend to a constant when the film thickness @xmath10 increases above the vortex spacing @xmath260 .
these thick - film values of @xmath113 and @xmath255 are of the same order , and they are approximately equal for large @xmath261 and also at large reduced inductions @xmath247 .
this is so since the order parameter @xmath173 in the film is nearly independent of @xmath24 , and thus @xmath255 is virtually only of magnetic origin , i.e. , it is the energy of the magnetic field change caused inside the film by the surface .
when the magnetic penetration depth is large , @xmath262 , this `` inner stray field '' is symmetric to the outer stray field .
this equality applies also near @xmath263 , since inside the superconductor the magnetic screening is reduced by the reduction of the order parameter and thus the effective penetration depth @xmath264 increases .
@xcite the dependence of @xmath113 and @xmath255 on the film thickness is shown in fig . 7 for @xmath185 and @xmath265 , 0.15 , and 0.4 . with increasing @xmath10 both energies increase from zero and
saturate to constant values at about @xmath216 for all @xmath0 and @xmath170 .
figure 8 shows these thick - film limits of @xmath113 and @xmath255 as functions of the reduced induction @xmath170 .
note that @xmath113 is slightly larger than @xmath255 ; this reflects the fact that the stray field inside the film is screened by @xmath266 , while outside the film there is no screening ( @xmath267 ) .
both @xmath113 and @xmath255 vanish at @xmath268 and at @xmath247 and have a maximum in between . at @xmath268
one has @xmath269 since each vortex contributes separately . at @xmath270 one
has @xmath271 ( in units @xmath272 ) since the amplitude of the periodic @xmath254 decreases as @xmath273 and the depth of the stray field is @xmath274 .
therefore , when plotted in units @xmath275 , all curves @xmath113 and all @xmath255 practically coincide for all @xmath276 and @xmath277 , see fig .
note that the total surface energy @xmath259 is very small , never exceeding the value @xmath278 reached at @xmath261 and @xmath279 .
= .981.0 = .981.0 = .981.0 = .981.0 = .981.0 finally , the elastic shear modulus @xmath2 of the vortex lattice will be discussed . as shown in ref . , the shear modulus of the triangular vortex lattice can be expressed with high accuracy by the difference of the free energies of the rectangular lattice , @xmath280 ( with @xmath281 and @xmath282 ) , and the triangular lattice , @xmath283 ( with @xmath284 and same @xmath282 ) , @xmath285 this is so since the free energy for constant unit cell height @xmath286 varies practically sinusoidally with @xmath287 : @xmath288 ( f_{\rm rect } -f_{\rm tr})/2 $ ] , thus the definition @xmath289 at small shear angle @xmath290 $ ] yields eq .
( 30).@xcite expressed in units @xmath34 , the shear modulus depends on three variables : @xmath291 .
there are thus several ways to present the numerical data for @xmath2 , each yielding different physical insight .
one result is that in the limit of small film thickness @xmath238 the shear modulus for a film with any @xmath0 tends to the bulk shear modulus at @xmath292 , as already obtained by conen and schmid .
@xcite this finding may be understood from the fact that in thin films the vortices are pearl vortices that have a long interaction range @xmath293 exceeding the london penetration depth @xmath11.@xcite this argument yields the correct limit @xmath294 for @xmath295 , but for @xmath296 the @xmath2 of films does not quantitatively coincide with the bulk @xmath2 for an effective @xmath297 , since @xmath2 is determined not only by the range but by the full shape of the interaction potential between vortices , which differs for abrikosov@xcite and pearl@xcite vortices . if this potential is @xmath298 with @xmath299 and the vortex density is @xmath300 , one has for a 2d triangular lattice with positions @xmath301 defined below eq .
( 3 ) : @xcite @xmath302 \,,\ ] ] see also eqs . ( 9 ) and ( 11 ) of ref . .
figures 9 and 10 show @xmath2 as a function of the reduced induction @xmath170 for different film thicknesses @xmath10 expressed in units of the gl coherence length @xmath202 , @xmath303 , @xmath69 , 32 , and for gl parameters @xmath304 , 1 , and 2 . for large @xmath305 ,
the curves @xmath306 for various thicknesses are very close to each other and to the bulk @xmath2 . in general , the curves for finite film thickness all fall between the two limiting cases @xmath307 coinciding with the bulk @xmath308 , and @xmath309 coinciding with the bulk @xmath310 .
this interval is very small for large @xmath0 and not too small @xmath170 since @xmath311 .
this means that for large @xmath305 the shear modulus is nearly the same for thin and thick films .
note that for the bulk and @xmath305 one has @xmath312 for @xmath313 ( see fig .
9 top ) ; this applies also to films .
for @xmath314 we confirm the finding of ref .
@xcite that @xmath315 for @xmath19 , but this law applies only to intermediate film thicknesses @xmath316 at @xmath317 . an interesting feature can be seen from fig .
the upper part shows the _ bulk _
@xmath318 for values @xmath319 to @xmath320 , i.e. , also for @xmath321 corresponding to type - i superconductors , in which the vortex lattice is energetically unfavorable in the bulk . for @xmath322
one finds _ negative _ @xmath323 .
this means the bulk rectangular and square vortex lattices @xcite have lower energy than the triangular lattice ( the meissner state without vortices has even lower energy in this case ) .
however , as can be seen in the lower plot for films with @xmath304 , in sufficiently thin type - i superconductor films the triangular vortex lattice can be stable ( i.e. , @xmath324 ) when the induction is sufficiently small .
this behavior was seen also in ref . .
the dependence of @xmath2 on the film thickness @xmath10 is visualized in fig .
11 by plotting the difference @xmath325 \cdot ( d/\xi ) \cdot \kappa\ ] ] ( an energy per unit area ) versus @xmath10 at various @xmath0 values for @xmath326 ( near the maximum of @xmath2 ) .
one can see that this function saturates when the film thickness exceeds a few coherence lengths @xmath202 . for all values of @xmath0 and @xmath327
one can fit these curves closely by @xmath328 \,.\ ] ] in fig .
11 ( at @xmath326 ) the length @xmath329 of the best fit accidentally coincides with @xmath202 , but in general @xmath329 depends on @xmath170 and is proportional to the vortex spacing @xmath260 : @xmath330 , thus @xmath331 , yielding @xmath332 at @xmath326 .
this saturation means that the additional rigidity of the vortex lattice caused by the film surfaces and measured by @xmath333 , becomes independent of @xmath10 in films thicker than a few coherence lengths , and thus one has @xmath334 . for thin films
with @xmath238 one has @xmath335 since @xmath336 is a constant independent of @xmath10 .
figure 12 visualizes the @xmath0 dependence of the shear modulus of the triangular lattice by plotting @xmath337\cdot\kappa^2 $ ] versus @xmath338 at @xmath339 , 0.3 , and 0.7 , for film thicknesses @xmath340 @xmath69 10 .
also shown are the limits @xmath341 ( the @xmath130-axis ) , proving that @xmath342 for any @xmath0 , and @xmath343 ( upper dashed line ) that coincides with @xmath344\cdot\kappa^2 $ ] .
these plots prove that the differences of any two @xmath2 values vanish as @xmath345 when @xmath292 .
one can see that this asymptotic law is a good approximation even for not so large @xmath346 , and at large @xmath347 it practically applies to all @xmath348 , and even for type - i superconductors with not too small @xmath349 . in fig .
13 the numerical @xmath350 is compared with the analytical result of conen and schmid , fig . 1 of ref . ,
valid at large inductions @xmath20 .
their result was obtained from an elegant expression derived by lasher @xcite for the free energy of films with vortex lattice of arbitrary symmetry at @xmath247 .
lasher @xcite implicitly used the fact that the magnetic stray field inside the film is not screened in this limit of @xmath247 .
= .981.0 for infinitely extended films the average magnetic induction @xmath5 equals the applied field , @xmath351 , and the magnetization @xmath352 is defined as the magnetic moment per unit volume of the film .
the demagnetization factor of infinite films is @xmath353 , and thus the effective lower critical field at which the first vortices penetrate is @xmath354 . for the magnetization of superconductors with general demagnetizing factor @xmath355 see , e.g. , refs . .
noting that the total free energy per volume @xmath112 , eq .
( 9 ) , equals the work done by the applied field , which may be written as @xmath356 , one obtains the relation @xmath357 .
figure 14 shows magnetization curves for films of various thicknesses @xmath239 = 0.1 , 1 , 3 , 10 , and @xmath320 for gl parameters @xmath304 , @xmath358 , 1 , and 1.5 .
for thick films ( @xmath359 ) at @xmath360 one has @xmath361 and thus @xmath362 ; for larger @xmath363 the thick film @xmath364 is concave ( has positive curvature ) ; and for smaller @xmath322 ( type i superconductors ) @xmath364 is convex ( has negative curvature ) and the vortex lattice is not stable . however , for sufficiently thin films , even when @xmath322 the curvature of @xmath364 can be positive and even may change sign at a certain value of @xmath365 .
note that the plotted curves @xmath364 for various @xmath239 cut each other approximately at @xmath366 .
for thick films the initial slope is @xmath367 for all @xmath0 , and @xmath368 .
this is so since when surface contributions may be disregarded , one has for superconducting ellipsoids @xmath369 at @xmath370 where @xmath371 .
the enhancement of @xmath364 at small @xmath372 for thin films in fig .
14 , originates from the energy of the magnetic stray field , which enhances the self energy of a vortex beyond its bulk value @xmath373 .
more details about this will be published elsewhere .
it is shown how the ginzburg - landau equations can be solved for a periodic lattice of abrikosov vortices in superconducting films in a perpendicular magnetic field . as illustration
how well this iteration method works , some results are presented .
the widening of the magnetic field lines as they exit the film surface is correctly obtained , figs . 1 , 2 , but this leads only to a very small correction of the order parameter near the surface , fig . 1 .
the variance of the transverse component of the magnetic induction is sharply peaked at the surface and vanishes deep inside and far outside the film , figs . 4 , 5 .
the variance of the periodic order parameter ( abrikosov parameter @xmath230 ) varies very little across the film thickness , by at most a factor 1.0033 , fig .
the surface energy saturates for large film thickness @xmath10 and vanishes linearly at small @xmath10 , fig . 7 . for not too thin films the surface energy originates mainly from the magnetic stray field , which comes in approximately equal parts from outside and inside the film , in particular for large @xmath0 or large @xmath170 , fig . 8 . for very thin films
the stray field energy may be disregarded and the very small surface energy comes mainly from the small modification of the order parameter , fig .
the shear modulus @xmath350 of the triangular vortex lattice in thin films approaches the @xmath374 of thick films ( bulk limit ) at @xmath292 , fig .
10 , the difference being proportional to @xmath345 , fig . 12 .
while the bulk @xmath375 is negative in type - i superconductors ( @xmath322 ) , the @xmath2 of sufficiently thin films can be positive and may change sign at some value of @xmath170 , fig .
the magnetization curves of films with @xmath322 may have positive or negative curvature , depending on the film thickness , fig .
more results will be published elsewhere .
extensions of this method to vortices with several flux quanta and to the periodic lattice of curved vortices in superconducting films in a tilted magnetic field are underway .
a practical question is how to plot the magnetic field lines of this 3d problem such that they have the correct orientation of @xmath174 and their 1d density ( reciprocal distance ) in the plotted plane is proportional to the magnitude @xmath376 .
a simple consideration shows that this is possible only for 2d planar problems , when the field lines coincide with the contour lines of the vector potential , e.g. , @xmath377 .
but for 3d magnetic fields , and even for cylindrically symmetric fields , such 2d plots of the field lines are not possible since the magnitude @xmath376 here is proportional to the 2d density of the 3d field lines , but not to the 1d density of the plotted 2d field lines . for our 3d problem of a thick film with a 2d periodic vortex lattice
we have two possibilities to plot field lines that approximately have the above mentioned properties . _
first method : _ one may use numerical programs that plot the field lines ( stream lines ) of the 2d planar field @xmath378 [ or any other planar cross section of @xmath174 ] starting from equidistant points ( @xmath380 ) far away from the film surface so that @xmath381 const . such field lines have the correct slope , but their density is only approximately proportional to @xmath376 .
_ second method : _ in this paper the 2d plots of the magnetic field lines show the contour lines of the function @xmath382 + which ranges from @xmath383 at @xmath384 ( vortex center ) to @xmath385 ( middle plane between two vortices ) and has a periodic derivative .
such field lines have a 1d density along @xmath130 proportional to @xmath386 , and a density perpendicular to these lines which is close to @xmath376 , since their orientation is close to the orientation of @xmath174 .
figure 2 shows that these two types of field lines are very similar .
in particular , the contour lines of @xmath387 , eq . ( 1 ) , have slopes that are close to the correct slope .
o. fritz , m. wlfert , h. j. hug , h. thomas , and h .- j .
gntherodt , phys .
b * 47 * , 384 ( 1993 ) .
d. yu . irz , v. n. ryzhov , and e. e. tareyeva , phys .
lett . a * 207 * , 374 ( 1995 ) .
wei and t .- j .
yang , jpn .
. phys . * 35 * , 5696 ( 1996 ) .
g. carneiro and e. h. brandt , phys .
b * 61 * , 6370 ( 2000 ) .
m. tinkham , phys .
rev . * 129 * , 2413 ( 1963 ) .
m. tinkham , rev .
phys . * 36 * , 268 ( 1964 ) .
k. maki , ann . phys .
( n.y . ) * 34 * , 363 ( 1965 ) .
g. lasher , phys . rev . * 154 * , 345 ( 1967 ) .
a. fetter and p. c. hohenberg , phys .
* 159 * , 330 ( 1967 ) . in eq .
( 5 ) of ref .
erroneously @xmath390 ( square lattice ) was written in place of @xmath280 .
this form of eq .
( 30 ) becomes numerically correct if the factor @xmath391 is replaced by 23.1 , i.e. , @xmath390 lies approximately in the middle between @xmath283 and @xmath280 . | the ginzburg - landau equations are solved for ideally periodic vortex lattices in superconducting films of arbitrary thickness in a perpendicular magnetic field .
the order parameter , current density , magnetic moment , and the 3-dimensional magnetic field inside and outside the film are obtained in the entire ranges of the applied magnetic field , ginzburg landau parameter @xmath0 , and film thickness .
the superconducting order parameter varies very little near the surface ( @xmath1 ) and the energy of the film surface is small .
the shear modulus @xmath2 of the triangular vortex lattice in thin films coincides with the bulk @xmath2 taken at large @xmath0 . in thin type - i superconductor films with @xmath3
, @xmath2 can be positive at low fields and negative at high fields .
the magnetization of thin films at small applied fields is enhanced beyond its bulk value @xmath4 due to the energy of the magnetic stray field . |
the discovery of high - t@xmath1 superconductors based on cuo - compounds @xcite has led to a large amount of theoretical work about the peculiar properties of these materials .
a major effort has focussed on the normal - state properties of these compounds .
this research was largely motivated and substantiated by experiments that revealed striking anomalies in the normal - state properties @xcite .
most prominent among these are the linear resistivity , a linear nmr - relaxation of the cu - spins , and a hall angle that behaves @xmath2 over a rather wide temperature region .
furthermore , the optical conductivity shows a drude peak with a width@xcite @xmath3 , consistent with the linearity of the resistivity , and a drude weight which grows linearly with doping , consistent with the notion of holes acting as the charge carriers .
in addition , there is a pronounced temperature and doping dependent mid - ir peak at frequencies above the drude peak .
it was argued from the beginning @xcite that most of these anomalous properties can be explained by two special features appearing simultaneously in these materials : ( i ) they are strongly correlated , i.e. their ( effective ) local coulomb interaction is comparable to or larger than the characteristic kinetic energy of the relevant carriers ; and ( ii ) they are highly anisotropic with the electrons being in principle confined to the cuo - planes characteristic for these compounds .
furthermore , it has been argued that the cuo planes can be accurately described by a planar single - band hubbard model@xcite .
the great interest in this class of materials has led to a number of new theoretical conjectures , that , although based on the assumption of strongly correlated carriers , focussed mainly on the 2d - character of the cuo - planes@xcite . in two earlier publications@xcite
, we found that several of these anomalies can be understood from a kondo - like effect in the infinite - dimensional hubbard model .
in particular , the density of states develops a sharp peak at the fermi surface as the temperature is lowered .
the development of this quasiparticle peak coincides with the screening of the effective local moments and anomalies in the transport .
for example , the resistivity of the model displays a distinct linear in @xmath4 behavior with a slope that increases in inverse proportion to the doping @xmath5 , consistent with experiment@xcite .
the nmr relaxation rate @xmath6 displays a pronounced linear in @xmath4 region , with a slope that increases with doping , also consistent with experiment@xcite .
finally , the qualitative features of the hall resistivity are consistent with experiment@xcite , including a hall angle that increases quadratically with the temperature .
recently , a sharp kondo - like peak at the fermi surface has also been seen in the two - dimensional hubbard model@xcite as the temperature is lowered .
_ this suggests that these anomalous normal - state properties are intrinsic to the hubbard model , independent of dimensionality ! _ in this contribution , we address the anomalous normal - state properties of the hubbard - model optical conductivity .
the optical conductivity @xmath7 is an important probe of the excitations of a strongly correlated system .
it measures the rate at which electron - hole pairs are created by photons of frequency @xmath8 . in a perfect ( translationally invariant ) metal ,
photons couple only to electron - hole pairs with vanishing momentum and energy ; @xmath9 is proportional to a dirac delta function [ @xmath10 with drude weight @xmath11 ( we set @xmath12 ) .
electron - electron correlations modify this picture at zero temperature : the charge and spin fluctuations induce a dynamic disorder to the lattice potential which reduces the free - carrier drude weight by the inverse of the quasiparticle renormalization factor @xmath13 ( @xmath14 ) and transfers the remaining spectral weight to a frequency - dependent component of @xmath9 reflecting the incoherent charge and spin fluctuations ; the total spectral weight is , however , conserved .
finite - temperature effects will broaden the zero - frequency delta function into a lorentzian and can modify both the quasiparticle renormalization and the higher - frequency excitations .
this simple picture is further modified when restriction is made to a single ( or finite number ) of electronic bands . in this case
, the total spectral weight can vary as a function of temperature or interaction strength because the projection onto a restricted basis set disregards all excitations to electronic bands that are higher in energy than those that have been kept in the model .
if the on - site coulomb repulsion @xmath15 is large enough , then the system will be a mott insulator at half - filling .
the single - particle density of states @xmath16 consists of two symmetric bands , separated by @xmath15 ( called the lower and upper hubbard bands ) with the fermi level lying in the middle of the gap .
the renormalization factor @xmath13 diverges since there are no quasiparticles at the fermi level .
the optical conductivity will consist of a charge - excitation peak centered at @xmath17 , whose width is the order of the bandwidth .
as the system is doped away from half filling , a quasiparticle resonance appears at the fermi energy as @xmath0 .
the weight of the quasiparticle peak increases with doping . in this case
, the optical conductivity will have a drude peak ( from the `` free '' quasiparticles at the fermi energy ) and a mid - ir band ( because of excitations between the lower hubbard band and the quasiparticle peak ) in addition to the charge - excitation peak .
the drude weight can be estimated as follows : when @xmath5 holes are doped into the half - filled band , the total electron concentration is @xmath18 , and the quasiparticle renormalization factor behaves roughly as @xmath19 , so the drude weight is expected to be @xmath20 which increases linearly for small and large doping , and is peaked near quarter filling @xmath21 . in addition , the weight of the mid - ir peak should also behave qualitatively like eq .
( [ eq : drude_approx ] ) because it involves excitations between the quasiparticle peak and the lower hubbard band , which also should grow as the density of holes in the half - filled band .
the `` hole - like '' nature of the charge excitations near half filling results from the strong coulomb renormalizations that create the mott insulator at exactly half filling . in this contribution
, we investigate the optical conductivity of the hubbard model in infinite dimensions .
our methodology is detailed in section ii , where we discuss the formalism and calculational techniques .
section iii contains our results , which include a drude peak whose width @xmath22 grows linearly with temperature , a mid - ir peak that becomes more visible at low temperatures and dopings , and a charge - excitation peak that decreases with doping and is weakly dependent upon @xmath4 .
comparison of our results with experiment is presented in section iv , and conclusions are given in section v.
motivated by the observations of anderson@xcite and zhang and rice@xcite we will study the single - band hubbard model@xcite in d - dimensions , with @xmath23 as a model for the cuo system .
our notation is the following : @xmath24 is an electron creation operator for an electron in a localized state at lattice site @xmath25 with spin @xmath26 ; @xmath27 is the corresponding electron number operator ; @xmath28 is the rescaled hopping matrix element ; and @xmath15 is the coulomb interaction strength . we choose @xmath29 as a convenient unit of energy .
this model may be solved in the limit of high dimensions , using the observations of metzner and vollhardt @xcite that with increasing coordination number of the underlying lattice the many - body renormalizations due to a two - particle interaction like the hubbard-@xmath15 in ( [ eq : hubmod ] ) become purely local .
the solution of the model ( [ eq : hubmod ] ) may be mapped onto the solution of purely local correlated system coupled to an effective ( self - consistently determined ) bath @xcite .
the quantum monte carlo ( qmc ) scheme based on the work of hirsch and fye @xcite has proven to be the most effective method for solving this strongly correlated local problem .
dynamical properties of the model are then obtained using numerical analytic continuation ( employing maximum - entropy techniques)@xcite .
this method requires a default model . to obtain the single - particle density of states we use the finite - u non - crossing approximation ( nca)@xcite result for the infinite - dimensional hubbard model as a default model at high temperatures , where it is essentially exact @xcite . at low temperatures , where the nca fails , we use the numerically continued result of the next higher temperature as a default model .
the posterior probability of the final result is employed to determine which default model should be used . generally , we find that the crossover temperature between using the nca default model and the higher - temperature continuation lies at @xmath30 , where @xmath31 is the kondo - like scale for this model@xcite
. once the density of states is determined , the self energy @xmath32 may then be found by ( numerically ) inverting the fadeev function @xmath33 in the relation @xmath34\right\}/\sqrt{\pi}\,.\ ] ] with the knowledge of the one - particle self energy , one can calculate transport quantities . for example , the optical conductivity can be calculated exactly in the local approximation .
it is given by the simple bubble only@xcite , whose evaluation leads to @xmath35 where @xmath36 is the lattice volume , @xmath37 and defines the unit of the conductivity , the spectral weight satisfies @xmath38 $ ] , and the noninteracting density of states is @xmath39 .
as the temperature is lowered , the hubbard model in infinite - d is found to always be a fermi liquid@xcite , except for the region of phase space where it is magnetic@xcite . a fermi liquid
is defined by a self - energy that has the following structure : @xmath40 with @xmath41 for temperatures @xmath42 the characteristic fermi temperature .
[ the fermi temperature @xmath31 decreases to zero as half filling is approached@xcite , and the fermi - liquid - theory form of eq .
( [ eq : flt ] ) still holds for moderate temperatures , with the only change being @xmath43 for @xmath44 .
] the spectral weight , then assumes the form @xmath45 with the fermi level defined by @xmath46 and @xmath47 denoting the ( rather structureless ) incoherent contributions to the spectral function .
the spectral function includes a delta function at zero temperature [ @xmath48 because the broadening @xmath49 vanishes in that limit . if the lorentzian form for the fermi - liquid - theory spectral function [ eq . ( [ eq : spec_fun_flt ] ) ] is substituted into the expression for the optical conductivity found in eq .
( [ eq : optcond ] ) , and the temperature satisfies @xmath50 , then the optical conductivity becomes @xmath51 with drude weight @xmath52 , relaxation time @xmath53 , and @xmath54 containing the contributions from the incoherent pieces of the spectrum . the noninteracting drude weight for the single - band model @xmath55 satisfies @xmath56 .
as the interaction @xmath15 is turned on in a single - band model , the total integrated spectral weight is not conserved but becomes @xmath15-dependent @xcite because spectral weight that would be shifted to higher bands is `` lost '' in any single - band model .
the total spectral weight satisfies @xmath57 where @xmath58 is the kinetic energy per site , divided by the number of lattice dimensions .
it should be stressed that the modification of the sum rule in eq .
( [ eq : sumrule ] ) will produce some systematic modifications to the behavior of the optical conductivity as a function of temperature , interaction strength , and doping .
these systematic effects must be kept in mind when one is comparing the results of a single - band calculation to experiment .
the drude weight @xmath59 for the interacting system , may also be determined ( at @xmath60 ) by extrapolation of the matsubara - frequency current - current correlation function using the method proposed by scalapino et al.@xcite ( @xmath59 measures the `` free '' quasiparticles in the system ) .
this method determines the drude weight of a metal by examining the asymptotic form of the current - current susceptibility in the x - direction , @xmath61 .
more specifically , @xmath59 is given by @xmath62\ , , \label{eq : scal_drude}\ ] ] where the limit @xmath0 is taken after first setting the momentum transfer to zero ( @xmath63 ) .
note that this latter method of determining the drude weight is a much better defined procedure than trying to fit the optical conductivity to the generic form of eq .
( [ eq : sigma_flt ] ) because of the uncertainty left in trying to fit @xmath54 .
we present here our results for the optical conductivity of the single - band hubbard model in infinite dimensions with @xmath64 .
figure [ optical_mir](a ) shows the optical conductivity obtained from eq .
( [ eq : optcond ] ) when @xmath65 for a variety of temperatures .
one finds the drude peak at @xmath66 developing with decreasing temperature .
in addition , a shoulder develops adjacent to the drude peak at @xmath67 which is strongly temperature dependent and clearly visible only for the lowest temperatures .
the last feature in @xmath68 is a roughly temperature - independent peak at @xmath17 . in order to compare our results to experiment ,
the three features in @xmath68 are fit to a lorentzian plus ( asymmetric ) harmonic - oscillator forms for the higher - energy peaks @xmath69 with @xmath70 the relaxation time of the quasiparticles , and the constants @xmath71 , @xmath72 , and @xmath73 ( @xmath74 , @xmath75 , and @xmath76 ) the weight , center , and width , respectively of the mid - ir ( charge - transfer ) peak .
the drude width @xmath22 obtained from this fitting procedure is plotted in the inset to fig .
[ optical_mir](b ) . note that for temperatures on the order of @xmath31 , it is well approximated by a straight line ( the line is a guide to the eye ) , while for @xmath42 the drude width must change its behavior to @xmath77 according to the general properties of a fermi liquid . by subtracting off the fit drude portion from the optical conductivity , we were able to isolate the mid - ir portion , as shown in fig .
[ optical_mir](b ) .
note that the `` double - peak '' structure emerging in the mid - ir peak at the lowest temperature is an artifact of the fitting procedure which is not perfect in extracting the parameters for the drude peak .
the mid - ir peak is temperature dependent , growing in size and moving to slightly lower frequencies as the temperature is lowered .
comparison of the drude weight @xmath59 determined by the fitting procedure in eq .
( [ eq : opt_cond_fit ] ) and the independent method of calculation in eq .
( [ eq : scal_drude ] ) produces only qualitative agreement .
we have tried more sophisticated fitting routines in which we fit the fermi - liquid - theory parameters ( @xmath13 , @xmath49 , and @xmath78 ) in the spectral function of eq .
( [ eq : spec_fun_flt ] ) and then determine the drude contribution to the optical conductivity by employing the full expression in eq .
( [ eq : optcond ] ) , and the fit improves , as does the comparison of the drude weight , but the quantitative agreement still has a systematic error on the order of @xmath79 ( where the fitting procedure always overestimates the drude weight ) , because of the simple form chosen for the mid - ir peak .
our conclusion is that the method of ref .
is superior to any _ ad hoc _ fitting procedure in determining the low - temperature drude weight .
note , however , that the charge - transfer peak is uniquely determined by the above fitting procedure ( in the sense that the results are independent of what fitting procedure is employed ) , indicating that the harmonic - oscillator form is a reasonable approximation to that peak .
the existence of the drude and charge - excitation peaks have been reported previously @xcite .
these results were obtained with the nca and were thus restricted to temperatures @xmath80 .
the mid - ir bump , however , becomes clearly visible only for temperatures below this scale .
we can presently resolve this feature because of the refined numerics employed in the qmc and the numerical analytic continuation at lower temperatures . in fig .
[ optical_fill ] , the optical conductivity is plotted as a function of doping for fixed temperature @xmath81 .
the different dopings are best identified by their decreasing charge - transfer peaks ( @xmath82 , 0.0928 , 0.1358 , 0.1878 , 0.2455 , 0.3 , 0.35 , 0.4 , and 0.45 ) .
the solid lines indicate the electron concentrations where the optical conductivity increases with doping in the lower - hubbard - band region of @xmath83 @xmath84 and the dotted lines are where @xmath68 decreases with doping in the same region @xmath85 .
one can see that the mid - ir peak is strongly doping dependent , being most distinct from the drude peak at low dopings , and merging with it as the doping and the width of the drude peak increases .
in the inset , we show that the drude weight @xmath59 increases linearly with @xmath5 .
this latter result shows that the `` free '' carriers in a doped mott insulator are holes in the half - filled band . as the doping is increased further , the drude weight eventually saturates , and then decreases with doping , as the character of the `` free '' carriers changes from being hole - like to being electron - like .
the optical conductivity clearly displays an isobestic point , where @xmath68 is independent of doping ( @xmath86 . the isobestic point marks the boundary between the regions where the weight of the optical conductivity increases as a function of doping @xmath87 , and the regions where the weight decreases with doping ( @xmath88 ) .
however , it is incorrect to assume that all of the spectral weight that lies below the isobestic point was transferred from above . an effective carrier number @xmath89 , defined by the integral of the optical conductivity @xmath90 is plotted in figure [ optical_int ] ( the normalization is chosen here to give the number of carriers for a model with a complete set of basis states in order to make contact with experiment ) .
the integrated optical conductivity increases rapidly with doping for small values of @xmath5 , but then increases more slowly , eventually saturating , and decreasing with doping ( at @xmath91 ) . once again
, solid lines denote the electron concentrations where @xmath92 increases with doping @xmath93 ) and the dotted line is where @xmath92 decreases with doping @xmath94 .
note that there is no frequency where the integrated spectral weight is independent of doping , which would be required for the above scenario ( of the drude and mid - ir weight coming entirely from the charge - transfer peak ) to occur .
in addition , we can examine how the sum rule for the total spectral weight , and for it s component pieces ( drude , mid - ir , and charge - transfer ) evolve as a function of doping at fixed low temperature ( in fact , the following analysis is an approximation to the zero - temperature behavior of the different components of the spectral weight ) .
this is shown in figure [ optical_weights ] .
the total spectral weight is found from eq .
( [ eq : sumrule ] ) ( which agrees with the integral of the optical conductivity to within 3% ) , the charge - transfer weight from the fitting procedure in eq .
( [ eq : opt_cond_fit ] ) , the drude weight from eq .
( [ eq : scal_drude ] ) , and the weight of the mid - ir peak is then determined by subtraction ( the charge - transfer weight is neglected for @xmath95 ) .
note how the mid - ir peak initially increases linearly with doping until @xmath96 where it saturates and then decreases as the system becomes less correlated , and how the system changes from a hole - like metal to an electron - like metal at about quarter filling .
in addition , one can easily see that the drude and mid - ir weights both grow quite rapidly as the system is doped away from half filling ( the initial growth rate of the mid - ir weight is about a factor of 2 faster than that of the drude weight ) .
this fast growth of the drude and mid - ir weights explains why the optical conductivity for @xmath97 increases so rapidly with doping as shown in fig .
[ optical_fill ] .
these results are qualitatively similar to those found in the strong - coupling limit of the hubbard model in one dimension@xcite , indicating that the optical properties of the hubbard model are not strongly dependent upon the dimensionality ( with the exception that there is no mid - ir peak in one dimension )
. one can also inquire into the temperature dependence of the drude , mid - ir , and charge - transfer peaks at fixed doping .
the analysis , in this case , relies on trying to fit the optical conductivity to the form in eq .
[ eq : opt_cond_fit ] , as a function of temperature .
such a procedure is ill - defined , because the drude and mid - ir peaks merge at high temperatures , and it is difficult to separate them into their component pieces , without explicit knowledge of the proper fitting forms .
however , the qualitative physics can be discussed : we find that as the temperature is raised , that spectral weight is lost from the system ( i. e. , the magnitude of the expectation value of the kinetic - energy operator decreases as @xmath4 increases ) , that the drude and charge - transfer weights slightly increase , and that the mid - ir weight is sharply reduced . in addition ,
spectral weight rapidly shifts to higher frequency , since the drude width ( which is easy to determine by the fitting procedure ) increases linearly with @xmath4 for temperatures larger than @xmath31 .
the origin of all of these features in the optical conductivity can be understood by studying the single - particle density of states , plotted as a function of both temperature ( a ) and doping ( b ) in fig .
[ optical_dos ] .
clearly the drude peak results from the quasiparticle band which develops at low temperatures . we have shown previously that while this peak is developing ( _ i .
e. _ in the region @xmath98 ) , the scattering rate ( as measured by the resistivity ) increases in proportion to the temperature@xcite .
thus , the drude width also increases in proportion to the temperature .
we attribute the mid - ir peak to excitations from the lower hubbard band to the quasiparticle band at the fermi energy .
note that the quasiparticle resonance is only sharp and distinct from the lower hubbard band at low @xmath4 and small @xmath5 .
thus , the mid - ir peak will only be visible in the optical conductivity at low @xmath4 and small @xmath5 as well .
on the other hand , the upper peak at @xmath99 is due to the charge excitations from both the lower hubbard band and the quasiparticle band to the upper hubbard band . since the upper hubbard band is distinct from the quasiparticle weight at the fermi surface , and
since the weight in the upper hubbard band does not change as significantly with doping , the upper peak in @xmath68 has a weaker temperature and doping dependence , but is expected to disappear as the system becomes uncorrelated in the low - density limit .
many experimental measurements have been made of the optical conductivity in doped cuprates .
kramers - kronig analysis is used to determine @xmath68 from reflectivity measurements in@xcite la@xmath100sr@xmath101cuo@xmath102 ( lsco ) , in@xcite yba@xmath103cu@xmath104o@xmath105 ( ybco ) , in@xcite bi@xmath103sr@xmath103cacu@xmath103o@xmath101 ( bscco ) , and in@xcite nd@xmath103cuo@xmath106 ( nco ) ; reviews have also been published@xcite .
photoinduced absorption is also employed to measure the optical conductivity in@xcite ybco and in both@xcite lsco and nco .
it is interesting to compare the theoretical calculation of the optical conductivity with these experimental results .
the experiments yield six trends for the cuprate superconductors : 1 ) the mid - ir peak maximum moves to lower frequency , and merges with the drude peak as the doping increases ; its spectral weight grows very rapidly with doping near half filling ; 2 ) at a fixed value of the doping , spectral weight rapidly moves to lower frequency as @xmath0 , but the total weight in the drude and mid - ir peaks remains approximately constant ; the width of the drude peak decreases linearly with @xmath4 ; 3 ) the insulating ( or undoped ) phase has a charge - transfer gap ; when doped , the optical conductivity initially increases within the gap region , but eventually saturates and then decreases with doping ; 4 ) there is an isobestic point or nearly isobestic behavior ( in that the optical conductivity is nearly independent of doping ) at a frequency that is approximately one half of the charge - transfer gap ; 5 ) the effective charge has a constant value with respect to doping near the charge - transfer edge ; and 6 ) more than one peak is observed within the mid - ir region .
most of these trends are reproduced by our theoretical model .
in particular , fig . [ optical_fill ] illustrates how the mid - ir peak moves to lower frequency and joins the drude peak ( 1 ) , how the optical conductivity initially increases with doping at low frequency , but then saturates and decreases ( 3 ) , and how there is an isobestic point ( 4 ) .
[ optical_mir ] shows how the spectral weight is transferred to lower frequencies as the temperature is lowered , but we find that the total ( drude plus mid - ir ) spectral weight does not remain constant as in ( 2 ) because the temperature dependence of the expectation value @xmath107 produces some temperature dependence to the total drude plus mid - ir spectral weight ( it actually decreases as @xmath4 increases ) .
[ optical_int ] does not display the trend of point ( 5 ) , possibly because the restriction to a single - band model reduces the optical conductivity at higher frequencies to such an extent that the effective charge must depend on the doping level for the theoretical model .
we do not see the multiple peaks in the mid - ir region of point ( 6 ) .
these experimental features are usually attributed@xcite to phonons or impurities present in the system , but judging from our results , the low - energy feature may also be due to excitations from the lower hubbard band to a dynamically generated quasi - particle band at the fermi energy ( which is generated by kondo - like screening of the moments ) .
naturally , the single - band model fits the experminental data much better at lower frequencies , where the single - band approximation is relevant , but fails in reproducing some of the higher - energy trends found in the cuprates , because of the neglected bands .
a comprehensive theory should include both the effects of strong electron correlation ( which reproduce the insulator at half filling , and give a hole - like character to the charge excitations near half filling ) with the effect of electron - phonon coupling ( which are necessary to explain similar mid - ir features in nonsuperconducting perovskites ) .
a quantum monte carlo and maximum entropy calculation of the optical conductivity of the infinite - dimensional hubbard model has been presented .
the mott - insulating character of the ground state at half filling drives many anomalous behaviors in the normal state near half filling that are similar to those observed in the cuprate superconductors .
in particular the system is always a fermi liquid away from half filling , but the fermi temperature vanishes , and the quasiparticle renormalization factor diverges as half filling is approached . as a result ,
the free carriers in the system initially have a hole - like character ( that changes to an electron - like character at approximately quarter filling ) .
the drude width for these carriers grows linearly with temperature for temperatures above @xmath31 , the drude weight grows linearly with doping , and there is a doping and temperature dependent mid - ir peak .
these anomalies arise naturally from the presence of a strongly temperature - dependent quasiparticle peak , whose origin is a kondo - like screening of the magnetic moments , and which appears to occur in the hubbard model for all dimensions greater than 1 .
the anomalous features in the experimentally measured optical conductivity for the cuprates are usually attributed to either polarons or impurities .
however , any purely polaronic theory has difficulty in explaining the magnetic insulating character of the ground state at half filling .
the hubbard model naturally describes such an insulating state , and appears to also describe many of the anomalous features present in the experimental data . in light of this fact ,
it is worthwhile to try to incorporate both the effects of strong electron correlation , and the electron - phonon interaction into a comprehensive theory for the normal state of the cuprate materials .
work along these lines is in progress .
we would like to acknowledge useful conversations with w. chung , j. keller , y. kim , d. scalapino , r. scalettar , d. tanner , and g. thomas .
this work was supported by the national science foundation grant number dmr-9107563 , the nato collaborative research grant number crg 931429 and through the nsf nyi program . in addition
, we would like to thank the ohio supercomputing center , and the physics department of the ohio state university for providing computer facilities .
j.g . bednorz and k.a .
mller , z. phys . * 64 * , 189 ( 1986 ) . for reviews of relevant experiments
see c. h. pennington and c. p. slichter , in _ physical properties of high temperature superconductors _ edited by d. m. ginsberg vol.2 ; n. p. ong ibid . ;
y. iye ibid .
3 . d.b . romero , c. d. porter , d. b. tanner , l. forro , d. mandrus , l. mihaly , g. l. carr , and g. p. williams , phys .
lett . * 68 * , 1590 ( 1992 ) ; see also l. forro , g. l. carr , g. p. williams , d. mandrus , and l. mihaly , phys .
65 * , 1941 ( 1990 ) . for a review
, please see t. timusk and d. b. tanner , in _ physical properties of high temperature superconductors _ , vol . 1 , edited by d. m. ginsberg ( world scientific , singapore , 1989 ) , pp . 339407 , g. a. thomas , in _ proceedings from the 39th scottish universities summer school in physics _ , edited by d.p .
tunstall , w. barford , and p. osborne ( adam hilger , new york , 1991 ) , pp . 169206 , and references contained therein .
anderson , _ frontiers and borderlines in many particle physics _ , proceedings of the international school of physics `` enrico fermi '' ( north holland , amsterdam , 1987 ) , p. 1 ; p.w .
anderson , science * 235 * , 1196 ( 1987 ) .
f. c. zhang and t. m. rice , phys .
b * 37 * , 3759 ( 1988 ) .
g. baskaran , p.w .
anderson , phys .
* b37 * , 850 ( 1988 ) . j. slyom , adv .
* 28 * , 201 ( 1979 ) .
anderson , physica c * 185 * , 11 ( 1991 ) .
varma , p.b .
littlewood , s. schmitt - rink , e. abrahams and a.e .
ruckenstein , phys .
* 63 * , 1996 ( 1989 ) . m. jarrell and th .
pruschke , phys .
* b49 * , 1458 ( 1993 ) ; th .
pruschke and m. jarrell , physica b * 199&200 * , 217(1994 ) .
christoph quitmann , ph.d .
thesis , aachen 1992 ( unpublished ) .
n. bulut , d. j. scalapino , and s. r. white , phys .
lett . * 72 * , 705 ( 1994 ) .
j. hubbard , proc . r. soc .
( london ) a * 276 * , 238 ( 1963 ) ; m.c .
gutzwiller , phys . rev
. lett .
* 10 * , 159 ( 1963 ) ; j. kanamori , prog .
* 30 * , 257 ( 1963 ) . w. metzner and d vollhardt , phys . rev
. lett . * 62 * , 324 ( 1989 ) .
u. brandt and c. mielsch , z. phys .
b * 75 * , 365 ( 1989 ) ; z. phys .
b * 79 * , 295 ( 1990 ) ; z. phys .
b * 82 * , 37 ( 1991 )
. v. jani , z. phys .
b * 83 * , 227 ( 1991 ) . c. kim , y. kuramoto and t. kasuya , j. phys .
japan * 59 * , 2414 ( 1990 ) .
v. jani and d. vollhardt , int .
j. mod .
phys . , 713 ( 1992 ) .
m. jarrell phys .
lett . * 69 * , 168 ( 1992 ) .
a. georges and g. kotliar , phys . rev . * b45 * , 6479 ( 1992 ) . j.e .
hirsch and r.m .
fye , phys .
* 56 * , 2521 ( 19 86 ) . m. jarrell , h. akhlaghpour , and thomas pruschke , _ quantum monte carlo methods in condensed matter physics _ , edited by m. suzuki , ( world scientific , 1993 ) .
gubernatis , m. jarrell , r.n .
silver , and d.s .
sivia , phys .
b * 44 * , 6011 ( 1991 ) ; and m. jarrell and j.e .
gubernatis , to appear .
h. keiter and j.c .
kimball , intern .
j. magnetism * 1 * , 233(1971 ) ; n.e .
bickers , d.l .
cox and j.w .
wilkins , phys .
b*36 * , 2036 ( 1987 ) ; th .
pruschke and n. grewe , z. phys .
b * 74 * , 439 ( 1989 ) . th .
pruschke , d.l .
cox and m. jarrell , phys .
, 3553 ( 1993 ) ; th .
pruschke , d.l .
cox and m. jarrell , europhys .
* 21 * , 593 ( 1993 ) .
a. khurana , phys .
64 * , 1990 ( 1990 ) .
j. k. freericks and m. jarrell , phys .
( to appear ) ; p. g. j. van dongen , ( submitted to phys . rev .
p. f. maldague , phys .
b * 16 * , 2437 ( 1977 ) .
scalapino , s. white and s. zhang , phys .
b , * 47 * , 7995 ( 1993 ) .
h. eskes and a. m. ole ' s , phys
lett . * 73 * , 1279 ( 1994 ) .
s. uchida , t. ido , h. takagi , t. arima , y. tokura , s. tajima , phys .
b * 43 * , 7942 ( 1991 ) ; s. uchida , j. phys . chem . solids , * 53 * , 1603 ( 1992 ) .
j. orenstein , g. a. thomas , a. j. millis , s. l. cooper , d. h. rapkine , t. timusk , l. f. schneemeyer , and j. v. waszczak , phys .
b * 42 * , 6342 ( 1990 ) .
romero , et al .
( unpublished ) .
g. a. thomas , d. h. rapkine , s. l. cooper , s - w .
cheong , and a. s. cooper , phys .
lett . * 67 * , 2906 ( 1991 ) ; g. a. thomas , d. h. rapkine , s. l. cooper , s - w .
cheong , a. s. cooper , l. f. schneemeyer , and j. v. waszczak , phys .
b * 45 * , 2474 ( 1992 ) .
s. uchida , h. takagi , y. tokura , s. koshihara , and t. arima , in _ strong correlation and superconductivity _ , edited by h. fukuyama , s. maekawa , and a. p. malozemoff ( springer , tokyo , 1989 ) pp .
194203 . g. yu , c. h. lee , d. mihailovic , a. j. heeger , c. fincher , n. herron , and e. m. mccarron , phys .
b * 48 * , 7545 ( 1993 ) .
kim , s .- w .
cheong , and z. fisk , phys . rev .
lett . , 2227 ( 1991 ) .
a. s. alexandrov , v. v. kabanov , and d. k. ray , physica c * 224 * , 247 ( 1994 ) . | a monte carlo - maximum entropy calculation of the optical conductivity of the infinite - dimensional hubbard model is presented .
we show that the optical conductivity displays the anomalies found in the cuprate superconductors , including a drude width which grows linearly with temperature , a drude weight which grows linearly with doping , and a temperature and doping - dependent mid - ir peak .
these anomalies arise as a consequence of the dynamical generation of a quasiparticle band at the fermi energy as @xmath0 , and are a generic property of the strongly correlated hubbard model in all dimensions greater than one . |
one of the very rare exactly solvable non - equilibrium systems is the totally asymmetric simple exclusion process ( tasep ) , see @xcite and references therein .
the model is defined on a 1d discrete lattice with @xmath0 sites that are either occupied by a single particle or empty . in the latter case
they can be thought of as occupied by a single hole . in its version with parallel update
the tasep is equivalent to a special case of the nagel - schreckenberg model for traffic flow @xcite . here
all bonds are updated simultaneously while in the original formulation , the random - sequential update , only one particle move can occur per infinitesimal time - step . in general , particles enter the lattice on site @xmath3 with probability @xmath4 and leave the lattice from site @xmath0 with probability @xmath5 . in between particles on sites
@xmath6 ( @xmath7 ) may move with probability @xmath8 if the target site @xmath9 is empty . for later use
we introduce also the symbol @xmath10 which denotes the probability that any of the possible moves is not executed .
note that for random - sequential update those probabilities are replaced by rates . in the thermodynamic limit @xmath11
there are three different phases : a low - density phase , a high - density phase and a maximum - current phase . here , the maximum - current phase is special , since all physical quantities become independent of the input and output probabilities @xmath4 and @xmath5 the system behaves as if @xmath12 .
the average density in the maximum - current phase is @xmath13 . in the limit @xmath14
only one move per time - step occurs and one recovers the well - studied tasep in continuous time .
+ while for random updating many results are known there is a definite lack of results for the parallel updating scheme @xcite .
one reason is that the structure of the exact solution @xcite , @xcite appeared to be much more involved .
based on a simplified formulation of the exact solution @xcite the present article tries to close this gap a little more since especially the parallel update is important for practical modeling of traffic @xcite . in this article
we calculate as the main result the weight for @xmath1 particles and @xmath15 holes in the open system . for continuous time
this quantitiy has been calculated by derrida et .
twenty years ago in @xcite .
it further has a natural interpretation as the normalization of a related system on a ring with one second - class particle .
this normalization is given by so - called narayana numbers @xcite .
we will derive the analogous results for the parallel case here . recently
a connection between equilibrium lattice walks and the normalization of the tasep in different variants has been established .
one of those interpretations involves so - called motzkin paths .
a motzkin path is a path defined on the triangular lattice .
it starts at @xmath16 and ends at @xmath17 with never going below the horizontal axis .
possible are the steps @xmath18 , where @xmath19 is an up - step , @xmath20 is a level step and @xmath21 is a down step .
let @xmath22 denote the set of those motzkin paths .
then their number @xmath23 is given by the @xmath24-th motzkin number .
the problem can also be formulated as a random walk in 1d that starts and ends at site 0 ( or 1 ) .
if the horizontal axis represents time , the two - dimensional walk corresponds to the trajectory of the walker .
a level step means that the walker has not moved during the time - step .
an up - step ( down - step ) means that the walker increases ( lowers ) its coordinate by one .
a brownian excursion @xcite , @xcite is a special motzkin path that never even touches the horizontal axis apart from the first and last vertex .
it can simply be constructed from a motzkin path by adding an up - step at the beginning and a down - step at the end @xcite . in this paper
we will find a motzkin path interpretation of the generating function for @xmath1 particles on an open lattice of @xmath0 sites .
for this interpretation we use a coloring of those different steps as it was considered in @xcite .
+ the paper is organized as follows .
first , we consider the generating function of the open tasep and show how it is related to the generating function of weighted motzkin paths . after taking the thermodynamic limit
we obtain expressions for the generating function of weighted motzkin paths at given length .
we calculate the tasep - generating function for the weight of configurations with given number of particles .
then we see how this function is interpreted in terms of a coloring of the corresponding motzkin path .
we write an expression for the weight that the open tasep system contains @xmath1 particles .
finally we introduce a second - class particle dynamics on the ring that has conserving dynamics and relate it to the one obtained for the tasep in the thermodynamic limit .
the matrix representation that derrida et .
@xcite presented for the known recursion relations of the tasep came as a very elegant and compact formulation of a non - equilibrium steady state . since then , this technique has become very successful to calculate many stationary properties for the tasep and related models , see @xcite for a recent review .
the authors of @xcite found that the stationary weight for a lattice configuration @xmath25 of the tasep with random - sequential update can be written as @xmath26 in this notation the matrix @xmath27 represents occupied sites ( @xmath28 ) and the matrix @xmath29 represents empty sites @xmath30 .
the boundary vectors @xmath31 and @xmath32 ensure that the resulting matrix product is reduced to a scalar @xmath33 .
those weights @xmath33 are stationary if the matrices and vectors involved satisfy the set of relations @xmath34 , @xmath35 and @xmath36 , now widely known as the dehp - algebra .
the matrices and vectors are half - infinite and there are different representations that have certain combinatorial interpretations @xcite .
note that @xmath37 is given by catalan numbers .
it was shown in @xcite for the case of a single second - class particle on the ring that the weight for a fixed number of particles @xmath1 and @xmath38 holes is given by @xmath39\langle 1|(xd+ye)^{l}|1 \rangle = \frac{1}{l+1}{l+1 \choose n}{l+1 \choose n+1}.\ ] ] here @xmath40 and @xmath41 are the fugacities for particles and holes respectively .
the combinatorial term on the right - hand side is known as the narayana numbers @xcite . among many other applications
they are known to count dyck paths those are motzkin paths without level steps . in @xcite this formula ( or more precisely the equivalent formula of @xcite ) was used for the tasep with open boundaries , too .
the reason why in both models the weights for @xmath1 particles are the same is that the dehp algebra holds for both models under the same ansatz . now to the parallel - update tasep .
the exact stationary state is known to be of a matrix form , too @xcite . from the findings in @xcite
it follows that the weight of a configuration in the maximum - current phase can be written as @xmath42 with @xmath10 . in this paper
we interpret @xmath43 as the number of particle - hole domain walls in the configuration @xmath44 .
this is one possible choice following @xcite .
again the vectors @xmath31 and @xmath32 reflect the boundaries and the matrices @xmath27 and @xmath29 represent particles and holes respectively but are different from those for the random - sequential case .
the operators obey the relations found by evans et .
al . in @xcite
@xmath45,\\ \label{dv } d|v\rangle & = & q |v\rangle\end{aligned}\ ] ] which generalize the dehp algebra .
it was shown in @xcite that the normalization can be expressed as @xmath46 in contrast to ( [ ra ] ) the term on the right - hand side is the @xmath47th narayana polynomial .
note that a possible representation is @xmath48 along with @xmath49 and @xmath50 .
this representation is a direct generalization of one known representation for the continuous - time case @xmath51 @xcite .
to calculate the weight for @xmath1 particles on a lattice of size @xmath0 , our aim is to study @xmath52 . for details of the following steps ,
the reader is referred to appendix [ prelim ] .
first , we note ( [ zs ] ) that @xmath53 is given by @xmath54 our intention is to study the generating function @xmath55 which follows from ( [ ab ] ) .
the formula for the generating function shows its two faces. on the one hand @xmath56 has an interpretation as contributing to the grand - canonical normalization @xmath57 for taseps of size @xmath0 ( in terms of the tasep fugacity @xmath58 ) and on the other hand @xmath56 will be interpreted as the generating function of motzkin paths of length @xmath24 with fugacity @xmath59 .
we could also take @xmath60 since the exponent of @xmath58 can already be worked out by replacing @xmath61 and @xmath62 .
however at this stage it will turn out useful .
the use of the explicit representation ( [ repres ] ) yields the tri - diagonal matrix @xmath63 this is the actual starting point of the problem with ( [ mxy ] ) being interpreted in terms of a weighted motzkin path with the transition matrix ( [ tildec ] ) with @xmath64 , @xmath65 and @xmath66 being the weights for level - steps , up - steps and down - steps , respectively .
related tri - diagonal matrices appear in the tasep with a second - class particle @xcite and in the partially asymmetric exclusion process @xcite for example .
for the first element of the @xmath24-th power of this matrix we find due to the tri - diagonal structure @xcite @xmath67}\frac{1}{r+1}{n \choose 2r}{2r \choose r}w_l^{n-2r}w_u^rw_d^r.\ ] ] with @xmath68 note that for random update ( [ mn ] ) turns into the well - known formula of @xcite and can be extracted to ( [ ra ] ) . hence ( [ mn ] ) is the total weight for all motzkin paths of length @xmath24 with weights ( [ lud ] ) .
from the first - return decomposition one finds @xcite that @xmath56 fulfills @xmath69 .
we see that the polynomials @xmath70 and @xmath71 only appear as a product @xmath72 .
this is because a path from height @xmath73 to @xmath73 has the same number of up- and down - steps .
hence up-/down - step pairs have weight @xmath74 and the generating function reads @xcite , @xcite in terms of the fugacity @xmath59 counting motzkin steps : @xmath75 in appendix [ eq ] , @xmath56 is calculated alternatively , following calculations in @xcite .
this is done without use of an explicit representation just by manipulation of the general matrix @xmath76 in ( [ mxy ] ) to calculate @xmath77 and finally taking the matrix element . the resulting expression which involves a function that appears in @xcite for several lattice paths
is shown to agree with ( [ mxy2 ] ) .
+ the radius of convergence of @xmath78 reads with ( [ mxy2 ] ) @xmath79 and the thermodynamic contribution of @xmath80 can be calculated by standard techniques @xcite . here , we find @xmath81 the average number of level steps fulfills @xmath82 .
this leads for @xmath24 large to @xmath83 and @xmath84 . in the following we consider paths at finite length
a realization of a path of length @xmath24 with @xmath85 up - down pairs has the total weight @xmath86 .
using the explicit expressions ( [ lud ] ) and expanding leads to the weight @xmath87 now working out the power of @xmath88 and inserting into ( [ mn ] ) leads lll .enumeration of the ( 2,2,2)-colored motzkin path with red ( left table ) and black steps ( right table ) . [
cols="^,^,^",options="header " , ] by comparison with ( [ ab ] ) to @xmath89 with @xmath90}\frac{1}{r+1}{n \choose 2r}{2r \choose r}\sum\limits_{s=0}^{r}\sum\limits_{t=0}^{r}{n-2r \choose c - s - t}{r\choose s}{r \choose t}(1+q)^{c - s - t}(qx+qy)^{n-2r - c+s+t}q^r x^{r - s } y^{r - t}.\ ] ] to apply this formula to the tasep , we first remark that the tasep system size @xmath0 obviously enters as @xmath91 .
the result for the tasep then reads ( [ mncompl ] ) @xmath92 where the @xmath93 are @xmath94 what we actually did in working out ( [ f ] ) is dividing the three weights @xmath95 , @xmath70 and @xmath71 in two terms each .
@xmath64 was split in @xmath96 and @xmath97 , @xmath65 in @xmath98 and @xmath99 and @xmath66 equivalently .
the following section shows how that appears naturally in coloring motzkin paths .
a @xmath2-colored motzkin path @xcite has up - steps , level - steps and down - steps in @xmath100 , @xmath6 and @xmath101 colors respectively .
splitting the weights ( [ lud ] ) into two parts suggests to use two colors which results in a @xmath102-colored motzkin path .
we distinguish each of the two distinct steps by the colors red and black , see table [ tab ] .
@xmath93 is the number of ( 2,2,2)-colored weighted motzkin paths with @xmath0 steps , @xmath103 of those being black steps .
further @xmath104 is the generating function of this ensemble of paths weighting each ensemble with @xmath105 ( a fugacity @xmath8 for each black step as well as a factor @xmath106 ) .
there is no need for @xmath58 to the explanation thus we set here @xmath60 .
+ to actually count the paths we have to split the contribution @xmath107 to red level steps further : a @xmath108-colored motzkin path is the most appropriate interpretation .
we take a continuous red line @xmath109 and a red dashed line @xmath110 as further coloring .
[ allowed ] shows the possible steps .
all red up - steps are drawn as continuous lines and all down - steps as a dashed line while black steps are drawn as dotted lines , for a better reading in black and white .
figure [ motzkin232 ] shows an example of a ( 2,3,2)-colored motzkin path of length @xmath111 ( see figure capture ) .
note that the path could also touch the horizontal axis .
+ let us consider the density of steps that are associated with weight @xmath40 .
those are continuous red level steps , red up - steps and black steps .
we can rescale the radius of convergence @xmath112 of @xmath113 to one , solve for @xmath40 and identify the physical root .
then we calculate the density of steps associated with @xmath40 , namely @xmath114 .
the resulting expression can be written as a series @xmath115.\ ] ] the case @xmath116 corresponds to @xmath117 and the density decreases monotonically to @xmath73 where @xmath118 .
the density takes the value @xmath13 for @xmath119 .
note that the right - hand side is equal to ( [ zsimeq ] ) and equals the asymptotic value of the lattice fugacity @xcite , @xcite .
the function @xmath93 can be interpreted with ( [ mncompl ] ) and ( [ zs ] ) as contributing to the weight for configurations of length @xmath0 with at least @xmath103 particle - hole domain - walls .
choosing @xmath103 of @xmath120 available particle - hole domain - walls happens naturally at every update step , where a fraction @xmath103 of @xmath120 bonds adds to the flow .
the @xmath108 colored motzkin path in figure [ motzkin232 ] has @xmath121 and is one realization of a tasep configuration with @xmath122 since the red peak at @xmath123 contributes a domain wall , too .
note that in the tasep the case @xmath124 corresponds to the @xmath73 order in @xmath8 .
then one has @xmath125 and the continuous - time result is recovered .
+ we continue by working out the weight @xmath126 for @xmath1 particles and @xmath15 holes in tasep that reads due to ( [ recm ] ) @xmath127 the probability that the tasep contains @xmath1 particles then is prob@xmath128 . here
, the denominator is the normalization constant ( [ norm ] ) .
first one shall expand @xmath129 .
the weight of a path with @xmath103 black steps and @xmath85 up - down pairs of a total of @xmath130 steps then reads @xmath131 now introducing the particle number @xmath1 and picking out the coefficient of @xmath132 yields @xmath133 then one obtains @xmath134 note that one has due to the particle - hole symmetry that @xmath135 .
further the @xmath136 entering the normalization ( [ norm ] ) read @xmath137 .
in @xcite it was argued that second - class particle dynamics in the parallel tasep are not as natural to define as for the generic random update . in this section
a model system is presented that mimics second - class particle dynamics and for which preliminary results were published in @xcite .
we change the boundary conditions and consider a closed periodic chain with one of the particles carrying a cargo .
the unit of the particle and the cargo is referred to as the second - class particle . under the parallel update every particle ( with or without cargo )
moves forward with probability @xmath8 if the target site is empty . in the most simple case the cargo
is carried to the next site .
however if the site behind it is occupied , the cargo can actively jump onto this particle which happens independently with probability @xmath8 .
thus the cargo can jump to the left while its carrier moves at the same time to the right .
this dynamics reminds a bit of someone standing on a train either standing still and moving with the train or actively jumping to the wagon behind .
while in the continuous - time case a move corresponds to an interchange of occupation numbers @xmath138 obviously the situation here is more complex .
one has the transitions @xmath139 with @xmath140 being either @xmath73 or @xmath3 . in the thermodynamic limit
the generating function @xmath141 for the second - class particle process reads @xmath142 the expression for @xmath56 is the corresponding one for the open - boundary tasep and @xmath143 is defined in ( [ def_z ] ) , see appendix [ appc ] . .
the corresponding tasep size is @xmath144.,height=170 ] in the thermodynamic limit , only the denominator @xmath145 of @xmath56 in ( [ simeq ] ) is physically relevant .
the other factors contribute just a finite number of excursions to the corresponding motzkin path @xcite .
therefore the second - class particle process has the same thermodynamic physics as the open tasep .
note that for random update both expressions for @xmath56 and @xmath141 are even the same .
this is due to the fact that for random update the weights for @xmath1 particles on a lattice with @xmath0 sites are the same for open boundaries and a second - class particle on the ring .
+ reference @xcite shows the equivalence of this process and the one studied in @xcite .
it focuses on the thermodynamic limit and simplifies some results of @xcite .
the velocity of the 2nd - class particle is shown to be calculated through @xcite @xmath146 here @xmath147 and @xmath148 are the densities directly behind and in front of the 2nd - class particle @xmath149 .
the dynamics of the 2nd - class particle is obtained from ( [ rel1]-[rel6 ] ) .
it moves either forward if it has a hole in front while at the same time there is no particle directly behind that simultaneously catches the cargo : first term in ( [ 2nd_v1 ] ) . or
it moves backwards if it has a particle behind : second term in ( [ 2nd_v1 ] ) . with the help of appendix [ asy2 ]
one finds @xcite that the neighboring densities are @xmath150 which yields @xmath151 hence the velocity of the 2nd - class particle vanishes only at half filling @xmath152 .
note that this has the form of a group velocity thus the 2nd - class particle is travelling with the velocity of the density disturbance .
this feature should of course be ensured by a 2nd - class particle dynamics and it is a further underpinning of the fact that the model is able to describe the tasep properly on hydrodynamic scale .
( while its carrier either moves to the right ( @xmath153 ) or not ( @xmath154 ) ) ; if not , it stays on its carrier with probability @xmath155 which means that it either moves to the right with its carrier ( @xmath154 ) or stays with it on the current site ( @xmath156).,height=75 ] to conclude , we have studied the parallel tasep with open boundaries where every particle move occurs with probability @xmath8 .
this case defines the physics of the maximum - current phase for all @xmath157 . the main result is the generating function of arbitrary particle numbers @xmath1 in a system of size @xmath0 for parallel update .
we established an analogy to weighted motzkin paths and derived an equivalence to ( @xmath158)-colored motzkin paths appearing in @xcite .
already known interpretations of tasep normalizations with colored motzkin paths @xcite , @xcite for random - sequential update correspond in this terminology to @xmath159-colored paths , i.e. where only level steps appear in two colors . here
, we found the most natural interpretation as @xmath108-colored paths .
the reason for this increase in the necessary number of colours is the complexity of the parallel update .
it is known , that driven - diffusive systems under parallel dynamics typically have a quartic matrix algebra @xcite , @xcite , @xcite rather than a quadratic one as under random update , ordered - sequential or sublattice parallel update @xcite .
in fact , we could alternatively use @xmath160-colored paths for the parallel update so that in contrast to the other updates the number of colours is doubled .
this nicely reflects the increase from a quadratic to a quartic algebra .
it is expected that the structure found here also holds for similar models with parallel update while the other possible updates mentioned above lead to a ( @xmath161)-colored path . in the @xmath108 motzkin path ,
red steps are associated with particles or holes while black steps correspond to particle - hole pairs .
therefore paths with the same length do not necessarily correspond to tasep configurations of the same system size .
one could alternatively represent particle - hole pairs by a succession of two steps , see also motzkin paths with higher rank considered in @xcite .
however those paths would not end at height @xmath73 , in general .
the fact that a fraction of the particle - hole domain walls corresponds to motzkin steps in a different colour somewhat reflects the fact that under parallel update such a fraction contributes to the flow . in this paper
it was shown how known lattice - path interpretations for the parallel tasep @xcite , @xcite are mapped onto the present one that keeps information of the particle number . closely related to those problems concerning a given number of particles
is the second - class particle process @xcite . for parallel update
such a process is not as straightforward to define @xcite , though .
here we presented one possible choice of second - class particle dynamics in terms of a cargo process @xcite for which the generating function shares the thermodynamics with the generating function that has been derived here for the tasep .
+ since the parallel tasep is equivalent to the nagel - schreckenberg model with maximal velocity 1 , the results also apply to a bottleneck situation in traffic flow .
the average density is @xmath13 , however , the density fluctuates around this value . with the help of the generating function obtained in this paper one finds the probability that the road section contains any density of cars .
further investigations shall consider the fluctuations of the particle number as in @xcite using the findings of the present paper .
this is planned to be done in the near future
. additionally it would be interesting to compare the joint current - density distribution with the one for random - sequential update @xcite . to finally generalize present results to arbitrary @xmath4 and
@xmath5 one proceeds as in @xcite : denote by @xmath162 the sum of all possible products of @xmath1 @xmath27-matrices and @xmath15 @xmath29-matrices ( with fugacities @xmath40 and @xmath41 ) in which particle - hole domain - walls are weighted by @xmath163 .
one observes that @xmath162 obeys the recursion @xmath164 . in order to obtain the weight for @xmath1 particles and @xmath15 holes one considers @xmath165 and obtains a recursion that can be solved .
once having @xmath165 one uses the representation of the boundary vectors @xmath166 , @xmath167 to find @xmath168 .
the result is no longer a motzkin path but a sum over lattice paths .
walks that start at height @xmath24 and end at height @xmath169 each contain an additional factor @xmath170 .
in @xcite a slightly different model for fixed particle number was considered but results can be used here .
application of those results ( compare equation ( a.2 ) in @xcite ) to the tasep yields @xmath171 note that the @xmath172 recover the @xmath136 appearing in ( [ norm ] ) for @xmath173 and therewith @xmath174 recovers @xmath175 .
the matrices @xmath176 are defined recursively through the matrices @xmath177 and @xmath178 , namely @xmath179 with the initial condition @xmath180 and @xmath181 .
iterating this equation yields @xmath182 eventually taking the matrix element one is drawn to @xmath183 where we defined @xmath184 the term in the parenthesis denotes the sum over all matrix products of @xmath177 and @xmath185 with exactly @xmath103 @xmath186s and @xmath187 @xmath188s so that @xmath124 recovers @xmath189 .
the generating function becomes with ( [ mncompl ] ) @xmath190 to obtain ( [ mxy ] ) , ( compare equation ( a.2 ) in @xcite ) one changes the summation . instead of summing ( [ gliste ] ) line - wise and taking the matrix element as in ( [ mf ] ) , one sums diagonal - wise , so to speak : multiply ( [ gliste ] ) by @xmath191 and sort the results depending on the lengths of words in @xmath188 and @xmath186 that occur to arrive at the matrix element @xmath192
in @xcite it was argued that the second - class particle process can be mapped onto the ( single species ) traffic model in @xcite which is defined by the transitions @xmath193 and @xmath194 .
we repeat shortly the arguments here .
if the total number of holes in the traffic model is an even number , then in the stationary state every particle @xmath59 has an even number of holes @xmath195 in front .
thus by mapping @xmath196 one has @xmath197 and the usual tasep is recovered . if the total number of holes is an odd number , the dynamics tries to do the same , however one odd - valued headway remains .
the stationary configurations are of the form @xmath198 .
if one changes the numbering of particles after every update , one is able to achieve that the first particle always has the single excess hole behind .
now one defines the second - class particle @xmath149 as the unit @xmath199 , i.e. the excess hole plus the following particle .
the rules ( [ rel1]-[rel6 ] ) then are just a consequence of the simultaneous updating .
the normalization found in @xcite is @xmath200\end{aligned}\ ] ] which simplifies for @xmath201 to @xmath202\end{aligned}\ ] ] with @xmath203 and @xmath204 is one of the roots of the equation @xmath205 note that rewriting this expression as @xmath206 shows that @xmath143 is in fact the generating function for the number of weighted - peak walks in @xcite .
those are excursions on the rotated square lattice with a fugacity @xmath207 for each peak and a fugacity @xmath208 for each up - down pair so that in total @xmath208 is conjugate to the length of the excursion .
the parameter @xmath208 in ( [ def_z ] ) is in fact a function @xmath209 : @xmath210}\ ] ] that is rewritten with ( [ lud ] ) as @xmath211 in the thermodynamic limit the process is dominated by the square - root singularity in @xmath143 .
this can be written as a condition for @xmath208 : @xmath212 the asymptotics of the model can be obtained from the grand - canonical partition function @xmath213 in ( [ zasymp ] ) , @xmath40 and @xmath58 are the thermodynamic particle and lattice fugacities that were first found in @xcite . they can further be simplified to @xcite @xmath214 ^ 2},\end{aligned}\ ] ] with the density @xmath215 in the second - class particle process and @xmath216 the current - density relation .
the identity of ( [ 2nd_fug ] ) and the expressions given in @xcite can be verified most easily using a computer algebra system .
[ eq ] in this appendix the generating functions @xmath141 and @xmath56 of the second - class particle process and the tasep respectively are derived that share the same thermodynamics . for simplicity
we take @xmath60 without loss of generality as explained in the main text . to calculate @xmath217
the idea in @xcite is to transform the matrices similar to @xcite .
define @xmath218 one can check that this transformation leaves the algebra invariant , i.e. @xmath219 in this notation @xmath220 becomes finally @xmath221 it can be checked that the motzkin - path transfer matrix @xmath76 is thus changed in a way so that ( while level steps have same weight @xmath222 ) one finds for the weights of up - steps and down - steps @xmath223 and @xmath224 , respectively so that @xmath225 .
one now executes the sum in ( [ f ] ) to find with @xmath208 from ( [ zw ] ) that @xmath226 one proceeds by using @xmath227 . with the help of @xmath143 being determined by ( [ peak ] )
one writes @xmath228 so that @xmath229 using ( [ deprimed ] ) the matrix element of @xmath77 can be taken that recovers @xmath56 : @xmath230 which can be written with ( [ def_z ] ) and ( [ zw ] ) as @xmath231 and finally as @xmath232 this is compared to the second - class particle process .
its algebra differs slightly from the tasep .
instead of @xmath233 one has @xcite @xmath234 and @xmath235 .
therefore we find from ( [ f_end ] ) a slightly different result that can be written as @xmath236 now using ( [ peak ] ) and the asymptotic expression ( [ zsimeq ] ) gives ( [ simeq ] ) .
we now prove the identity of ( [ mxy2 ] ) and ( [ erzmshort ] ) involving the function @xmath237 that appears in @xcite in several lattice - path interpretations for the parallel tasep , see appendix [ asy2 ] .
first , note that ( [ peak ] ) can be rearranged to @xmath238 . now solving ( [ erzmshort ] ) for @xmath143 and inserting the resulting expression yields @xmath239 with @xmath240 .
finally using ( [ zw ] ) leads to @xmath241 which completes the proof .
4 k mallick : some exact results for the exclusion process .
mech . , p01024 ( 2011 ) nagel , k. , schreckenberg , m. : a cellular automaton model for freeway traffic.j .
i france 2 , 22212229 ( 1992 ) b derrida , s a janowsky , j l lebowitz , and e r speer : exact solution of the asymmetric exclusion process : shock profiles .
73 , 5/6 , 813843 ( 1993 ) r a blythe , w janke , d a johnston , and r kenna : dyck paths , motzkin paths and traffic jams
. j. stat .
mech . , p10007 ( 2004 ) r a blythe , w janke , d a johnston , and r kenna : continued fractions and the partially asymmetric exclusion process .
a 42 325002325023 ( 2009 ) mansour , t. , schork , m and sun , y. : motzkin numbers of higher rank : generating function and explicit expression .
journal of integer sequences 10 article 07.7.4 , 11pp .
( 2007 ) b derrida , m r evans , v hakim , and v pasquier : exact solution of a 1d asymmetric exclusion model using a matrix formulation .
j. phys . a 26 , 1493 - 1517 ( 1993 ) blythe , r.a . , evans , m.r .
: nonequilibrium steady states of matrix product form : a solver s guide .
a 40 , r333 ( 2007 ) r brak and j essam : asymmetric exclusion model and weighted lattice paths .
a 37 , 4183 4217 ( 2004 ) b derrida , c enaud , and j l lebowitz : the asymmetric exclusion process and brownian excursions .
115 , 365 - 382 ( 2004 ) m r evans , n rajewsky , and e r speer : exact solution of a cellular automaton for traffic . j. stat .
95 , 4596 ( 1999 ) j de gier and b nienhuis : exact stationary state for an asymmetric exclusion process with fully parallel dynamics .
e 59 , 48994911 ( 1999 ) m woelki and m schreckenberg : exact matrix - product states for parallel dynamics : open boundaries and excess mass on the ring . j. stat .
mech p05014 ( 2009 ) e duchi and g schaeffer : a combinatorial approach to jumping particles : the parallel tasep .
random structures and algorithms , 33 no .
4 , 434451 ( 2008 ) arita , c. and yanagisawa d. exclusive queuing process with discrete time . j. stat .
phys . 141 : 829 - 847 ( 2010 ) m woelki : phase transitions in cellular automata for cargo transport and kinetically constrained traffic . cellular automata lecture notes in computer science vol
. 6350 , pp 637 - 645 ( 2010 ) mallick , k. shocks in the asymmetry exclusion model with an impurity .
j. phys . a 29 , 53755386 ( 1996 ) | in this paper the totally asymmetric exclusion process ( tasep ) with parallel update on an open lattice of size @xmath0 is considered in the maximum - current region .
a formal expression for the generating function for the weight of configurations with @xmath1 particles is given .
further an interpretation in terms of @xmath2-colored weighted motzkin paths is presented . using previous results ( woelki and schreckenberg 2009 _ j. stat .
mech _
p05014 , woelki 2010 _ cellular automata _ , pp 637 - 645 ) the generating function is compared with the one for a possible 2nd - class particle dynamics for the parallel tasep .
it is shown that both become physically equivalent in the thermodynamic limit . |
it has been a longstanding goal in semiconductor spintronics to inject , transport , manipulate , and detect spin - polarized carriers in silicon - based devices.@xcite despite great success in the field over the past ten years using other semiconductors such as gaas,@xcite the goal of achieving the same with si has been reached only recently , using all - electrical hot - electron methods with undoped single - crystal silicon transport layers.@xcite later , spin injection into silicon was realized as well in an epitaxially - grown silicon n - i - p diode structure using circular polarization analysis of weak electroluminescence spectra for spin detection across a transport layer of 80 - 140nm.@xcite although our previous studies demonstrate electron spin manipulation in undoped silicon - even over a very long distance ( 350 microns)@xcite - it is necessary to investigate magnetic- and electric - field control of electron spin in _ doped _
silicon for integration of spintronics into present - day silicon - based microelectronic technology , where impurity doping plays a critical role . in this report
, we present spin injection , transport and detection in an n - type doped silicon device using our all - electrical methods .
unlike previous studies with undoped si , the presence of ionized impurities in the depletion regions of these doped transport layers gives rise to conduction band bending that for sufficient biasing conditions confines injected electrons for long dwell times . by modeling transport with drift and diffusion in the inhomogeneous electric fields provided by the band bending with a monte - carlo method , we simulate both spin precession and spin decay , showing that the transit time distribution of spin - polarized electrons can be controlled over a very wide range with an applied voltage , and can yield a measurement of spin lifetime .
fig . [ fig1 ] illustrates the structure of our device .
fabrication consists of ultra - high vacuum metal film wafer bonding to assemble a semiconductor - metal - semiconductor hot - electron spin detector ; a silicon - on insulator ( soi ) wafer including a 3.3@xmath0 m single - crystal ( 100 ) nominally 1 - 20 @xmath1 cm phosphorus - doped n - type silicon spin transport layer is bonded to an n - type bulk silicon collector wafer with a ni@xmath2fe@xmath3 ( 4nm)/ cu ( 4 nm ) bilayer .
conventional wet - etching techniques expose the soi device layer , onto which a ferromagnetic - emitter tunnel junction hot - electron spin injector is built .
the final device structure is al ( 40nm)/co@xmath4fe@xmath5 ( 10nm)/al@xmath6o@xmath7/al ( 5nm)/cu ( 5nm)/n - si ( 3.3 @xmath0m)/ni@xmath2fe@xmath3 ( 4nm)/cu ( 4nm)/n - si substrate , as displayed in fig .
1 . further details on fabrication of similar devices can be found in previous reports@xcite .
an applied emitter voltage @xmath8 on the tunnel junction ( larger than the cu / n - si injector schottky barrier ) injects hot electrons tunneling from the ferromagnetic co@xmath4fe@xmath5 cathode through the thin - film al / cu anode base and into the doped silicon transport layer conduction band . the first collector voltage ( @xmath9 )
controls the voltage drop across the transport layer and modifies the spatially nonlinear conduction band potential energy .
electrons escaping the transport layer are ejected over a schottky barrier at the detector side into hot - electron states in a buried ni@xmath2fe@xmath3 thin film .
the final spin polarization is detected by measuring the ballistic component of this hot electron current ( second collector current , @xmath10 ) in the n - type si wafer below ; spin - dependent scattering in the ferromagnetic ni@xmath2fe@xmath3 makes this current dependent on the projection of final spin angle on the ni@xmath2fe@xmath3 detector magnetization .
the spin - detection current @xmath10 was first measured with an external magnetic field parallel to the device plane .
a spin - valve effect , resulting from the different in - plane coercive fields of injector and detector ferromagnetic layers , is displayed in fig .
the measurements were done with @xmath8 = -1.6v applied , using different values of @xmath9 between 4.5v and 8v at temperature @xmath11 = 152k .
because of the @xmath10-@xmath9 dependence , we normalize the data for comparison between different @xmath9 values . after this normalization
, it can be seen that the measurement is only weakly dependent on accelerating voltage @xmath9 over this range .
the in - plane magnetic field was swept between -4 koe to + 4 koe for this measurement .
since the coercive fields of both ferromagnetic ( fm ) layers are smaller than 200 oe , the data obtained from the @xmath9 = 5v measurement is magnified in the inset of fig . 2 and the field sweep direction is specified by correspondingly colored arrows . when the in - plane magnetic field reaches approximately + 20 oe from the negative saturation field ( below -300 oe ) , the ni@xmath2fe@xmath3 layer switches its magnetization , causing an anti - parallel ( ap ) configuration in the two fm layers , which lowers the @xmath10 current relative to a parallel ( p ) configuration , because in this case spin `` up '' is injected , but spin `` down '' is detected .
if the magnetic field increases further , the co@xmath4fe@xmath5 layer reverses magnetization , resulting in a p configuration and restoration of the higher @xmath10 .
this happens as well in the opposite sweeping field direction due to the symmetric but hysteretic coercive fields of each fm layer .
the magnetocurrent ( mc ) ratio ( @xmath10@xmath12@xmath10@xmath13)/@xmath10@xmath13 calculated from the spin - valve plot , where the superscripts refer to p and ap magnetization configurations in the two fm layers , is approximately 6% .
as the magnetic field reaches up to @xmath144 koe after the magnetization reversal of both fm layers , @xmath10 monotonically rises because of domain magnetization saturation in the direction of the external field .
= -1.6v , with different @xmath9 values applied as indicated in the plot .
inset : data measured with @xmath9 = 5v plotted over a smaller field range .
field sweep directions are indicated by red ( increasing ) and blue ( decreasing ) arrows.,width=302,height=264 ] to unambiguously confirm spin transport through the doped silicon layer , we have performed measurements of @xmath10 in an external magnetic field perpendicular to the device plane , which allows us to examine spin precession and dephasing ( hanle effect ) during transport.@xcite depending on the magnitude of the applied magnetic field and the transit time ( subject to drift and diffusion through the conduction band from injector to detector ) , the polarized electron spin ( initially parallel to the injector fm layer magnetization ) can arrive at the detector having rotated through precession angle @xmath15 , where @xmath16 is the transit time , @xmath17 is the magnetic field , @xmath18 is the electron spin g - factor , @xmath0@xmath19 is the bohr magneton , and @xmath20 is the reduced planck constant .
our measurements in a perpendicular magnetic field , using the same experimental conditions as were applied in the spin - valve effect measurement ( v@xmath21 = -1.6v and t = 152k ) , are shown in fig .
[ fig3 ] for the same varied values of @xmath9 as in fig .
the measured @xmath10 was normalized for data comparison at different accelerating voltages @xmath9 , as in the spin - valve effect experiment .
again , the inset of fig .
[ fig3 ] shows the data for @xmath9 = 5v with magnetic field sweep directions indicated by correspondingly colored arrows . when a perpendicular magnetic field sweeps from -4 koe ( or from + 4 koe ) , @xmath10 exhibits a minimum before the field reaches 0 oe and then it suddenly drops and slowly moves up between 0 and + 1koe .
the former minima is induced by a full spin flip due to spin precession ( average @xmath22 rad rotation ) during transport through the doped silicon layer , and the latter is induced by the in - plane magnetization switching of the two fm layers by a residual in - plane component of the largely perpendicular magnetic field , causing an antiparallel injector / detector magnetization configuration and reduction in signal as seen in previously - discussed in - plane spin - valve measurements .
this argument is further upheld by changing @xmath9 ; minima attributed to precession appear at higher magnitude of applied perpendicular magnetic field as @xmath9 increases due to the shorter transit time , while the fm switching fields clearly do not change.@xcite electron spin precession in doped si spin - transport devices .
emitter voltage and temperature are same as in fig .
minima , corresponding to @xmath22 rad precession angle appear at higher magnetic field as @xmath9 increases and transit time decreases .
inset : data measured with @xmath9 = 5v .
field sweep directions are indicated by red ( increasing ) and blue ( decreasing).,width=321,height=283 ]
average spin transit times @xmath16 on the order of 45 - 180ps can be determined from the magnetic field values at @xmath22 rad precession minima @xmath17@xmath23 in fig .
3 ( @xmath24 1 koe 4 koe ) using@xcite @xmath25 , where @xmath26 is the planck constant . correlating spin polarization from spin - valve measurement to these transit times can , in principle , be used to determine spin lifetime . however , these transit times are very short , so direct correlation as in ref .
@xcite is unable to independently determine the ( long ) spin lifetime of conduction electrons in doped si .
we have previously measured spin lifetime of 73ns at similar temperature using a 350 micron - thick undoped si transport layer device ; this lifetime increases to over 500ns at 60k.@xcite in the undoped silicon transport layers used in previous works,@xcite the schottky depletion region was much larger than the layer thickness .
therefore , the conduction band was linear , resulting in a spatially constant induced electric field , and relatively `` ohmic '' spin transport where the spin transit time was inversely proportional to the injector - detector voltage drop . in this work , however , carrier depletion of the doped silicon due to schottky contacts and the resulting space - charge from ionized impurities causes a nonlinear conduction band that can have a potential energy minimum between depletion regions unless the voltage drop is very large . since injected electrons may sit in this potential well for a long time before escaping over the detector barrier , their spins will depolarize and the observed mc ratio will be suppressed . to significantly reduce this dwell time ,
an accelerating voltage ( induced by applied voltage @xmath9 , which adds to approximately 0.3v of the applied emitter voltage due to resistive tunnel junction electrodes@xcite ) can be used to alter the confining potential energy . in particular , for sufficient voltage the confining potential can be eliminated .
it is therefore expected that the spin signal is strongly sensitive to applied voltage and `` non - ohmic '' spin transport results .
modeling this non - ohmic behavior is necessary . in previous works using undoped si transport layers where the electric field is constant from injector to detector , a modeling technique using the arrival - time distribution given by the green s function solution to the drift - diffusion equation
can be easily implemented.@xcite however , the electric field in these doped si devices is highly inhomogeneous , making it difficult to implement the standard method here because the drift velocity is spatially dependent , requiring green s function solution of a nonlinear partial differential equation . in general , this procedure is non - trivial .
monte - carlo simulated transit - time distributions for injector - detector voltage drops of ( a ) 3.0v ; ( b ) 4.46v ; and ( c ) 6.0v .
note timescale changes over 4 orders of magnitude from ( a ) to ( c ) .
, width=283,height=566 ] results of hanle effect simulation , eq .
( [ mchanle ] ) .
similar to the experimental results in fig .
[ fig3 ] , minima corresponding to @xmath22 rad precession angle appear at higher magnetic field as @xmath9 increases and transit time decreases .
magnetization switching is simulated by signal sign reversal between the coercive field values of the injector and detector ferromagnets.,width=283,height=283 ] comparison of experimental and monte - carlo simulated voltage dependence of ( a ) hanle peak full width at half maximum ; and ( b ) spin polarization @xmath27 at 152k .
the simulation shown in ( a ) is insensitive to a choice of @xmath28 over the range 10 - 100ns .
, width=283,height=453 ] to overcome this problem and simulate spin transport behavior in these doped devices , we use a monte - carlo technique which translates electrons a distance @xmath29 ( due to drift ) , and @xmath30 ( due to diffusion ) in a timestep @xmath31 , where @xmath32 is the drift velocity at the position @xmath33 and @xmath34 is the diffusion constant .
( the sign on the latter expression is randomly chosen to simulate the stochastic nature of 1-dimensional diffusion . )
the spatially - dependent electric field is calculated within the depletion approximation .
using a doping density of @xmath35 @xmath36 , injector schottky barrier height of 0.6ev ( for cu / si ) and detector schottky barrier height of 0.75ev ( for nife / si ) results in a band diagram whose dependence on injector - detector voltage drop is shown in fig .
[ banddiagram ] .
this figure illustrates that the voltage drop across the si transport layer can be used to alter the dominant transport mode : at low bias a wide neutral region exists between depletion regions and electrons must diffuse _ against an electric field _ to escape to the detector , whereas for biases greater than 6v , the potential minimum is annihilated by the boundary so the internal electric field carries electrons toward the detector everywhere and drift is expected to dominate . a realistic empirical mobility model using eq . 10 from ref .
@xcite is used to evaluate @xmath37 .
the diffusion coefficient at each point in space is then calculated from the einstein relation @xmath38 , where mobility @xmath39 and @xmath40 is electric field .
we simulate transport for @xmath41 electrons at each value of injector - detector voltage drop and the arrival time at the detector for each is recorded .
the distribution of arrival times @xmath42 is constructed from a histogram of this data and are used to calculate the expected output due to spin precession in a perpendicular magnetic field ( hanle effect ) : @xmath43 where @xmath44 is the angle between the injector / detector magnetization and the device plane , @xmath28 is effective spin lifetime and the spin precession angular frequency @xmath45 .
the tilting angle @xmath44 is caused by the external magnetic field partially overcoming the finite geometric anisotropy of the magnetic thin films@xcite and is necessary to correctly model the experimental results . as in ref .
@xcite , we use @xmath46 . in addition , the final spin polarization after transport can be calculated from @xmath47
as can be seen in fig .
[ banddiagram ] , an electric field opposing transport to the detector is present at low voltage .
electrons must therefore diffuse against this electric field to escape the confining potential in the bulk of the si transport layer . under these conditions of diffusion - dominated transport
, the arrival - time distribution has a very wide exponential shape with average transit time of approximately 500ns , as shown in fig .
[ distributions ] ( a ) .
although the width of the distribution can be reduced significantly by increasing the voltage drop to the point where the si transport layer is fully depleted as shown in fig .
[ distributions ] ( b ) , the confining electric field remains and the exponential shape is maintained .
this indicates that diffusion is still strong . for sufficiently high voltage drops ,
the potential energy minimum is annihilated by the detector boundary as indicated in fig .
[ banddiagram ] and drift - dominated transport occurs .
this is reflected in the gaussian - like shape of the distribution in fig .
[ distributions ] ( c ) for a voltage drop of 6v . at this voltage ,
the average transit time is only approximately 50ps , consistent with the analysis of experimental hanle effect measurements .
therefore , as a result of our monte - carlo modeling we see that the average electron transit time in our doped si spin transport devices can be controlled over approximately 4 orders of magnitude by changing the injector - detector voltage drop by only several volts ( from 3v to 6v ) . using eq .
( [ mchanle ] ) , we simulate the hanle effect in our devices using @xmath48ns ( choice of this value will be discussed later ) .
[ hanle ] shows hanle effect simulations for voltages corresponding to the same @xmath9 values as in fig .
[ fig3 ] ( again , a shift of 0.3v due to a portion of the emitter bias dropping across the resistive tunnel junction base@xcite is accounted for to make a direct comparison ) in wide agreement to those experiments .
in particular , the qualitative shape and precession minima positions are well modeled .
the most salient feature of the hanle effect simulation is the magnetic - field width of the central ( zero precession - angle ) peak , plotted as a function of injector - detector voltage drop in fig .
[ hanlewidthpol](a ) and compared to the experimental values .
note that the width is constant for voltages greater than 6v ( due to drift velocity saturation at high electric field in si ) , and the presence of a threshold near 5v ( due to appreciable lowering of the confining potential barrier at the detector side of the transit layer once full - depletion occurs at approximately that voltage ) .
this sudden collapse of the hanle peak width is not seen in the voltage dependence of spin precession measurements using undoped drift - dominated spin transport devices . in fig .
[ hanlewidthpol ] ( b ) , we show the voltage dependence of the measured spin polarization @xmath27 from experimental data using in - plane magnetic field spectroscopy as described in fig .
again , a threshold is seen in the experimental data . however , the position of the spin polarization threshold in fig .
[ hanlewidthpol ] ( b ) near 3.5v is at much smaller bias voltage as compared to the hanle width collapse threshold near 5v shown in fig [ hanlewidthpol](a ) .
this indicates that the electrons maintain their spin despite a long dwell time which causes strong spin dephasing in the confining conduction band potential minimum at low voltages . comparing this behavior to the model results from eq .
( [ mcpol ] ) with different values of spin lifetime @xmath28 shows that this discrepancy in threshold position is consistent with a long spin lifetime of 10 - 100ns .
this can be compared to a spin lifetime of approximately 73ns measured in undoped si at the same temperature using a different technique.@xcite
in summary , we have demonstrated spin transport through n - type doped si . using a monte - carlo algorithm to model drift and diffusion
, we simulated electron transport through the inhomogeneous internal electric field and make quantitative comparisons to experimental values of spin polarization and hanle peak width without any free fitting parameters .
analysis of the arrival - time distribution indicates that in doped transport layers , the spin - polarized electron transit time can be controlled over several orders of magnitude with applied voltage .
the resulting non - ohmic behavior seen here is in contrast to spin transport measurements using undoped silicon transport layers , and is expected to influence future semiconductor spintronic device designs utilizing current - sensing spin detection methods in n - type doped semiconductors . | we demonstrate the injection and transport of spin - polarized electrons through n - type doped silicon with in - plane spin - valve and perpendicular magnetic field spin precession and dephasing ( `` hanle effect '' ) measurements . a voltage applied across the transport layer is used to vary the confinement potential caused by conduction band - bending and control the dominant transport mechanism between drift and diffusion . by modeling transport in this device with a monte - carlo scheme , we simulate the observed spin polarization and hanle features , showing that the average transit time across the short si transport layer can be controlled over 4 orders of magnitude with applied voltage . as a result
, this modeling allows inference of a long electron spin lifetime , despite the short transit length . |
liquid helium system has a total hamiltonian as @xmath0 where @xmath1 is the mass of a helium atom , @xmath2 and @xmath3 respectively signify the creation and annihilation operators .
we examine the general form of the total energy via the unitary transformation @xmath4 diagonalizing the hamiltonian @xmath5 .
all eigenstates are described as @xmath6 where @xmath7 denotes the vacuum state .
new creation and annihilation operators are defined as @xmath8 which indicate the creation and annihilation operators of a quasi - particle .
we designate this quasi - particle as a `` dressed boson '' .
the dressed boson number operator is defined as @xmath9 the total number conservation and the total momentum conservation are expressed as @xmath10 that is to say , the total number of helium atoms is equal to the total number of dressed bosons and the total momentum of helium atoms is equal to the total momentum of dressed bosons .
the total energy of the system is a sum of the kinetic energy @xmath11 of the center of mass and galilean invariant terms @xmath12 : @xmath13 where @xmath14 is the total mass of liquid helium .
@xmath15 where galilean invariant terms are described only by relative momenta of dressed bosons : ( galilean invariant terms)= @xmath16 + substitution of eq.([eq : c ] ) into eq.([eq : b ] ) yields @xmath17 where we abbreviate higher terms because three - particle collision is a rare case for diluteness of liquid helium .
the single excitation state has a distribution of @xmath18 and therefore its total energy is derived from eq.([eq : d ] ) as follows : @xmath19 where we have used @xmath20 and the spherical symmetric property of the function @xmath21 .
therein , the latent heat at zero kelvin is equal to @xmath22 . accordingly
the elementary excitation energy at zero kelvin is given by @xmath23 .
this relation engenders a function form of the nonlinear term as @xcite @xmath24
the energy of one dressed boson is an increase value of the total energy when one dressed boson is added to the system . accordingly
the dressed boson energy is defined as @xmath25 .
the calculation result for the derivative of eq .
( [ eq : d ] ) shows @xmath26 where we have used @xmath27 .
the distribution function is determined as @xmath28 we can obtain approximate solutions of the coupled equations of ( [ eq : g ] ) and ( [ eq : h ] ) via the iteration method @xcite .
we adopt the landau distribution function as the zero - th order distribution : @xmath29 the j - th order solutions are derived from the ( j-1)-th distribution function as follows : @xmath30 this j - th energy form produces the j - th distribution function : @xmath31 therein the excitation energy from the bose - einstein condensate of dressed bosons is expressed as @xmath32 we can evaluate the second order solutions @xmath33 and @xmath34 via the iteration processes from the zero - th order distribution .
using the second order excitation energy @xmath33 and the distribution function @xmath34 , we can calculate the second order approximation values of specific heat as follows @xcite ; + @xmath35 + + the evaluated results are shown in fig.[f : 1 ] and fig.[f : 2 ] .
figure [ f : 1 ] and [ f : 2 ] indicate the second order results of specific heat via the nonlinear theory .
the curves express the calculated values .
the dots with red indicate experimental data @xcite . as shown in fig.[f : 1 ] and fig.[f : 2 ] , the theoretical values of the second order are in good agreement with the experimental data for @xmath36 .
it is noteworthy that the present calculation uses the experimental values of excitation energy only for the temperature 1.1 k. of course the iteration method is insufficient in close vicinity of the @xmath37 transition temperature .
we have discussed origin of the logarithmic divergence at the @xmath37 point in the previous paper @xcite . it is clarified that the logarithmic divergence is caused by the nonlinear mechanism of the total energy .
the calculation results are shown in fig .
[ f : 3 ] .
the large dots colored with blue indicate the experimental data @xcite , and the small dots with red are measured by lipa et al @xcite .
bd theory indicates the results in the reference of @xcite .
bcy theory indicates the results in the reference of @xcite .
the curve of the nonlinear theory is the results of the reference @xcite ] thus the nonlinear theory has well explained the temperature dependence of the specific heat of superfluid helium for all temperature region . accordingly
the nonlinear mechanism of total energy is important for understanding the properties of liquid helium .
landau l d 1941 _ zh . eksp .
fiz . _ * 11 * 592 ; landau l d 1941 _ j. phys .
moscow _ * 5 * 71 ; ibid .
1947 * 11 * 91 .
khalatnikov i m 1965 _ an introduction to the theory of superfluidity _
( w. a. benjamin inc .
new york , amsterdam ) .
sasaki s and hori h 2008 _ bose - einstein condensation and superfluidity _
( jaist press ) .
sasaki s 1987 _ proc .
int . conf . on low temp .
, j. j. a. p. _ * 26 * ( 1987 ) 23 .
sasaki s 1990 _ physica _ b * 165 * 507 .
sasaki s 2003 _ physica _ b * 329 * 232 .
sasaki s 2007 _ journal of low temperature physics _ * 148 * 103 .
sasaki s 1994 _ physica _ b * 194 * 503 .
sasaki s 1989 _ springer series in solid - state sciences _ * 79 * 160 .
sasaki s 1994 _ physica _ b * 194 * 497 . | the specific heat of liquid helium was calculated theoretically in the landau theory @xcite .
the results deviate from experimental data in the temperature region of 1.3 - 2.1 k. many theorists subsequently improved the results of the landau theory by applying temperature dependence of the elementary excitation energy @xcite , @xcite . as well known , many - body system has a total energy of galilean covariant form .
therefore , the total energy of liquid helium has a nonlinear form for the number distribution function .
the function form can be determined using the excitation energy at zero temperature and the latent heat per helium atom at zero temperature .
the nonlinear form produces new temperature dependence for the excitation energy from bose condensate .
we evaluate the specific heat using iteration method .
the calculation results of the second iteration show good agreement with the experimental data in the temperature region of 0 - 2.1 k , where we have only used the elementary excitation energy at 1.1 k. |
in recent years higher dimensional gravity is attracting much interest .
one reason is the possibility that these higher dimensions could be detectable at cern .
the possibility that space time may have more than four dimensions is initiated by high energy physics and inspired by d - brane ideology in string theory .
our 4-dimensional space time ( brane ) is embedded in the 5-dimensional bulk . it is assumed that all the standard model degrees of freedom reside on the brane , where as gravity can propagate into the bulk @xcite .
the effect of string theory on classical gravitational physics is investigated by the low - energy effective action . if our 5-dimensional space time is obtained as an effective theory , the matter fields , for example the u(1 ) field , can exists in the bulk . in general relativity(gr ) , gravitating non - abelian gauge field , i.e. , the yang - mills(ym ) field , can be regarded as the most natural generalization of einstein - maxwell(em ) theory .
in particular , particle - like , soliton - like and black hole solutions in the combined einstein - yang - mills(eym ) models , shed new light on the complex features of compact object in these models . see @xcite for an overview . the reason for adding a cosmological constant to these models , was inspired by the study of the so - called ads / cft correspondence @xcite , since the 5-dimensional einstein gravity with cosmological constant gives a description of 4-dimensional conformal field theory in large n limit .
brane world scenarios predict a negative cosmological constant .
there is a relationship between the frw equations controlling the cosmological expansion and the formulas that relate the energy and entropy of the cft @xcite , indicating that both sets of equations may have a common origin . in the ads - brane
cosmological models , the ads / cft model describes a cft dominated universe as a co - dimension one brane , with fixed tension , in the background of an ads black hole @xcite .
the brane starts out inside the black hole , passes through the horizon and keeps expanding until it reaches a maximal radius , after which it contracts and falls back into the black hole . at these moments of horizon
crossing , it turns out that the frw equation turns into an equation that expresses the entropy density in terms of the energy density and coincides with the entropy of the cft .
however , in these models , one adds on an artificial way tension into the equations .
more general , one could solve the equations of einstein together with the matter field equations , for example , the ym field and try to obtain the same correspondence .
string theory also predicts quantum corrections to classical gravity theory and the gauss - bonnet(gb ) term is the only one leading to second order differential equations in the metric . in the 4-dimensional eym - gb model with a dilaton field ( eymd - gb ) @xcite
, it was found that the gb contribution can lead to possible new types of dilatonic black holes .
further , for a critical gb coupling @xmath0 the solutions cease to exist .
the ads / cft correspondence can also be investigated in the einstein - gb gravity .
for a recent overview , see @xcite . from the viewpoint of ads
/ cft correspondence , it is argued that the gb term in the bulk corresponds to the next leading order corrections in the @xmath1 expansion of a cft .
further , it is argued that the entropy of an einstein - gb black hole and the cft entropy induced on the brane are equal in the high temperature limit . in this paper
we investigate the possibility of regular and singular solutions in the 5-dimensional eym - gb model and the effect of a cosmological constant on the behaviour of the solutions .
the action of the model under consideration is @xcite @xmath2,\ ] ] with @xmath3 the gravitational constant , @xmath4 the cosmological constant , @xmath5 the gauss - bonnet coupling and @xmath6 the gauge coupling .
the coupled set of equations of the eym - gb system will then become @xmath7 @xmath8 with the einstein tensor @xmath9 and gauss - bonnet tensor @xmath10 further , with @xmath11 the ricci tensor and @xmath12 the energy - momentum tensor @xmath13 and with @xmath14 , and @xmath15 where @xmath16 represents the ym potential .
consider now the spherically symmetric 5-dimensional space time @xmath17 with the ym parameterization @xmath18 where @xmath19 and @xmath20 are functions of and @xmath21 and @xmath22 .
it turns out that no time evolution of the metric component @xmath23 can be found from the equations , so @xmath23 depends only on r. the equations become ( we take @xmath24 ) @xmath25 @xmath26 @xmath27 and @xmath28
the independent field equations then read @xmath29 @xmath30 while the equation for @xmath23 decouples and can be integrated : @xmath31 the equations are easily solved with an ode solver and checked with maple .
we will take for the initial value of @xmath20 the usual form @xmath32
. then the other variables can be expanded around @xmath33 : @xmath34 so we have 3 initial parameters and 4 fundamental constants @xmath35 , @xmath6 and @xmath5 .
we solved the equations with a two point boundary value solver .
epsf fig.1 solution for @xmath36 for @xmath37 and @xmath38 respectively .
+ fig . 2 solution for @xmath39 for @xmath37 and @xmath38 respectively .
3 solution for @xmath40 for @xmath37 and @xmath38 respectively .
+ from our numerical solutions , we see that the gb term increases the number of nodes of the yang - mills field .
further , we see that for positive @xmath4 the solution develops a singularity , while for negative @xmath4 it remains singular free .
a matter field term in the action will lead to an extra term inside the square root of eq.(16 ) , for example in the case of the 5-dimensional einstein - maxwell - gb model @xcite and the 5-dimensional einstein - yang - mills - gb model with the wu - yang ansatz @xcite . in these models , however , there are no additional equations for the maxwell field and ym field respectively .
so an analytic solution for @xmath41 is obtained .
when one simultaneously tries to solve the einstein equation and matter field equations , then it is not easy to obtain an analytic expression for @xmath41 , as is the case of our eym - gb model .
however we can analyse the equation for @xmath41 when @xmath20 becomes a constant : @xmath42 the solution is @xmath43 with @xmath44 an integration constant .
since horizons occur where @xmath45 , we can expect cosmological- and event horizons .
one can easily check that the zero s of @xmath41 are @xmath46 so the horizon radius depends only on suitable combinations of @xmath4 and @xmath44 .
the expression inside the square root becomes negative ( and hence @xmath41 is singular ) for @xmath47 depending on the parameters , this singular surface can be shielded by the event horizon ( otherwise , it will be naked ) .
this is well known behaviour in the models where the equation for @xmath41 decouples from the matter field equation .
one should like to prove that for negative @xmath4 that f(r ) has no zero s and is regular everywhere in our model .
this is currently under study .
a 5-dimensional spherically symmetric particle - like solution is found in the einstein - yang - mills gauss - bonnet model . as in other studies in higher dimensional cosmological models , a negative cosmological
constant seems to favor for stability and results in most cases in asymptotically anti de sitter space time . in our 5-dimensional eym - gb model
, we also find a profound influence of a negative cosmological constant on the behaviour of horizons .
the appearance of horizons in e - gb models is not surprising .
these gb black holes are found by many authors
. however , the lacking of horizons in the eym - gb model for suitable negative cosmological constant is quite new . the explanation for this behaviour must come from the ym term on the right hand side of eq.(9 ) .
the zero s of @xmath41 will depend on the behaviour @xmath48
. there could be a connection of the solution presented here with the ads / cft correspondence . as mentioned before , no analytic expression for @xmath41 available .
moreover , to obtain the ( n-1)-dimensional entropy on the brane , one needs the junction conditions at the brane ( @xcite ) , which becomes very complicated in the eym - gb model .
the junction condition also introduces a brane tension .
this tension must cancel the cosmological constant , in order to obtain the desired cft correspondence .
the contribution of ym field on the junction could have profound impact on the tension of the brane and the role of a cosmological constant could be different . so the strong influence of a small cosmological constant on the eventually formed gb black hole in our model is quite clear from the consideration mentioned above .
10 randall l and sundrum r 1999 _ phys .
rev . lett . * 83 * 3370 , 4690 _
volkov m and galtsov d v 1999 _ phys .
rep . * 319 * 1 _ maldcena j 1999 _ adv .
phys * 2 * 3370 , 231 _ bjoraker j and hosotani y 2000 _ phys .
lett . * 84 * 1853 _ savonije i and verlinde e 2001 _ phys .
lett . b*507 * 305 _ gubser s s 2001 _ phys .
d_63 _ 084017 _ donets e e and galtsov d v 1995 _ phys
b * 352 * 261 _ torii t , yajima h and maeda k 1997 _ phys
d * 55 * 739 _ ogushi s and sasaki m 2005 _ prog .
. phys . * 113 * 979 _ okuyama n and maeda k 2003 _ phys .
d * 67 * 104012 _ boulware d g and deser s 1985 _ phys .
. lett . * 55 * 2656 _ cai r and guo q 2004 _ phys
. rev . d*69 * 104025 _ clunan t , ross s f and smith d j 2004 _ class .
* 21 * 3447 _ thibeault m simeone c and eiroa e f 2005 _ preprint gr - qc/0512029 _ mazharimousavi s and halilsoy m 2008 _ preprint gr - qc/08011562 _ slagter r j 2006 _ in the proceedings of the conference on general relativity and gravitation , paris , 2005 _ | we present a numerical solution on a 5-dimensional spherically symmetric space time , in einstein - yang - mills - gauss - bonnet theory using a two point boundary value routine .
it turns out that the gauss - bonnet contribution has a profound influence on the behaviour of the particle - like solution : it increases the number of nodes of the ym field . when a negative cosmological constant in incorporated in the model , it turns out that there is no horizon and no singular behaviour of the model . for positive cosmological constant
the model has singular behaviour . |
consider a system of @xmath0 identical bosons in one dimension , with an attractive delta - function interaction . there is a bound state for all @xmath0 , having the form of a well - localized `` molecule''@xcite .
this result is interesting , since it is the unique example of an exactly solved localized many - body system
. a possible experimental realization of this system would be in a @xmath1 vapor @xcite confined to an atomic trap which is so tight in two directions , that the system can be regarded as one - dimensional with respect to the third direction .
such traps are realistic prospects in the near future @xcite , and high - aspect ratio cigar - shaped traps approximating quasi - one - dimensional systems are already available experimentally @xcite .
we seek the localized eigenfunctions @xmath2x@xmath3 of the hamiltonian @xmath4 wherein @xmath5 is the particle mass and @xmath6 is a positive parameter representing the strength of interparticle attraction .
the wavefunction and many derived quantities can be given exactly and in closed form ; however , it is useful to first consider the mean - field version of this theory . the gross - pitaevskii ( gp ) approximation @xcite assumes that @xmath7 can be written as a simple product of wavefunctions @xmath8 where the @xmath9 are eigenfunctions for a single particle moving in the average potential of all the others @xmath10 where @xmath11 .
thus the ground - state wavefunction @xmath12 is actually determined by solving a nonlinear schrdinger equation .
we will show that this approximation generally captures the physics of the large @xmath0 limit ; this is in contrast to the case of repulsive bosons in one dimension , for which the interaction terms must be treated more carefully @xcite .
the solution to eq.(3 ) is @xmath13 where @xmath14 ; the single particle energy is @xmath15 .
this has the implications for the many - body problem that there can be a state that is localized in the sense that it is unlikely that any particle be found far from the center of mass , and provides an estimate for the particle density ( with the center of mass fixed at the origin ) @xmath16 ; the corresponding ground state energy is @xmath17 .
it can be shown that there are no other bound states of eq .
( 3 ) for the potential v(x ) constructed from the ground state wavefunction .
this suggests that we can not construct approximate localized excited states of the boson molecule by symmetrizing @xmath18 ( for some appropriate choice of @xmath19 ) .
the gp wavefunction is not translationally invariant , and thus violates momentum conservation .
every translation of it is a solution to the same self - consistent equation with the same energy ; then by forming the linear combination @xmath20 we can restore this symmetry .
for the case @xmath21 , the wavefunction can be found by changing to relative coordinates , giving a problem equivalent to a single particle and a fixed potential well .
the ( unnormalized ) ground state wavefunction is @xmath22 where @xmath14 ; the corresponding energy is @xmath23 . for @xmath24
the delta function vanishes , and the wavefunction is an eigenfunction of the kinetic energy operator ; for a bound state this is the product of exponential functions with real argument . the potential generates the discontinuity in the derivative of @xmath25 at the crossing point @xmath26 .
these two properties hold for all @xmath0 , so that we can write down the general ( unnormalized ) wavefunction as @xcite @xmath27 for the sector @xmath28 this can also be written in the form @xmath29 in view of the exchange symmetry of bosons , this completely describes the wavefunction .
the ground state energy is proportional to the sums of the squares of the coefficients of @xmath30 in the argument of the exponential ; explicitly , @xcite @xmath31 this agrees with the gp result only in leading order in @xmath0 .
the implications of the wavefunction are more clearly revealed when reduced to the probability that there is a particle at position @xmath32 relative to the center of mass . for the case
@xmath33 , there can only be a particle at @xmath32 if the other is at @xmath34 , and then the probability is proportional to the square of the wavefunction : @xmath35 this has been normalized so that the integral over all @xmath32 gives @xmath21 . for larger @xmath0
we integrate over the unconstrained degrees of freedom ; for example for @xmath36 the probability of finding a particle at @xmath37 is proportional to @xmath38 where the integrals describe the case that it is particle 3 or particle 2 that is at @xmath32 ( particle 1 is necessarily in the region @xmath39 ) .
the resulting normalized probability distribution is @xmath40 = 3\cdot 2\alpha [ z - { \frac 1 2 } z^{2}]\ ] ] where in the second representation we have introduced the abbreviation @xmath41 . in a similar way @xmath42 can be constructed for other small @xmath0 .
the results for @xmath43 suggest the general expression ( given previously by yoon and negele@xcite ) @xmath44 this has been normalized so that the integral over all @xmath32 gives @xmath0 .
the density at @xmath45 is @xmath46 .
an interesting feature of the distributions is revealed when we represent them as a power series in @xmath47 : the coefficients of @xmath48 vanish when @xmath49 is odd and less than @xmath50 .
thus the discontinuity in slope at @xmath45 that is present in @xmath51 is replaced by a discontinuity in a much higher derivative for larger @xmath0 ( see footnote 7 ) .
this property is a demonstrable consequence of the representation ( 12 ) ; in fact , combined with the assumption that @xmath42 is a polynomial in @xmath52 of order @xmath53 , this is sufficient to determine the form of @xmath42 up to normalization . in the limit of large @xmath0 ,
( 12 ) reduces to @xmath54 which has form similar to the gp result @xmath55 sech@xmath56 figure 1 shows @xmath57 as a function of @xmath58 , for various @xmath0 .
we may also characterize @xmath59 through its moments .
starting from ( 12 ) it can be shown that @xmath60 @xmath61 the gp theory would give @xmath62 ( and similarly for @xmath63 ) they agree at large @xmath0 , but the exact result approaches the asymptotic limit from below , while gp approaches from above .
the single - particle green function @xmath64 is calculated from the wavefunction by constructing the product @xmath65 and integrating out the spectator coordinates @xmath66 .
explicit construction of the @xmath64 for @xmath67 suggests the result @xmath68 \ } z^{k(m ) } \ ] ] where @xmath69 and @xmath70/4 $ ] . with the choice @xmath71
, @xmath72 is normalized so that @xmath73 ; a physically more meaningful normalization would be to choose @xmath74 , but we have not succeeded in finding the closed form representation for the corresponding @xmath75 . since @xmath76 , the sum contains every term twice , except for @xmath77 when @xmath0 is odd . in the large @xmath0 limit , @xmath78 and then @xmath79 which is the green function that one would construct from the translationally invariant gp wavefunction @xmath80 .
the momentum distribution is the fourier transform of g@xmath81 @xmath82^{2 } } [ n^{2}-1 - 4k ] \alpha \}\ ] ] where @xmath83 is the wavevector .
figure 2 gives a graph of @xmath84 versus @xmath85 , for various @xmath0 .
as in the case of the repulsive boson system@xcite , the momentum distribution is smooth : there is no bose condensation .
the width of the momentum distribution is set by the parameter @xmath86 , so that the momentum distribution for the noninteracting case ( @xmath87 ) is attained in a natural way . in a highly localized wavefunction
we would hardly expect to find _ long - range _ off - diagonal order !
the two - point correlation function is calculated from the squared wavefunction by choosing one of the particles to be @xmath88 and another to be at @xmath89 ( in all possible ways ) , and integrating out all other @xmath90 .
we have chosen to normalize @xmath91 so that the integral over all @xmath32 gives unity .
it proved harder to analyze , because its representation is more irregular : for example , @xmath92\ ] ] where again @xmath69 .
based on the explicit construction of @xmath93 for @xmath43 , we find @xmath94 k(m ) + \alpha |x| [ n^{2 } - 4 k(m ) ] k(m)^{2 } \ } w^{k(m)}\ ] ] where @xmath95 . since @xmath96 , the sum contains every term twice , except for @xmath97 when @xmath0 is even . the special value at @xmath45 ( which is the probability of having two particles at the same place ) is @xmath98 .
figure 3 gives a graph of @xmath99 for various @xmath0 . in the limit of large @xmath0
, we may approximate @xmath100 in the prefactors and then @xmath101 , with the result @xmath102 the structure factor is given by the fourier transform of @xmath91 : @xmath103 \over k^{2}+ k(m)^{2 } } - { [ n^{2 } - 4 k(m)][k(m)^{2}-k^{2}]\over [ k^{2}+ k(m)^{2}]^{2 } } \}\ ] ]
in this section we will show that for any @xmath0 there is only one localized state with zero total momentum .
there are other states of negative energy , but these can all be interpreted as uncorrelated smaller molecules . for indistinguishable bosons we need only specify the wavefunction for the sector @xmath104 , and in the interior of this region ( where no two coordinates are equal ) , the hamiltonian operator reduces to the kinetic energy .
the bound eigenstates of the kinetic energy have the form @xmath105 and the energy is determined to be @xmath106 m . to be normalizable
it is necessary that @xmath107 ( otherwise a uniform translation of all the particles could lead to indefinite growth of @xmath25 ) .
the wavefunction for the interacting system is a linear combination of degenerate functions of this form . at the sector boundaries ( where @xmath108 for some j ) , the interaction gives rise to a discontinuity in the derivative of the wavefunction , which implies@xcite @xmath109 this must hold for arbitrary choices of the remaining @xmath110 variables
; then in the sum of terms of the form ( [ eq19 ] ) the only terms that are coupled by ( [ eq20 ] ) are those for which the lists @xmath111 differ only at @xmath112 and @xmath113 but then the two sum constraints on the @xmath111 ensure that there are only two such lists , which differ by having the values for @xmath112 and @xmath113 interchanged .
thus the general form of the wavefunction would appear to be constructable from a single list @xmath111 ( which we will take to be ordered , so that @xmath114 for all @xmath115 in the form @xmath116 where the first sum is over all permutations of @xmath0 objects , and @xmath117 is the @xmath118 member of the permuted list .
the boundary condition ( [ eq20 ] ) implies relations among the coefficients @xmath119 @xmath120 ( it should be noted that the permutations @xmath121 and @xmath122 are related so that @xmath123 and @xmath124 .
the foregoing differs only slightly from the case of bosons with repulsive delta - potential interactions@xcite the main difference is that the @xmath125 are real instead of purely imaginary .
however , the remaining boundary conditions are different : for repulsive interactions , the wavefunction is delocalized and must be confined within a periodic box , while for attractive interactions , the space can be infinite if the wavefunction itself is localized .
the relevant ( and interesting ) condition is that @xmath126 should become exponentially small if the extremal particle at @xmath127 is moved to large positive @xmath32 , or if the particle at @xmath128 is moved to large negative @xmath32 .
this requires that @xmath129 be positive and @xmath130 be negative , for every term in ( [ eq21 ] ) that has nonzero coefficient .
the sum itself is over all permutations , which certainly will generate terms in which the k s have the wrong sign ; but the coefficients will vanish ( according to ( [ eq22 ] ) ) for all permutations such that in the list @xmath131 it happens there are members @xmath132 and @xmath133 such that @xmath134 and @xmath135 . the ground state wavefunction is a simple example of the working of this rule : the ordered list is of the form of a sequence @xmath136 and every possible permutation of it gives rise to a vanishing coefficient .
the ground state is not the only possibility : we could have a list which is made of several distinct sequences , so that in one of its permutations it takes the form @xmath137 .
this gives some freedom in the choice for the starting elements @xmath138 and sublist lengths @xmath139 ; this generates the wavefunction for an excited state of the molecule .
however , it does not represent a localized state .
suppose we simultaneously displace the particles @xmath140 a large distance to negative values .
if @xmath141 , this term will become exponentially large , giving a nonnormalizable wavefunction .
however , if @xmath142 there will be a different permutation of the list in which this sublist is at the end , again giving a nonnormalizable wavefunction when these particles are moved to large positive values .
the remaining possibility is @xmath143 but now the wavefunction remains finite if it is separated into two widely separated pieces , one of which contains @xmath144 particles : this describes states in which the particles have formed several separate and uncorrelated molecules . in the foregoing argument
we have assumed that the @xmath111 are real . allowing them to be complex does not change the situation much :
the argument above implies that the k@xmath145 having a real part must be organizable into sublists as above , all having the same imaginary part ; each sublist represents a separated molecule , now with finite momentum . from this point of view
we can readily see that when two molecules collide , nothing happens
they pass through each other leaving each undisturbed@xcite .
e. lieb and w. liniger , phys
130 * , 1605 ( 1963 ) .
j. b. mcguire , j. math .
phys . * 5 * , 439 ( 1966 ) .
b. thacker , rev .
mod . phys . * 53 * , 253 ( 1981 ) .
c. c. bradley , c. a. sacket , j. j. tollet , and r. g. hulet , phys .
* 75 * , 1687 ( 1995 ) .
a. d. jackson , g. m. kavoulakis , and c. j. pethick , phys .
a , * 58 * , 2417 ( 1998 ) .
m. r. andrews , m. -o .
mewes , n. j. van druten , d. s. durfee , d. m. kurn , and w. ketterle , science , * 273 * , 84 ( 1996 ) .
l. p. pitaevskii , zh eksp .
. fiz . * 40 * , 646 ( 1961 ) [ sov .
jetp * 13 * , 451 ( 1961 ) ] ; e. p. gross , nuovo cimento * 20 * , 454 ( 1961 ) .
e. b. kolomeisky , t. j. newman , j. p. straley , and x. qi , phys . rev . lett . *
85 * , 1146 ( 2000 ) .
b. yoon and j. w. negele , phys .
a * 16 * , 1451 ( 1977 ) .
it s a little surprising that there is any singularity , since placing one particle at the center of mass would not seem to be significantly different from placing it nearby .
however , for this special position the configurations with differing numbers of particles to right and left of the origin ( for @xmath0 even , _ all _ configurations have this property ) are equivalent and coupled by the possibility of moving another particle from one side to the other . | we give the green function , momentum distribution , two - particle correlation function , and structure factor for the bound state of @xmath0 indistinguishable bosons with an attractive delta - function interaction in one dimension , and an argument showing that this boson `` molecule '' has no excited states other than dissociation into separated pieces . |
k " ahler@xcite has shown how to pack the dirac wave function into the language of differential forms in the sense that the dirac equation is an equation in which a linear operator acts on a linear combination @xmath0 of @xmath1-forms ( p=0 , 1 , ... , d ; here d = dimension = 4 ) .
this is the dirac - k " ahler formalism .
one of us@xcite has long developed an a priori rather different formalism in an attempt to unify spin and charges . in this approach
the spin degrees of freedom come out of canonically quantizing certain grassmannian odd ( position analogue in the sense of being on an analogue footing with @xmath2 ) variables @xmath3 .
these variables are denoted by a vector index @xmath4 , and there are at first to see no spinors at all ! one of the main purposes of the present article is to point out the analogy and nice relations between the two different ways of achieving the - almost miraculous - appearance of spin one half degrees of freedom in spite of starting from pure vectors and tensors . of course it is a priori impossible that vectorial and tensorial fields ( or degrees of freedom ) can be converted into spinorial ones without some cheat . the
cheat consists really in exchanging one set of lorentz transformation generators by another set ( which indeed means putting strongly to zero one type of grassmann odd operators fulfilling the clifford algebra and anticommuting with another type of grassmann odd operators , which also fulfill the clifford algebra@xcite ) .
in fact one finds on page 512 in the k " ahlers@xcite article that there are two sets of rotation generators ; one set for which the @xmath0 field ( in the k " ahler s notation ) transforms as a spinor field and another one for which it transforms as superpositions of vector and ( antisymmetric ) tensor fields .
analogously in the approach of one of us the apriori lorentz transformation generators @xmath5 have the wave function transform as vectors and antisymmetric tensors , while @xmath6 ( @xmath7 $ ] ) or @xmath8 ( @xmath9 $ ] and @xmath10 $ ] means the commutator ) used alone are also possible lorentz generators for which now the wave function transforms as a spinor wave function . by putting @xmath11 ( which has the property that @xmath12 = 0 $ ] ) equal strongly to zero
is the same as replacing @xmath13 by @xmath14 . in both approaches to get spinors out of vectors and antisymmetric tensors , as start , you get not only one but several copies , families , of dirac fields .
this is a fundamental feature in as far as these different families are connected by the generator parts not used : if one for instance uses @xmath6 as the lorentz generator to get spinors , then the not used part @xmath15 transforms the families ( of the same grassmann character ) into each other .
it will be a major content of the present article to bring about a dictionary relating the two formalisms so that one can enjoy the simplicity of one also working on the other one .
we also shall generalize the k " ahler operators for @xmath16 , comment on the discrete symmetries , which in the one of us approach show up clearly and use the @xmath17 dimensions to describe spins and charges@xcite .
in the following section we shall put forward the little part of the formalism of the work of one of us needed for the comparison with the dirac - k " ahler formalism . in the next section again - section 3 - we
shall then tell the ( usual ) dirac - k " ahler formalism as far as relevant . the comparison which should now be rather obvious
is performed in section 4 . in section 5
we shall analyse in the two approaches in parallel how the remarkable finding of the dirac equation inside a purely tensorial - vectorial system of fields occurs .
in section 6 we shall comment on the evenness of the @xmath18 matrices , which have to transform grassmann odd wave functions into grassmann odd wave functions .
in section 7 we shall comment on discrete symmetries for either k " ahler or the one of us approach , also discussing the realization of the discrete symmetries pointed up clearly by weinberg in his book@xcite on pages 100 - 105 .
in section 8 we want to investigate how unavoidable the appearance of families is to this type of approaches . in section 9
we shall look at how the ideas of one of us of extra dimensions generalizes the k " ahler approach .
in section 10 we discuss the nielsen and ninomija @xcite no go theorem for spinors on a lattice and a possible way out . in section 11 we shall resume and deliver concluding remarks .
what we can call the manko approach@xcite , and which is the work of one of us , is a rather ambitious model for going beyond the standard model with say 10 ( or more ) extra dimensions , but what we need for the present connection with the dirac - k " ahler@xcite formalism is only the way in which the spin part of the dirac particle fields comes about .
the total number of dimensions in the model is ( most hopefully ) 13 + 1 bosonic degrees of freedom , i.e. normal dimensions , and the same number of fermionic ones .
let us call the dimension of space - time @xmath19 and then the dirac spinor degrees of freedom shall come from the odd grassmannian variables @xmath20 . in wanting to quantize or just to make poisson brackets out of the @xmath19 @xmath3 s we have two choices since we could either decide to make the different @xmath3 s their own conjugate , so that one only has @xmath21 degrees of freedom - this is the approach of ravndal and divecchia@xcite - or we could decide to consider the @xmath3 s configuration space variables only . in the latter case - which is the mankov
c case - we have then from the @xmath3 s different conjugate variables @xmath22 .
in this latter case we are entitled to write wave functions of the form @xmath23 this is the only form a function of the odd grassmannian variables @xmath3 can take .
thus the wave function space here has dimension @xmath24 .
completely analogously to usual quantum mechanics we have the operator for the conjugate variable @xmath3 to be @xmath25 the right arrow here just tells , that the derivation has to be performed from the left hand side .
these operators then obey the odd heisenberg algebra , which written by means of the generalized commutators @xmath26 where @xmath27 takes the form @xmath28 here @xmath29 is the flat metric @xmath30 . for later use we shall define the operators @xmath31 for which we can show that the @xmath32 s among themselves fulfill the clifford algebra as do also the @xmath11 s , while they mutually anticommute : @xmath33 note that the linear combinations ( [ eq6 ] ) presuppose a metric tensor , since otherwise only @xmath3 and @xmath34 but not @xmath35 and @xmath36 are defined .
we could recognize formally @xmath37 as the dirac - like equation , because of the above generalized commutation relations . applying either the operator @xmath38 or @xmath39 on the two equations ( eqs.([d ] ) ) we get the klein - gordon equation @xmath40 .
here of course we defined @xmath41 however , it is rather obvious that these equations ( [ d ] ) are not dirac equations in the sense of the wave function transforming as a spinor , w.r.t . to the generators for the lorentz transformations ,
if taken as usual @xmath42 however it is easily seen that we can write these generators as the sum @xmath43 where we have defined @xmath44 , \quad \tilde{\tilde{s } } { } ^{ab } : = -\frac{i}{4}[\tilde{\tilde{a } } { } ^a,\tilde{\tilde{a } } { } ^b ] , \label{s}\ ] ] with @xmath45:=ab - ba$ ] .
one can now easily see that the solutions of the two equations ( [ d ] ) now transform as spinors with respect to either @xmath46 or @xmath47 it is of great importance for the trick of manipulating what we shall consider to be the lorentz transformations and thus to be able to make the miraculous shifts of lorentz representation that is the somewhat remarkable characteristic of the k " ahler type of shift in formulation interpretation , that both - untilded , the single tilded and the double tilded - @xmath13 obey the @xmath19-dimensional lorentz generator algebra @xmath48 , when inserted for @xmath49
. really the cheat
consist -as we shall return to - in replacing the lorentz generators by the @xmath6 , say . this
means indeed that for this choice the operators @xmath50 have to be put strongly to zero in the generators of the lorentz transformations ( eq.([vecs ] , [ vecsp ] , [ s ] ) ) as well as in all the other operators , representing the physical quantities .
we shall present the one of us approach in further details in section 4 pointing out the similarities between this approach and the k " ahler approach and generalizing the k " ahler approach .
the k " ahler formulation@xcite takes its starting point by considering p - forms in the @xmath19-dimensional space , @xmath51 elegantly , the 1-forms say are defined as dual vectors to the ( local ) tangent spaces , and the higher p - forms can then be defined as antisymmetrized cartesian ( exterior ) products of the one - form spaces , and the 0-forms are the scalars ; but we can perhaps more concretely think about the p - forms as formal linear combinations of the differentials of the coordinates @xmath52 : a general linear combination of forms is then written @xmath53 where the p - form is of the form @xmath54
@xmath55 then one can define both the presumably most well known exterior algebra denoted by the exterior product @xmath56 and the clifford product @xmath57 among the forms .
the wedge product @xmath56 has the property of making the product of a p - form and a q - form be a ( p+q)-form , if a p - form and a q - form have no common differentials .
the clifford product @xmath58 on a p - form is either a @xmath59 form , if a p - form does not include a one form @xmath52 , or a @xmath60 form , if a one form @xmath52 is included in a p - form .
actually k " ahler found how the dirac equation could be written as an equation@xcite ( eq .
( 26.6 ) in the k " ahler s paper ) @xmath61 where the symbol @xmath0 stands for a linear combination of @xmath1-forms ( eq.([u ] ) ) and @xmath62 with d being the dimension of space - time , namely d = 4 for the k " ahler s case .
further in the notation of k " ahler the symbol , as k " ahler does in his paper , but @xmath63 of k " ahler is in the becher and joos paper replaced by @xmath64 , which means that in their paper @xmath65 . ] @xmath63 denotes inner differentiation , which means the analogue of the exterior differential @xmath19 but with the use of the clifford product @xmath57 instead of the exterior product @xmath66 @xmath67 the symbol @xmath68 determines the coupling of the charge with the electromagnetic field @xmath69 and @xmath70 means the electron mass , the symbol @xmath71 transforms a @xmath1-form into @xmath72-form , if the @xmath1-form includes @xmath73 , otherwise it gives zero . for a free massless particle living in a d dimensional space - time -
this is what interests us in this paper since the mass term brings no new feature in the theory - eq.([dk1 ] ) can be rewritten in the form @xmath74 where the symbol @xmath0 stands again for a linear combination of @xmath1-forms ( p = 0,1,2,3,5, ...
,d ) . that is to say that the wave function describing the state of the spin one half particle is packed into the exterior algebra function @xmath0 .
more about the k " ahler s approach will come in section 4 giving the correspondence between that and the one with the grassmann @xmath3 s , where we shall also give some generalizations .
we demonstrate the parallelism between the k " ahler@xcite and the one of us@xcite approach in steps , paying attention on the becher - joos@xcite paper as well .
first we shall treat the spin @xmath75 fields only , as k " ahler did .
we shall use the simple and transparent definition of the exterior and interior product in grassmann space to generalize the k " ahler approach to two kinds of @xmath63 ( eq.([dk0 ] ) ) operators on the space of p - forms and then accordingly to three kinds of the generators of the lorentz transformations , two of the spinorial and one of the vectorial character , the first kind transforming spinor @xmath75 fields , the second one transforming the vector fields .
we comment on the hodge star product for both approaches , define the scalar product of vectors in the vector space of either p - forms or of polynomials of @xmath3 s and comment on four replications of the weyl bi - spinor .
we also discuss briefly the vector representations in both approaches .
we present here , side by side , the operators in the space of differential forms and in the space of polynomials of @xmath76 s .
we present the exterior product @xmath77 the operator of differentiation @xmath78 and
the two superpositions of the above operators @xmath79 @xmath80 here @xmath81 stays instead of @xmath57 of eq.([dk1 ] ) , used by k " ahler . introducing the notation with @xmath82 and @xmath83 we not only point out the similarities between the two approaches but also the two possibilities for the clifford product - only one of them used by k " ahler .
both @xmath84 and @xmath85 are clifford products on p - forms , while @xmath32 @xmath11 are the corresponding linear operators operating on the space of polynomials of @xmath3 s .
one easily finds the commutation relations , if for both approaches the generalized form of commutators , presented in eq.([gen ] ) , are understood @xmath86 @xmath87 here @xmath29 is the metric of space - time .
the vacuum state @xmath88 is defined as @xmath89 @xmath90 now we can define the dirac - like equations for both approaches : @xmath91 @xmath92 since @xmath93 and @xmath94 , while @xmath95 and @xmath96 , it is obvious that @xmath97 plays in the p - form formalism the role of the derivative with respect to a differential @xmath98form , similarly as @xmath99 does with respect to a grassmann coordinate .
taking into account the above definitions , one easily finds that @xmath100 @xmath101 both vectors , the @xmath0 , which are the superpositions of differential p - forms and the @xmath102 , which are polynomials in @xmath103 s are defined in a similar way ( eqs .
( [ * ] , [ up ] ) ) , as we shall point out in the following subsection .
we see that either @xmath104 or @xmath105 similarly as either @xmath106 or @xmath107 can represent the dirac - like equation . the superpositions of p - forms on which the dirac - k " ahler equation is defined and the superpositions of polynomials in grassmann space , on which the dirac - like equations are defined , are @xmath108 @xmath109 the coefficients @xmath110 depend on coordinates @xmath2 in both cases and are antisymmetric tensors of the rank @xmath111 with respect to indices @xmath112 .
the vector space is in both cases @xmath113 dimensional .
both , @xmath115 and @xmath116 define the algebra of the @xmath18 matrices and so they do both @xmath117 and @xmath118 .
one would thus be tempted to identify @xmath119 but there is a large freedom in defining what to identify with the gamma - matrices , because except when using @xmath120 as a parity operation you have an even number of gamma matrices occurring in the physical applications . then you may multiply all the gamma matrices by some factor provided it does not disturb their algebra nor their even products .
we shall comment this point in section 6 .
this freedom might be used to solve , what seems a problem : having an odd grassmann character , neither @xmath32 nor @xmath11 should be recognized as the dirac @xmath18 operators , since they would change , when operating on polynomials of @xmath3 , polynomials of an odd grassmann character to polynomials of an even grassmann character .
one would , however , expect - since grassmann odd fields second quantize to fermions , while grassmann even fields second quantize to bosons - that the @xmath18 operators do not change the grassmann character of wave functions .
one can notice , that similarly to the grassmann case , also the two types of the clifford products defined on p - forms , change the oddness or the evenness of the p - forms : an even p - form , @xmath121 , is changed by either @xmath115 or @xmath122 to an odd @xmath123-form , with either @xmath124 , if @xmath125 is not included in a p - form , or @xmath126 , if @xmath52 is included in a @xmath1-form , while an odd p - form , @xmath127 , is changed to an even @xmath59-form or @xmath128-form .
we shall later therefore propose that accordingly @xmath129 are recognized as the dirac @xmath18 operators operating on the space of @xmath1-forms or polynomials of @xmath3 s , respectively , since they both have an even grassmann character and they both fulfill the clifford algebra @xmath130 of course , the role of @xmath131 and @xmath132 can in either the k " ahler case or the case of polynomials in grassmann space , be exchanged .
whether we define the gamma - matrices by ( [ eq25 ] ) or ( [ naive ] ) makes only a difference for an odd products of gamma - matrices , but for applications such as construction of currents @xmath133 or for the lorentz generators on the spinors @xmath134 $ ] only products of even numbers of gamma - matrices occur , except for the parity representation on the dirac fields , where the @xmath120-matrix is used alone .
this @xmath120-matrix has to simulate the parity reflection which is either @xmath135 the ugly gamma - matrix identifications ( [ eq25 ] ) indeed perform this operation . and
as long as the physical applications are the ones just mentioned - and that should be sufficient - the choice ( [ eq25 ] ) is satisfactory : living in the grassmann odd part of the hilbert space , we do nt move into the grassmann even part of it .
the canonical quantization of grassmann odd fields , that is the procedure with the hamiltonian and the poisson brackets , then automatically assures the anticommuting relations between the operators of the fermionic fields .
the simplest solution to the problem with the evenness and oddness is to use the naive gamma - matrix identifications ( [ naive ] ) and simply ignore that the even - odd - ness does not match .
this is what k " ahler did , we can say , in as far as he did not really identify the even - odd - ness of the p - forms with the statistics of dirac fields . if one - along the lines of the becher s and joos s paper ( @xcite ) - will make a second quantized theory based on the k " ahler trick one does not proceed by insisting on taking p - forms to be fermionic only when p is odd . becher and joos take all the forms as fermion fields and assume then anticommuting relations for operators of fields .
this simplest solution can thus be claimed to be the one applied by k " ahler and used by becher and joos : they simply do not dream about in advance postulating that the p - forms should necessarily be taken to be boson or fermion fields depending on whether p is even or odd .
it is only when one as one of us in her model has the requirement of canonical quantization saying that the @xmath3 s should be grassmann odd objects , which indeed they are , that the problem occurs .
again , we are presenting the generators of the lorentz transformations of spinors for both approaches @xmath136 differing among themselves in the definition of @xmath13 only , which define the generators of the lorentz transformations in the internal space , that is in the space of @xmath1-forms or polynomials of @xmath3 s , respectively . while k " ahler suggested the definition @xmath137 in the grassmann case @xcite the operator @xmath138 is one of the two generators defined above ( eq .
( [ s ] ) ) , that is @xmath139 = -\frac{i}{4 } [ \tilde{\gamma}^a , \tilde{\gamma}^b ] , \quad { \rm or } \qquad { \cal s}^{ab } = \tilde{\tilde{s } } { } ^{ab } = -\frac{i}{4 } [ \tilde{\tilde{a } } { } ^a , \tilde{\tilde{a } } { } ^b ] .
\label{eq29}\ ] ] one further finds @xmath140 = i(\eta^{ac } \tilde{a}^b - \eta^{bc } \tilde{a}^a ) , \qquad [ \tilde{\tilde{s } } { } ^{ab } , \tilde{\tilde{a } } { } ^c ] = i(\eta^{ac } \tilde{\tilde{a } } { } ^b - \eta^{bc } \tilde{\tilde{a } } { } ^a ) , \quad { \rm while } \quad [ \tilde{s}^{ab } , \tilde{\tilde{a } } { } ^c ] = 0 = [ \tilde{\tilde{s } } { } ^{ab } , \tilde{a}^c].\ ] ] one can also in the k " ahler case define two kinds of the lorentz generators , which operate on the internal space of p - forms , according to two kinds of the clifford products , presented above .
following the definitions in the one of us @xcite approach , one can write the @xmath13 for the k " ahler case @xmath141 = -\frac{i}{4 } [ \tilde{\gamma}^a , \tilde{\gamma}^b ] , \quad { \rm or } \quad \tilde{\tilde{\cal s } } { } ^{ab } = \frac{i}{4 } [ dx^a\ ; \wedge -\ ; e^a , \ ; dx^b \ ; \wedge - \;e^b ] . \label{eq31}\ ] ] not only are in this case the similarities between the two approaches more transparent , also the definition of the generators of the lorentz transformations in the space of p - forms simplifies very much .
one further finds for the spinorial case @xmath142 = 0 , \quad { \rm for } \quad m^{ab } = l^{ab } + \tilde{s}^{ab},\ ] ] which demonstrates that the total angular momentum for a free massless particle is conserved .
the above equation is true for both approaches and the generators of the lorentz transformations @xmath49 fulfill the lorentz algebra in both cases .
in addition , the operators of the lorentz transformations with the vectorial character can also be defined for both approaches in an equivalent way , that is as a sum of the two operators of the spinorial character @xmath143 which again fulfill the lorentz algebra .
the operator @xmath144 if being applied on differential p - forms , transforms vectors into vectors , correspondingly @xmath145 , if being applied to polynomial of @xmath3 s transforms vectors into vectors @xcite .
elements of the lorentz group can be written for both approaches , for either spinorial or vectorial kind of the generators as @xmath146 where @xmath147 are parameters of the group .
if @xmath49 are equal to either @xmath148 or @xmath149 , the period of transformations is @xmath150 either in the space of differential forms or in the grassmann space , demonstrating the spinorial character of the operator .
if @xmath49 is the sum of @xmath151 and @xmath152 , the period of transformation is @xmath153 , manifesting the vectorial character of the operator . in the way how we have defined the operators in the space of @xmath1-form , the definition of the `` hodge star '' operator , defined by k " ahler working in the space of p - forms and the space of @xmath3 polynomials
, will be respectively @xmath154 with @xmath155 equal to either @xmath156 in the k " ahler case , or to @xmath157 in the one of us @xcite approach .
for an even d the factor with double tilde ( @xmath158 ) can be in both cases omitted ( @xmath159 , again we could distinguish the operators @xmath160 and @xmath161 in both cases , according to the elements , which define the casimir ) .
it follows that @xmath162 are the two operators , which when being applied on wave functions defined either on p - forms or on polynomials in grassmann space , project out the left or right handed component , respectively .
one easily recognizes that when being applied on a vacuum state @xmath163 , the operator @xmath160 behaves as a `` hodge star '' product , since one finds for @xmath19 even @xmath164 in the manko s approach @xcite the scalar product between the two functions @xmath165 and @xmath166 is defined as follows @xmath167 here @xmath168 is the weight function @xmath169 which operates on the first function , @xmath170 , only while @xmath171 no summation over repeated index is meant and @xmath172 since @xmath173 , @xmath174 means the complex conjugation and @xmath175 means the hermitian conjugation , then with respect to the above defined scalar product the operator @xmath176 @xmath177 while @xmath178 and @xmath179 .
again no summation over repeated index is performed .
accordingly the operators of the lorentz transformations of spinorial character are self - adjoint ( if @xmath180 and @xmath181 ) or anti - self - adjoint ( if a = 0 or b = 0 ) .
according to eqs.([scalpt ] , [ funv ] ) the scalar product of two functions @xmath182 and @xmath183 can be written as follows @xmath184 in complete analogy with the usual definition of scalar products in ordinary space .
k " ahler @xcite defined in eq .
( 15.11 ) and on page 519 the scalar product of two superpositions of p - forms @xmath185 and @xmath186 as follows @xmath187 which ( for real coefficients @xmath188 ) agrees with eq.([grasscal ] ) .
we shall limit ourselves in @xmath16 and in spinorial case ( as indeed k " ahler did ) .
the representations for higher d , analyzed with respect to the groups @xmath189 , and some other groups , in grassmann space are ( only for grassmann even part of the space belonging to the groups which does not include @xmath190 ) presented in ref.@xcite . in the case of @xmath191 one
may arrange the space of @xmath192 vectors into four times two weyl spinors , one left ( @xmath193 ) and one right ( @xmath194 ) handed .
we are presenting this vectors , which are at the same time the eigenvectors of @xmath195 and @xmath196 , as polynomials of @xmath197 s , @xmath198 .
the two weyl vectors are connected by the operation of @xmath199 operators ( eq.([eq25 ] ) ) .
taking into account that @xmath200 , where @xmath201 is the vacuum state ( eq.([vac ] ) ) , we find [ cols=">,>,^,>,>,>",options="header " , ] table i .- irreducible representations of the two subgroups @xmath202 of the group @xmath190 as defined by the generators of the spinorial character @xmath203 and the operator of handedness @xmath204 .
the four copies of the weyl bispinors have either an odd or an even grassmann character .
the generators @xmath205 , @xmath206 , transform the two copies of the same grassmann character one into another .
similarly also the k " ahler spinors can be arranged into four copies .
we find them by only replacing in table i. @xmath32 by @xmath207 .
we shall discuss this point also in the next section .
analysing the irreducible representations of the group @xmath190 in analogy with the spinor case but taking into account the generator of the lorentz transformations of the vector type ( eqs.([vecs ] , [ vecsk ] ) ) one finds @xcite for d = 4 two scalars ( a scalar and a pseudo scalar ) , two three vectors ( in the @xmath208 representation of @xmath190 usually denoted by @xmath209 and @xmath210 representation , respectively , with @xmath211 equal to @xmath212 ) and two four vectors ( in the @xmath208 representation of @xmath190 both denoted by @xmath213 and differing among themselves in the grassmann character ) all of which are eigenvectors of @xmath214 . using eq.([vecsk ] ) and analyzing the vector space of p - forms in an analogous way as the space of the grassmann polynomials , one finds the same kind of representations also in the k " ahler case .
both , in the spinor case and in the vector case one has @xmath215 dimensional vector space .
one may quite strongly wonder about how it is at all possible that there appear the dirac equation - usually being an equation for a _ spinor _ field - out of models with only scalar , vector and tensor objects ! immediately one would say that it is of course sheer impossible to construct spinors such as dirac fields out of the integer spin objects such as the differential one forms and their external products or of the @xmath3 s and their products @xmath216 .
let us say immediately that it also only can be done by a
this cheat really consists in _ replacing _ the lorentz transformation concept ( including rotation concept ) by exchanging the lorentz generators @xmath13 by the @xmath6 say ( or the @xmath15 if we choose them instead ) , see equations ( [ eq29 ] , [ eq31 ] ) .
this indeed means that one of the two kinds of operators fulfilling the clifford algebra and anticommuting with the other kind - it has been made a choice of @xmath217 in the k " ahler case and @xmath218 in the approach of one of us - are put to zero in the operators of lorentz transformations ; as well as in all the operators representing the physical quantities .
the use of @xmath219 or @xmath220 in the operator @xmath221 is the exception only used to simulate the grassmann even parity operation @xmath222 and @xmath223 respectively .
the assumption , which we call cheat was made in the k " ahler approach @xcite and in its lattice version @xcite , as well as in the approach of one of us @xcite . in ref .
@xcite the @xmath11 s are argued away on the ground that with a certain single particle action @xmath224 ( with @xmath2 being ordinary coordinates , @xmath3 grassmann coordinates , @xmath225 , @xmath226 an ordinary time parameter and @xmath227 an anticommuting time parameter and assuming @xmath228 and making a choice for @xmath229 ) with which we shall not go in details here , the @xmath11 appear to be zero as one of the constraints .
this constraint has been used to put @xmath230 s equal zero in the further calculations in this reference and it was used as argument for dropping the @xmath231-part of the lorentz generator @xmath232 .
is an odd grassmann operator - can not at all be fulfilled .
] let us stress that once the @xmath11 or @xmath233 is dropped and accordingly _ the @xmath15 is dropped _ - for whatever reason - _ one is no longer asking for the representation under the same lorentz transformations ( including rotations ) and one shall not expect to find say integer spin even if the field considered is purely constructed from scalars , vectors and tensors _ !
let us point out further that what happens is that as well the @xmath3 polynomials of one of us as the linear combinations of p - forms in the k " ahler approach can be formulated as _
double spinors _ , i.e. expressions with two ( dirac ) spinor indices , @xmath234 and @xmath235 say , and that the cheat consists in _ dropping _ from the concept of lorentz transformations the transformations in _ one of these indices_. in fact we can rewrite : for the even d case one has @xmath236
@xmath237 while for the odd d case one has : @xmath238 @xmath239 with the convention @xmath240 here the sums run over the number @xmath111 of factors in the products of @xmath241 or @xmath242 coordinates , a number , which is the same as the number of gamma - matrix factors and it should be remarked that we include the possibility @xmath243 which means no factors and is taken to mean that the product of zero @xmath244 or @xmath3-factors is unity and the product of zero gamma matrices is the unit matrix .
the indices @xmath245 are the spinor indices and taking the product of gamma - matrices conceived of as matrices the symbol @xmath246 stands for an element in the @xmath234-th row and in the @xmath235-th column .
there is an understood einstein convention summation over the contracted vector indices @xmath247 , k=1,2, ... ,i .
the gamma - matrices are in the even dimension case @xmath248 by @xmath248 matrices and in the odd dimension case @xmath249 by @xmath249 matrices . in the odd case
we have worked with two ( slightly ) different gamma - matrix choices - and thus have written the gamma - matrices as depending on the sign @xmath250 as @xmath251 - namely gamma matrix choices obeying @xmath252 the @xmath253 matrices should be constructed of course so that they obey the clifford algebra @xmath254 and we could e.g. choose @xmath255 for an even dimension @xmath256 , while for an odd dimension @xmath257 the gamma matrix @xmath258 has to be included @xmath259 with @xmath260 .
( see e.g. ( @xcite ) ) .
the above metric is supposed to be euclidean . for
the minkowski metric @xmath261 has to be taken , if the index @xmath262 is recognized as the `` time '' index .
we shall make use of the minkowski metric , counting the @xmath18 from @xmath263 , and assuming the metric @xmath264 . in this notation we can see that for fixed values of the index @xmath235 we obtain one of the four bispinors in table i. conceived of as a spinor in the index @xmath234 and with the understanding that the @xmath32 in the table lead to the corresponding @xmath3 , when acting on the vacuum state .
the equivalent table for the k " ahler approach follows by replacing @xmath265 by @xmath241 .
it is our main point to show that the action by the operators @xmath266 or @xmath32 and @xmath267 or @xmath11 in the representation based on the basis @xmath268 or @xmath269 with @xmath270 transforms the index @xmath234 and @xmath235 , respectively , of the basis @xmath271 or equivalently @xmath272 as follows : @xmath273 @xmath274 @xmath275 @xmath276 which demonstrates the similarities between the spinors of the one of us approach and the k " ahler approach : the operators @xmath115 and @xmath32 transform the left index of the basis @xmath277 , or correspondingly of the basis @xmath278 , while keeping the right index fixed and the operators @xmath279 and @xmath230 transform the right index of the basis @xmath280 , or correspondingly of the basis @xmath281 and keep the left index fixed . under the action of either @xmath282 and @xmath32 or @xmath283 and @xmath230 the basic functions transform as spinors .
the index in parentheses @xmath284 is defined for only odd d. we can count that the number of spinors is @xmath24 either in the manko s approach or in the k " ahler s approach ; the d dimensional grassmann space or the space of p - forms has @xmath24 basic functions .
we shall prove the above formulas for action of the @xmath32 and @xmath230 .
the proof is also valid for the k " ahler case if @xmath32 is replaced by @xmath285 and @xmath230 by @xmath233 .
let us first introduce the notation @xmath286 with @xmath287 we recognize that @xmath288 and @xmath289 using the first equation we find @xmath290 the index @xmath284 has the meaning for only an odd @xmath19 .
that is why we put it in parenthesis .
we may accordingly write @xmath291 with @xmath292 in ascending order and with @xmath293 in descending order
. then we find , taking into account that @xmath294 , @xmath295 , where @xmath296 is a vacuum state and eq.([eq7 ] ) @xmath297 @xmath298 using the above relations we further find @xmath299 where @xmath300 is @xmath301 , which depends on the operator @xmath302 and the dimension of the space .
we find in a similar way @xmath303 @xmath304 which finally gives @xmath305 with the signum @xmath306 depending on the dimension of the space and the operator @xmath307 .
we have therefore proven the two equations which determine the action of the operators @xmath32 and @xmath230 on the basic function @xmath308 .
according to the eqs.([t],@xmath309[tt ] ) it is obvious that the @xmath18 * matrices , entering into the dirac - k " ahler approach or one of us approach for spinors , have an odd grassmann character * since both , @xmath310 and @xmath32 as well as @xmath311 and @xmath230 , have an odd grassmann character .
they therefore transform a grassmann odd basic function into a grassmann even basic function changing fermion fields into boson fields .
it is clear that such @xmath18 matrices are not appropriate to enter into the equations of motion and lagrangeans for spinors .
there are several possibilities to avoid this trouble @xcite .
one of them was presented in section 4 .
if working with @xmath310 or @xmath32 alone , putting @xmath312 or @xmath11 in the hamiltonian , lagrangean and all the operators equal to zero , the @xmath313 matrices of an even grassmann character can be defined as proposed in eq.([eq25 ] ) @xmath314 or @xmath315 which fulfill the clifford algebra @xmath316 , while as we already have said @xmath317 .
$ ] we then have @xmath318 one can check that @xmath313 have all the properties of the dirac @xmath18 matrices .
( exchanging @xmath319 or @xmath32 by @xmath320 or @xmath321 , respectively , the gamma - matrices defined as @xmath322 have again all the properties of the dirac @xmath18 matrices . )
we shall comment in this section the discrete symmetries of spinors and vectors in the hilbert space spanned over either the grassmann coordinate space or the space of differential forms from the point of view of the one particle states of massless dirac ( that is the weyl ) particles . in oder to define the discrete symmetries of the lorentz group we introduce the space inversion @xmath323 and the time inversion @xmath324 operator in ordinary space - time in the usual way .
we shall assume the case @xmath16 .
@xmath325 with the metric @xmath326 already defined in section 2 . since one wants the time reversal operator to leave @xmath327 , that is the zero component of the ordinary space - time momentum operator ( @xmath328 ) , unchanged ( @xmath329 ) , while the space component @xmath330 should change sign ( @xmath331 ) , one also requires @xmath332 we first shall treat spinors . having the representation of spinors expressed in terms of polynomials of @xmath3 s in table i. , which also represents the corresponding superpositions of p - forms
if @xmath3 is accordingly substituted by @xmath333 , we expect each of the four copies of dirac massless spinors to transform under discrete symmetries of the lorentz transformations in an usual way .
the parity operator @xmath323 should transform left handed spinors with @xmath334 to right handed spinors with @xmath335 , without changing the spin of the spinors .
this is what @xmath221 ( eq.([eq25 ] ) ) does for any of four copies of the dirac massless spinors , which are the weyl bispinors of table i. , separately .
the time reversal operator @xmath324 should transform left handed spinors with @xmath334 and spin @xmath75 to left handed spinors with @xmath334 and spin @xmath336 , what the operator @xmath337 does when applied to any of four copies of the dirac spinors of table i. this transformation involves only members of the same copy of the dirac bispinor .
the operators @xmath313 which are defined in eq.([eq25 ] ) , have due to the appropriate choice of phases of the spinors of table i , the usual chiral matrix representation ( for both approaches - the k " ahler and the one of us ) .
one would , however , expect that the time and the space reversal operators should work in both spaces - that is in the ordinary space - time and in the space of either grassmann polynomials or in the space of p - forms - in an equivalent way @xmath338 @xmath339 @xmath340 @xmath341 and changing equivalently the momenta conjugate to coordinates in either the one of us or the k " ahler approach . applying the transformation @xmath323 of eq.([ctn ] ) on any of four copies of the dirac bispinors of table i. ,
one obtains the same result as in the above , that is the standard definition of the space - reversal operation . applying the transformation @xmath324 of eq.([ctn ] ) on ,
let us say , the first spinor of the first copy of the dirac bispinors of table i. ( that is on @xmath342 ) , one obtains the last spinor of the last copy ( that is @xmath343 ) .
the left handed spinor with spin @xmath336 transformed to the left handed spinor of spin @xmath75 , just as it did under the usual time - reversal transformation , except that in this case the copy of spinors has been changed .
one can write down the matrix representation for this second kind of the time - reversal transformation .
if we choose for the basis the first copy of bispinors of table i. and the fourth copy of bispinors of table i , we obtain the matrix : @xmath344 where the @xmath345 , due to the choice of the phase of the spinors in table i. , and @xmath346 means that the complex conjugation has to be performed on the phase coefficients only , which in our case have again been chosen to be one .
this is the time - reversal operation discussed by weinberg @xcite in appendix c of the weinberg s book .
when vectors and scalars are treated in the similar way for either of the two approaches , it turns out that the time - reversal operators do not transform one copy into another one .
we payed attention in this section on only spin degrees of freedom .
the complex conjugation affects , of course , the higher part of the internal space as well , affecting the charges of spinors , vectors and tensors , if one thinks of the extension @xcite as discussed in section 9 .
we want to look at the funny shift of the spin compared to the a priori spin for a field by shifting a priori generators @xmath347 out by anther set @xmath348 as a general nice idea .
a prerequisite for that working is that the difference between the two proposals for lorentz generators @xmath349 is also a conserved set of quantities . in the notation above of course
we find @xmath350 assuming that there is indeed such two lorentz generator symmetries in a model , we can ask for the representation under both for a given set of fields , and we can even ask for representation under the difference algebra @xmath351 . in order to shift in going from @xmath49 to @xmath352 from integer spin to half integer spin
the representation for the fields in question must at least be spin 1/2 for @xmath351 . actually in the cases we discussed the @xmath351 were in the dirac spinor representation .
but that means that the representation of the fields which shift representation going from @xmath49 to @xmath352 have to belong under @xmath351 to at least a spin 1/2 which means at least the weyl spin representation of the lorentz group , and that has @xmath353 dimensions . but
that means then that a given representation of the final @xmath352 lorentz group always must occur in at least @xmath354 families .
we have discussed the connection between the grassmann @xmath3 formulation and the k " ahler formalism for general dimension @xmath19 and thus we could apply it simply in the d=4 case , or we could use it in extended models with extra dimensions .
one should note that the connection between the spinor and the forms is such that for each extra two dimensions the number of components of a dirac - spinor goes up by a factor 2 , and at the same time the number of families also doubles .
this agrees with that adding one extra @xmath3 doubles the number of terms in the @xmath3 polynomials and thus adding two would make this number four times as big .
let us now study the application of the extra degrees of freedom which consists in supposing the k " ahler degrees of freedom or equivalently the grassmann @xmath3 s we discussed to the case where the @xmath19 dimensional space is used in a kaluza - klein type model .
that is to say we here look at a kaluza - klein model extended with @xmath3 s or the forms , much more rich than usual kaluza - klein .
it has long been suggested @xcite that special kinds of rotations of the spins especially in the extra @xmath355 dimensions manifest themselves as generators for charges observable at the end for the four dimensional particles . since both the extra dimension spin degrees of freedom and the ordinary spin degrees of freedom originate from the @xmath3 s or the forms we have a unification of these internal degrees of freedom .
we can say then that the generators rotating these degrees of freedom , namely the just mentioned charges acting as hinger dimensional spins ( at high energy ) and the 4-dimensional spin , are unified . such rotations of the internal spin degrees of freedom would in order to correspond to a kaluza - klein gauge fields with massless gauge bosons have to represent full symmetries of the vacuum state , i.e. they should as in usual kaluza - klein correspond to killing - vectors , but with the further degrees of freedom also corresponding to symmetry for the latter .
so at the end we may consider also the charges associated with the internal spin as ordinary klauza - klein charges , of course in the sense of being for the very rich model considered here .
but of course unless we have the @xmath3 or forms degrees of freedom one could risk that the gauge field from such symmetry could be practically decoupled .
let us now look at what the families found in the dirac - k " ahler will develop into in case we use it for a kaluza - klein type model , as just proposed : usually the number of surviving massless fermions into the ( 3 + 1 ) space consists only of those which are connected with
zero - modes .
this is to be understood so that we imagine weyl particles in the high ( @xmath19 ) dimensional space because of an atiah - singer theorem in @xmath355 dimensional staying compactified
space ensures some modes with the extra dimension part of the dirac operator gets zero for some number of modes - for each @xmath19-dimensional family . if the model had a strength for compact space atiah - singer theorem @xmath356 and if the dimension of the full space , the number of @xmath3 s , is @xmath19 , so that the number of families at the @xmath19-dimensional level becomes @xmath248 , the total number of at low energy observable families should be @xmath357 as an example take the model @xcite which has @xmath358 and at first - at the high energy level - @xmath359 lorentz group , but which should be broken to ( in two steps ) to first @xmath360 and then to @xmath361 .
we may see the appearance of equally many ( namely @xmath362 ) right handed and left handed
flavours in the k " ahler model as an expression for the no go theorem@xcite for putting chiral charge conserving fermions on the lattice in as far as we could make attempts to make lattice fermions along the lines of becher and joos @xcite .
in fact it would of course have been a counterexample to the no go @xcite theorem if there had been a different number of right and of left weyl particle species in the becher - joos model , because in the free model the number of particles functions as a conserved charge .
as is very well known the becher - joos model really is just the kogut - susskind @xcite lattice fermion model , it is also well known that it does not violate the no go theorem @xcite and this is because there is this species doubling , which can be interpreted as the flavours . becher and joos show that the kogut - susskind lattice description of dirac fields is equivalent to the lattice approximation of the dirac - k " ahler equation .
( see page 344 in the becher - joos@xcite article ) .
this kogut - susskind model is one that gives us dirac particles , but we can seek to get to weyl particles in a naive @xmath363 ( or @xmath204 or @xmath364 in the usual notation ) projecting way , but of course now such a projection would have to be translated into the language with the vector and scalar fields in the k " ahler s formulation , and it is rather easy to see @xcite that requiring only one @xmath363 projection implies that the coefficient to one p - form say @xmath365 should relate ( just by a sign @xmath366 ) to that of the by the hodge star @xmath367 associated @xmath368 ( see subsection 4.5 ) . actually we easily see that requiring the restriction that @xmath369 in the language of k " ahler becomes @xmath370 if we want like joos and becher to put the theory on the lattice there is a difficulty in just imposing this constraint , because the natural relation imposed by the hodge star @xmath367 on the lattice would go from lattice to the dual lattice and we could not identify without a somewhat ambiguous choice the @xmath367dual of a given lattice element , so as to impose the self duality condition .
could we possibly invent a way to circumvent the no go theorem @xcite for chirality conserving fermions on the lattice by making the species doublers bosons instead of fermions , both having though spin 1/2 ? in the formulation by one of us which we have related to the k " ahler formulation there is ( naturally ) assigned different grassmanian character to different components of the wave function .
in fact the wave function with coefficients to monomial terms that are products of different sets of ( mutually different ) @xmath3-variables - in the sense of course that a polynomial is given by its coefficients - , and thus the coefficients to the products with an even number of factors have different grassmannian character from those of the odd number of factors .
that actually is in the theory of one of us somewhat of an embarrassing reason for a super selection rule , which though may be overcome by taking into account the charges related to extra dimensions appearing in that model .
but here we now want to point out the hope that these very grassmann character problems may be used as a new idea to circumvent the no go theorem .
in fact we could hope for that spin @xmath371 and say left handed flavour appear with fermionic statistics ( the grassmann odd character ) while spin @xmath371 flavour with bosonic statistics would appear as right handed , and that even on the lattice .
the way that manko @xcite chooses to quantize the system , that is a particle moving in ordinary and grassmann coordinate space , is to let the wave function be allowed to be any function of the @xmath19 gassmann variables @xmath3 , so that any such function represents a state of the system . but in this quantization the @xmath11 s can not be put weakly to zero . in other words that quantization turned out not to obey the equation expected from expression for the canonical momentum @xmath22 , being proportional to the coordinate @xmath3 as derived from the lagrangian . if , however , in the operators such as the hamiltonian and the lorentz transformation operators @xmath11 s are just put strongly to zero , so that all the operators only depend on @xmath32 , while either @xmath117 or @xmath372 fulfill the clifford algebra : @xmath373 and @xmath374 , the expressions obtained after having put the @xmath11 s to zero describe spinor degrees of freedom . in particular , only the operators @xmath6 are used as the lorentz generator .
one has accordingly the new lorentz transformations instead of the a priori one in the wave function on grassmann space quantization used . in that case
the argument for only having integer spin breaks down , what the calculations indeed confirm to happen .
we should now attempt to get an understanding of what goes on here by using a basis inspired from the dirac - k " ahler - construction , which is a way often used on lattices to implement fermions on the lattice .
the dirac - k " ahler construction starts from a field theory with a series of fields which are 0-form , 1-form , 2-form , ... , d - form .
they can be thought of as being expanded on a basis of all the wedge product combinations of the basis @xmath375 , @xmath376 , ... , @xmath377 for the one - forms , including wedge products from zero factors to d factors . in the dirac - k "
ahler construction one succeeds in constructing out of these `` all types of forms '' @xmath248 dirac spinor fields .
this construction is without a cheat
impossible in much the same way as manko s approach ought to be .
we have pointed out clearly in this paper how this cheat
occurs in both approaches , showing up all the similarities of the two approaches and using the simple presentation of the quantum mechanics in grassmann space to not only simplify the dirac - k " ahler approach but also to generalize it .
we have shown in particular that in both approaches besides the ( two kinds of ) generators for the lorentz transformations for spinors also the generators for vectors and tensors exist .
there are four copies of the weyl bispinors .
one kind of the spinorial type of the lorentz transformations defines the weyl spinors , another kind transforms one copy of weyl spinors into another of the same grassmann character .
we also have shown the two kinds of the time reversal operators , as well as the fact that in grassmann space or space of differential forms of d dimensions , @xmath378 , spins and charges unify .
we pointed out the necessity of defining the gamma - matrices of an even grassmann character .
this work was supported by ministry of science and technology of slovenia as well as by funds nbi - he - 99 - 35 , chrx - ct94 - 0621 , intas 93 - 3316 , intas - rfbr 95 - 0567 .
n. manko bortnik , phys . lett .
* b 292 * , 25 ( 1992 ) ; n. cimento * a 105 * , 1461 ( 1992 ) ; j. math . phys . * 34 * , 3731 ( 1993 ) ; int .
phys . * a 9 * 1731 ( 1994 ) ; j. math . phys . * 36 * , 1593 ( 1995 ) ; mod . phys .
lett . * a 10 * , 587 ( 1995 ) ; hep - th9408002 ; hep - th9406083 ; n. manko bortnik , s. fajfer , n. cimento , * 112 b * 1637(1997 ) proceedings of the international conference quantum systems , new trends and methods , minsk , 23 - 29 may , 1994 , p. 312
, ed . by a.o .
barut , i.d .
feranchuk , ya.m .
shnir , l.m .
tomilchik , world scientific , singapore 1995 ; proceedings of the us - polish workshop physics from plank scale to electroweak scale , warsaw , 21 - 24 sept .
1994 , p. 86
, ed . by p. nath , t. taylor , s. pokorski , world scientific , singapore 1995 ; proceedings of the @xmath379 adriatic meetings on high energy physics , brijuni , croatia , 13 - 22 sept.1994 , p. 296
d. klabuar , i. picek , d. tadi ' c , world scientific , singapore 1995 ; proceedings of the barut memorial conference on group theory in physics , tr . j. of phys.*21 * 321 ( 1997 ) , proceedings to the international workshop on what comes beyond the standard model , bled , slovenia , 29 june-9 july 1998 , ed . by n. manko bortnik , h. b. nielsen , c. froggatt , dmfa zalonitvo 1998 , p. 20 ; proceedings of the @xmath380 international conference symmetry methods in physics , dubna 10 - 18 july 1995 , p. 385
, ed . by sissakian , g.p .
pogosyan , publ .
, joint institute for nuclear research , 14 19 80 dubna , isbn 5 85165 453 8 .
h. b. nielsen , n. manko bortnik , proceedings to the international workshop on what comes beyond the standard model , bled , slovenia , 29 june-9 july 1998 , ed . by n.
manko bortnik , h. b. nielsen , c. froggatt , dmfa zalonitvo 1998 , p. 68 , ijs.tp.99/17 or nbi - he-99 - 35 or cern - th/99 - 288 , hep - th/9909169 . | we compare the way one of us got spinors out of fields , which are a priori antisymmetric tensor fields , to the dirac - k " ahler rewriting .
since using our grassmann formulation is simple it may be useful in describing the dirac - k " ahler formulation of spinors and in generalizing it to vector internal degrees of freedom and to charges .
the cheat
concerning the lorentz transformations for spinors is the same in both cases and is put clearly forward in the grassmann formulation .
also the generalizations are clearly pointed out .
the discrete symmetries are discussed , in particular the appearance of two kinds of the time - reversal operators as well as the unavoidability of four families . |
magnetic tunnel junctions ( mtj ) have gained considerable interest in recent years due to their high potential in various applications , e.g. as reads heads @xcite , angle @xcite or strain sensors @xcite and as programmable resistance in data storage ( mram ) @xcite or even magnetic logic devices @xcite .
.,width=264 ] the underlaying concept is a spin valve consisting of a hard magnetic reference electrode separated from the soft magnetic sense or storage layer by a tunnel barrier like al@xmath1o@xmath2 .
the reference layer usually is an artificial ferrimagnet ( afi ) exchange biased by a natural antiferromagnet , in which the afi consists of two ferromagnetic layers coupled antiparallel via a thin non - magnetic spacer . for the soft electrode ,
mostly single layers of polycrystalline material , e.g nife and cofe , have been used @xcite . recently , soft electrodes of polycrystalline afis , based on ferromagnetic materials like cofe and nife , have been investigated .
they show a further reduction of the stray field due to the reduced net moment , smaller switching field distribution @xcite and an easier establishment of a single domain structure in patterned elements with small aspect ratio @xcite .
additionally , the concept of an afi allows one to further adjust the magnetic properties of the soft layer .
compared to the coercivity of a continuous single layer , @xmath3 , the coercivity of the afi , @xmath4 , is enhanced by a factor @xmath5 : @xmath6 where @xmath7 and @xmath8 are the saturation magnetization and the thickness of the two composite ferromagnetic layers.@xcite as shown before for unpatterned afi films consisting of two amorphous , ferromagnetic cofeb layers , separated by a thin ru spacer , the coercivity can be tailored in a wide range and is approximately by a factor of nine smaller than in systems of polycrystalline cofe / ru / cofe.@xcite furthermore the coupling of the amorphous cofeb - afi shows an oscillating behavior in dependence of the thickness of the nonmagnetic ru spacer and achieves a coupling strength of the order of @xmath9 at the second antiferromagnetic maximum , which is about a factor of ten smaller than in polycrystalline cofe / ru / cofe trilayers.@xcite it was the purpose of this study to investigate the switching behavior of the amorphous cofeb - afi at sub-@xmath0 m sizes , where additional shape anisotropy and the magnetostatic edge coupling have to be taken into account .
these contributions lead to an _ effective anisotropy _ which is for patterned elements different from the anisotropy of continuous films ( eqn .
[ qvalue ] ) . as the amorphous alloy we chose co@xmath10fe@xmath11b@xmath11 due to a high tunnel magnetoresistance ( tmr ) @xcite and an enhanced temperature stability of the tmr @xcite .
a thin ru spacer was used to mediate the coupling between the two ferromagnetic layers of the afi .
.[tab : table ] investigated samples and parameters extracted from the agm measurements of the unpatterned layer systems . [ cols="^,^,^,^,^ " , ]
samples have been deposited by magnetron sputtering on thermally oxidized sio@xmath1 wafers at a base pressure of @xmath12 mbar . a magnetic field of approximately 4 ka / m was applied during deposition in order to induce the easy axis in the magnetic layers .
the afi was grown on a 1.2 nm thick al layer , oxidized in an ar / o@xmath13 plasma for 0.8 min without breaking the vacuum , to have similar growth conditions as in a mtj . in order to investigate the switching behavior in dependence on the q - value ,
three different samples have been prepared where the thickness of the ferromagnetic layer in contact with the al@xmath1o@xmath2 layer , and thus the net magnetic moment , has been varied ( see table [ tab : table ] , samples a to c ) .
additional a single cofeb layers with a thickness of 4 nm ( sample d ) has been deposited for comparison .
all samples were capped with a ta layer to protect the multilayers from oxidation .
the samples have been patterned by a single step e - beam lithography and ar - ion etching process . for lithography
a positive e - beam resist was used , leading to patterns with a small edge roughness and high reproducibility across the whole array . on each sample different arrays of ellipses with a nominal width , @xmath14 , of 250 nm and varying length , @xmath15 , have been defined , leading to different aspect ratios , @xmath16 , between 1.3 and 8 .
the lateral distances have been chosen three times the dimension of the elements , and therefore a dipolar coupling between the individual ellipses of an array can be neglected.@xcite each of the arrays had the dimension of @xmath17 .
after the arrays have been coated by a ta layer of appropriate thickness ( ranging from 8 to 15 nm ) , the capping was removed in a lift - off process in a bath of solvent under application of ultrasonic agitation . during etching with a 80 @xmath18 ar ion current
the samples were tilted by approximately 30 degrees and rotated to obtain a uniform etch profile over the sample .
the etching depth was monitored by a secondary ion mass spectrometer ( sims ) attached to the etching facility .
the sizes and the uniformity of all patterns have been characterized by scanning electron microscopy ( sem ) after the patterning process .
the sem image in figure [ fig : sem ] shows one of the arrays of sample d with ellipses in the size of @xmath19 , confirming the high uniformity of the patterns .
the width for samples with small aspect ratios ( @xmath20 ) varied between 250 and 270 nm . due to a tendency of over - exposure , ellipses with larger aspect ratios show a slightly larger widths of approximately 300 nm .
after patterning all samples have been field annealed for 20min . at 150@xmath21 and
@xmath22 applied parallel to the long axis of the ellipses using a vacuum annealing furnance .
magnetization loops of all arrays have been taken by a commercial magneto - optical kerr effect magnetometer ( moke ) with a typical spot diameter of 4@xmath0m.@xcite , and the plateau field , @xmath23 , are indicated for sample a.,width=302 ]
a room temperature magnetization curve , @xmath24 , of the antiferromagnetically ( af ) coupled systems is shown in figure [ fig : agmloops ] . from the @xmath24 loops , obtained by alternating gradient field magnetometery ( agm )
, one can extract the saturation field , @xmath25 , the total , @xmath26 , and the net magnetic moment , @xmath27 , allowing to calculate the measured q - value , @xmath28 , the individual magnetization of the layers , @xmath29 , and the coupling energy @xcite @xmath30 the coupling is @xmath31 for all investigated afi samples .
the values are in accordance to the oscillating coupling in dependence on the ru spacer thickness around the second antiferromagnetic maximum as presented elsewhere.@xcite the measured q - values vary between 3.7 and 7.7 for the investigated afi samples , and are significantly smaller than the @xmath5 value , @xmath32 , calculated from the nominal thickness of the ferromagnetic layers .
this discrepancy is supposed to result from a thicker magnetically dead layer of the upper cofeb layer in comparison to the bottom layer . since the samples are well protected from oxidation , as confirmed by auger depth profiling
, this would indicate a stronger intermixing of the upper cofeb interfaces .
all data extracted from the agm measurements are given in table [ tab : table ] .
figure [ fig : majorloopspatterned ] shows the magnetization loops of patterned arrays of all afi samples as obtained by moke measurements .
the strong af coupling is maintained after the patterning and annealing steps .
additionally , the saturation field is increasing with decreasing aspect ratio ( i.e. length or size ) of the ellipses , due to an increase in stray field coupling between the layers within the afi system . for large aspect ratios @xmath25 achieves the values of the unpatterned samples .
the saturation field can be expressed by two contributions , one originating from the antiferromagnetic interlayer coupling , the other resulting from the stray field coupling . whereas the first should depend on @xmath33 , as derived from eqn .
[ coupling ] , the latter should depend on @xmath34,@xcite where @xmath35 is the total thickness of the afi .
the second contribution only depends on the demagnetization factor @xmath36 , since the @xmath37-components in case of an af coupled system are compensated for external fields larger than the plateau field , @xmath23 .
since the width of the elements was hold constant for the investigated samples , the dependence of @xmath25 on the sample geometry is only given by the demagnetization factor@xcite @xmath38 this model is fitted to the measured data as shown in figure [ fig : hsathc](a ) .
the fitting shows a high accordance between the experimental data and the model , thus verifying the dependence of the saturation field on @xmath39 for small aspect ratios . from the fitting parameters one
can further extract the saturation field for an infinite elongated ellipse to 28.6 ka / m for sample a , 24.2 ka / m for sample b and 22.8
ka / m for sample c , respectively .
this values are in good agreement to the measured saturation fields at the unpatterned afi samples given by the interlayer coupling ( see table [ tab : table ] ) . for all samples.,width=264 ] for external fields smaller than @xmath23 , the af coupling remains stable . as can be seen from figure [ fig : majorloopspatterned ] , the afi should reverse its magnetization like a single layer sample with a reduced net moment and enhanced effective anisotropy
. therefore it is possible to measure minor magnetization loops in a small field window ( @xmath40 to @xmath41 ka / m , depending on @xmath23 of the sample ) and extract the coercivity , @xmath42 , of the arrays ( see figure [ fig : hsathc](b ) ) . , and ( b ) coercivity , h@xmath43 , of patterned samples vs. the aspect ratio @xmath44 .
the width of approximately 250 to 270 nm was kept constant for all investigated samples .
the dashed lines in ( a ) represent a fit with a function proportional to the demagnetization factor @xmath39 .
( c ) numerical calculated demagnetization factors for ellipses against the aspect ratio.,width=302 ] for small aspect ratios ( @xmath45 ) , @xmath42 increases with @xmath46 by approximately 3.5 ka / m and remains constant for large aspect ratios ( see figure [ fig : hsathc](b ) ) .
the slight decrease in coercivity for the largest aspect ratios is most likely attributed to the slightly larger width of the ellipses .
this behavior of the coercivity vs. aspect ratio holds also for the single layer sample .
the increase of @xmath42 for small aspect ratios is caused by the increase in shape anisotropy , which for an ellipse is given by@xcite @xmath47 where the demagnetization factor ( @xmath48 ) depends on the aspect ratio .
therefore the experimental results of figure [ fig : hsathc ] ( b ) are in qualitative accordance with the ( @xmath48 ) dependence on aspect ratio ( see figure [ fig : hsathc](c ) )
. deviations from the calculated dependence for large aspect ratios are likely to result from micromagnetic differences : small elements are stabilized by a non vanishing stray field , arising from the magnetic poles of the elements .
these stray fields stabilize the overall magnetization of the elements , so that the reversal can be more accurately approximated by a single domain mechanism . for large aspect ratios
the magnetic poles are larger separated , thus minimizing the stray field coupling and as a result a nucleation driven magnetization reversal , most likely initiated by edge domains , is more favourable , leading to a almost constant coercivity.@xcite unlike elliptic elements with a single ferromagnetic layer , the coercivity of the patterned afi samples does not scale proportional to the net magnetic moment . with higher net moment
, the coercivity is _ decreasing _ , but remaining always larger than for elements of a 4 nm thick single layer .
the reason is a superposition of the effective anisotropy due to the af coupling ( @xmath49 ) , as expressed by eqn .
[ qvalue ] for the unpatterned films , and the dependence of coercivity on the net magnetic moment as described by eqn .
[ sun ] ( @xmath50 ) .
overall the influence of the af coupling is dominating for the investigated cofeb afis and the coercivity is increased by approximately a factor of 1.4 when decreasing the net thickness of the afi , and therefore the net moment , from 1.5 to 0.5 nm .
, on the net thickness , @xmath51 , of the afi and a single layer , respectively .
the experimental data points are for elliptical shaped elements with the dimension of @xmath52 , the simulated data are evaluated for the same sample geometry .
lines are guides to the eye.,width=264 ] the dependence of coercivity on the net thickness , @xmath53 , of the afi and for single layer samples with layer thicknesses of @xmath54 have been simulated within a micromagnetic model using landau - lifshitz - gilbert equations ( see figure [ fig : simulation]).@xcite for the simulation of the afis the thickness of the second fm layer , @xmath55 , was varied between 3.5 and 6 nm , whereas the thickness of the first fm layer was kept fixed to @xmath56 , and the thickness of the nonmagnetic spacer was chosen to 1 nm . the saturation moment of the fm layers were assumed to @xmath57 , the uniaxial anisotropy to @xmath58 , the exchange stiffness constant to @xmath59 and for the af coupling constant to @xmath60 .
the coercivity of the simulated elliptic afi elements of @xmath52 decreases with net thickness for @xmath61 nm , and increases for larger @xmath62 , asymptotically reaching the values of the single layer coercivity .
the experimental data of the afi arrays with the same geometry show a similiar behavior , therefore verifying the above described model for the samples under study .
, illustrating also the gain in effective anisotropy by introduction of the antiferromagnetic coupling and the strong influence of the increased q - value on the anisotropy of the system.,width=325 ] figure [ fig : asteroids ] shows the bias field dependence of the coercivity for samples patterned with an aspect ratio of @xmath63 . due to the increase of
effective anisotropy with the q - value ( and therefore basically with the inverse net magnetic moment ) , the asteroid gets stretched along the hard axis field .
if used in a conventional writing scheme as a soft magnetic electrode for applications like mram or magnetic logic , the broad asteroid shape of the afi storage layer cells limits the choice of q - value due to field limitations , reducing the proposed advantages of an afi structure as reported by others.@xcite
the afi based on an amorphous cofeb alloy shows a strong increase in the effective ansiotropy due to the af coupling , mediated by a thin ru interlayer and by the stray field coupling , respectively .
this is reflected by a higher coercivity and an increase of the asteroid width compared to a single cofeb layer .
the dependence of the switching field on the net magnetic moment can not be explicitely explained within the model that considers the afi as a rigid ferromagnetic layer with a reduced moment .
one has further to take into account the increase of effective anisotropy , which basically scales with the inverse net moment for the investigated combinations of layer thicknesses .
this last factor appears to be dominating in the system under study and leads to the decrease of coercivity with net moment .
finally , it has been found , that the saturation field of the patterned afis is decreasing with larger aspect ratio and is asymptotically reaching the saturation field of the unpatterned films . the origin for this behavior for small aspect ratios can be found in the additional contribution of the stray field coupling of the two ferromagnetic layers within the afi .
therefore the antiparallel alignment of the system is additionally favoured by the stray field coupling , depending basically on the demagnetization factor @xmath39 vs. aspect ratio .
the authors wish to thank j. bangert and g. gieres for fruitful discussions , h. mai and k. rott for experimental support . financial support of the german ministry for education and research is gratefully acknowledged ( grant no .
13n8208 ) . | this work reports on the magnetic reversal of sub-@xmath0 m sized elements consisting of an cofeb / ru / cofeb artificial ferrimagnet ( afi ) .
the elements were patterned into ellipses having a width of approximately 250 to 270 nm and a varying aspect ratio between 1.3 and 8 .
the coercivity was found to decrease with an increasing imbalance of the magnetic moment of the two antiferromagnetically coupled layers and is therefore strongly affected by an increase of effective anisotropy due to the antiferromagnetic coupling of the two layers . with respect to a single layer of amorphous cofeb , patterned in comparable elements , the afi has an increased coercivity .
switching asteroids comparable to single layers were only observed for samples with a high net moment . |
efforts to implement quantum computation have recently intensified .
the application of the quantum bit ( qubit ) to solid - state materials , such as superconducting josephson junctions @xcite and quantum dots @xcite , is particularly promising , because these implementations have the advantage of scalability . in a coupled - dot system , for example ,
the localized states in the left and right dots are treated as a basic two - level system , in which the tunnel coupling between the two dots constructs a quantum superposition of the dots .
this superposition manifests itself in coherent quantum oscillation ( rabi oscillation ) , and transition can be induced between the superposed states @xcite . a nanometer - scale superconducting electrode connected to a reservoir via
a josephson junction constitutes an artificial two - level system in which the charge states , coupled by tunneling , differ by @xmath0 , where @xmath1 is the electronic charge .
this system has shown clear rabi oscillation @xcite .
quantum coherence must be maintained during quantum gate operation .
dephasing , characterized by the dephasing time , originates from various couplings between the qubit and the environment .
when the qubit is implemented in a solid - state system , the effects of phonons and electromagnetic and background charge fluctuations ( bcfs ) are important in the dephsing process .
the effect of phonons has been examined in semiconductor quantum dots as the source of the dephasing accompanying dissipation @xcite ; the effect of electromagnetic fluctuation in josephson junction qubits has been extensively studied @xcite .
however , bcfs have not yet been examined systematically , in spite of their importance in the dephasing process .
bcfs have been observed in many systems @xcite . in nanoscale systems , they are the electrostatic potential fluctuations due to the dynamics of electrons or holes trapped at impurity sites . in particular , the charge of a single impurity fluctuates with the lorentzian spectrum form , which is called `` random telegraph noise '' in the time domain @xcite .
the random distribution of the positions of such impurities and their time constants lead to bcfs or 1/f noise @xcite . in solid - state charge qubits , these bcfs lead to a dynamical electrostatic disturbance and hence the dephasing .
the effect of 1/f noise on a charge josephson qubit has been examined theoretically - the interaction between the qubit and environment has been treated by the perturbation method @xcite , by gaussian approximation @xcite and by the path integral method within a spin - boson model @xcite .
when fluctuating impurities exist in the substrate , not in the junction @xcite , the coupling between the qubit and bcfs is weak , and the perturbation method is sufficient .
however , when the interaction between the qubit and environment is strong , methods that go beyond perturbation are needed . in this study , we investigated how the electrostatic disturbance of time constant @xmath2 coming from a single impurity affects the quantum coherence of a qubit irrespective of the strength of the qubit - impurity coupling .
we also examine the effect of many impurity sites .
this approach is in clear contrast with previous ones @xcite , in which the phenomenological spectral weight of the boson thermal bath was used to characterize the effect of bcfs .
we consider two types of couplings between the qubit and the environment : pure dephasing and dephasing accompanied with relaxation of the population @xcite . in symmetrical coupled - dot systems ,
the former corresponds to a fluctuation in the tunnel coupling constant , and the latter to that in the asymmetric bias @xcite .
the mapping from a coupled - dot system to the josephson charge qubit is discussed in sec .
vi . by using the method of stochastic differential equations ,
we obtain analytically the dephasing rate , which is shown to be always smaller than @xmath3 .
it should be noted that this dephasing process does not mean the qubit becomes entangled with the environment , but rather it means the stochastical evolution of an external classical field , suppressing the off - diagonal density matrix elements of the qubit after being averaged over statistically distributed samples .
section ii defines the hamiltonian of the system .
section iii explains the method of stochastic differential equations .
the fluctuations in tunnel coupling and asymmetric bias are examined in secs .
iv and v , respectively .
section vi is devoted to discussion , including the effect of many impurities .
section vii summarizes the paper .
the qubit and the effect of a single impurity are examined in terms of the following hamiltonian : @xmath4 where @xmath5 and @xmath6 are the creation and annihilation operators of the left and the right dots , assuming a single level for each dot , as shown in fig .
the @xmath7 is the tunnel coupling between the dots , and @xmath8 is the asymmetric bias between them .
the interaction between a qubit and the charge at the impurity site is described by the following hamiltonian : @xmath9 where @xmath10 is the magnitude of the fluctuation in the tunnel coupling , @xmath11 is the magnitude of the fluctuation in the asymmetrical bias , and @xmath12 and @xmath13 are the creation and annihilation operators of the charge at the impurity site , respectively .
we assume that the time evolution of statistical variable @xmath14 is a poisson process . assuming a strong coupling between the charge at the impurity site and the nearby electron reservoir , the dynamics of the charge induces not gaussian white noise , but random telegraph noise @xcite .
we therefore consider the effect of random telegraph noise with characteristic time constant @xmath2 , where the statistical variable @xmath15 takes the value 1 or -1 ( fig .
1(b ) ) . the time constant is determined by the barrier height of the electron or hole trap and the temperature , like @xmath16 , where @xmath17 , @xmath18 , @xmath19 , and @xmath20 are the activation energy of the impurity potential , boltzmann s constant , the temperature , and the temperature - independent prefactor , respectively @xcite .
when the temperature decreases , @xmath2 becomes longer .
the telegraph noise has been experimentally observed to take a value of 1 or -1 with asymmetric probabilities , which arises from the difference between the fermi energy of the electron reservoir and the energy level of the impurity sites .
@xcite . to include this asymmetric weight , we introduce asymmetric probabilities @xmath21 and @xmath22 , which relate the asymmetric transition rates for the process from -1 to 1 ( @xmath23 ) to those of the opposite process ( @xmath24 ) .
we neglect the backaction from the qubit to the charge at the impurity site , so this environment reduces to a classical stochastic external field .
we also assume that the temperature is high enough for the effect of the quantum fluctuation of the charge between the impurity site and electron reservoir to be neglected . for compactness , we rewrite the hamiltonian in terms of the pauli matrices while rotating basis @xmath25 from the basis of localized states in the left and right dots to bonding - antibonding basis : @xmath26 in the following , we consider only the case of a symmetrical coupled - dot system , @xmath27 , in which the effect of the bias fluctuation due to the dephasing starts only from the second order , @xmath28 , as shown in the following , and is less effective in the perturbation regime . as reduced hamiltonian eq .
( [ eqn : hm ] ) suggests , the present results can also be applied to other quantum two - level systems in which telegraph - type fluctuation exists ( see sec .
we are interested in the time - evolution of the qubit s two - by - two density matrix , @xmath29 , with an arbitrary initial condition at @xmath30 , @xmath31 . if bcf is absent , starting with the initial condition that the left dot is occupied , for example , the density matrix at @xmath30 is given by @xmath32 where @xmath33 is the matrix of rotation from the localized basis to the bonding - antibonding basis . at time t , @xmath34 where the bases of the inner matrix are the left and right occupancy states .
therefore , the density matrix shows rabi oscillation with frequency @xmath7 . in the following ,
we keep the matrix indices in the bonding - antibonding basis . to examine the instantaneous potential change
, we use the method of stochastic differential equations @xcite .
the density matrix averaged over all possible sequences of telegraph noise can be represented as a series : @xmath35 where @xmath36 is the distribution of @xmath37 with the probability of @xmath21 for @xmath38 and that of @xmath22 for @xmath39 , with the constraint @xmath40 .
the density matrix before the ensemble average , @xmath41 , is given by @xmath42 where @xmath43 is the unitary time evolution operator , which is determined by @xmath44 the explicit form of @xmath45 is given by @xmath46 where @xmath47 .
equation ( [ eqn : emeqn ] ) can be rewritten in terms of the integral equation @xmath48 using @xmath49 , we define matrix @xmath50 as follows : @xmath51 we can then reduce eq .
( [ eqn : prorho ] ) to the following compact form : @xmath52 \nonumber \\ + & \frac{1}{\tau_0 } \int_0^{\tau } & dt { \rm exp } [ - \frac{\tau - t}{\tau_0 } ] { \rm tr } [ r^{im } ( \tau , t ) \rho ( t)].\end{aligned}\ ] ]
first , we consider the case of fluctuation in tunnel coupling ( @xmath53 ) . since the interaction hamiltonian commutes with @xmath54 , the environment leads to pure dephasing without energy dissipation .
we derive @xmath45 and @xmath55 from eqs .
( [ eqn : smatrix ] ) and ( [ eqn : rmatrix ] ) as follows : @xmath56 \delta_{kl } , \nonumber \\
\\ r_{lk}^{nm } ( \tau - t ) & = & \int_{x } { \rm exp } [ i \delta ( \tau - t ) + i j_t x ( \tau - t ) ] \nonumber \\ & \times & \delta_{nk } \delta_{lm } d w ( x ) .
\nonumber \\\end{aligned}\ ] ] as a result , the off - diagonal element of the density matrix obeys the following integral equation : @xmath57 \sigma ( t)dt,\ ] ] where @xmath58 is a normalized off - diagonal element of the qubit density matrix measured in a rabi oscillation frame , and @xmath59 + p_d { \rm exp } [ - i j_t ( \tau - t ) ] .\ ] ] equation ( [ eqn : rint ] ) can be rewritten as a differential equation : @xmath60 the initial conditions are @xmath61 and @xmath62 . as a result of coupling with the environment , the off - diagonal element of the density matrix decays as a function of time . when the real part of the two roots of the characteristic equation of eq .
( [ eqn : odepure ] ) almost completely degenerates , the short - time behavior for @xmath63 ( initial regime ) is not a simple exponential decay . in this initial regime ,
the off - diagonal element of the density matrix shows gaussian decay , @xmath64 , irrespective of the asymmetric probabilities , @xmath21 and @xmath22 .
the decay of the off - diagonal element of the density matrix becomes exponential for the asymptotic regime , @xmath65 when @xmath66 . for @xmath67 , this criterion
is obtained when one of the two exponential decay terms becomes negligibly small . for @xmath68 ,
the time constants of the envelope of the two dumped oscillating terms are the same , @xmath69 .
exponential decay appears after the inverse of the oscillation frequency : @xmath70 .
when @xmath66 and @xmath71 , one obtains , @xmath72 , where the dephasing can never be a simple exponential decay . the time constant of this exponential decay corresponds to the dephasing time , @xmath73 . for @xmath74 @xcite , @xmath75.\ ] ] figure 2 shows the @xmath76 dependence of dephasing rate @xmath77 for @xmath78 , 0.6 , 0.7 , 0.8 , and 0.9 . in the two limits of weak and strong coupling
, we have @xmath79 namely , for a fixed @xmath80 , when @xmath2 increases from 0 , @xmath77 first increases and then decreases .
it has a single maximum of @xmath81 when @xmath82 ; therefore , for any parameters , @xmath83 .
changing weight @xmath21 to make it more asymmetric reduces @xmath77 . in the limit of @xmath84 or 1 ,
the environment is nearly frozen , so the dephasing time becomes infinity .
next we examine the effect of bias fluctuation ( @xmath85 ) .
we consider only the case of symmetrically weighted telegraph noise ( @xmath86 ) for simplicity . in this model , @xmath54 and the interaction hamiltonian do not commute and the dephasing process is accompanied by a relaxation of the population .
the unitary operator is thus @xmath87 where @xmath88 is the nutation frequency . after averaging , the few non - vanishing elements of matrix @xmath89
have the form @xmath90 , % r_{12}^{12 } \end{aligned}\ ] ] where @xmath91 the time evolution of the diagonal element of this matrix is determined by the following differential equation @xcite : @xmath92 here , @xmath93 , and the initial conditions are @xmath94 and @xmath95 ; and @xmath96 depends on the initial population having an arbitrary value between 1 and -1 . the time constant of the exponential decay of @xmath97 is the relaxation time of the population , @xmath98 .
when @xmath99 ( weak coupling ) , @xmath100 .
when @xmath101 ( strong coupling ) , @xmath102 .
the obtained differential equation for the off - diagonal element is @xmath103 where @xmath104 is the normalized off - diagonal element of the qubit density matrix .
the initial conditions are @xmath105 where @xmath106 is the initial phase of the off - diagonal density matrix element .
differential equation ( [ eqn : eqt2 ] ) with initial condition eqs .
( [ eqn : eqt2ini1 ] - [ eqn : eqt2ini3 ] ) has explicit solutions : @xmath107 \nonumber \\ & & \times ( c_{ic } \cos \phi + i c_{is } \sin \phi ) + i e^{-u } \sin \phi , \end{aligned}\ ] ] where @xmath108 . for @xmath109 ,
the coefficients are given by @xmath110 and the @xmath111 are the three solutions of @xmath112 where @xmath113 , @xmath114 , and @xmath115 .
we will show one special case and three asymptotes .
@xmath116 * isolated system*:since eq .
( [ eqn : lambda ] ) has solutions @xmath117 and @xmath118 , the coefficients are determined as @xmath119 , @xmath120 and @xmath121 .
therefore , we get @xmath122 , which is simply a natural rotation of the off - diagonal element of the density matrix . @xmath123 * fast modulation*:in this asymptotic case , the roots of eq .
( [ eqn : lambda ] ) are @xmath124 and @xmath125 . after determining coefficients @xmath126
, we obtain @xmath127 therefore , the off - diagonal element of the qubit density matrix decays exponentially at a rate of @xmath128 .
@xmath129 * slow modulation*:the roots are @xmath130 and @xmath131 , so @xmath132 and @xmath133 . in particular , for @xmath134 ( strong coupling limit ) , @xmath135 these apparent different time dependences between the real and imaginary parts stem from the choice of coupling in the form @xmath136 .
if we choose the form @xmath137 instead , the time dependences of the real and imaginary parts are interchanged . in this strong coupling limit
, the time evolution of @xmath138 explicitly depends on its initial phase , @xmath106 , not the simply like @xmath139 .
therefore , as will be discussed later , if there are several such impurities , the total time evolution of @xmath138 is not the simple product of each impurity s contribution . for @xmath140 ( weak coupling ) , we have @xmath141 in this case , we have exponential decay with @xmath142 . for @xmath143 * weak coupling or preservative regime * , have @xmath144 , and @xmath145 , so @xmath146 .
therefore , we again have exponential decay with @xmath147 .
this coincides with the redfield result , which was obtained by perturbation theory and is justified in the weak coupling case , @xmath148 @xcite . taking the limit @xmath149 further , we restore the result for @xmath123 , fast modulation . to summarize , the dephasing rate is given by @xmath150 in all regimes , @xmath83 . here
, we show the results of solving eqs .
( [ eqn : eqt1 ] ) and ( [ eqn : eqt2 ] ) numerically .
figure 3 shows the @xmath151 dependence of @xmath152 and @xmath153 when @xmath154 ( weak coupling ) along with the asymptotic curves obtained analytically .
figure 4 shows the @xmath151 dependence of @xmath97 and @xmath155 in the case of strong coupling .
it also shows the asymptotic envelope for @xmath97 .
figure 5 shows the @xmath151 dependence of @xmath156 in the case of strong coupling .
it also shows the asymptotic curve . as shown in figs . 3 , 4 , and 5 , in the two contrasting limits ,
the numerical and analytical results coincide very well .
it should be noted that we do not find gaussian decay of the off - diagonal element of the density matrix for the initial regime , in contrast to the fluctuation in the tunneling coupling constant .
next , we examine the @xmath2 dependence of the relaxation rate , particularly for @xmath157 .
figure 6 shows the @xmath2 dependences of @xmath158 and @xmath77 when @xmath159 and @xmath160 .
the limits of the long and short @xmath2 fit well with the analytical asymptotic given by eq .
( [ eqn : t2b ] ) .
the @xmath2 dependence of the relaxation time is fitted as @xmath161 when @xmath162 , the relaxation rates increase with @xmath2 .
when @xmath163 , the rates decrease with an increasing @xmath2 .
the shape of the dephasing rate as a function of @xmath164 is explained as follows .
when @xmath165 , many dephasing events occur during one rabi oscillation cycle , each event leads to collective disturbance . because the long time constant of telegraph noise leads to large fluctuations in the variance of the rotating angle in the bloch sphere during rabi oscillation , the dephasing time decreases with an increasing @xmath2 .
when @xmath163 , rabi oscillation occurs over more than one cycle in time @xmath2 and , in this regime , each dephasing event is independent .
hence , the dephasing time increases with @xmath2 .
the maximum around @xmath166 is a kind of resonance .
we considered the effect of electrostatic disturbance due to background charge fluctuations ( bcfs ) . to summarize our findings , for pure dephasing ( @xmath167 ) , @xmath168 and @xmath169 . for dephasing with relaxation of the population ( @xmath170 ) , @xmath171 .
next , we discuss the relationship between the experimentally observed @xmath73 and our results . in the present study ,
the dephasing time with a single background charge was found to be longer than the time constant of the telegraph noise for both tunneling and bias fluctuations .
the observed time constant of a dominant random telegraph noise is about 30 @xmath172 or longer @xcite , so a rather long dephasing time is expected .
however , in another experiment , the dephasing time was about 1 ns @xcite .
therefore , a single telegraph noise source may not be enough to explain the experimental results ; we should thus consider the effect of many impurity sites or other additional effects .
we consider the effect of many impurities for the case in which there is no correlation between background charges .
flucuation in the tunnel coupling arises from the modulation of the wave function in the coupled dots .
the gradient of the electrostatic potential around the tunneling barrier , which comes from an electron or hole located at an impurity site , leads to a change in the tunnel coupling as well as a change in the asymmetric bias @xcite . for a charge state that couples with a coupled - dot system symmetrically
, the pure dephasing event is critical . for dephasing accompanied with relaxation of the population , the unitary operators of each impurity , which lead to dephasing of the qubit , are not commuting .
however , by neglecting the higher order @xmath173 s in the dephasing rate , we can take the ensemble sum of the effect of each charge state .
in such a weak coupling case , @xmath174 , we can use eq .
( [ eqn : t2bl ] ) , and the simple summation of the dephasing rate is expressed by @xcite @xmath175 where @xmath176 is the distribution width of the thermal activation energy of the charge states , @xmath177 is the distribution function of @xmath11 , which depends on the relative position between the qubit and impurity site , and @xmath178 is the sum over the random impurities . in the second equation , we assume uniform distribution of the activation energies of the background charges , @xmath179 ; for typical cases , @xmath180 is approximately 23 @xcite .
use of the perturbation method showed that the dephasing rate of a josephson charge qubit in terms of @xmath173 is proportional to the inverse of @xmath181 in the limit of @xmath182 @xcite , where @xmath181 and @xmath183 are the josephson coupling constant and charging energy , respectively .
this is similar to the estimate of @xmath184 in eq .
( [ eqn : t2 t ] ) , where @xmath185 corresponds to @xmath7 . with a larger @xmath7 and lower temperature ,
the dephasing rate is lower .
we next estimate the magnitude of the fluctuations , @xmath11 .
the asymmetric bias fluctuation comes from asymmetric coupling between the two dots and the background charge , which is in the form of a dipole interaction , @xmath186 for @xmath187 , where @xmath13 is the distance between the two dots , @xmath188 is the distance between the coupled - dot system and the background charge , and @xmath189 is the angle between them @xcite .
therefore , for a smaller qubit or a charge located far from the qubit , the effect of bias fluctuation should be less important .
the dephasing rate is proportional to @xmath190 , which is estimated as @xmath191 for the impurity sites where the dipole approximation is appropriate , where @xmath192 is the radius beyond which the dipole approximation is valid , which depends on @xmath13 , @xmath193 is the density of impurity sites , and @xmath194 is the relative dielectric constant .
therefore , the total dephasing time is well defined .
the quality factor of a quantum logic gate is defined by the ratio of @xmath7 to @xmath184 : @xmath195 which represents how many gate operations can be done before the quantum coherence vanishes . from eq .
( [ eqn : q ] ) , we conclude that a large @xmath196 is needed for quantum computation .
we estimated q using @xmath197 @xmath198 , @xmath199 @xmath198 , @xmath200 @xmath201 ( characteristic parameter values for an experiment in which the quantum mechanical coupling of the dots was observed in the frequency domain @xcite ) , and @xmath202 ( for gaas ) . to enable quantum error correction ,
the lower bound of the necessary gate quality factor was roughly estimated as @xmath203 @xcite .
thus , density of charge states should be less than @xmath204 for fabrication .
if there is a correlation between impurities ( a screening effect ) , dephasing will be suppressed in general @xcite .
it should be noted that there might be strong dephasing from the nearby impurities for which dipole approximation is not adequate , even if these are only a few impurities , ( a few in this case ) .
the non - commutativity between the qubit hamiltonian and environment hamiltonian and the qubit backaction make it difficult to evaluate the dephasing rate for strongly coupled background charge fluctuations in the asymmetric bias case @xcite .
finally , we discuss the josephson charge qubit system @xcite . under an appropriate condition ( single - electron charging energy @xmath205 much larger than josephson coupling energy @xmath181 and temperature @xmath206 ) only two charge states in the cooper pair box ( cpb ) are important , and
the hamiltonian is given by @xmath207 where @xmath208 is the energy difference between the two charge states , and @xmath209 is the total gate - induced charge in the box .
the two - charge - state basis is expressed using pauli matrices , and @xmath210 is the coupling strength between the qubit and the background charge , which induces fluctuation in the charging energy .
the @xmath211 corresponds to the asymmetric bias @xmath8 , and @xmath212 corresponds to @xmath7 . here ,
@xmath213 @xmath214ev , and @xmath215 @xmath214ev @xcite ; if we can neglect @xmath181 , the pure dephasing event is critical . in pure dephasing ,
the effect of a large number of impurities is obtained by simply summing the dephasing rates , because @xmath54 and interaction hamiltonian commute .
when the background charge and cpb interact , the charging energy in the cpb fluctuates .
the spectrum of the fluctuation is given by @xmath216 where @xmath217 , and we take an ensemble average over the activation energy , as was done in the coupled - dot system .
the spectrum of the charging energy fluctuation was experimentally found to be @xmath218 , where @xmath219 @xcite . from this estimation , @xmath220 for @xmath221 @xmath222 .
for an initial regime , the envelope of rabi oscillation shows gaussian decay , namely , the off - diagonal element of the density matrix is given by @xmath223 .
the rate of gaussian decay is given by @xmath224 .
for the above value of @xmath225 , the time constant of the gaussian decay is given by @xmath226 @xmath227 , which is consistent with the experimental finding of @xmath228 @xmath227 , @xcite .
note that in the gaussian regime , the time constant does not depend on the temperature .
numerical calculation @xcite also suggests this type of gaussian decay for the pure dephasing case . at the charge degeneracy point , namely @xmath229
, dephasing with relaxation of the population occurs . in this case
, the dephasing rate is estimated using the same value of @xmath225 : @xmath230 so @xmath231 @xmath172 for @xmath232 @xmath222 . in a recent experiment ,
a longer coherence time of @xmath233 @xmath172 was found when the saddle point of the ground state energy was used as a function of @xmath209 and the flux @xcite .
we examined the effect of the fluctuation of a single charge in an impurity site on a qubit . using the method of stochastic differential equations , we calculated the time evolution of the ensemble averaged density matrix of the qubit and obtained analytical results for various conditions .
the dephasing time , @xmath73 , was always longer than the time constant of the random telegraph noise for both tunneling and bias fluctuations . for bias fluctuation , @xmath73 was twice the relaxation time of the population in the weak coupling case . to suppress the bias fluctuation ,
the coupled dots should be positioned closer together or the tunnel coupling should be made stronger .
we also investigated the case in which many impurity sites are distributed and examined the gate quality factor .
for pure dephasing , which corresponds to a josephson charge qubit experiment , the gaussian decay of the off - diagonal element of the density matrix dominated .
the present results can be applied to other quantum two - level systems in which there is telegraph - type fluctuations .
* acknowledgements * we thank j. s. tsai , t. tanamoto , n. imoto , m. koashi , h. nakano , t. fujisawa , y. hirayama , t. hayashi , c. uchiyama and k. harmans for their stimulating discussions .
this work was partly supported by crest - jst , interacting carrier electronics project .
y. nakamura , yu .
a. pashkin and j. s. tsai : nature * 398 * 786 ( 1999 ) .
d. loss and d. p. divincenzo : phys .
a * 57 * 120 ( 1998 ) .
t. tanamoto : phys .
a * 61 * 22305 ( 2000 ) .
a. miranowicz , s. k. zdemir , m. koashi , n. imoto and y. hirayama : phys .
a * 65 * 062321 ( 2002 ) . t. h. oosterkamp , t. fujisawa , w. g. van der wiel , k. ishibashi , r. v. hijman , s. tarucha and l. p.
kouwenhoven : nature * 395 * 873 ( 1998 ) .
t. fujisawa , t. h. oosterkamp , w. g. van der wiel , b. w. broer , r. aguado , s. tarucha and l. p. kouwenhoven : science * 282 * 932 ( 1998 ) .
a. shnirman , g. schn and z. hermon : phys .
* 79 * 2371 ( 1997 ) .
p. l. lafarge , p. joyez , h. pothier , a. cleland , t. holst , d. esteve , c. urbina and m. h. devoret : c. r. acad .
paris , * 314 * 883 ( 1992 ) .
g. zimmerli , t. m. eiles , r. l. kautz and j. m. martinis : appl .
. lett . * 61 * 13 ( 1992 ) . c. kurdak ,
c. j. chen , d. c. tsui , s. parihar , s. lyon and g. w. weimann : phys .
b * 56 * 9813 ( 1997 ) .
a. b. zorin , f .- j .
ahlers , j. niemeyer , t. weimann , h. worf , v. a. krupenin and s.v .
lotkhov : phys .
b * 53 * 13682 ( 1996 ) t. fujisawa and y. hirayama : appl
lett * 77 * 543 ( 2000 ) .
p. dutta and p. h. horn : rev .
phys . * 53 * 497 ( 1981 ) .
e. paladino , l. faoro , g. falci and r. fazio : phys .
lett . * 88 * 228304 ( 2002 ) .
a. shnirman , y. makhlin and g. schn : cond - mat/0202518 . y. nakamura , yu a. pashkin , t. yamamoto and j. s. tsai : phys . rev
. lett . * 88 * 047901 ( 2002 ) in the present study , dissipation means the relaxation of the population of bonding - antibonding states . on large time scales ,
the qubit state becomes mixed with equal populations of bonding and anti - bonding states , this relaxation is not an energy relaxation .
y. m. galperin and k. a. chao : phys .
b * 52 * 12126 ( 1995 ) .
r. bauernschmitt and yuli v. nazarov : phys .
b * 47 * 9997 ( 1992 ) .
y. imry : _ introduction to mesoscopic physics , _ oxford university press .
i. burshtein : soviet phys .
jetp * 22 * 939 ( 1965 ) .
a. i. burshtein : soviet phys .
jetp * 21 * 567 ( 1965 ) .
a. j. leggett , s. chakravarty , a. t. dorsey , m. p. a. fisher , a. garg and w. zwerger : rev .
phys . * 59 * 1 ( 1987 ) c. l. slichter : _ principals of magnetic resonance 3th ed , _ springer - verlag .
t. fujisawa , w. g. van der wiel and l. p. kouwenhoven : physica e * 7 * 413 ( 2000 ) . g. burkard , d. loss and d. p. divincenzo : phys .
b * 59 * 2070 ( 1999 ) . in the present study
, we assume the interaction between the coupled - dot system and charge state to be dipole - monopole coupling . when the electron reservoir screens the charge state effectively , the charge state and the reservoir act on the system as a dipole .
the interaction type is dipole - dipole , and the interaction strength decreases more rapidly as the distance between the charge state and system increases . a longer dephasing time is thus expected compared with that of a dipole - monopole interaction
. j. preskill : proc .
london a * 454 * 385 ( 1998 ) .
t. itakura and y. tokura : unpublished .
d. vion , a. aassime , a. cottet , p. joyez , h. pothier , c. urbina , d. esteve and m. h. devoret : science * 296 * 886 ( 2002 ) . | in quantum computation , quantum coherence must be maintained during gate operation .
however , in physical implementations , various couplings with the environment are unavoidable and can lead to a dephasing of a quantum bit ( qubit ) .
the background charge fluctuations are an important dephasing process , especially in a charge qubit system .
we examined the dephasing rate of a qubit due to random telegraph noise . solving stochastic differential equations
, we obtained the dephasing rate of a qubit constructed of a coupled - dot system ; we applied our results to the charge josephson qubit system .
we examined the dephasing rates due to two types of couplings between the coupled - dot system and the background charge , namely , fluctuation in the tunnel coupling constant and fluctuation in the asymmetric bias . for a strong coupling condition ,
the dephasing rate was inversely proportional to the time constant of the telegraph noise .
when there is fluctuation in the tunnel coupling constant , gaussian decay occurs in the initial regime .
we also examined the rate of dephasing due to many impurity sites . for a weak coupling condition with fluctuation in the asymmetric bias , the obtained dephasing rate coincided with that obtained by the perturbation method using the spectral weight of a boson thermal bath , which is proportional to the inverse of the frequency . 2 |
in the study of atoms with at least two electrons , bound states and resonances are of interest .
the resonances evolve into states with one free electron after a very short time .
this phenomenon is called auto - ionization of the atom . with a revival of interest in the auto - ionization
, fano published an appealing theoretical paper @xcite comprising an analysis of the excitation of the @xmath0 level of helium by electrons .
he argued that the natural line shape contains a zero .
later , the optical absorption spectra of the rare gases have been analyzed @xcite , while the paper @xcite is one of many studies dealing with the mechanism of atomic auto - ionization .
a unified approach to the configuration interaction and the influence of strong lasers have been expounded in @xcite . in this framework , the studies @xcite have been realized .
the quantum laser field has been taken into account in @xcite and the effect of the squeezed state has been studied in @xcite . the fano resonances can occur also in other physical settings .
the fano resonances in nanoscale structures can be mentioned @xcite .
the treatment of auto - ionization and the influence of laser may be extended to a simultaneous auto - ionization , the influence of laser , and to the interaction with a neighbor two - level atom @xcite .
the presence of a neighbor system may also considerably increase photo - ionization and recombination rates @xcite . in the analysis ,
the assumption of weak optical pumping is usually used and leads to a simpler behavior , cf .
, @xcite . in @xcite ,
the entanglement between an auto - ionization system and a neighbor atom is studied for a classical driving field . besides the possibility to calculate a measure of entanglement for the two atomic systems ,
a somewhat arbitrary , but systematic , filtering is adopted .
two frequencies can be selected in the auto - ionization system and the study of entanglement reduces to the well - known two - qubit problem . in this paper
, we modify this analysis by including the quantal nature of the field . in sec .
ii , we describe the model . in sec .
iii , we discuss photoelectron spectra and the density plots of entanglement measure .
iv provides conclusion .
we consider two mutually interacting atoms , @xmath1 and @xmath2 , in the presence of an electromagnetic field ( for the scheme , see fig . 1 ) .
to quantize the electromagnetic field , we have to add to the usual model annihilation and creation operators of the modes which participate in the radiative interactions . indeed ,
although only the frequency @xmath3 of optical field is considered , an infinity of modes at this frequency can be introduced .
we may suppose that the atom @xmath4 interacts with the mode @xmath5 and the atom @xmath6 interacts with the mode @xmath7 .
we complete the levels of the atomic system by the photon - number states , @xmath8 where @xmath9 is a photon number in the mode @xmath5 and @xmath10 is a photon number in the mode @xmath7 . in eq .
( [ 1 ] ) , @xmath11 ( @xmath12 ) is the ground state of the atom @xmath4 ( @xmath6 ) , @xmath13 is the excited state of the atom @xmath4 , @xmath14 is the auto - ionization state of the atom @xmath6 , @xmath15 @xmath16 @xmath17 is the continuum state of the atom @xmath6 , and @xmath18 is an energy difference between the ground state @xmath12 and the state @xmath15 . here
we have used the photon - number states @xmath19 , @xmath20 , @xmath21 , and @xmath22 simultaneously to indicate that the hilbert space of the states can be decomposed into invariant subspaces . for @xmath23 ,
these subspaces have a dimension equal to 6 .
each invariant subspace is a tensorial product of the subspaces corresponding to the jaynes
cummings model ( dimension 2 ) and the model due to leoski and buek ( dimension 3 ) @xcite .
the hamiltonian has the form @xmath24 where @xmath25 with @xmath26 and @xmath27 ( @xmath28 and @xmath29 ) being the photon annihilation ( creation ) operators .
the hamiltonian @xmath30 of atom @xmath2 with auto - ionizing level in eq . ( [ 2 ] ) is written as @xmath31 where @xmath32 means an energy difference between the ground state @xmath12 and the state @xmath14 .
symbol @xmath33 gives the strength of optical excitation from the ground state @xmath34 into the auto - ionization state @xmath35 , @xmath36 is the strength of optical excitation from the ground state @xmath34 of the atom @xmath6 into the continuum state @xmath37 , and @xmath38 describes the coulomb configuration interaction between the excited states of atom @xmath6 .
the hamiltonian of the neighbor two - level atom in eq .
( [ 2 ] ) reads @xmath39 where @xmath40 means an energy difference between the ground state @xmath11 and the state @xmath13 , @xmath41 is the strength of optical excitation from the ground state @xmath42 into the excited state @xmath43 . in eq .
( [ 2 ] ) , the hamiltonian @xmath44 characterizes the dipole
dipole interaction between the atoms @xmath4 and @xmath6 , @xmath45 where @xmath46 @xmath47 characterize energy transfer from the ground state @xmath12 into the state @xmath14 @xmath48 at the cost of the decay from the state @xmath13 into the state @xmath11 .
we note that if @xmath49 and @xmath50 , the hamiltonian @xmath51 describes uncoupled atoms .
we will treat the situation where the atoms @xmath4 and @xmath6 interact with a single mode @xmath52 , @xmath53 . in this case , the levels written in eq .
( [ 1 ] ) simplify , @xmath54 and @xmath55 is the number of photons in the mode @xmath52 . here
we have used the photon - number states @xmath56 , @xmath57 , and @xmath58 simultaneously to indicate that the hilbert space of the states can be decomposed into invariant subspaces . but in the case of a single mode , an invariant subspace can not be investigated as a tensorial product .
we can see from eq .
( [ 7 ] ) , that the atom @xmath4 at the level @xmath11 interacts with the field in the state @xmath56 or @xmath57 in the dependence on the state of the atom @xmath6 and the atom @xmath4 at the level @xmath13 interacts with the field in the state @xmath57 or @xmath58 in the dependence on the state of the atom @xmath6 .
the hamiltonian has the form @xmath59 where @xmath60 and @xmath61 following @xcite , we modify the schrdinger picture by considering the state vector @xmath62 in the form @xmath63 \nonumber\\ & & \mbox { } \times \bigl[c_{00}^{(n)}(t)|n\rangle_l|0,0\rangle_{ab}+ c_{10}^{(n-1)}(t)|n-1\rangle_l|1,0\rangle_{ab } \nonumber\\ & & \mbox { } + c_{01}^{(n-1)}(t)|n-1\rangle_l|0,1\rangle_{ab } \nonumber\\ & & \mbox { } + c_{11}^{(n-2)}(t)|n-2\rangle_l|1,1\rangle_{ab } \nonumber\\ & & \mbox { } + \int d_0^{(n-1)}(e_d , t)|n-1\rangle_l|0,e_{d}\rangle_{ad}\,de_d \nonumber\\ & & \mbox { } + \int d_1^{(n-2)}(e_d , t)|n-2\rangle_l|1,e_d\rangle_{ad}\,de_d\bigr ] , \label{11}\end{aligned}\ ] ] where @xmath64 .
the prime indicates that for @xmath65 some of the components must be omitted .
the components with @xmath66 have to be omitted for @xmath65 and those with @xmath67 have to be left out for @xmath68 .
it holds that @xmath69 $ ] @xmath70 @xmath71 for @xmath72 an invariant subspace @xmath73 is the eigenspace of the operator @xmath74 related to an eigenvalue @xmath55 .
we assume that @xmath75 . in this invariant subspace ,
the composite system is described by the equations @xmath76 where @xmath77 further @xmath78 @xmath79 we introduce the matrix @xmath80 and let @xmath81 and @xmath82 denote the eigenvalues of the matrix @xmath83 and @xmath84 , @xmath85 , be the eigenvalues of the matrix @xmath86 .
let us recall the possibility of decompositions @xmath87 where @xmath88 , @xmath89 are solutions of the equations @xmath90 similarly , @xmath91 , @xmath85 , are solutions of the equations @xmath92 in eqs .
( [ 22 ] ) and ( [ 23 ] ) , @xmath93 and @xmath94 @xmath70 @xmath95 are @xmath96 and @xmath97 unit matrices , respectively . the first vector of the components of the solution of eqs .
( [ 13 ] ) has the very simple form @xmath98 we introduce a @xmath99 matrix @xmath100 as the solution of the sylvester equation @xmath101 the solution has the form @xmath102 the dependence of the components of the amplitude spectrum on the initial state of the system with @xmath103 is @xmath104 { \bf t}^{(n)}(e_d ) \nonumber\\ & & \mbox { } - { \bf t}^{(n)}(e_d )
\exp\left[-\frac{i}{\hbar}{\bf m}^{(n)}t \right ] \biggr\}{\bf c}^{(n)}(0 ) .
\label{27}\end{aligned}\ ] ] we observe that @xmath105 where @xmath106 { \bf t}^{(n)}(e_d){\bf c}^{(n)}(0 ) .
\label{29}\ ] ] the increase of the diagonal terms by @xmath107 means that also the eigenvalues @xmath108 , @xmath109 , @xmath110 are raised by the same amount . in the relations like ( [ 26 ] ) ,
these increments mutually cancel and elsewhere they already correspond to the relation ( [ 11 ] ) .
all the previous exposition beginning with ( [ 14 ] ) should be modified for @xmath111 .
let us note that the initial vacuum field and the ground states of the atoms @xmath4 and @xmath6 , @xmath68 , do not lead to any transitions to the continuum states .
it holds that @xmath112 @xmath70 @xmath113 and the description reduces to the equation @xmath114 @xmath70 @xmath115 .
the reduction for @xmath116 is a consequence of non - existence of @xmath58 as in ( [ 11 ] ) and need not be made explicit .
let us note that transition to a continuum state can occur for @xmath116 , but not simultaneously with an excitation of the atom @xmath4 .
the long - time behavior is characterized by a complete ionization of the atom with an auto - ionizing level and by both the levels of the two - level atom @xmath1 being occupied .
the long - time behavior is periodical due to the dynamics of the two - level atom in the cw laser field . at all times
the spectra can be determined as the probability distribution of the two - level atom at its levels and of the atom with the auto - ionizing level in the continuum of levels . by the normalization , conditional spectra are defined .
the difference between the conditional spectra is an effect of the atomic quantum correlation .
the difference between the conditional spectra can be seen even in the case where the dipole dipole interaction of the atoms is missing .
an important case of the quantum correlation is the entanglement .
we measure this entanglement using the negativity .
the entanglement is conserved , even though we restrict the continuum of levels to two of them , in the most of the pairs of the selected frequencies .
we calculate the negativity of the partially transposed statistical matrix for two levels of the two - level atom and selected continuum levels of the atom with an auto - ionizing level . in the previous work ,
the functions @xmath117 of the parameters of the atom @xmath6 without a neighbor have been conveniently introduced . also the functions @xmath118 of the parameters of both the atoms have been defined .
here we conveniently introduce a one - photon version of the usual excitation parameter @xmath119 , @xmath120 where @xmath121 by the replacements @xmath122 , @xmath123 in the function @xmath124 , the function @xmath125 originates . for @xmath126 ,
the parameters of the model can be expressed in the forms @xmath127 from this , @xmath128 are new parameters . in what follows
, we will assume @xmath129 and four different physically interesting cases that elucidate the behavior of the analyzed system : \(a ) @xmath130 ; @xmath131 ; @xmath132 ; @xmath133 ; @xmath134 , ( a ) @xmath135 ; @xmath131 ; @xmath132 ; @xmath133 ; @xmath134 , \(b ) @xmath135 ; @xmath136 ; @xmath132 ; @xmath137 ; @xmath134 , \(c ) @xmath135 ; @xmath136 ; @xmath132 ; @xmath137 ; @xmath138 .
whereas atom @xmath2 is alone in ( a ) , it feels the presence of atom @xmath1 due to the quantized optical field in ( a ) . in ( b ) , both atoms interact by the dipole - dipole interaction that includes only the continuum of states at the atom @xmath2 .
finally , also the dipole - dipole interaction between the discrete levels of both atoms is taken into play in ( c ) .
we note that detuning of energy levels of both atoms from the laser frequency does not qualitatively modify the behavior of the system ( for more details , see @xcite for semiclassical model ) .
also , we analyze the system at time @xmath139 bellow . for the considered values of parameters , the behavior of the system at time @xmath139 already corresponds to that appropriate to the long - time limit . in the model , atoms @xmath1 and @xmath2
are in fact mutually coupled by two types of interactions .
side by side with the discussed dipole - dipole interaction , the interaction mediated by photons in the quantized field also occurs .
this interaction qualitatively distinguishes the presented fully quantum model from the common semiclassical models that assume a classical predefined optical pump field @xcite .
the long - time limit of the two atomic systems is described by a statistical matrix @xmath140 , with two discrete indices @xmath141 and two continuous arguments @xmath18 , @xmath142 . here
@xmath143 as usual , the photoelectron spectra are identified with the distributions @xmath144 this joint description may be reduced to the marginal probability distribution of the levels of the atom @xmath4 , @xmath145 we consider also the conditional distributions or spectra @xmath146 the closed formula for @xmath147 is rather complicated , @xmath148 where @xmath149 with @xmath150 } { i\left(\xi_{k'}^{(n)}-\xi_k^{(n)}- \lambda_{{\bf m}^{(n)}j'}^*+\lambda_{{\bf m}^{(n)}j}\right ) } \nonumber\\ & & \mbox { } \times { \bf k}_k^{(n)}{\bf b}_2^{(n)}{\bf m}_j^{(n ) } { \bf c}^{(n)}(0){\bf c}^{(n)\dagger}(0 ) { \bf m}_{j'}^{(n)\dagger}{\bf b}_1^{(n)}{\bf k}_{k'}^{(n)}. \label{42}\end{aligned}\ ] ] the long - time total photoelectron spectrum is time - independent , @xmath151 in figs . 2 and 3
the case ( a ) of data with @xmath137 and for the initial coherent state @xmath152 is illustrated .
the unconditioned and conditional photoelectron spectra have a multi - peak structure and the peak positions are about the same for both the values of the subscript @xmath153 . therefore the plot is restricted to an interval which includes a single peak of a spectrum . in fig .
2 , it is seen that the unconditioned photoelectron spectra coincide and can not be discerned in the chosen interval .
in contrast , in fig . 3 it is obvious that the conditional spectra differ significantly in the selected interval .
it proves the dependence of the occupation of the atom @xmath6 s level on the atom @xmath4 s level . to reveal the role of atom @xmath1 in ionization of atom @xmath2
, we compare the long - time ionization spectra of atom @xmath2 for atom @xmath154 present and absent .
we consider a greater value of single - photon rabi frequency @xmath119 to emphasize quantum features of the model ( @xmath155 ) .
ionization of isolated atom @xmath2 in a quantized field leads , in general , to the occurrence of sharp peaks in the ionization spectra ( see fig . 4 ) .
these peaks arise from the ionization caused by individual fock states of the optical field .
this is documented in figs .
4 , 5 , in which the ionization spectra corresponding to the coherent and fock states are shown .
it holds that the greater the fock number @xmath156 is , the narrower is the corresponding spectral peak and also the closer is the peak to the position of energy of the fano zero ( see fig . 4 ) .
such behavior qualitatively resembles that of an ionization spectrum caused by a classical strong pump field @xcite .
the presence of atom @xmath1 in the quantized pump field considerably modifies the ionization spectra of atom @xmath2 due to the mutual indirect interaction of both atoms through the quantized pump field .
contrary to the spectra of isolated atom @xmath157 , the ionization spectra have contributions both below and above the pump - field frequency .
moreover , the spectral peaks above and below the pump - field frequency occur in pairs which results in nearly symmetric ionization spectra ( see figs . 6 , 7 ) .
this symmetry is inherent to the fock states from which it transfers into the coherent states , as documented in figs . 6 , 7 .
it holds that the greater the fock number @xmath156 , the closer the two peaks to the pump - field frequency .
if the pump - field intensity increases , the spectrum of isolated atom @xmath2 is built more and more from contributions of higher - number fock states and it moves to lower energies crossing the energy of fano zero .
the more intense the pump field , the more suppressed ( smoothed ) the spectral structure of individual fock states ( see figs . 8 , 9 ) . also the narrowing of the overall ionization spectrum in the vicinity of the energy of fano zero is observed . when atom @xmath1 is present ,
the ionization spectra also gradually lose their peaked structure with the increasing pump - field intensities ( see figs . 10 , 11 ) .
for sufficiently high pump - field intensities , the spectrum approaches that of the isolated atom @xmath1 . when the interaction mediated by the quantized field is weaker , the behavior of ionization spectra with the increasing pump - field intensities is qualitatively similar to the usual one discussed in the fano model .
the spectra move towards lower energies with the increasing pump - field intensities and cross at certain intensity the energy of fano zero , as documented in figs .
12 , 13 .
whereas the isolated atom @xmath2 has a one - peak spectrum , the spectrum of atom @xmath2 influenced by atom @xmath1 consists of two peaks that form a spectral dublet at greater pump - field intensities clearly visible in fig . 13 .
the dipole - dipole interaction between atoms @xmath1 and @xmath2 , in general , splits the peaks in the ionization spectra of isolated atom @xmath2 into two parts ( see figs .
14 , 15 , 16 ) . as a consequence ,
there occur two major peaks in the ionization spectra for greater pump - field intensities .
these spectral peaks are broken into many sub - peaks for low pump - field intensities as a consequence of quantum character of the pump field ( see fig .
individual sub - peaks can be connected with the appropriate fock states , similarly as in the previous section .
two major peaks approach each other with the increasing pump - field intensity and form a spectral dublet at certain moment ( see fig .
if only the dipole - dipole interaction between the discrete level of atom @xmath1 and the continuum of states of atom @xmath2 ( @xmath158 ) is considered , the fano zero of isolated atom @xmath2 is inevitably lost .
however , when also the dipole - dipole interaction between the discrete levels of atoms @xmath1 and @xmath2 occurs ( @xmath159 ) , the fano zero can be preserved under certain conditions found in @xcite .
the two mentioned dipole - dipole interactions compete in ionizing atom @xmath2 in certain sense . if the strengths of two interactions equal for the energy of fano zero formed at atom @xmath2 , the fano zero is preserved .
the appropriate condition was derived in @xcite for the semi - classical model in the form @xmath160 numerical computations have revealed that the condition in eq .
( [ 44 ] ) is valid also in the analyzed quantum model ( see fig .
17 ) . here
, we would like to note that the original fano zero of isolated atom @xmath2 is usually replaced by a broad deep minimum in the ionization spectra provided that the condition in eq .
( [ 44 ] ) is not fulfilled ( see fig .
such behavior originates in the weakness of dipole - dipole interactions compared to the coulomb and optical dipole interactions that form the fano zero of isolated atom @xmath2 .
we have assessed the entanglement by the computable measure of entanglement , i. e. , the negativity @xcite .
it is recommended as such in the case of two parties ( components ) each possessing a finite number of levels .
we mark a difference , because in our analysis one of the two parties has an infinite number of levels .
the straightforward approach was successful on the assumption of a classical light field @xcite , because the two components are in a joint pure quantum state .
to our knowledge , such an approach can not be based on simple formulas on inclusion of the quantum nature of the field which leads to a mixed quantum state describing the involved parties .
numerical calculation would be a challenging task .
recently , a selection of the frequencies has been realized in a somewhat arbitrary , but systematic , way @xcite .
two states with these frequencies are just the levels needed for producing a qubit . in such a way
, we return to the well - known two - qubit problem . for @xmath161 @xmath162 @xmath163 $ ] , @xmath164,[-5,10 ] , [ -1.5,1.5]$ ]
, we generate a density plot of the negativity at @xmath165 that is @xmath166 where @xmath167 are eigenvalues of the partially transposed statistical matrix @xmath168 here @xmath169 } \nonumber\\ & & \mbox { } \times\left(\begin{array}{cc } \rho_{jk}^{\rm out}(e_d , e_d , t)&\rho_{jk}^{\rm out}(e_d , e'_d , t)\\ \rho_{jk}^{\rm out}(e'_d , e_d , t)&\rho_{jk}^{\rm out}(e'_d , e'_d ,
t)\end{array } \right ) , \label{47}\end{aligned}\ ] ] with @xmath170 both the dipole - dipole interaction and the interaction mediated by the quantized pump field create the entanglement between the bound electron at atom @xmath1 and the ionized electron at atom @xmath2 .
suitable conditions for creating highly entangled states have been revealed in @xcite concerning a classical pump field .
it holds that the stronger the dipole - dipole interaction , the more entangled state is reached .
however , also a weaker dipole - dipole interaction can provide highly entangled states provided that the ionization process is sufficiently slow .
this can be reached when the strengths of the direct ionization path ( connecting the ground state of atom @xmath2 with the continuum ) and the indirect ionization path ( that ionizes an electron from the ground state of atom @xmath2 through the auto - ionizing discrete state of atom @xmath2 ) are balanced .
similarly as in the semiclassical model analyzed in @xcite , the overall negativity can roughly be composed of negativities of qubit - qubit systems obtained from the qubit of atom @xmath1 and all possible qubits found in the continuum of atom @xmath2 .
such densities of negativity give us information about the spectral distribution of entanglement .
the density of negativity for the ionization spectrum shown in fig . 6 and appropriate for the interaction mediated by the quantized field is plotted in fig .
we can see in fig .
18 that the negativity is distributed in the whole area of energies present in the ionization spectrum .
it is remarkable that the values of density of negativity are very low for the degenerate energies of qubits inside the continuum of atom @xmath2 ( @xmath171 ) .
this behavior can be explained by the long - time energy conservation that does not allow to entangle such qubits in the continuum with the qubit of atom @xmath172 .
the densities of negativity appropriate to the coherent and fock states completely differ , as demonstrated in figs . 18 , 19 .
we note that a pump field in the fock state with one photon can not create entanglement due to the energy conservation
. however , higher - number fock states are already suitable for the entanglement creation . the densities of negativities formed by the dipole - dipole interaction
behave similarly as those created by the interaction mediated by the quantized field .
it holds also here that appreciable values of the density of negativity are found for energies appreciably present in the ionization spectra .
also very low values of the density of negativity occur around the degenerate energies @xmath171 ( see figs . 20 , 21 ) .
thus , the spectral concentration of negativity is observed as the pump - field intensity increases ( compare figs .
20 , 21 ) .
when the spectrum forms a spectral dublet , the entanglement is encoded between the two peaks of the dublet . as follows from the above results , effects stemming from quantum features of the pump optical field are clearly visible both in ionization spectra and entanglement provided that the one - photon excitation parameter @xmath119 is greater or comparable to 0.1 and the mean number of photons is smaller or comparable to 10 .
both coherent laser fields and highly - nonclassical fock - state fields are suitable for the observation of quantum signatures of auto - ionization process . as for the fock - state fields ,
they can be generated , e.g. , in heralded single - photon sources @xcite or their generalizations @xcite and in qed cavities @xcite .
greater values of the one - photon excitation parameter @xmath119 represent experimental challenge as the values reached in current ionization experiments are much smaller . however , modern photonic band - gap structures @xcite give a hope here .
they allow to dramatically increase electric - field amplitudes inside due to constructive interference on one side . on the other side
, they form photonic bands with continuum of states which are similar to those participating in ionization .
we have studied quantum correlations of two atoms . we have assumed that one atomic system contains an auto - ionizing level whereas the other atom does not comprise any auto - ionizing level .
both the atoms interact with the same mode of the quantized field .
we have concentrated ourselves to the long - time behavior of the atomic systems .
the long - time behavior exhibits quantum correlations of the two atoms even in the case where the atoms do not interact directly .
we have illustrated quantum correlations comparing the one - peak spectrum appropriate for the neighbor atom without optical excitation with the two - peak spectrum occurring for the optically - excited neighbor atom . in the classical limit of strong field
the differences vanish .
* we have identified conditions for the observation of quantum features in long - time electron ionization spectra . *
also the fano zero has been found in these spectra for the quantized optical field considering the same conditions as for the classical optical field .
the authors acknowledge the financial support by the project operational program research and development for innovations - european social fund ( project cz.1.05/2.1.00/03.0058 ) of the ministry of education , youth and sports of the czech republic and the project iga no .
prf-014 - 05 .
a. luk , v. peinov , j. peina , jr . , j. kepelka , and w. leoski , in _ wave and quantum aspects of contemporary optics _ , edited by j. mllerov , d. senderkov , and s. jureka , proceedings of spie vol .
7746 ( bellingham : spie ) p 77460w ( 2010 ) .
g. brida , i. p. degiovanni , m. genovese , f. piacentini , p. traina , a. della frera , a. tosi , a. bahgat shehata , c. scarcella , a. gulinatti , m. ghioni , s. v. polyakov , a. migdall , and a. giudice , appl .
* 101 * , 221112 ( 2012 ) . | quantum correlations between two neighbor atoms are studied .
it is assumed that one atomic system comprises a single auto - ionizing level and the other atom does not contain any auto - ionizing level .
the excitation of both atoms is achieved by the interaction with the same mode of the quantized field .
it is shown that the long - time behavior of two atoms exhibits quantum correlations even when the atoms do not interact directly .
this can be shown using the optical excitation of the neighbor atom .
also a measure of entanglement of two atoms can be applied after reduction of the continuum to two levels . |
in heavy ion reactions at energies between 20 a.mev and several a.gev , many clusters are formed .
this cluster formation presents a big challenge for transport models in which nucleons are the degrees of freedom which are propagated .
identifying clusters in a transport code which transports nucleons is all but simple and therefore in many approaches the fragment formation is simply omitted .
this invalidates the prediction of single particle observables as well , because the cluster formation and therefore the modification of the single particle spectra due to the fragment formation depends on the phase space region and , as a consequence , can not be approximated by a momentum independent scaling factor .
the simplest way to identify clusters is by employing coalescence or a minimum spanning tree procedure .
the first needs a multitude of free parameters , whereas the second allows only for an identification at the end of the reaction which excludes any study on the physical origin @xcite .
in addition , quantum effects , like additional binding energies due to closed shells or pairing energies , are not supplied by the underlying transport theory which is semi - classical .
if one wants to identify fragments early , while the reaction is still going on , one has to use the momentum as well as the coordinate space informations .
an idea how to do this has been launched by dorso et al .
it has been further developed into the simulated annealing clusterisation algorithm ( saca ) @xcite in the late 1990 s and has been successfully applied to understand the measured fragment charge distribution and spectra as well as bimodality @xcite . starting from the positions and momenta of the nucleons at a given time during the reaction , nucleons are combined in all possible ways into fragments or single nucleons applying a simulated annealing technique . neglecting the interaction among nucleons in different clusters , but taking into account the interaction among the nucleons in the same fragment
, this algorithm identifies that combination of fragments and free nucleons which has the _ highest binding energy_. if applied after the time when the energetic initial collisions are over , this most bound configuration has been proven to be close to the final distribution of fragments identified by the minimum spanning tree method at the end of the reaction@xcite .
the reason for this is the fact that fragments are not a random collection of nucleons at the end , but an initial - final state correlation .
saca can be applied at any moment during the reaction and allows therefore for a detailed study of the fragment production mechanism . in saca , for accounting the interaction in - between nucleons , only the bulk skyrme interaction supplemented by a yukawa potential is used which is also the potential used for the propagation of the nucleons in the qmd transport model . to obtain more realistic fragment observables and to be able to predict observables for isotopes and hyper - nuclei , we employ in our new approach a more realistic interaction and
add the secondary decay because the fragments , when identified , have a ( moderate ) excitation energy .
this new approach is dubbed friga ( `` fragment recognition in general application '' ) .
r0.49 in order to predict the absolute multiplicity of the isotope yields , we have added new features to the saca cluster identification .
they include the asymmetry energy , pairing and quantum effects . for the asymmetry energy ,
we adopt the parametrisation from iqmd @xcite , a transport code which we use in the present article for the transport of nucleons . for a proton
the single particle energy thus reads : @xmath0 where @xmath1=23.3 mev , and @xmath2 , @xmath3 , @xmath4 , @xmath5 are the neutron , proton , baryonic and saturation densities , respectively . in the present work ,
we take @xmath6=1 ( `` stiff '' asymmetry potential ) . +
another significant part of the binding energy of light isotopes are the shell structure and odd - even effects ( pairing ) . in the conditions of high pressure and temperature where friga is used to determine the pre - fragments ,
these structure effects are not well known .
e. khan et al .
@xcite showed that there are some indications that they affect the primary fragments .
the authors demonstrate that the pairing vanishes above a nuclear temperature @xmath7 ( pairing energy ) . at normal density
the pairing energy tends to be negligible for heavy nuclei , with @xmath8 , whereas it is strong for light isotopes , like @xmath9 and @xmath10 with 12 mev and 6.9 mev , respectively . in friga ,
the primary fragments are usually produced slightly below the saturation density ( typically around half of it ) and quite cold , with @xmath11 , and hence below @xmath12 .
therefore , one can not neglect the pairing energy .
the same is true for shell effects which produce experimentally a visible enhancement of the fragment yield for closed shell nuclei . in order to determine the contribution of all structure effects to the binding energy of clusters , we make two hypotheses independent of the density and the average kinetic energy of the fragment environment .
first , the relative ratio of this nuclear structure contribution to the overall binding energy remains unchanged at the moderate temperatures and at the density at which clusters are formed which is not far away from the saturation density .
introducing among the nucleons , initialised with the right root - mean - square radius , two body interactions , which corresponds in infinite matter to the skyrme equation of state , the total fragment energy @xmath13 where @xmath14 is the single - particle wigner density @xmath15 reproduces very well the nuclear binding energy given by the weizscker mass formula for ground state nuclei , @xmath16 @xcite .
( fig.12 ) .
our second hypothesis is that eq .
[ binding ] remains the right description of the binding energy if the nuclei are deformed or excited when the fragments are identified by the friga algorithm .
taken both assumptions together , we can express the nuclear structure contribution to the binding energy of a deformed cluster with z protons and n neutrons in the following way : @xmath17 where @xmath18 and @xmath19 are the experimentally measured binding energy ( which contains the structure contribution and the sum of the volume and surface terms of the bethe - weizscker formula ) .
( hyper-)isotopes which are not stable at all in nature , are discarded in friga by assigning to them a very repulsive @xmath20 .
the total binding energy of a cluster with n and z , which is used in the annealing algorithm , will then be : @xmath21 in contradistinction to saca in which only the first term is used .
[ fig1 ] shows the influence of the asymmetry energy and of the structure energy on the isotope yield in the reaction @xmath22 at 100 a.mev .
we display here the results for central collisions ( @xmath23 . this figure illustrates as well how the various ingredients influence the fragments yield obtained in friga , assuming an early clusterisation at t=60 fm / c . from that time on , the size of the pre - fragments does not change anymore .
we see that the charge distributions are not strongly modified for the different options , whereas details of the isotopic yield are strongly influenced : the asymmetry energy tends to narrow the distributions towards the valley of stability , whereas the structure effects contribute to restore the natural abundances , particularly strong for the @xmath9 clusters .
the pre - fragments , called also `` primary '' fragments , created in friga , are often produced non relaxed in shape and density .
when turning to their ground state , the shape surface energy is converted into excitation energy .
using qmd simulations , for beam energies between 50 a.mev - 1 a.gev , friga obtains for central heavy ion collisions a mean excitation energy of the intermediate mass fragments between 0 and 3 a.mev , depending on the fragment size and very similarly to the experimental measurements of @xcite .
this excitation energy is sufficiently large that the secondary decay of the pre - fragments causes a significant contribution to the yield of small clusters .
for this reason , we optionally allow in friga the excited cluster to undergo sequential secondary decays , using the gemini algorithm @xcite . another interesting feature of the primary clusters in friga is their internal density . although the medium is close to @xmath24 , at the stage of the collision when the primary cluster formation is stabilised , just after the colliding system begins to separate , the fragments predicted by friga are produced quite dilute , typically around @xmath25 for intermediate mass fragments , and around @xmath26 for the light @xmath27 isotopes .
this is explained by the fact that the dense clusters are disfavoured , because they would contain nucleons which are moving against each other . in this case
the nucleons have a too high relative momenta to form a cluster .
therefore , in the friga approach , fragment formation tests only the low density behaviour of the potentials , which are contributing to the binding energy .
r0.49 an hyper - nucleus is a nucleus which contains at least one hyperon ( @xmath28 , ... ) in addition to nucleons .
extending friga to the strange sector requires the knowledge of the @xmath29n potential . in this first study
, we consider the strange quark as inert and use @xmath30 for protons as well as for neutrons .
similarly , we consider the case of multiple strange nuclei as well , in which more than one hyperon is part of the fragment .
there , the coupling of 2 @xmath31 contributes with the potential @xmath32 . in the present approach we neglect @xmath33 for the hyperons , and take the contribution of the core nucleus ( partner of the hyperons ) as if it were decoupled from the hyperon .
since the pairing and shell contributions in the binding energy are not yet well known for hyper - nuclei , we neglect the @xmath20 contribution . using these modifications of the potentials
, friga produces hyper - nuclei with the same procedures as non strange fragments . in the underlying qmd - like programs , which propagate the hadrons , @xmath29 s
are produced in different reactions : @xmath34 , @xmath35 , @xmath36 , @xmath37 .
their abundance , position and momentum distributions are strongly influenced by the reaction kinematics , the nuclear equation of state and the in - medium properties of the @xmath38 ( kaon potential , etc . ) which are implemented in the transport model @xcite .
r0.49 due to their composition , the yields of hyper - nuclei are produced when a cluster in coordinate and momentum space absorbs a hyperons . in heavy ions collisions at relativistic energies ,
the hyperon distributions are strongly peaked around the mid - rapidity region whereas the large fragments have rapidities close to the beam or target rapidity .
the closer the rapidity of the hyperon approaches by production or by subsequent collisions the target / beam rapidity , the larger is the probability that it can be absorbed by one of this larger clusters .
heavy hyper - nuclei are therefore observed not far away from beam / target rapidity . at the same time hyperons
can also form with other nucleons light clusters at mid - rapidity . there , the probability decreases with the cluster size because it is increasingly difficult to form large cluster out of a gas of nucleons . whereas the large clusters in the beam / target rapidity regime can be identified quite early , the light clusters at mid - rapidty are formed later and many of them decay due to the interactions with the surrounding nucleons which form a gas of a large temperature as compared to the cluster binding energy .
this is illustrated in fig .
[ fig2 ] . as seen previously , in fig .
[ fig1 ] , the ingredients of the cluster binding energy influence the light isotope yields in friga .
the same is observed for hypernuclei . adopting the factor @xmath39 in @xmath40 has a strong effect , decreasing on the average the hypernucleus yields by around 20 percent .
the asymmetry energy in the cluster can have a similar effect , depending on the isotope ( z , n ) asymmetry . in order to illustrate the predicting power of the friga algorithm
, we confront it to experimental observations of light hypernuclei produced in the spectator region in collisions of the light system @xmath41 at 2 agev incident energy , measured by the hyphi collaboration at the sis synchrotron of gsi darmstadt .
the data presented here are taken from @xcite .
[ fig3 ] compares the iqmd - friga predictions for the rapidity distributions of @xmath42 and @xmath43 with the experiment .
the best agreement in the experimentally resolved rapidity region ( close to the projectile spectator , @xmath44 ) has been obtained while excluding the most central collisions ( taking b@xmath453 fm ) .
this procedure is a very basic approach to the simulate the effect of the complex experimental trigger .
the chosen rapidity region has the highest hadronic yield and contains still the tail of the @xmath29 distribution , as predicted by iqmd - friga . at these rapidities ,
the experiment has measured a yield ratio @xmath46 , with which iqmd - friga agrees within the experimental uncertainty with @xmath47 . including the asymmetry contribution @xmath33 in the cluster binding energy in friga
, we obtain a yield ratio of @xmath48 which is still within this uncertainty . therefore , at this level , the role of the asymmetry energy is difficult to judge .
[ fig4 ] shows that the transverse momentum distributions in the spectator region agree as well , in the slopes and the absolute yield .
just a slight discrepancy has to be noticed with a shift of 0.1 gev / c in @xmath49 . here ,
we present the results without @xmath33 contribution , but the conclusion is similar when including it .
we present here the first step towards an understanding of the production of isotopic yields and hypernuclei in heavy ion reactions .
our clusterisation algorithm friga , an improved version of the saca approach , which includes pairing and asymmetry energies as well as other structure effects is able to describe more precisely the nuclear binding energy and allows for realistic predictions of absolute ( hyper-)isotope yields .
we have seen that the asymmetry and pairing potentials can have a strong influence on both , the yields and momentum anisotropies for the ( hyper-)isotopes . according to this model , the nucleons which form fragments have initially a fairly low density .
they contract and form finally slightly excited fragments which may undergo secondary decays .
therefore , the fragment formation is sensitive to the density dependence of the asymmetry energy and the pairing energy .
however , fragments test this dependence only for densities below the saturation density . 99 p.b .
gossiaux , r. puri , ch .
hartnack , j. aichelin , nucl .
phys . a * 619 * ( 1997 ) 379 - 390 . c. o. dorso and j. randrup , phys .
b * 301 * , 328 ( 1993 ) .
r. k. puri and j. aichelin , j. comput . phys . *
162 * , 245 ( 2000 ) .
k. zbiri , a. le fvre , j. aichelin et al .
c * 75 * ( 2007 ) 034612 .
a. le fvre et al . , phys .
c * 80 * ( 2009 ) 044615 .
j. aichelin .
reports * 202 * , 233 ( 1991 ) .
w. cassing , e.l .
bratkovskaya , nucl .
phys . a * 831 * ( 2009 ) 2 .
hartnack et al . , eur .
j. a * 1 * ( 1998 ) 151 .
e. khan , nguyen van giai , n. sandulescu , nucl .
a * 789 * ( 2007 ) 94 .
e. khan , m. grasso and j. margueron , phys .
c * 80 * ( 2009 ) 044328 .
a. le fvre et al . , nucl .
phys . a * 735 * ( 2004 ) 219 .
s. hudan et al .
( indra collaboration ) , phys .
c * 67 * ( 2003 ) 064613 .
charity et al . , nucl .
phys . a * 476 * ( 1988 ) 516 .
@xcite c. hartnack , h. oeschler , y. leifels , e. l. bratkovskaya and j. aichelin , phys .
* 510 * ( 2012 ) 119 [ arxiv:1106.2083 [ nucl - th ] ] .
rappold et al . , phys .
b * 747 * ( 2015 ) 129 . | we present a new algorithm to identify fragments in computer simulations of relativistic heavy ion collisions .
it is based on the simulated annealing technique and can be applied to n - body transport models like the quantum molecular dynamics .
this new approach is able to predict isotope yields as well as hyper - nucleus production . in order to illustrate its predicting power
, we confront this new method to experimental data , and show the sensitivity on the parameters which govern the cluster formation . |
gravitational lensing can be used as a powerful astrophysical tool for probing the density profiles of galaxies , and is one of the few ways in which dark matter can be detected ( e.g. * ? ? ?
in addition , it often magnifies source objects by one to two orders of magnitude .
this allows us to use the intervening gravitational lens as a kind of natural telescope , magnifying the source so that we can observe more detail than we would have been able to without the lens .
this extra magnification provided by lensing has been very beneficial to studies of star formation and galaxy morphology at high redshift .
regions of the galaxy size and luminosity distribution that are inaccessible in unlensed observations are made ( more ) visible by lensing ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the properties of the lens galaxies ( typically elliptical galaxies ) can also be inferred from their lensing effect ( e.g. * ? ? ?
* ; * ? ? ?
of course , gravitational lensing distorts the image of the source , as well as magnifying it .
thus , techniques have been developed that aim to infer the mass profile of the lens galaxy and the surface brightness profile of the source , given observed images ( e.g. * ? ? ?
* ; * ? ? ?
the aim of this paper is to carry out this process with the recently discovered gravitationally lensed quasar / host galaxy system rxs j1131 - 1231 .
this system consists of a quadruply imaged quasar at redshift @xmath0 lensed by a galaxy at @xmath2 . at the time of its discovery , it was the closest known lensed quasar , with some evidence for an extended optical einstein ring - the image of the quasar host galaxy .
initial simple modelling suggested that the quasar source was magnified by a factor of @xmath1 50 .
thus , subsequent observations with the acs aboard the hubble space telescope allow the recovery of the morphology of the quasar s host galaxy down to a resolution of about 0.01 arc seconds - at least in principle , for the parts of the source nearest the caustic .
indeed , c06 presented a wide array of results based on hst observations ( at 555 nm and 814 nm with acs , and 1600 nm with nicmos ) , including a detailed reconstruction of the extended source . the source reconstruction method used by c06
is based on lensing the image back to a pixellated grid in the source plane , setting the source surface brightnesses to equal the image surface brightness , and using a decision rule ( in this case , the median ) to decide on the value of a source pixel whenever two or more image pixels disagree about the value of the same source pixel .
if the point spread function ( psf ) is small or the image has been deconvolved ( in c06 , the deconvolution step was neglected for the purposes of the extended source reconstruction ) and the lens model is correct , this method can expect to be quite accurate .
however , in principle , the uncertainty in the lens model parameters and the deconvolution step should always be taken into account . in this paper , we focus our attention on the 555 nm acs image ( the drizzled images , as reduced by c06 , were provided to us ) , and the process by which we reconstruct the original , unlensed source from it .
any differences between our reconstruction and the c06 one can be attributed to the following advantages of our approach : psf deconvolution , averaging over the lens parameter uncertainties , simultaneous fitting of all parameters , and the prior information that bayesian methods are capable of taking into account : in the case of our model , that is the knowledge that the majority of pixels in an astrophysical sources should be dark @xcite .
the 555 nm image is also of particular interest because its rest wavelength ( 335 nm ) probes regions of recent star formation in a galaxy with an agn . in the case of the einstein ring 0047 - 2808 @xcite ,
our method was able to resolve structure down to scales of @xmath1 0.01 arcsec , a factor of five smaller than that obtainable in an unlensed observation with the hubble space telescope and about double the resolution obtained by @xcite using adaptive pixellation and least squares _ applied to exactly the same data_. this was possible because we used a prior distribution over possible sources that is more realistic as a model of our knowledge of an unknown astrophysical source , that is , we took into account the fact that it should be a positive structure against a dark background , a fact many methods ( such as least squares and some popular regularisation formulas ) implicitly ignore @xcite .
these differences between methods are likely to be most significant when data are sparse or noisy , and all methods tend to agree as the data quality increases and we approach the regime where the observed image uniquely determines the correct reconstruction .
the conceptual basis of the bayesian reconstruction method was presented in @xcite .
the idea is to fit a complex model to some data , but rather than simply optimising the parameters of the model to achieve the best fit , we try to explore the whole volume of the parameter space that remains viable after taking into account the data .
the effect of data is usually to reduce the volume of the plausible regions of the parameter space considerably .
the exploration of the parameter space can be achieved by using markov chain monte carlo ( mcmc ) algorithms , which are designed to produce a set of models sampled from the posterior distribution . in the case of modelling the background source and lens mass distribution of a gravitational lensing system , this allows us to obtain a sample of model sources and lenses that , when lensed and blurred by a psf , match the observational data .
the diversity of the models gives the uncertainties in any quantity of interest .
the reader is referred to @xcite for an introduction to bayesian inference and mcmc .
the first step of a bayesian analysis is to assign a likelihood function , or the probability density we would assign to the observed data if we knew the values of all of the parameters . to assign this , we need a noise frame , a measure of how uncertain we are about the noise level in each pixel .
this is typically done by assuming that the observational error at pixel @xmath3 is from a normal distribution with mean 0 and known standard deviation @xmath4 .
we extended this to include two `` extra noise parameters '' @xmath5 and @xmath6 , such that the standard deviation for the error in the @xmath3th pixel is @xmath7 , where @xmath8 is the predicted flux in the @xmath3th pixel .
@xmath5 and @xmath6 then become extra model parameters to be estimated from the data .
the inclusion of @xmath5 and @xmath6 implies that the extra noise level sigma varies with the predicted brightness of the pixel , with a square root dependence expected from poisson photon counting .
we chose the @xmath9 values to be zero for most of the image , but infinite for the brightest regions of the quasar images , effectively masking out those parts of the image ; this mask can be seen in figure [ data ] .
a model psf was obtained using the tinytim software @xcite . as noted by c06
, the tinytim simulations did not successfully perform the geometric distortion correction , and the output psf had slightly non - orthogonal diffraction spikes , whereas the spikes in the image are perpendicular . to correct this ,
the psf was `` straightened out '' by evaluating it with respect to non - orthogonal axes ; the resulting psf is shown in figure [ tinytim ] .
whilst this process is imperfect , the extra noise sigma protects against serious consequences resulting from slight inaccuracies in this process .
while our choice of @xmath9 was designed to block out the brightest parts of the quasar images , since they are so bright , light from the quasar images still extends out past the masked regions and overlaps with interesting einstein ring structures .
thus , when modelling the image , we still required a flux component due to the quasar images .
the four quasar images were modelled as being proportional to the corrected tinytim profiles with unknown fluxes and central positions .
the surface brightness profile of the lens galaxy was modelled as the sum of two elliptical gaussian - like profiles ( sersic profiles , one for the core and one for the extended emission ) proportional to @xmath10 , where @xmath11 with unknown peak surface brightness , ellipticity @xmath12 , length scale @xmath13 , and angle of orientation .
the central position was also considered initially unknown , but for mcmc purposes the starting point was to have both profiles centred near the observed centre of the lens galaxy core .
the slopes @xmath14 were restricted to the range @xmath15 $ ] and assigned a uniform prior distribution , along with all of the other free parameters .
although elliptical galaxies are well modelled by a sersic profile with @xmath16 , we are modelling this galaxy by _ two _ such profiles .
this was done because the wings of the lens galaxy light profile ( where it overlaps the einstein ring ) are of great significance for our source reconstruction , and we do not want the core of the lens galaxy to be relevant to the wings . note that there are parts of the image where flux is present from three sources : the lensed einstein ring , the wings of the psf from the quasar images , and the foreground lens galaxy .
the fact these all overlap suggests that the optimal approach ( in all senses apart from cpu time ) involves simultaneously fitting all of these components . throughout this paper ,
any modelling has included all of the lens galaxy profile parameters as free parameters , where computational restrictions required that we fix the lens parameters .
] , as well as the source , lens model parameters and positions and brightnesses of the four tinytim psf profiles , to model the contribution from the quasar images that remains even after masking out their central regions .
the particular lens model we assumed for this system was a pseudo - isothermal elliptical potential ( piep ) @xcite , primarily for computational speed but also because it is fairly general and realistic , at least for single galaxy lenses that are not too elliptical .
this model has five parameters : strength @xmath17 , ellipticity @xmath18 ( actually the axis ratio : @xmath19 implies a circularly symmetric lens ) , orientation angle @xmath20 , and two parameters @xmath21 for the central position ( measured relative to the central pixel of the images in figure [ data ] ) .
although any bayesian modelling can only explore a particular slice through the full hypothesis space we might have in our minds , using a simplified analytical model is often sufficient to illuminate the general properties of the true lens mass distribution .
also , it is typically the case that inferences about the source of an einstein ring are insensitive to the specific parameterisation for the lens model ( e.g. * ? ? ?
* ) , as long as the model is able to fit the observed image at all .
all einstein rings can be expected to reside in an environment where the external shear due to neighbouring galaxies is nonzero ( kochanek , private communication ) , and thus , external shear was also included in the lens model .
there are two parameters for the external shear : @xmath22 , its magnitude , and its orientation angle @xmath23 . @xcite
have observed the flux ratios of the quasar images ( via integral field spectroscopy ) and found that most of these ratios are consistent with a model of this type ( elliptical potential plus external shear ) .
a similar model was used by c06 ( they used a singular isothermal ellipse+@xmath22 ) , where they find that it is the simplest parameterised model that can reproduce the observations . in principle , we could adopt ever less restrictive parameterisations for the lens , to hunt for substructures in its density profile . however , such an approach is extraordinarily computationally expensive ( unless simplifying assumptions about the source are also made ) and is beyond the scope of this paper . in terms of all of these parameters ,
the deflection angle formula , relating a point @xmath24 in the image plane to a corresponding point @xmath25 in the source plane , is @xmath26 where the deflection angles @xmath14 are given by the gradient of the potential @xmath27 and @xmath28 are the ray coordinates in the rotated coordinate system whose origin is @xmath29 and is aligned with the external shear ( i.e. rotated by an angle @xmath23 ) , and @xmath30 are the ray coordinates in the rotated and translated coordinate system centred at @xmath21 and oriented at an angle @xmath20 .
the physical interpretation of each of these parameters suggests a plausible prior range for their values . to represent this knowledge we used the following prior distributions ( table [ lenspriors ] ) .
since these are broad distributions , and the data are good , the influence of these choices is negligible ; they are included for completeness . [ cols="<,^,^",options="header " , ] + a preliminary investigation of the time delays predicted by our lens model suggests that it does not exactly reproduce the time delays measured by @xcite .
since this system exhibits significant microlensing @xcite , the time delay measurements are uncertain , but it is possible that the piep+@xmath22 lens model will be ruled out by further observations
. this would not be catastrophic for the present study , for several reasons .
firstly , source reconstructions tend to be insensitive to slight misspecification of the lens model ( e.g. * ? ? ?
secondly , all parameterisations are false .
we already know from prior information that the lens is not _ really _ a piep+@xmath22 model .
all modelling can only consider a single `` slice '' through a full hypothesis space , and the conclusions reached on that slice may or may not be representative of the full space .
they often are , but there are never any guarantees .
in this paper , we have presented a detailed gravitational lens reconstruction of the optical extended source in the einstein ring rxs j1131 - 1231 .
the source is a medium sized galaxy ( @xmath1 8 kpc in visible extent ) with several compact bright emission regions .
the substructures we found are in general agreement with those found by c06 in terms of their position , but we have shown that they are brighter and more compact than was previously determined .
in addition , our reconstruction provides a clearer view of the substructures , including near the central regions of the source .
the quasar resides in a bright emission region with an extent of about @xmath1 0.15 arcseconds or 1 kpc .
it should be noted that the wavelength of the observations in the rest frame is 335 nm , so this reconstruction traces regions of recent star formation in the source galaxy .
we have also directly compared point images vs extended images with regard to how well each is able to constrain the lens model .
we found that there is a significant gain to be made in taking into account all of the information from the extended images .
it has been suggested that this is not true in general @xcite , although it really depends on the resolution and number of extended images , which in this case is high .
certainly , in using both , there is nothing to lose but cpu cycles .
this system has the potential to become one of the most well - constrained gravitational lenses , with multiple images of the extended ring , quasar image positions and flux ratios in multiple bands , and time delay measurements available @xcite .
hence , it should be possible to carry out a detailed kiloarametric study of its mass profile to shed some light on the dark matter halo of the lens galaxy .
this paper was based on a single image of this system , the 555 nm acs image .
other hst images at different wavelengths ( 814 nm , 1.6@xmath31 m ) are available ( c06 ) and can further constrain the lens model .
simultaneous multi - wavelength reconstructions are now becoming routine ( e.g. * ? ? ? * ) .
however , all of the structures in these images are in the same locations , and so a multi - wavelength reconstruction would not produce significantly different conclusions to those reached here .
c06 note that in the near infrared image , the compact bright images are less pronounced compared to the diffuse background , which is what would be expected if the substructures are regions dense in hot young stars .
this study has relied on a number of common assumptions that future research will seek to relax . extending lens reconstruction techniques to incorporate kiloparametric models of the source and
the lens simultaneously is an ambitious task , but some steps are already being taken in that direction @xcite .
flexible lens modelling plus information from time delay measurements and other sources would be extremely valuable for studies of galaxy dark matter haloes . also , explicit modelling of dust absorption by the lens galaxy is proving to be an important ingredient in the inversion of einstein rings and
would be an essential part of future work on this system .
bjb thanks olivia ross for encouragement , and dennis stello for allowing me to use his fast computer .
this research is supported under the australian research council s discovery funding scheme ( project number dp0665574 ) and is part of the commonwealth cosmology initiative ( cci ) .
the authors would like to thank jean - francois claeskens and dominique sluse for providing the acs data .
the constructive comments of the anonymous referee helped us to improve the paper significantly . | we present a source and lens reconstruction for the optical einstein ring gravitational lens system rxs j1131 - 1231 .
we resolve detail in the source , which is the host galaxy of a @xmath0 quasar , down to a resolution of 0.045 arc seconds ( this is the size of the smallest conclusively resolved structures , rather than the pixel scale ) , using a bayesian technique with a realistic model for the prior information .
the source reconstruction reveals a substantial amount of complex structure in the host galaxy , which is @xmath1 8 kpc in extent and contains several bright compact substructures , with the quasar source residing in one of these bright substructures .
additionally , we recover the mass distribution of the lensing galaxy , assuming a simply - parameterised model , using information from both the quasar images and the extended images .
this allows a direct comparison of the amount of information about the lens that is provided by the quasar images in comparison to the extended images . in this system
, we find that the extended images provide significantly more information about the lens than the quasar images alone , especially if we do not include prior constraints on the central position of the lens .
[ firstpage ] gravitational lensing methods : statistical galaxies : structure |
global - scale magnetic fields and cycles of magnetic activity in sun - like stars are generated by the interplay of rotation and convection . at rotation rates greater than that of the current sun , such as when our sun was younger , observations tend to show increased magnetic activity indicating a strong global dynamo may be operating ( pizzolato et al .
2003 ) . here
we explore large - scale dynamo action in sun - like stars rotating at three times the current solar rate , or @xmath0 , with a rotational period of 9.32 days . as shown by helioseismology ,
the solar interior is in a state of prominent differential rotation in the convection zone ( roughly the outer 30% by radius ) whereas the radiative interior is in uniform rotation .
a prominent shear layer , or tachocline , is evident at the interface between the convective and radiative regions .
motivated by these observations , a number of theoretical models have been proposed for the solar dynamo .
the current paradigms for large - scale solar dynamo action favor a scenario in which the generation sites of toroidal and poloidal fields are spatially separated ( e.g. , charbonneau 2005 ) .
poloidal fields generated by cyclonic turbulence within the bulk of the convection zone , or by breakup of active regions , are pumped downward to the tachocline of rotational shear at its base .
the differential rotation there stretches such poloidal fields into strong toroidal structures , which may succumb to magnetic buoyancy instabilities and rise upward to pierce the photosphere as curved structures that form the observed active regions .
similiar dynamo processes are believed to be active in sun - like stars rotating several times faster than the current sun . here
we explore a variation to this paradigm by excluding the tachocline and the photosphere from our simulated domain , which extends from @xmath1 to @xmath2 , in order to see if magnetic cycles can be realized in the bulk of the convection zone itself .
radial velocity in global mollweide projection at @xmath3 with fast , narrow downflows in dark tones and broad , slow upflows in light tones .
differential rotation profile , with lines of constant angular velocity @xmath5 largely along cylinders , as expected for rapidly rotating systems .
some deviation toward conical contours is seen at low latitudes .
magnetic wreaths tend to form in the regions of strong shear near the equator . ] using massively - parallel supercomputers , we solve the nonlinear anelastic mhd equations in rotating 3-d spherical shells using the anelastic spherical harmonic ( ash ) code ( brun et al .
the anelastic approximation filters out fast - moving sound and magneto - acoustic waves , allowing us to follow the decidedly subsonic flows in the solar convection zone with overturning times of days to months . in large - eddy simulation ( les ) such as those using ash , the effects of small , unresolved scales on larger scales must be parameterized using a turbulence closure model .
previous ash simulations of convective dynamos in sun - like stars rotating at @xmath0 have yielded large - scale wreaths of strong toroidal magnetic field in the bulk of their convection zones ( brown et al 2010 ) .
these wreaths persist for decades of simulation time , remarkably coexisting with the strongly turbulent flows . here
we explore the effects of decreased levels of diffusion on these wreaths in two simulations , labeled case b and case s. case b uses an eddy viscosity that varies with depth as the square root of the mean density .
case s uses the dynamic smagorinsky model of germano et al .
( 1991 ) , which is based on the assumption of self - similarity in the inertial range of the velocity spectra .
case s has 50 times less diffusion on average than case b. figure 1a shows the radial velocity field for case s near the top of the convection zone with columnar cells at low latitudes and smaller - scale helical convection at higher latitudes .
figure 1b shows the differential rotation profile for case s with roughly 20% ( 250 nhz ) contrast in rotation rate between the equator and poles .
the radial velocity patterns and differential rotation for case b are qualitatively similar to figure 1 . longitudinal magnetic field @xmath6 for case b at @xmath7 in mollweide projection , showing two strong but patchy magnetic wreaths of opposite polarity with peak field strengths of 38 kg . @xmath4 time - latitude plot of @xmath6 averaged over longitude @xmath8 at the same depth over 15 years in case b , with strong negative - polarity wreaths shown in dark tones and strong positive - polarity wreaths shown in light tones , clearly indicating cyclic behavior and reversals in magnetic polarity .
the most remarkable feature of case b is a cyclic variation in the toroidal wreaths of magnetic field . with significantly less diffusion than the simulation of brown et al .
( 2010 ) that produced persistent wreaths with no reversals , case b creates strong toroidal bands of magnetic field as shown in figure 2 with peak field strengths of about 38 kg .
these wreaths of magnetic field vary strongly with time in both polarity and amplitude .
figure 2a shows @xmath6 in the lower convection zone when there are strong wreaths of opposite polarity in each hemisphere with significant longitudinal variation , which we term patchy wreaths .
if we average over longitude , figure 2b shows a time - latitude map of the @xmath9 in the lower convection zone .
the simulation clearly goes through reversals in the magnetic polarity of the wreaths in each hemisphere . at times
the hemispheres are out of phase with each other , occasionally yielding wreaths of the same polarity in both hemispheres .
such behavior might be termed irregular magnetic activity cycles . from case
s , 3-d volume visualization of magnetic field lines in the core of a wreath - segment with the inner and outer simulation boundaries shown as lined surfaces .
view is looking at low latitudes along the rotation axis .
@xmath4 radial location of the top of a buoyant loop as a function of time .
magnetic field strength at the top of the loop is indicated at representative times .
time corresponding to @xmath10 is indicated by circular plotting symbol at day 13.7 . ]
as we move to even less diffusive simulations , case s shows additional features in the strong toroidal wreaths , most notably buoyant loops of magnetic field .
the wreaths are again patchy in longitude and roughly cyclic in time .
the peak magnetic field strength rises to about 45 kg inside the wreaths .
these strong fields combine with the very low levels of diffusion to allow regions of very strong field to coherently move upward without changing the magnetic topology via reconnection or simply diffusing away the strong fields .
such magnetic loops rise due to a combination of magnetic buoyancy and advection by convective giant cells that span the layer .
figure 3a shows a magnetic loop near its maximum size , extending from @xmath11 to @xmath12 .
examination reveals that there is a significant amount of twist present in the loops and that there is a significant deflection poleward as they rise .
the radial location of the top of one buoyant loop as a function of time is shown in figure 3b .
initially the buoyancy of the wreath due to evacuation of fluid from magnetic pressure dominates over the advective force of the convective upflows , but within 6 days advection becomes dominant . after about 10 days
the magnetic tension force begins to balance the advection , causing the top of the loop to stall near @xmath13 .
these simulations suggest that stars rotating slightly faster than the current sun may produce dynamos capable of cycles of magnetic activity and buoyant magnetic structures in the bulk of their convective envelopes despite the absence of a tachocline of shear .
this both challenges and informs the interface dynamo paradigm for sun - like stars .
the essential questions are what drives the magnetic reversals in these simulations and what are the conditions necessary to generate buoyant magnetic loops that can survive transit through the convection zone .
+ this work is supported by nasa heliophysics theory program grants nng05g124 g and nnx08ai57 g and major supercomputing support through nsf teragrid resources .
the presentation of this paper in iau symposium 273 was aided by nsf grants atm 0548260 and ast 0968672 , and nasa grant 09-lwstrt09 - 0039 .
browning is supported by the jeffrey l. bishop fellowship at cita .
brown , b.p . , browning , m.k . , brun , a.s . , miesch , m.s . ,
& toomre , j. , 2010 , `` persistent magnetic wreaths in a rapidly rotating sun '' _ astrophys .
j. _ * 711 * 424 | observations of sun - like stars rotating faster than our current sun tend to exhibit increased magnetic activity as well as magnetic cycles spanning multiple years . using global simulations in spherical shells to study the coupling of large - scale convection , rotation , and magnetism in a younger sun ,
we have probed effects of rotation on stellar dynamos and the nature of magnetic cycles .
major 3-d mhd simulations carried out at three times the current solar rotation rate reveal hydromagnetic dynamo action that yields wreaths of strong toroidal magnetic field at low latitudes , often with opposite polarity in the two hemispheres .
our recent simulations have explored behavior in systems with considerably lower diffusivities , achieved with sub - grid scale models including a dynamic smagorinsky treatment of unresolved turbulence .
the lower diffusion promotes the generation of magnetic wreaths that undergo prominent temporal variations in field strength , exhibiting global magnetic cycles that involve polarity reversals . in our least diffusive simulation
, we find that magnetic buoyancy coupled with advection by convective giant cells can lead to the rise of coherent loops of magnetic field toward the top of the simulated domain . |
given data @xmath1 , where @xmath2 is the response and @xmath3 is the @xmath4-dimensional covariate , the goal in many analyses is to approximate the unknown function @xmath5 by minimizing a specified loss function @xmath6 [ a common choice is @xmath0-loss , @xmath7 . in trying to estimate @xmath8 ,
one strategy is to make use of a large system of possibly redundant functions @xmath9 .
if @xmath9 is rich enough , then it is reasonable to expect @xmath8 to be well approximated by an additive expansion of the form @xmath10 where @xmath11 are base learners parameterized by @xmath12 . to estimate @xmath8 , a joint multivariable optimization over @xmath13 may be used .
but such an optimization may be computationally slow or even infeasible for large dictionaries .
overfitting may also result . to circumvent this problem , iterative descent
algorithms are often used .
one popular method is the gradient descent algorithm described by @xcite , closely related to the method of `` matching pursuit '' used in the signal processing literature [ @xcite ] .
this algorithm is applicable to a wide range of problems and loss functions , and is now widely perceived to be a generic form of boosting . for the @xmath14th step , @xmath15 ,
one solves @xmath16 where @xmath17 ^ 2\ ] ] identifies the closest base learner to the gradient @xmath18 in @xmath0-distance , where @xmath19 is the gradient evaluated at the current value @xmath20 , and is defined by @xmath21_{f_{m-1}(\mathbf{x}_i ) } = -l'(y_i , f_{m-1}(\mathbf{x}_i)).\ ] ] the @xmath14th update for the predictor of @xmath8 is @xmath22 where @xmath23 is a regularization ( learning ) parameter . in this paper
, we study friedman s algorithm under @xmath0-loss in linear regression settings assuming an @xmath24 design matrix @xmath25 $ ] , where @xmath26 denotes the @xmath27th column . here
@xmath28 represents the @xmath27th base learner ; that is , @xmath29 where @xmath30 and @xmath31 .
it is well known that under @xmath0-loss the gradient simplifies to the residual @xmath32 .
this is particularly attractive for a theoretical treatment as it allows one to combine the line - search and the learner - search into a single step because the @xmath0-loss function can be expressed as @xmath33 .
the optimization problem becomes @xmath34 it is common practice to standardize the response by removing its mean which eliminates the issue of whether an intercept should be included as a column of @xmath35 .
it is also common to standardize the columns of @xmath35 to have a mean of zero and squared - length of one .
thus , throughout , we assume the data is standardized according to @xmath36 the condition @xmath37 leads to a particularly useful simplification : @xmath38 thus , the search for the most favorable direction is equivalent to determining the largest absolute value @xmath39 .
we refer to @xmath40 as the _ gradient - correlation _ for @xmath27 .
we shall refer to friedman s algorithm under the above settings as 2boost .
algorithm [ a : l2boost ] provides a formal description of the algorithm [ we use @xmath41 for notational convenience ] .
initialize @xmath42 for @xmath43 @xmath44 , where @xmath45 @xmath46 , where @xmath47 properties of stagewise algorithms similar to 2boost have been studied extensively under the assumption of an infinitesimally small regularization parameter .
@xcite considered a forward stagewise algorithm @xmath48 , and showed under a convex cone condition that the least angle regression ( lar ) algorithm yields the solution path for @xmath49 , the limit of @xmath48 as @xmath50 .
this shows that @xmath48 , a variant of boosting , and the lasso [ @xcite ] are related in some settings .
@xcite showed in general that the solution path of @xmath49 is equivalent to the path of the monotone lasso .
however , much less work has focused on stagewise algorithms assuming an arbitrary learning parameter @xmath23 .
an important exception is @xcite who studied 2boost with componentwise linear least squares , the same algorithm studied here , and proved consistency for arbitrary @xmath51 under a sparsity assumption where @xmath4 can increase at an exponential rate relative to @xmath52 .
as pointed out in @xcite , the @xmath48algorithm studied by @xcite bears similarities to 2boost .
it is identical to algorithm [ a : l2boost ] , except for line 4 , where @xmath53 is used in place of @xmath51 and @xmath54.\ ] ] thus , @xmath48 replaces the gradient - correlation @xmath55 with the sign of the gradient - correlation @xmath56 . for infinitesimally small @xmath51
this difference appears to be inconsequential , and it is generally believed that the two limiting solution paths are equal [ @xcite ] . in general , however , for arbitrary @xmath23 , the two solution paths are different .
indeed , @xcite indicated certain unique advantages possessed by 2boost .
other related work includes @xcite , who described a bias - variance decomposition of the mean - squared - error of a variant of 2boost . in this paper
, we investigate the properties of 2boost assuming an arbitrary learning parameter @xmath23 . during 2boost s descent along a fixed coordinate direction ,
a new coordinate becomes more favorable when it becomes closest to the current gradient .
but when does this actually occur ? we provide an exact simple closed form expression for this quantity : the number of iterations to favorability ( theorem [ criticalpoint.theorem ] of section [ s : fixeddescent ] ) .
this core identity is used to describe 2boost s solution path ( theorem [ full.path.solution.general ] ) , to introduce new tools for studying its path and to study and characterize some of the algorithm s unique properties .
one of these is active set cycling , a property where the algorithm spends lengthy periods of time cycling between the same coordinates when @xmath51 is small ( section [ s : cyclingbehavior ] ) .
our fixed descent identity also reveals how correlation affects 2boost s ability to select variables in highly correlated problems .
we identify a _
repressible condition _ that prevents a new variable from entering the active set , even though that variable may be highly desirable ( section [ s : repressibility ] ) .
using a data augmentation approach , similar to that used for calculating the elastic net [ @xcite ] , we describe a simple method for adding @xmath0-penalization to 2boost ( section [ s : elasticboost ] ) . in combination with decorrelation , this reverses the repressible condition and improves 2boost s performance in correlated problems . because 2boost is known to approximate forward stagewise algorithms for arbitrarily small @xmath51 , it is natural to expect these results to apply to such algorithms like lar and lasso , and thus our results provide a new explanation for why these algorithms may perform poorly in correlated settings and why methods like the elastic net , which makes use of @xmath0-penalization , are more adept in such settings .
all proofs in this manuscript can be found in the supplemental article [ @xcite ] .
to analyze 2boost we introduce the following notation useful for describing its solution path . let @xmath57 be the @xmath58 nonduplicated values in order of appearance of the selected coordinate directions @xmath59 .
we refer to these ordered , nonduplicated values as _ critical directions _ of the path . for example , if @xmath60 , the critical directions are @xmath61 and @xmath62 . to formally describe the solution path we introduce the following nomenclature.=-1 [ path.def ] the descent length along a critical direction @xmath63
is denoted by @xmath64 .
the critical point @xmath65 is the step number at which the descent along @xmath63 ends .
thus , following step @xmath66 , the descent is along @xmath63 for a total of @xmath64 steps , ending at step @xmath65 .
the set of values @xmath67 can be used to formally describe the solution path of 2boost : the algorithm begins by descending along direction @xmath68 ( the first critical direction ) for @xmath69 steps , after which it switches to a descent along direction @xmath70 ( the second critical direction ) for a total of @xmath0 steps .
this continues with the last descent along @xmath71 ( the final critical direction ) for a total of @xmath72 steps .
see figure [ figure1 ] for illustration of the notation . .
the @xmath73 critical directions are @xmath74 with critical descent step lengths @xmath75 and critical points @xmath76 . _
a key observation is that 2boost s behavior along a given descent is deterministic except for its descent length @xmath64 ( number of steps ) .
if we could determine the descent length , a quantity we show is highly amenable to analysis , then an exact description of the solution path becomes possible as 2boost can be conceptualized as collection of such fixed paths .
imagine then that we are at step @xmath77 of the algorithm and that in the following step a new critical direction @xmath27 is formed .
let us study the descent along @xmath27 for the next @xmath78 steps .
thus , in the @xmath14th step of the descent along @xmath27 , the predictor is @xmath79 consider then algorithm [ a : incrementall2boost ] which repeatedly boosts the predictor along the @xmath27th direction for a total of @xmath80 steps .
@xmath81 @xmath82 , where @xmath83 the following result states a closed form solution for the @xmath14-step predictor of algorithm [ a : incrementall2boost ] and will be crucial to our characterization of 2boost .
[ incremental.operator.theorem ] @xmath84 , where @xmath85 and @xmath86 .
theorem [ incremental.operator.theorem ] shows that taking a single step with learning parameter @xmath87 yields the same limit as taking @xmath14 steps with the smaller learning parameter @xmath51 .
the result also sheds insight into how @xmath51 slows the descent relative to stagewise regression .
notice that the @xmath14-step predictor can be written as @xmath88 the first term on the right is the predictor from a greedy stagewise step , while the second term represents the effect of slow - learning .
this latter term is what slows the descent relative to a greedy step .
when @xmath89 this term vanishes , and we end up with stagewise fitting , @xmath90 . theorem [ incremental.operator.theorem ] shows how to take a large boosting step in place of many small steps , but it does not indicate how many steps must be taken along @xmath27 before a new variable enters the solution path . if this were known , then the entire @xmath27-descent could be characterized in terms of a single step . to determine the descent length ,
suppose that 2boost has descended along @xmath27 for a total of @xmath14 steps . at step @xmath91
the algorithm must decide whether to continue along @xmath27 or to select a new direction @xmath92 .
to determine when to switch directions , we introduce the following definition .
[ favorable.def ] a direction @xmath92 is said to be more favorable than @xmath27 at step @xmath91 if @xmath93 and @xmath94 .
thus , if @xmath92 is more favorable at @xmath91 , the descent switches to @xmath92 for step @xmath91 . to determine when @xmath92 becomes more favorable , it will be useful to have a closed form expression for @xmath95 and @xmath96 . by theorem [ incremental.operator.theorem ] , @xmath97 \\ & = & \rho_{j,1 } -\nu_{m}\rho_{k,1}r_{j , k},\end{aligned}\ ] ] where @xmath98 . setting @xmath99 yields @xmath100 .
therefore , @xmath94 if and only if @xmath101 dividing throughout by @xmath102 , with a little bit of rearrangement , this becomes @xmath103 ^ 2,\ ] ] where @xmath104 .
notice importantly that @xmath105 because @xmath27 is the direction with maximal gradient - correlation at the start of the descent .
it is also useful to keep in mind that @xmath106 is the sample correlation of @xmath107 and @xmath28 due to , and thus @xmath108 .
the following result states the number of steps taken along @xmath27 before @xmath92 becomes more favorable .
[ criticalpoint.theorem ] the number of steps @xmath109 taken along @xmath27 so that @xmath92 becomes more favorable than @xmath27 at @xmath110 is the largest integer @xmath14 such that @xmath111 it follows that for @xmath112 @xmath113,\hspace*{-30pt}\ ] ] where @xmath114 is the largest integer less than or equal to @xmath115 .
[ repressible.remark ] in particular , notice that @xmath116 when @xmath117 [ adopting the standard convention that @xmath118 and assuming that @xmath119 .
we call @xmath117 the repressible condition
. section [ s : repressibility ] will show that repressibility plays a key role in 2boost s behavior in correlated settings .
when @xmath90 we obtain @xmath120 from which corresponds to greedy stagewise fitting .
because this makes the @xmath90 case uninteresting , we shall hereafter assume that @xmath121 .
theorem [ criticalpoint.theorem ] immediately shows that the problem of determining the next variable to enter the solution path can be recast as finding the direction requiring the fewest number of steps @xmath109 to favorability . when combined with theorem [ incremental.operator.theorem ] , this characterizes the entire descent and can be used to characterize 2boost s solution path
. as before , assume that @xmath27 corresponds to the first critical direction of the path , that is , @xmath122 . by theorem [ criticalpoint.theorem ] , 2boost descends along @xmath27 for a total of @xmath123 steps , where @xmath124 and @xmath70 is the coordinate requiring the smallest number of steps to become more favorable than @xmath27 . by theorem [ incremental.operator.theorem ] ,
the predictor at step @xmath125 is @xmath126 applying theorem [ incremental.operator.theorem ] once again , but now using a descent along @xmath70 initialized at @xmath127 , and continuing this argument recursively , as well as using the representation for the number of steps from theorem [ criticalpoint.theorem ] , yields theorem [ full.path.solution.general ] , which presents a recursive description of 2boost s solution path . [ full.path.solution.general ] @xmath128 , where @xmath129 are determined recursively from @xmath130 , \\
l_r&=&m_{l_{r+1}}^{(r)},\qquad s_r = s_{r-1 } + l_r,\qquad s_0=0 , \\
d_j^{(r ) } & = & \frac{\rho_j^{(r)}}{\rho_{l_r}^{(r)}},\qquad \rho_{j}^{(r+1 ) } = { \mathbf{x}}_j^t({\mathbf{y}}-{\mathbf{f}}_{s_{r } } ) = \rho_{j}^{(r ) } - \nu_{l_r}\rho_{l_r}^{(r)}r_{j , l_r}.\end{aligned}\ ] ] [ step.number.tie.remark ] a technical issue arises in theorem [ full.path.solution.general ] when @xmath131 is not unique .
non - uniqueness can occur due to rounding which is caused by the floor function used in the definition of @xmath109 .
this is why line 1 selects the next critical value , @xmath132 , by maximizing the absolute gradient - correlation @xmath133 and not by minimizing the step number @xmath131 .
this definition for @xmath132 is equivalent to the two - step solution @xmath134 [ equal.gradient.remark ] another technical issue arises when there is a tie in the absolute gradient - correlation . in line 3 of algorithm [
a : l2boost ] it may be possible for two coordinates , say @xmath92 and @xmath27 , to have equal gradient - correlations at step @xmath135 .
theorem [ full.path.solution.general ] implicitly deals with such ties due to definition [ favorable.def ] .
for example , suppose that the first @xmath136 steps are along @xmath27 with the tie occurring at step @xmath14 .
in the language of theorem [ criticalpoint.theorem ] , because @xmath92 becomes more favorable than @xmath27 at @xmath91 , where @xmath137 , we have @xmath138 resolves the tie at @xmath14 by continuing to descend along @xmath27 , then switching to @xmath92 at step @xmath91 .
although algorithm [ a : l2boost ] does not explicitly address this issue , the potential discrepancy is minor because such ties should rarely occur in practice .
this is because for @xmath139 to hold , the value inside the floor function of used to define @xmath109 must be an integer ( a careful analysis of the proof of theorem [ criticalpoint.theorem ] shows why ) .
a tie can occur only when this value is an integer which is numerically unlikely to occur.=-1 theorem [ full.path.solution.general ] immediately yields a recursive solution for the coefficient vector , @xmath140 . the solution path for @xmath140 is the piecewise solution @xmath141 where @xmath142 is the vector with one in coordinate @xmath63 and zero elsewhere . aside from the technical issue of ties , theorem
[ full.path.solution.general ] and algorithm [ a : l2boost ] are equivalent . for convenience ,
we state theorem [ full.path.solution.general ] in an algorithmic form to facilitate comparison with algorithm [ a : l2boost ] ; see algorithm [ a : l2boostpath ] .
computationally , algorithm [ a : l2boostpath ] improves upon algorithm [ a : l2boost ] by avoiding taking many small steps along a given descent .
however , the difference is not substantial because the benefits only apply when @xmath51 is small , and as we will show later ( section [ s : cyclingbehavior ] ) , this forces the algorithm to cycle between its variables following the first descent , thus mitigating its ability to take large steps .
thus , strictly speaking , the benefit of algorithm [ a : l2boostpath ] is confined primarily to the first descent .
@xmath143 ; @xmath144 ; @xmath145 @xmath146 ; @xmath147 @xmath148 ; @xmath149 @xmath128 to investigate the differences between the two algorithms we analyzed the diabetes data used in @xcite .
the data consists of @xmath150 patients in which the response of interest , @xmath151 , is a quantitative measure of disease progression for a patient . in total
there are 64 variables , that includes 10 baseline measurements for each patient , 45 interactions and 9 quadratic terms . in order to compare results ,
we translated each iteration , @xmath152 , used by algorithm [ a : l2boostpath ] into its corresponding number of steps , @xmath14 .
thus , while we ran algorithm [ a : l2boostpath ] for @xmath153 iterations , this translated into @xmath154 steps .
as expected , this difference is primarily due to the first iteration @xmath155 which took @xmath156 steps along the first critical direction ( first panel of figure [ figure2 ] ; the rug indicates critical points , @xmath65 ) .
there are other instances where algorithm [ a : l2boostpath ] took more than one step ( corresponding to the light grey tick marks on the rug ) , but these were generally steps of length 2 .
the standardized gradient - correlation is plotted along the @xmath151-axis of the figure .
the standardized gradient - correlation for step @xmath14 was defined as ( using the notation of algorithm [ a : l2boost ] ) @xmath157 the middle panel displays the results using algorithm [ a : l2boost ] with @xmath158 steps . clearly
, the greatest gains from algorithm [ a : l2boostpath ] occur along the @xmath155 descent .
one can see this most clearly from the last panel which superimposes the first two panels . against step number @xmath14 for algorithms [ a : l2boostpath ] and [ a : l2boost ] , respectively .
only coordinates in the solution path are displayed ( a total of four ) .
the third panel superimposes the first two panels .
all analyses used @xmath159 . ] note a potential computational optimization exists in algorithm [ a : l2boostpath ] .
it is possible to calculate the correlation values only once as each new variable enters the active set , then cache these values for future calculations .
thus , when @xmath132 is a new variable in the active set , we calculate @xmath160 .
the updated gradient - correlation is calculated efficiently by using addition and scalar multiplication using the previous gradient - correlation and the cached correlation coefficients @xmath161 this is in contrast to algorithm [ a : l2boost ] which requires a vector multiplication of dimension @xmath4 at each step @xmath14 to update the gradient - correlation : @xmath162 .
[ active.set.remark ] above , when we refer to the `` active set , '' we mean the unique set of critical directions in the current solution path .
this term will be used repeatedly throughout the paper . throughout the paper
we illustrate different ways of utilizing @xmath109 of theorem [ criticalpoint.theorem ] to explore 2boost .
so far we have confined the use of theorem [ criticalpoint.theorem ] to determining the descent length along a fixed direction , but another interesting application is determining how far a given variable is from the active set . note that although theorem [ criticalpoint.theorem ] was described in terms of an active set of only one coordinate , it applies in general , regardless of the size of the active set .
thus , @xmath109 can be calculated at any step @xmath14 to determine the number of steps required for @xmath92 to become more favorable than the current direction , @xmath27 .
this value represents the distance of @xmath92 to the solution path and can be used to visualize it . to demonstrate this , we applied algorithm [ a : l2boost ] to the diabetes data for @xmath163 steps and recorded @xmath109 for each of the @xmath164 variables . figure [ figure3 ] records these values .
each `` jagged path '' in the figure is the trace over the 10,000 steps for a variable @xmath92 .
each point on the path equals the number of steps @xmath109 to favorability relative to the current descent @xmath165 .
the patterns are quite interesting .
the top variables have @xmath109 values which quickly drop within the first 1000 steps .
another group of variables have values which take much longer to drop , doing so somewhere between 2000 to 4000 steps , but then increase almost immediately .
these variables enter the solution path but then quickly become unattractive regardless of the descent direction . of each variable @xmath92 to favorability relative to the current descent @xmath27
( results based on algorithm [ a : l2boost ] where @xmath159 ) . for visual clarity the @xmath109 values have been smoothed using a running median smoother . ]
it has become popular to visualize the solution path of forward stagewise algorithms by plotting their gradient - correlation paths and/or their coefficient paths .
figure [ figure3 ] is a similar tool .
a unique feature of @xmath109 is that it depends not only on the gradient - correlation ( via @xmath166 ) , but also the correlation in the @xmath167-variables ( via @xmath106 ) and the learning parameter @xmath51 . in this manner ,
figure [ figure3 ] offers a new tool for understanding and exploring such algorithms .
it has been widely observed that decreasing the regularization parameter slows the convergence of stagewise descent algorithms .
@xcite showed that the @xmath48 algorithm tracks the equiangular direction of the lar path for arbitrarily small @xmath53 . to achieve what lar does in a single step ,
the @xmath48 algorithm may require thousands of small steps in a direction tightly clustered around the equiangular vector , eventually ending up at nearly the same point as lar .
we show that 2boost exhibits this same phenomenon .
we do so by describing this property as an active set cycling phenomenon . using results from the earlier fixed descent analysis
, we show in the case of an active set of two variables that 2boost systematically switches ( cycles ) between its two variables when @xmath51 is small . for an arbitrarily small @xmath51 this forces the absolute gradient - correlations for the active set variables to be nearly equal .
this point of equality represents a singularity point that triggers a near - perpetual deterministic cycle between the variables , ending only when a new variable enters the active set with nearly the same absolute gradient - correlation .
our insight will come from looking at theorem [ criticalpoint.theorem ] in more depth .
as before , assume the algorithm has been initialized so that @xmath27 is the first critical step . previously the descent along @xmath27 was described in terms of steps , but this can be equivalently expressed in units of the `` step size '' taken .
define @xmath168 recall that theorem [ incremental.operator.theorem ] showed that a single step along @xmath27 with @xmath51 replaced with @xmath169 yields the same limit as @xmath109 steps along @xmath27 using @xmath51 .
we call @xmath169 the step size taken along @xmath27 . because @xmath92 becomes more favorable than @xmath27 at @xmath110 , the gradient following a step size of @xmath169 along @xmath27 satisfies @xmath170 , and in particular holds for the second critical direction , @xmath70 , which rephrased in terms of step size , is the smallest @xmath169 value , @xmath171
although inequality is strict , it becomes arbitrarily close to equality with shrinking @xmath51 . with a little bit of rearranging , implies that @xmath172 we will show @xmath173 is the step size making the absolute gradient - correlation between @xmath92 and @xmath27 equal @xmath174 converges to the smallest @xmath173 satisfying ; thus , becomes an equality in the limit . for convenience ,
we define @xmath175 .
[ dynamic.nu.size ] let @xmath176 . then @xmath177 . furthermore ,
if @xmath178 and @xmath179 , then @xmath180 and @xmath181 as @xmath182 . therefore , for arbitrarily small @xmath51 , @xmath183 and @xmath27 and @xmath70 will have near - equal absolute gradient - correlations .
this latter property triggers two - cycling . to see why ,
let us assume for the moment that the active set variables have equal absolute gradient - correlations .
then by a direct application of theorem [ criticalpoint.theorem ] , one can show that the number of steps taken along @xmath70 before @xmath27 becomes more favorable is @xmath184 .
thus , following the descent along @xmath27 , the algorithm switches to @xmath70 , but then immediately switches back to @xmath27 . if @xmath51 is small enough , this process is repeated , setting off a two - cycling pattern .
the next result is a formal statement of these arguments .
define @xmath185 for notational convenience , let @xmath186 and @xmath137 . for technical reasons we shall assume @xmath187 .
recall remark [ repressible.remark ] showed that @xmath188 , the repressible condition , yields an infinite number of steps to favorability .
thus , for @xmath27 to be even eligible for favorability we must have @xmath187 .
[ long.descent.followed.two.cycle ] if the first two critical directions are @xmath189 and @xmath190 , then @xmath27 is favored over @xmath92 for the next step after @xmath92 if @xmath187 .
theorem [ long.descent.followed.two.cycle ] assumes that @xmath190 . while this only holds in the limit , the two values should be nearly equal for arbitrarily small @xmath51 , and thus the assumption is reasonable .
notice also that theorem [ long.descent.followed.two.cycle ] only shows that @xmath27 is more favorable than @xmath92 , and not that the algorithm switches to @xmath27 .
however , we can see that this must be the case . for arbitrarily small
@xmath51 , @xmath27 s gradient - correlation should be nearly equal to @xmath92 s , and by definition , @xmath92 has maximal absolute gradient - correlation along the second descent .
indeed , the following result shows that the absolute gradient - correlations for @xmath27 and @xmath92 can be made arbitrarily close for small enough @xmath51 for any step @xmath191 following the descent along @xmath27 .
the result also shows that the sign of the gradient - correlation is preserved when @xmath51 is arbitrarily small , a fact that we shall use later . [ gradient.correlation.equality ]
@xmath192 as @xmath182 for each @xmath191 . combining theorems [ long.descent.followed.two.cycle ] and [ gradient.correlation.equality ] , we see that if @xmath51 is small enough , the first three critical directions of the path must be @xmath193 with critical points @xmath194 . and
once the descent switches back to @xmath27 , it is clear from the same argument that the next critical direction , @xmath195 , will be @xmath92 , and so forth .
we present a numerical example demonstrating two - cycling . for our example , we simulated data according to @xmath196 where @xmath197 , and @xmath198 .
the first 10 coordinates of @xmath140 were set to 5 , with the remaining coordinates set to 0 .
the design matrix @xmath35 was simulated by drawing its entries independently from a standard normal distribution . .
top left panel details the path through the first three active variables , the remaining panels detail each active variable descent . ]
figure [ figure4 ] plots the standardized gradient - correlations from algorithm [ a : l2boostpath ] using @xmath199 . as done earlier , we have converted iterations @xmath152 into step numbers @xmath14 along the @xmath167-axis .
the plots show the behavior of each coordinate within an active set descent .
the rug marks show each step @xmath14 for clarity , and dashed vertical lines indicate the step @xmath109 where the next step adds a new critical direction to the solution path .
the top left panel shows the complete descent along the first three active variables .
the remaining panels detail the coordinate behavior as the active set increases from one to three coordinates .
the top right panel shows repeated selection of the @xmath68 direction shown in black .
the last step along @xmath68 occurs at @xmath109 marked with the vertical dashed line , where the next step is along the @xmath70 direction shown in red .
this point marks the beginning of the two - cycling behavior , which continues in the lower left panel . at each step
, the algorithm systematically switches between the @xmath68 and @xmath70 directions , until an additional direction becomes more favorable .
the cycling pattern is @xmath200 .
the lower right panel demonstrates three - cycling behavior .
here it is instructive to note that the order of selection within three - cycling is nondeterministic . in this panel
the order starts as @xmath201 , but changes near @xmath202 to @xmath203 . as discussed later , nondeterministic cycling patterns are typical behavior of higher order cycling ( active sets of size greater than two ) . here
we provide a formal limiting result of two - cycling .
the result can be viewed as the analog of theorem [ dynamic.nu.size ] when the active set involves two variables . using a slightly modified version of 2boost
we show that for arbitrarily small @xmath51 , if the algorithm cycles between its two active variables , it does so until a new variable enters the active set with the same absolute gradient - correlation .
assume the active set is @xmath204 and that @xmath27 and @xmath92 are cycling according to @xmath205 .
the @xmath14-step predictor for @xmath15 is @xmath206 where @xmath207 .
the cycling pattern is assumed to persist for a minimum length of @xmath208 .
it will simplify matters if the cycling is assumed to be initialized with strict equality of the gradient correlations : @xmath209 . with an arbitrarily small @xmath51 ,
this will force near equal absolute gradient - correlations at each step and by theorem [ gradient.correlation.equality ] will preserve the sign of the gradient - correlation .
we assume @xmath210 it should be emphasized that the above assumptions represent a simplified version of 2boost . in practice
, we would have @xmath211 where @xmath212 .
however , for convenience we will not concern ourselves with this level of detail here .
readers can consult @xcite for a more refined analysis .
one way to ensure @xmath209 is to initialize the algorithm with the limiting predictor @xmath213 of theorem [ dynamic.nu.size ] obtained by letting along the @xmath27-descent . with a slight abuse of notation
denote this initial estimator by @xmath214 .
however , the fact that this specific @xmath214 is used does not play a direct role in the results . under the above assumptions , the following closed form expression for the @xmath14-step predictor under two - cycling holds .
[ two.cycle.predictor ] assume that @xmath215 for @xmath216 .
if @xmath217 , then for any @xmath218 satisfying @xmath219 , we have for each @xmath216 , @xmath220 , & \quad if $ m$ is odd,\vspace*{2pt } \cr { \mathbf{f}}_{0 } + v_{m}\rho_{k,1 } [ { \mathbf{x}}_k+(s-\nu r_{j , k}){\mathbf{x}}_j ] , & \quad if $ m$ is even , } \ ] ] where @xmath221 $ ] and @xmath222 .
note that @xmath223 under the asserted conditions .
to determine the above limit requires first determining when a new direction @xmath224 becomes more favorable . for @xmath225 to be more favorable at @xmath91 , we must have @xmath226 when @xmath14 is odd , or @xmath227 when @xmath14 is even .
the following result determines the number of steps to favorability .
for simplicity only the case when @xmath14 is odd is considered , but this does not affect the limiting result .
[ criticalpoint.twocycle.theorem ] assume the same conditions as theorem [ two.cycle.predictor ] .
then @xmath225 becomes more favorable than @xmath92 at step @xmath91 where @xmath14 is the largest odd integer @xmath228 such that @xmath229 where @xmath230 and @xmath231 clearly shares common features with .
this is no coincidence .
the bounds are similar in nature because both are derived by seeking the point where the absolute gradient - correlation between sets of variables are equal . in the case of two - cycling , this is the singularity point where @xmath27 , @xmath92 and @xmath225 are all equivalent in terms of absolute gradient - correlation .
the following result states the limit of the predictor under two - cycling .
[ dynamic step.two.cycles ] under the conditions of theorem [ two.cycle.predictor ] , the limit of @xmath232 as @xmath182 at the next critical direction @xmath233 equals @xmath234,\ ] ] where @xmath235 , @xmath179 , @xmath236 and @xmath237 .
furthermore , @xmath238 , where for each @xmath225 , @xmath239 .
this shows that the predictor moves along the combined direction @xmath240 taking a step size @xmath241 that makes the absolute gradient - correlation for @xmath233 equal to that of the active set @xmath204 .
theorem [ dynamic step.two.cycles ] is a direct analog of theorem [ dynamic.nu.size ] to two - cycling .
not surprisingly , one can easily show that this limit coincides with the lar solution . to show this , we rewrite @xmath242 in a form comparable to lar , @xmath243.\ ] ] recall that lar moves the shortest distance along the equiangular vector defined by the current active set until a new variable with equal absolute gradient - correlation is reached .
the term in square brackets above is proportional to this equiangular vector .
thus , since @xmath242 is obtained by moving the shortest distance along the equiangular vector such that @xmath244 have equal absolute gradient - correlation , @xmath242 must be identical to the lar solution .
analysis of cycling in the general case where the active set @xmath245 is comprised of @xmath246 variables is more complex . in two - cycling
we observed cycling patterns of the form @xmath247 , but when @xmath248 , 2boost s cycling patterns are often observed to be nondeterministic with no discernible pattern in the order of selected critical directions .
moreover , one often observes some coordinates being selected more frequently than others .
a study of @xmath249-cycling has been given by @xcite .
however , the analysis assumes deterministic cycling of the form @xmath250 which is the natural extension of the two - cycling just studied . to accommodate this framework ,
a modified 2boost procedure involving coordinate - dependent step sizes was used .
this models 2boost s cycling tendency of selecting some coordinates more frequently by using the size of a step to dictate the relative frequency of selection . under constraints to the coordinate step sizes , equivalent to solving a system of linear equations defining the equiangular vector used by lar , it was shown that the modified 2boost procedure yields the lar solution in the limit .
interested readers should consult @xcite for details .
now we turn our attention to the issue of correlation .
we have shown that regardless of the size of the active set a new direction @xmath92 becomes more favorable than the current direction @xmath27 at step @xmath110 where @xmath109 is the smallest integer value satisfying @xmath251 using our previous notation , let @xmath173 and @xmath169 denote the left and right - hand sides of the above inequality , respectively .
generally , large values of @xmath109 are designed to hinder noninformative variables from entering the solution path .
if @xmath92 requires a large number of steps to become favorable , it is noninformative relative to the current gradient and therefore unattractive as a candidate .
surprisingly , however , such an interpretation does not always apply in correlated problems .
there are situations where @xmath92 is informative , but @xmath109 can be artificially large due to correlation . to see why , suppose that @xmath92 is an informative variable with a relatively large value of @xmath166 .
now , if @xmath92 and @xmath27 are correlated , so much so that @xmath252 , then @xmath253
. hence , @xmath254 and @xmath255 due to . thus , even though @xmath92 is promising with a large gradient - correlation , it is unlikely to be selected because of its high correlation with @xmath27 .
the problem is that @xmath92 becomes an unlikely candidate for selection when @xmath166 is close to @xmath106 .
in fact , @xmath116 when @xmath117 so that @xmath92 can never become more favorable than @xmath27 when the two values are equal . we have already discussed the condition @xmath117 several times now , and have referred to it as the _
repressible condition_. repressibility plays an important role in correlated settings .
we distinguish between two types of repressibility : weak and strong repressibility .
weak repressibility occurs in the trivial case when @xmath256 .
weak repressibility implies that @xmath257 .
hence the gradient - correlation for @xmath92 and @xmath27 are equal in absolute value and @xmath92 , and @xmath27 are perfectly correlated .
this trivial case simply reflects a numerical issue arising from the redundancy of the @xmath92 and @xmath27 columns of the @xmath35 design matrix . the stronger notion of repressibility , which we refer to as strong repressibility ,
is required to address the nontrivial case @xmath258 in which @xmath92 is repressed without being perfectly correlated with @xmath27 .
the following definition summarizes these ideas .
[ repressible.def ] we say @xmath92 has the strong repressible condition if @xmath117 and @xmath259 .
we say that @xmath92 is ( strongly ) repressed by @xmath27 when this happens . on the other hand , @xmath92 has the weak repressible condition if @xmath92 and @xmath27 are perfectly correlated ( @xmath256 ) and @xmath117 .
we present a numerical example of how repressibility can hinder variables from being selected . for our illustration
we use example ( d ) of section 5 from @xcite .
the data was simulated according to @xmath260 where @xmath261 , @xmath262 and @xmath263 .
the first 15 coordinates of @xmath140 were set to 3 ; all other coordinates were 0 .
the design matrix @xmath264_{100\times40}$ ] was simulated according to @xmath265 \\[-8pt ] { \mathbf{x}}_j & = & { \mathbf{z}}_3 + \tau{\boldsymbol{{{\varepsilon}}}}_j,\qquad j = 11,\ldots,15,\nonumber \\ { \mathbf{x}}_j & = & { \boldsymbol{{{\varepsilon}}}}_j,\qquad j > 15,\nonumber\end{aligned}\ ] ] where @xmath266 and @xmath267 were i.i.d.@xmath268 and @xmath269 . in this simulation , only coordinates 1 to 5 , 6 to 10 and 11 to 15 have nonzero coefficients .
these @xmath167-variables are uncorrelated across a group , but share the same correlation within a group . because the within group correlation is high , but less than 1
, the simulation is ideal for exploring the effects of strong repressibility .
for the first 5 coefficients from simulation : red points are iterations @xmath152 where the descent direction @xmath270 .
variables 2 and 3 are never selected due to their excessively large @xmath169 step sizes : an artifact of the correlation between the 5 variables .
the last panel ( bottom right ) displays @xmath271 for those iterations @xmath152 where @xmath272 . ]
figure [ figure5 ] displays results from fitting algorithm [ a : l2boostpath ] for @xmath273 iterations with @xmath274 .
the first 5 panels are the values @xmath275 against the iteration @xmath276 , with points colored in red indicating iterations @xmath152 where @xmath270 and @xmath27 is used generically to denote the current descent direction .
notationally , the descent at iteration @xmath152 is along @xmath27 for a step size of @xmath277 , at which point @xmath225 becomes more favorable than @xmath27 and the descent switches to @xmath225 , the next critical direction .
the value plotted , @xmath278 , is the step size for @xmath279 .
whenever the selected coordinate is from the first group of variables ( we are referring to the red points ) one of the coordinates @xmath280 achieves a small @xmath169 value .
however , coordinates @xmath281 and @xmath282 maintain very large values throughout all iterations .
this is despite the fact that the two coordinates generally have large values of @xmath166 , especially during the early iterations ( see the bottom right panel ) .
this suggests that 1 , 4 and 5 become active variables at some point in the solution path , whereas coordinates 2 and 3 are never selected ( indeed , this is exactly what happened ) .
we can conclude that coordinates 2 and 3 are being strongly repressed by @xmath272 .
interestingly , coordinate 4 also appears to be repressed at later iterations of the algorithm .
observe how its @xmath166 values decrease with increasing @xmath152 ( blue line in bottom right panel ) , and that its @xmath169 values are only small at earlier iterations .
thus , we can also conclude that coordinates @xmath283 eventually repress coordinate 4 as well .
we note that the number of iterations @xmath273 used in the example is not very large , and if 2boost were run for a longer period of time , coordinates 2 and 3 will eventually enter the solution path ( panels 2 and 3 of figure [ figure5 ] show evidence of this already happening with @xmath169 steadily decreasing as @xmath152 increases ) .
however , doing so leads to overfitting and poor test - set performance ( we provide evidence of this shortly ) . using different values of @xmath51
also did not resolve the problem .
thus , similar to the lasso , we find that 2boost is unable to select entire groups of correlated variables . like the lasso
this means it also will perform suboptimally in highly correlated settings . in the next section
we introduce a simple way of adding @xmath0-regularization as a way to correct this deficiency .
the tendency of the lasso to select only a handful of variables from among a group of correlated variables was noted in @xcite . to address this deficiency
, @xcite described an optimization problem different from the classical lasso framework .
rather than relying only on @xmath69-penalization , they included an additional @xmath0-regularization parameter designed to encourage a ridge - type grouping effect , and termed the resulting estimator `` the elastic net . '' specifically , for a fixed @xmath284 ( the ridge parameter ) and a fixed @xmath285 ( the lasso parameter ) , the elastic net was defined as @xmath286 to calculate the elastic net , @xcite showed that could be recast as a lasso optimization problem by replacing the original data with suitably constructed augmented values .
they replaced @xmath287 @xmath288 and @xmath35 @xmath289 with augmented values @xmath290 and @xmath291 , defined as follows : @xmath292_{(n+p)\times1},\qquad { { \mathbf{x}}^*}= \frac{1}{\sqrt{1+\lambda } } \left[\matrix { { \mathbf{x}}\cr \sqrt{\lambda } { \mathbf{i } } } \right]_{(n+p)\times p } = [ { { \mathbf{x}}^*}_1,\ldots,{{\mathbf{x}}^*}_p].\hspace*{-35pt}\ ] ] the elastic net optimization can be written in terms of the augmented data by reparameterizing @xmath140 as @xmath293 . by lemma 1 of @xcite , it follows that can be expressed as @xmath294 which is an @xmath69-optimization problem that can be solved using the lasso .
one explanation for why the elastic net is so successful in correlated problems is due to its decorrelation property .
let @xmath295 .
because the data is standardized such that @xmath296 [ recall ] , we have @xmath297 one can see that @xmath298 is a decorrelation parameter , with larger values reducing the correlation between coordinates .
@xcite argued that this effect promotes a `` grouping property '' for the elastic net that overcomes the lasso s inability to select groups of correlated variables .
we believe that decorrelation is an important component of the elastic net s success .
however , we will argue that in addition to its role in decorrelation , @xmath298 has a surprising connection to repressibility that further explains its role in regularizing the elastic net .
the argument for the elastic net follows as a special case ( the limit ) of a generalized 2boost procedure we refer to as elasticboost .
the elasticboost algorithm is a modification of 2boost applied to the augmented problem .
to implement elasticboost one runs 2boost on the augmented data , adding a post - processing step to rescale the coefficient solution path : see algorithm [ a : eboost ] for a precise description . for arbitrarily small @xmath51 ,
the solution path for elasticboost approximates the elastic net , but for general @xmath23 , elasticboost represents a novel extension of 2boost .
we study the general elasticboost algorithm , for arbitrary @xmath23 , and present a detailed explanation of how @xmath298 imposes @xmath0-regularization .
augment the data .
set @xmath299 for @xmath300 .
run algorithm [ a : l2boostpath ] for @xmath301 iterations using the augmented data .
let @xmath302 denote the @xmath301-step predictor ( discard @xmath302 for @xmath303 ) .
let @xmath304 denote the @xmath301-step coefficient estimate .
rescale the regression estimates : @xmath305 . to study the effect @xmath298 has on elasticboost s solution path we consider in detail how @xmath298 effects @xmath306 , the number of steps to favorability [ defined as in but with @xmath287 and @xmath35 replaced by their augmented values @xmath290 and @xmath291 ] .
at initialization , the gradient - correlation for @xmath308 is @xmath309 in the special case when @xmath299 , corresponding to the first descent of the algorithm , @xmath310 therefore , @xmath311 , and hence @xmath312.\ ] ] this equals the number of steps in the original ( nonaugmented ) problem but where @xmath35 is replaced with variables decorrelated by a factor of @xmath313 . for large values of @xmath298
this addresses the problem seen in figure [ figure5 ] .
recall we argued that @xmath109 can became inflated due to the near equality of @xmath166 with @xmath106 .
however , @xmath314 shrinks to zero with increasing @xmath298 , which keeps @xmath306 from becoming inflated .
this provides one explanation for @xmath298
s role in regularization , at least for the case when @xmath298 is large .
but we now suggest another theory that applies for both small and large @xmath298 .
we argue that regularization is imposed not just by decorrelation , but through a combination of decorrelation and reversal of repressibility .
s role is more subtle than our previous argument suggests . to show this ,
let us suppose that near - repressibility holds .
we assume therefore that @xmath315 for some small @xmath316 .
then , @xmath317}_{\mathrm{repressibility\ effect } } \\ & & \hphantom{\qquad= } \underbrace{- \log\biggl(1-\frac{r_{j , k}}{\sqrt{1+\lambda}}{\operatorname{sgn}}\biggl(r_{j , k } \biggl[\frac{1}{1+{\delta}}-\frac{1}{\sqrt{1+\lambda}}\biggr ] \biggr)\biggr)}_{\mathrm { decorrelation\ effect}}.\nonumber\end{aligned}\ ] ] the first term on the right captures the effect of repressibility .
when @xmath318 is small , @xmath298 plays a crucial role in controlling its size . if @xmath319 , the expression reduces to @xmath320 which converges to @xmath321 as @xmath322 ; thus precluding @xmath92 from being selected [ keep in mind that is divided by @xmath323 , which is negative ; thus @xmath324 . on the other hand , any @xmath284 , even a relatively small value ,
ensures that the expression remains small even for arbitrarily small @xmath318 , thus reversing the effect of repressibility . the second term on the right of is related to decorrelation .
if @xmath325 ( which holds if @xmath298 is large enough when @xmath326 , or for all @xmath284 if @xmath327 ) , the term reduces to @xmath328 which remains bounded when @xmath284 if @xmath329 . on the other hand , if @xmath330 , the term reduces to @xmath331 which remains bounded if @xmath329 and shrinks in absolute size as @xmath298 increases . taken together , these arguments show @xmath298 imposes @xmath0-regularization through a combination of decorrelation and the reversal of repressibility which applies even when @xmath298 is relatively small .
these arguments apply to the first descent .
the general case when @xmath332 requires a detailed analysis of @xmath333 . in general ,
@xmath334 we break up the analysis into two cases depending on the size of @xmath298 .
suppose first that @xmath298 is small .
then @xmath335 which is the ratio of gradient correlations based on the original @xmath35 without pseudo - data .
if @xmath92 is a promising variable , then @xmath333 will be relatively large , and our argument from above applies . on the other hand
if @xmath298 is large , then the third term in the numerator and the denominator of @xmath333 become the dominating terms and @xmath336 the growth rate of @xmath337 for the pseudo data is @xmath338 for a group of variables that are actively being explored by the algorithm .
thus @xmath339 and our previous argument applies .
as evidence of this , and to demonstrate the effectiveness of elasticboost , we re - analyzed using algorithm [ a : eboost ] .
we used the same parameters as in figure [ figure5 ] ( @xmath273 and @xmath274 ) .
we set @xmath340 .
the results are displayed in figure [ figure6 ] .
in contrast to figure [ figure5 ] , notice that all 5 of the first group of correlated variables achieve small @xmath341 values ( and we confirmed that all 5 variables enter the solution path ) .
it is interesting to note that @xmath333 is nearly 1 for each of these variables . ) .
now each of the first 5 coordinates are selected and each has @xmath333 values near one . ]
( top ) and @xmath342 ( bottom ) based on 250 independent learning samples .
the distribution of coefficient estimates are displayed as boxplots ; mean values are given in red . ] to compare 2boost and elasticboost more evenly , we used 10-fold cross - validation to determine the optimal number of iterations ( for elasticboost , we used doubly - optimized cross - validation to determine both the optimal number of iterations and the optimal @xmath298 value ; the latter was found to equal @xmath343 )
. figure [ figure7 ] displays the results .
the top row displays 2boost , while the bottom row is elasticboost ( fit under the optimized @xmath298 ) .
the minimum mean - squared - error ( mse ) is slightly smaller for elasticboost ( 217.9 ) than 2boost ( 231.7 ) ( first panels in top and bottom rows ) .
curiously , the mse is minimized using about same number of iterations for both methods ( 190 for 2boost and 169 for elasticboost ) .
the middle panels display the coefficient paths .
the vertical blue line indicates the mse optimized number of iterations . in the case of 2boost
only 4 nonzero coefficients are identified within the optimal number of steps , whereas elasticboost finds all 15 nonzero coefficients .
this can be seen more clearly in the right panels which show coefficient estimates at the optimized stopping time .
not only are all 15 nonzero coefficients identified by elasticboost , but their estimated coefficient values are all roughly near the true value of 3 .
in contrast , 2boost finds only 4 coefficients due to strong repressibility .
its coefficient estimates are also wildly inaccurate .
while this does not overly degrade prediction error performance ( as evidenced by the first panel ) , variable selection performance is seriously impacted .
the entire experiment was then repeated 250 times using 250 independent learning sets .
figure [ figure8 ] displays the coefficient estimates from these 250 experiments for elasticboost ( left side ) and 2boost ( right side ) as boxplots .
the top panel are based on the original sample size of @xmath261 and the bottom panel use a larger sample size @xmath342 .
the results confirm our previous finding : elasticboost is consistently able to group variables and outperform 2boost in terms of variable selection .
finally , the left panel of figure [ figure9 ] displays the difference in test set mse for 2boost and elasticboost as a function of @xmath298 over the 250 experiments ( @xmath261 ) .
negative values indicate a lower mse for elasticboost , which is generally the case for larger @xmath298 .
the right panel displays the mse optimized number of iterations for 2boost compared to elasticboost .
generally , elasticboost requires fewer steps as @xmath298 increases .
this is interesting , because as pointed out , this generally coincides with better mse performance .
a key observation is that 2boost s behavior along a fixed descent direction is fully specified with the exception of the descent length , @xmath64 . in theorem
[ criticalpoint.theorem ] , we described a closed form solution for @xmath109 , the number of steps until favorability , where @xmath344 is the currently selected coordinate direction and @xmath345 is the next most favorable direction .
theorem [ criticalpoint.theorem ] quantifies 2boost s descent length , thus allowing us to characterize its solution path as a series of fixed descents where the next coordinate direction , chosen from all candidates @xmath346 , is determined as that with the minimal descent length @xmath109 ( assuming no ties ) .
since we choose from among all directions @xmath346 , @xmath109 , and equivalently the step length @xmath169 , can be characterized as measures to favorability , a property of each coordinate at any iteration @xmath152 .
these measures are a function of @xmath51 and the ratio of gradient - correlations @xmath166 and the correlation coefficient @xmath106 relative to the currently selected direction @xmath27 .
characterizing the 2boost solution path by @xmath109 provides considerable insight when examining the limiting conditions .
when @xmath347 , 2boost exhibits active set cycling , a property explored in detail in section [ s : cyclingbehavior ] .
we note that this condition is fundamentally a result of the optimization method which drives @xmath348 when @xmath51 is arbitrarily small .
this virtually guarantees the notorious slow convergence seen with infinitesimal forward stagewise algorithms .
the repressibility condition occurs in the alternative limiting condition @xmath349 .
repressibility arises when the gradient correlation ratio @xmath166 equals the correlation @xmath106 .
when @xmath259 , @xmath92 is said to be strongly repressed by @xmath27 , and while descending along @xmath27 , the absolute gradient - correlation for @xmath92 can never be equal to or surpass the absolute gradient - correlation for @xmath27 .
strong repressibility plays a crucial role in correlated settings , hindering variables from being actively selected .
adding @xmath0 regularization reverses repressibility and substantially improves variable selection for elasticboost , an 2boost implementation involving the data augmentation framework used by the elastic net . | we consider @xmath0boosting , a special case of friedman s generic boosting algorithm applied to linear regression under @xmath0-loss .
we study @xmath0boosting for an arbitrary regularization parameter and derive an exact closed form expression for the number of steps taken along a fixed coordinate direction .
this relationship is used to describe @xmath0boosting s solution path , to describe new tools for studying its path , and to characterize some of the algorithm s unique properties , including active set cycling , a property where the algorithm spends lengthy periods of time cycling between the same coordinates when the regularization parameter is arbitrarily small .
our fixed descent analysis also reveals a _ repressible condition _ that limits the effectiveness of @xmath0boosting in correlated problems by preventing desirable variables from entering the solution path . as a simple remedy ,
a data augmentation method similar to that used for the elastic net is used to introduce @xmath0-penalization and is shown , in combination with decorrelation , to reverse the repressible condition and circumvents @xmath0boosting s deficiencies in correlated problems . in itself , this presents a new explanation for why the elastic net is successful in correlated problems and why methods like lar and lasso can perform poorly in such settings . . |
the purpose of this paper is to give , in the @xmath0-scenario of uniform hyperbolicity , a characterization of those invariant measures that satisfy pesin s entropy formula in terms of their physical - like properties .
our main result works , for @xmath0 anosov diffeomorphisms , as ledrappier - young characterization @xcite of the measures @xmath2 that satisfy pesin s entropy formula ( which holds in the @xmath3 context but not in the general @xmath0 context ) , by substituting the property of absolute continuity of the unstable conditional decomposition of @xmath2 , by the weak pseudo - physical property of its ergodic decomposition .
pesin theory @xcite gives relevant tools and results of the modern differentiable ergodic theory .
it works for @xmath3 ( or at least @xmath0 plus hlder ) dynamical systems .
for instance , for @xmath3 hyperbolic systems , pesin s entropy formula computes exactly the metric entropy of a diffeomorphism in terms of the mean value of the sum of its positive lyaypunov exponents . in the @xmath3 scenario ,
pesin s entropy formula holds if and only if the invariant measure has absolutely continuous conditional decomposition along the unstable manifolds .
. through the properties of absolute continuity of invariant measures , and mainly through the absolute continuity of the holonomy along the invariant foliations , pesin theory gives the tools to construct physically significant invariant measures for @xmath0-plus hlder systems . among these measures , the so called sinai - ruelle - bowen ( srb ) measures @xcite ,
have particular relevance to describe the asymptotic statistics of lebesgue - positive set of orbits , not only for @xmath0 plus hlder uniform and non - uniform hyperbolic systems , but also for @xmath0 plus hlder partially hyperbolic systems @xcite .
precisely , one of the most relevant properties of ergodic srb measures for @xmath0 plus hlder hyperbolic systems , is that they are _ physical ; _ namely , their basins of statistical attraction have positive lebesgue measure , even for lebesgue non - preserving systems . _ _ in particular , for transitive anosov @xmath0 plus hlder systems , the theorem of pesin - sinai ( see for instance @xcite ) states that there exists a unique physical measure : it is the unique invariant probability measure that satisfies pesin s entropy formula , and so , the only one with absolutely continuous conditional measures along the unstable foliation .
besides , its basin of statistical attraction covers lebesgue all the orbits . in other words , for a @xmath0 plus hlder anosov system , the definition of physical measure , srb measure , and measure that satisfies pesin s entropy formula , are equivalent . nevertheless , in the @xmath0-scenario , the above results do not work , because the theorems of pesin theory that ensure the absolute continuity of unstable conditional measures , and of the holonomies along invariant foliations , fail .
even the existence of the unstable manifolds , along which one could construct the conditional unstable measures , fails in the @xmath0 context @xcite . in the particular case of @xmath0-anosov diffeomorphisms ,
invariant @xmath4 foliations with @xmath0 leaves do exist ( see for instance @xcite ) , but the holonomies along the invariant foliations are not necessarily absolutely continuous @xcite . as a consequence , for @xmath0 systems ,
if one defined srb measures by the existence of their absolutely continuous unstable conditional measures , one would lack the hope to construct them .
nevertheless , one can still define srb or srb - like measures , if one forgets for a while the properties of absolute continuity , and focus the attention of the properties of statistical attraction . in other words , one can try to look directly at their physical properties , dodging the lack of conditional absolute continuity . for that reason , in the @xmath0-scenario , we look for the lebesgue abundance of points in their statistical basins ( or , more precisely , in the @xmath5-approach of their statistical basins ) .
this search was used in @xcite to construct the srb - like or pseudo - physical measures for @xmath4 dynamical systems on a compact riemannian manifold .
the notion of srb - like or pseudo - physical measures , even in a non differentiable context , translate to the space of probability measures the concept of statistical attraction defined by ilyashenko @xcite in the ambient manifold . in @xcite
it is proved that @xmath0 _ generically , _ transitive and uniformly hyperbolic systems do have a unique measure satisfying pesin s entropy formula . besides this measure is physical and its basin of statistical attraction covers lebesgue almost all the orbits . in the general @xmath0-scenario with non uniform hyperbolicity ,
pesin s entropy formula was first proved considering systems that preserve a smooth measure ( we call a measure @xmath2 smooth if @xmath6 , where @xmath7 is the lebesgue measure ) .
in fact , for @xmath0 generic diffeomorphisms that preserve a smooth measure @xmath2 , @xcite proved that @xmath2 satisfies pesin s entropy formula .
later , in @xcite , this formula was also proved for any @xmath0 partially hyperbolic system that preserves a smooth measure . _ _ if no smooth measure is preserved , in @xcite is proved that the pseudo - physical or srb - like measures still exist and satisfy pesin s entropy formula , provided that the system is @xmath0 partially hyperbolic .
recently , and also for @xmath0 partially hyperbolic systems , @xcite derived a proof of shub s entropy conjecture @xcite from their method of construction of measures that satisfy pesin s entropy formula .
in this paper we focus on @xmath0 anosov systems to search for a converse of the result in @xcite .
namely , our purpose is to characterize all the invariant measures that satisfy pesin s entropy formula .
first , we need to generalize the concept of pseudo - physical or srb - like measure .
so , we define the _ weak pseudo physical _ measures @xmath2 , by taking into account only the @xmath5-approach of its basin of statistical attraction up to time @xmath8 , which we denote by @xmath9 , and the exponential rate of the variation of the lebesgue measure of @xmath9 when @xmath10 ( definition [ definitionweakpseudophysicalmeasure ] ) . in theorem [ propositionweaklypseudophysicalmeasures ]
we study general properties of the weak pseudo physical measures , which do always exist .
we prove that for any @xmath0 anosov diffeomorphism , the weak pseudo - physical measures satisfy pesin s entropy formula ( part a ) of theorem [ maintheo ] ) . besides , we prove a converse result , to conclude that the set of invariant measures that satisfy pesin s entropy formula is the closed convex hull of the weak pseudo - physical measures ( theorem [ maintheo ] ) . _ _
so , theorem [ maintheo ] characterizes all the measures that satisfy pesin s entropy formula in terms of the statistical properties that define the weak pseudo - physical notion .
nevertheless , as far as we know , no example is still known of a @xmath11 anosov diffeomorphism for which weak pseudo - physical measures are not physical . in other words , there are not known examples of @xmath0-anosov systems such that an ergodic measure satisfies pesin s entropy formula and is non physical .
the proof of theorem [ maintheo ] is based on the construction of local @xmath0 pseudo - unstable foliations , which approach the local @xmath4 unstable foliation , and allow us to apply a fubini decomposition of the lebesgue measure of the @xmath5-basin @xmath9 , for any ergodic measure @xmath2 .
the pseudo - unstable foliations are constructed via hadamard graphs whose future iterates have bounded dispersion .
this method was introduced by ma in @xcite to prove pesin s entropy formula in the @xmath0 plus hlder context .
much later , it was applied also to @xmath0 systems in @xcite and @xcite . as said above , in theorem [ maintheo ] of this paper
, we prove that the weak pseudo - physical condition for the ergodic components of an invariant measure , is necessary and sufficient to satisfy pesin s entropy formula .
the sufficient condition is just a corollary of the results in @xcite . on the contrary ,
the proof of the necessary condition is new , although it is also strongly based on ma s method to construct , via hadamard graphs , the @xmath0-pseudo unstable foliations . as a subproduct of the proof of theorem [ maintheo ] , we also obtain an equality for any ergodic measure of a @xmath0-anosov diffeomorphism , even for measures that do not satisfy pesin s entropy formula .
this equality , which is stated in theorem [ maintheorem3 ] , considers the exponential rate @xmath12 according to which the lebesgue measure of the @xmath5-basin @xmath13 of each ergodic measure @xmath2 varies with time @xmath8 .
theorem [ maintheorem3 ] equals the exponential rate @xmath14 with the difference @xmath15 where @xmath16 is the metric entropy and @xmath17 are the positive lyapunov exponents .
in the particular case of ergodic measures @xmath2 satisfying pesin s entropy formula , the exponential rate @xmath14 is null , and conversely .
let @xmath18 be a compact , connected , riemannian @xmath0-manifold without boundary and let @xmath19 be continuous .
_ _ [ definitionempiricproba ] ( empiric probability ) for each @xmath20 , the _ empiric probability _
@xmath21 along the finite piece of the future orbit of @xmath22 up to time @xmath8 , is defined by @xmath23 where @xmath24 denotes the dirac delta probability measure supported on the point @xmath25 . _
_ _ _ we denote by @xmath26 the space of all the borel probability measures on @xmath18 , endowed with the weak@xmath1 topology .
we denote by @xmath27 the space of @xmath28-invariant borel probability measures .
it is well known that @xmath26 and @xmath29 are nonempty , weak@xmath1 compact , metrizable , sequentially compact and convex topological spaces .
we fix and choose a metric @xmath30 in @xmath26 that induces the weak@xmath1 topology .
_ _ [ definitionbasinofattraction ] ( basin and pseudo basin of attraction of a measure . ) let @xmath31 and @xmath32 .
we construct the following measurable sets in the manifold @xmath18 : @xmath33 @xmath34 @xmath35 _ _ we call @xmath36 the _ basin of attraction _ of @xmath2 .
we call @xmath37 the _ @xmath5-pseudo basin of attraction _ of @xmath2 .
we call @xmath9 the _ @xmath5-pseudo basin of @xmath2 up to time @xmath8 . _
_ _ _ _ _ _ in the sequel we denote by @xmath7 the lebesgue measure of @xmath18 , renormalized to be a probability measure . _ _ [ definitionweakpseudophysicalmeasure ] ( physical , pseudo - physical and weak pseudo - physical measures ) _ _ let @xmath38 .
we call @xmath2 _ physical _ if @xmath39 . _ _
we call @xmath2 _ pseudo - physical _ if @xmath40 for all @xmath32 . _ _
we call @xmath2 _ weak pseudo - physical _ if @xmath41 we denote @xmath42 _ _ on the one hand , it is standard to check that for continuous mappings @xmath19 the physical measures , if they exist , are @xmath28-invariant . also , pseudo - physical are @xmath28-invariant ( see @xcite ) , page 153 , and as proved in theorem 1.3 of @xcite , the set of pseudo - physical measures is never empty , weak@xmath1 compact and independent of the chosen metric @xmath30 that induces the weak@xmath1 topology of @xmath26 . besides , it is immediate to check that physical measures , if they exist , are particular cases of the always existing pseudo - physical measures . on the other hand ,
in this paper we will generalize the previous results that hold for pseudo - physical measures , by proving the following properties also for weak pseudo - physical measures : [ propositionweaklypseudophysicalmeasures ] let @xmath19 be a continuous map . then : \a ) weak pseudo - physical measures are @xmath28-invariant .
\b ) physical measures and pseudo - physical measures are particular cases of weak pseudo - physical measures .
\c ) weak pseudo - physical measures do always exist .
\d ) the set @xmath43 of weak pseudo - physical measures does not depend on the choice of the metric @xmath30 that induces the weak@xmath1 topology on @xmath26 .
\e ) @xmath43 is weak@xmath1-compact , hence sequentially compact .
\f ) @xmath44 for lebesgue almost all @xmath45 .
\g ) if the weak pseudo - physical measure @xmath2 is unique , then it is physical and its basin of attraction @xmath36 covers lebesgue a.e . @xmath45 .
_ _ [ remarkpseudophysicalnoergodica ] weak pseudo - physical measures _ are not necessaritly ergodic _
( see example 5.4 of @xcite ) .
_ _ _ _ now , let @xmath46 be a @xmath0 diffeomorphism on @xmath18 .
[ definitionanosov ] ( anosov diffeomorphisms ) the diffeomorphism @xmath28 is called _ anosov _ if there exists a riemannian metric of @xmath18 and asplitting @xmath47 which is continuous and non trivial ( i.e. @xmath48 ) , and a constant @xmath49 , such that @xmath50 we call @xmath51 and @xmath52 the _ stable and unstable subbundles _ respectively .
we call @xmath53 the _ ( uniform ) hyperbolicity constant . _ _ _ _ _ _ _ _ [ remarkanosov ] _ we observe that the condition of continuity of the unstable and stable subbundles is redundant in definition [ definitionanosov ] . besides ,
since the manifold is connected , from the continuity of @xmath54 and @xmath51 we deduce that they are uniformly transversal sub - bundles and @xmath55 and @xmath56 are constants . _ _ from inequalities ( [ eqnl11 ] ) , for any anosov diffeomorphism @xmath28 and for any regular point @xmath45 , the minimum lyapunov exponent along @xmath57 is not smaller than @xmath58 , and the maximum lyapunov exponent along @xmath51 is not larger than @xmath59 .
thus , for any regular point @xmath45 all the lyapunov exponents along @xmath60 are strictly negative and bounded away from zero , and all the lyapunov exponents along @xmath57 are strictly positive and bounded away from zero .
_ _ [ definitionpesinformula ] ( pesin s entropy formula ) let @xmath46 .
let @xmath61 .
we say that @xmath2 _ satisfies pesin s entropy formula _ if @xmath62 where @xmath16 is the metric entropy of @xmath28 with respect to @xmath2 ; for @xmath2-a.e .
@xmath45 the lyapunov exponents of the orbit of @xmath22 are denoted by @xmath63 and @xmath64 .
_ _ _ _ recall that for any @xmath0- anosov diffeomorphism @xmath28 , the set of measures that satisfy pesin s entropy formula is nonempty ( see for example theorems 4.2.3 and 4.5.6 of @xcite ) .
the main purpose of this paper is to prove the following result : [ maintheo ] for @xmath0 anosov diffeomorphisms , the set of ergodic weak pseudo - physical measures is nonempty , and the set of invariant probability measures that satisfy pesin s entropy formula is its closed convex hull . _
_ the following is an equivalent restatement of theorem [ maintheo ] : _ _
\a ) all the weak pseudo - physical measures satisfy pesin s entropy formula .
\b ) any invariant probability measure @xmath2 satisfies pesin s entropy formula if and only if its ergodic components @xmath65 are weak pseudo - physical @xmath2-a.e .
_ from theorem [ maintheo ] , we obtain the following consequence : [ corollaryc2-bonattidiazviana ] if @xmath66 is anosov , then for lebesgue - almost all @xmath45 any convergent subsequence of the empirical probabilities
@xmath21 converges to a measure that satisfies pesin s entropy formula . from assertion f ) of theorem
[ propositionweaklypseudophysicalmeasures ] , for lebesgue - almost all @xmath45 any convergent subsequence of @xmath67 converges to a weak pseudo - physical measure @xmath2 .
thus , applying part a ) of theorem [ maintheo ] @xmath2 satisfies pesin s entropy formula .
the arguments to prove theorem [ maintheo ] are based in the following more general result , which we will prove along the paper : [ maintheorem3 ] if @xmath68 is anosov , if @xmath54 denotes its unstable sub - bundle , and if @xmath2 is an ergodic probability measure for @xmath28 , then the @xmath5-pseudo basin @xmath13 of @xmath2 up to time @xmath8 satisfies the following equality : _
@xmath69 _ in section [ sectionpropertiesweaklypseudophysicalmeasures ] we prove theorem [ propositionweaklypseudophysicalmeasures ] , which states the general properties of weak pseudo physical measures for any continuous map @xmath19 . in section [ sectionsufficientcondition ] , for anosov diffeomorphisms , we prove part a ) of theorem [ maintheo ] and also the first part of b )
precisely , we prove that the weak pseudo - physical property of the ergodic components is a sufficient condition to satisfy pesin s entropy formula . in section [ sectionnecessarycondition ] , for anosov diffeomorphisms , we prove the converse statement in part b ) of theorem [ maintheo ] .
namely , the weak pseudo - physical property of the ergodic components is also a necessary condition to satisfy pesin s entropy formula . through the proof of theorem [ maintheo ]
, we obtain some stronger intermediate results that hold for any ergodic measure .
finally , at the end of section [ sectionnecessarycondition ] , we join those intermediate results to prove theorem [ maintheorem3 ] .
the purpose of this section is to prove theorem [ propositionweaklypseudophysicalmeasures ] . along this section , we assume that @xmath28 is only a continuous map from a compact riemannian manifold @xmath18 into itself .
let us divide the proof of theorem [ propositionweaklypseudophysicalmeasures ] into its assertions a ) to f ) : theorem [ propositionweaklypseudophysicalmeasures ] a ) _ any weak pseudo - physical measure @xmath2 is @xmath28-invariant .
_ _ _ from equality ( [ eqnl12 ] ) , for any fixed value of @xmath32 there exists @xmath70 such that @xmath71 .
thus , there exists @xmath72 such that @xmath73 since @xmath74 and @xmath26 is sequentially compact , it is not restrictive to assume that @xmath75 is weak@xmath1 convergent .
denote by @xmath76 its limit .
we assert that @xmath76 is @xmath28-invariant .
in fact , consider the operator @xmath77 defined by @xmath78 for any borel measurable set @xmath79 .
then @xmath80 for all @xmath25 ; hence @xmath81 for all @xmath82 .
it is well known that @xmath83 is continuous .
thus , taking limit in the weak@xmath1 topology , we obtain : @xmath84 since @xmath85 we deduce that the total variation of the signed measure @xmath86 is @xmath87 thus @xmath88 hence @xmath89 , or equivalently @xmath76 is @xmath28-invariant . from ( [ eqnl13 ] ) @xmath90 .
we have proved that for all @xmath32 there exists @xmath91 such that @xmath92 . since @xmath29 is sequentially compact
, we deduce that @xmath61 , as wanted .
theorem [ propositionweaklypseudophysicalmeasures ] b ) _ any physical or pseudo - physical measure is weak pseudo - physical . _ _ _ trivially any physical measure is pseudo - physical .
so , it is only left to prove that any pseudo - physical measure @xmath2 is weak pseudo - physical .
consider @xmath93 . from equality ( [ eqnl15 ] )
, there exists @xmath70 such that @xmath94 .
therefore , from ( [ eqnl16 ] ) @xmath95 since the latter assertions holds for all @xmath93 , we have proved that @xmath96 as @xmath2 is pseudo - physical , we deduce : @xmath97 now , assume by contradiction , that @xmath2 is not weak pseudo - physical .
taking into account that @xmath98 , from the contrary of equality ( [ eqnl12 ] ) , we deduce that there exist @xmath32 and @xmath99 such that @xmath100 therefore , there exists @xmath101 such that @xmath102 for all @xmath103 from where we deduce that @xmath104 finally , applying borell - cantelli lemma , we conclude that @xmath105 contradicting inequality ( [ eqnl17 ] ) .
theorem [ propositionweaklypseudophysicalmeasures ] c ) _ weak pseudo - physical measures do exist .
_ _ _ in theorem 1.3 of @xcite , it is proved for any continuous map @xmath28 on a compact manifold , that the pseudo - physical measures ( which in that paper are also called srb - like or observable ) do exist . since any pseudo - physical measure is weak pseudo - physical ,
these latter measures always exist .
theorem [ propositionweaklypseudophysicalmeasures ] d ) _ the set @xmath43 of weak pseudo - physical measures does not depend on the choice of the metric in @xmath26 that induces the weak@xmath1 topology .
_ _ take two metrics @xmath106 and @xmath107 , both inducing the weak@xmath1 topology on @xmath26 .
we assume that @xmath108 is weak pseudo - physical according to @xmath106 , and let us prove that it is also weak pseudo - physical according to @xmath107 . since both metric induce the same topology , for any @xmath32 there exists @xmath109 such that @xmath110 in the notation of equality ( [ eqnl16 ] ) , add a subindex 1 or 2 to denote the sets @xmath111 and @xmath112 , according to which metric ( @xmath106 and @xmath107 , respectively ) is used to define them .
so , from assertion ( [ eqnl19 ] ) we have : @xmath113 from where @xmath114 since we are assuming that @xmath2 is weak pseudo - physical according to @xmath115 , from equality ( [ eqnl12 ] ) we know that @xmath116 then , @xmath117 as @xmath32 was arbitrarily chosen , the latter inequality holds for all @xmath32 .
but the limit in the latter inequality is non positive because @xmath7 is a probability measure .
we conclude that @xmath118 ending the proof that @xmath2 is also weak pseudo - physical with respect to the metric @xmath107 .
theorem [ propositionweaklypseudophysicalmeasures ] e ) _ the set @xmath119 of weak pseudo - physical measures is weak@xmath1-compact . _ _ _ since @xmath120 and @xmath26 is weak@xmath1-compact , it is enough to prove that @xmath43 is weak@xmath1-closed .
assume @xmath121 and @xmath38 such that @xmath122 we will prove that @xmath123 . for any given @xmath32 , choose and fix @xmath124 such that @xmath125 .
thus , from equality ( [ eqnl16 ] ) and the triangle property , we obtain : @xmath126 from where @xmath127 since @xmath121 , we can apply equality ( [ eqnl12 ] ) to @xmath128 , which joint with inequality ( [ eqnl18 ] ) implies : @xmath129 finally , since @xmath7 is a probability measure , we deduce that the above limsup equals 0 , concluding that @xmath123 as wanted .
theorem [ propositionweaklypseudophysicalmeasures ] f ) _ @xmath44 for lebesgue almost all @xmath45 . _
_ _ theorem 1.5 of @xcite , states that the distance between @xmath21 and the set of pseudo - physical measures converges to zero with @xmath10 for lebesgue almost all @xmath45 .
since the pseudo - physical measures are contained in @xmath43 , we trivially deduce the wanted equality .
theorem [ propositionweaklypseudophysicalmeasures ] g ) _ if the weak pseudo - physical measure @xmath2 is unique , then it is physical and its basin of attraction @xmath36 covers lebesgue a.e .
_ _ _ it is an immediate consequence of part f ) .
in the sequel we assume that the map @xmath130 is anosov .
the purpose of this section is to deduce , as an immediate consequence from previous known results , part a ) of theorem [ maintheo ] , and the sufficient condition to satisfy pesin s entropy formula in part b ) of theorem [ maintheo ] .
namely , we will deduce that if all the ergodic components of an @xmath28-invariant measure @xmath2 are weak pseudo - physical , then @xmath2 satisfies pesin s entropy formula .
recall definition [ definitionbasinofattraction ] , which defines the @xmath5-pseudo basin @xmath9 up to time @xmath8 of a probability measure @xmath131 .
we will apply the following result : @xcite [ theoremcce1 ] let @xmath18 be a compact riemannian manifold of finite dimension .
let @xmath46 be anosov with hyperbolic splitting @xmath47 , where @xmath51 and @xmath54 are the stable and unstable sub - bundles respectively .
then , the following inequality holds for any @xmath28-invariant @xmath38 : @xmath132 proof .
see proposition 2.1 in @xcite . _ [ remarkruelle ] _ for the non negative lyapunov exponents , we adopt the notation @xmath133 as in definition [ definitionpesinformula ] . for any @xmath46 , margulis and ruelle inequality
@xcite states : @xmath134 thus , pesin s entropy formula holds for an invariant measure @xmath2 , if and only if the following inequality holds : @xmath135 besides , from definition [ definitionanosov ] , and from the formula of the integral of the volume form along the unstable sub - bundle @xmath54 , we obtain the following equality for anosov diffeomorphisms : @xmath136 joining the above assertions , we conclude : _ _ _ let @xmath68 be anosov , and @xmath54 be its unstable sub - bundle .
then , any @xmath28-invariant probability measure @xmath2 satisfies pesin s entropy formula if and only if @xmath137 _
we are ready to deduce part a ) of theorem [ maintheo ] , which is indeed a corollary of theorem [ theoremcce1 ] : part a ) of theorem [ maintheo ] : _ if @xmath68 is anosov and if @xmath2 is a weak pseudo - physical @xmath28-invariant measure , then @xmath2 satisfies pesin s entropy formula .
therefore , the set of invariant probability measures that satisfy pesin s entropy formula is nonempty . _ _ _ by contradiction , assume that @xmath2 does not satisfy pesin s entropy formula .
according to remark [ remarkruelle ] , inequality ( [ eqnl21 ] ) does not hold : @xmath138 therefore , applying inequality ( [ eqnl20 ] ) of theorem [ theoremcce1 ] , we conclude that there exists @xmath32 such that @xmath139 so , equality ( [ eqnl12 ] ) does not hold ; hence @xmath2 is not weak pseudo - physical , contradicting the hypothesis . we have proved that all the weak pseudo - physical measures for @xmath28 satisfy pesin s entropy formula . from part c ) of theorem [ propositionweaklypseudophysicalmeasures ] , weak pseudo - physical measures do exist .
so , the set of measures that satisfy pesin s entropy formula is nonempty .
we now recall the following well known result ( see for instance theorems 4.3.7 and 4.5.6 of @xcite ) : [ theoremconvexhull ] let @xmath68 be anosov .
an @xmath28-invariant measure @xmath2 satisfies pesin s entropy formula if and only if its ergodic components @xmath65 satisfy it for @xmath2-a.e .
@xmath45 . on the one hand , we recall that any anosov @xmath0 diffeomorphism @xmath28 is expansive , and for any expansive homeomorphism @xmath28 on @xmath140 the metric entropy @xmath16 depends upper semi - continuously on the @xmath28-invariant measure @xmath2 ( see for instance theorem 4.5.6 of @xcite ) .
so , we can apply the theorem of the affinity of the entropy function ( see theorem 4.3.7 of @xcite ) , which states that @xmath141 where the measures @xmath142 for @xmath143 are the ergodic components of @xmath2 . on the other hand ,
the ergodic decomposition theorem states that @xmath144 joining equalities ( [ eqnl22 ] ) and ( [ eqnl23 ] ) , and taking into account margulis and ruelle inequality ( [ eqnruelleinequality ] ) , we deduce that @xmath145 if and only if @xmath146 ending the proof of theorem [ theoremconvexhull ] . as a consequence
we obtain : part b ) of theorem [ maintheo ] , sufficient condition : _ if @xmath68 is anosov and if @xmath2 is an invariant measure whose ergodic components @xmath65 are weak pseudo - physical for @xmath2-a.e .
@xmath45 , then @xmath2 satisfies pesin s entropy formula . _ _ _ from the hypothesis , and applying part a ) of theorem [ maintheo ] , we deduce that the ergodic components @xmath65 of @xmath2 satisfy pesin s entropy formula for @xmath2-a.e .
so , from theorem [ theoremconvexhull ] , the measure @xmath2 also satisfies this formula .
in this section we will prove the necessary condition to satisfy pesin s entropy formula , as stated in part b ) of theorem [ maintheo ] .
precisely , we will prove that if @xmath46 is anosov , and if the @xmath28-invariant measure @xmath2 satisfies pesin s entropy formula , then the ergodic components of @xmath2 are weak pseudo - physical .
we will also prove the equality of theorem [ maintheorem3 ] for any ergodic measure @xmath2 .
recall that any anosov diffeomorphism @xmath28 is expansive ( see for instance lemma 3.4 in @xcite ) .
namely , there exists a constant @xmath147 , which is called the expansivity constant , such that @xmath148 metric entropy for expansive systems .
recall the following result , which follows from kolmorgorov - sinai theorem in the case of expansive homeomorphisms ( see for instance , proposition 2.5 of @xcite , or also theorem 3.2.18 and lemma 4.5.4 of @xcite ) : _ if @xmath150 is a finite partition whose pieces are borel measurable sets and have diameter smaller than the expansivity constant @xmath153 , then @xmath154 generates the borel @xmath155-algebra , and for any @xmath28-invariant measure @xmath2 , the metric entropy @xmath16 can be computed by : _
@xmath156 @xmath157 note : in ( [ eqn39 ] ) at right , @xmath158 denotes the number of elements of the finite set @xmath159 . _
_ rectangles .
recall the definition of rectangle @xmath160 in the manifold @xmath18 for the anosov diffeomorphism @xmath28 .
( see @xcite , page 78 . ) in particular a rectangle @xmath160 is proper if @xmath161 . for any proper rectangle @xmath160 and any @xmath162 denote @xmath163 @xmath164 where @xmath165 are the stable and unstable submanifolds of the point @xmath22 .
\(ii ) there exists a constant @xmath169 such that , if @xmath170 is a local embedded @xmath0-submanifold , with dimension equal to the unstable dimension , and such that @xmath171 intersects transversally the local stable manifolds @xmath172 for all @xmath162 , then @xmath173 where @xmath174 denotes the lebesgue measure along @xmath171 .
let @xmath179 be a markov partition of the manifold @xmath18 , let @xmath45 and denote by @xmath180 the rectangle of @xmath150 that contains @xmath22 .
let @xmath181 be a positive natural number .
the _ dynamical rectangle _ @xmath182 that contains @xmath22 is defined by @xmath183 _ _ [ lemmaentropia ]
let @xmath28 be an anosov diffeomorphism on a compact manifold .
then , for any finite markov partition @xmath150 , there exists a constant @xmath185 satisfying the following inequality for any @xmath28-invariant probability measure @xmath2 , for any @xmath186 , for any borel measurable set @xmath187 such that @xmath188 , and for any natural number @xmath189 : @xmath190 denote @xmath191 since @xmath192 we have @xmath193 . if @xmath194 then inequality ( [ eqnl40 ] ) holds trivially as a consequence of ( [ eqn39 ] ) .
so , let us prove lemma [ lemmaentropia ] in the case @xmath195 by definition : @xmath196 @xmath197 @xmath198 @xmath199 construct the probability measures @xmath200 and @xmath201 defined by the following equalities for all borelian set @xmath79 : @xmath202 we obtain @xmath203 @xmath204 applying inequality ( [ eqn39 ] ) : @xmath205 @xmath206 taking into account that @xmath207 and that @xmath208 is strictly increasing for @xmath209 and strictly decreasing for @xmath210 , we obtain : @xmath211 @xmath212 @xmath213 where @xmath214 therefore , to end the proof of lemma [ lemmaentropia ] it is enough to show that @xmath215 .
in fact , for a markov partition @xmath150 , any rectangle @xmath216 is obtained as a connected component of the intersection @xmath217 for some pair of rectangles @xmath218 .
fixing @xmath219 , the maximum number of connected components of the intersections of @xmath220 with the rectangle @xmath221 of the partition , is upper bounded the following quotient @xmath222 @xmath223 where @xmath224 thus , denoting @xmath225 , we have @xmath226 we conclude that @xmath227 which implies @xmath228 ending the proof of lemma [ lemmaentropia ] . in the following lemma we will construct a local @xmath0-foliation , whose leaves are _ pseudo - unstable manifolds _
@xmath5- approaching ( in the @xmath0-topology ) the true local unstable manifolds of any rectangle of a given markov partition .
_ _ [ lemmafoliation ] let @xmath68 be anosov .
denote the stable and unstable subbundles by @xmath51 and @xmath229 , respectively .
denote the expansivity constant by @xmath147 .
then , for all @xmath32 there exist @xmath230 and @xmath231 such that , for any finite markov partition @xmath175 into rectangles with diameter smaller than @xmath232 , there exists a finite family @xmath233 of local foliations @xmath234 , each one defined in an open neighborhood of each rectangle @xmath176 , satisfying the following properties for all @xmath235 , for all @xmath236 and for all @xmath237 : \d ) there exist a point @xmath241 and an open subset @xmath242 , in the topology of the stable submanifold @xmath243 , such that _ @xmath244 _ where _ @xmath245 _ denotes the lebesgue measure along the submanifold @xmath243 ; and besides , if @xmath246 , then _
@xmath247 _ where _ @xmath248 _ denotes the lebesgue measure along the submanifold @xmath249 . _
_ _ _ _ _ _ _ proposition 3.6 of @xcite states the existence of @xmath250 and the local @xmath0-foliation @xmath234 satisfying ( a ) , ( b ) and ( c ) .
so , it is enough to prove that if @xmath32 is small enough , then any local @xmath0-foliation @xmath234 defined in a neighborhood of the rectangle @xmath176 and satisfying ( a ) and ( b ) , also satisfies ( d ) for some constant @xmath251 .
in fact , choose and fix any point @xmath252 in the interior of the rectangle @xmath176 .
from the definition of rectangle , for each @xmath253 there exists a unique point in the transversal intersection @xmath254 by continuity of the transversal intersection between @xmath0-manifolds , there exists @xmath255 such that the following assertion holds : if @xmath256 and if @xmath234 is any local foliation whose leaves have dimension @xmath55 , are @xmath0 , and are @xmath257-near the unstable local leaves of @xmath176 in the @xmath0-topology , then for each @xmath253 the intersection @xmath258 is transversal and contains a single point . from the definition of the dynamical rectangle @xmath182 and from the properties of the markov partition , there exists a rectangle @xmath269 of the partition such that local stable manifold @xmath270 for all @xmath271 .
so , we deduce @xmath272 for all @xmath267 and for all @xmath273 in other words , the pseudo - unstable @xmath274 submanifold @xmath275 intersects transversally all the local stable submanifolds of the rectangle @xmath221 where it is contained . thus , applying inequality ( [ eqnl25 ] ) we have @xmath276 for any probability measure @xmath2 recall equality ( [ eqnl16 ] ) , defining the measurable set @xmath9 which we called the @xmath5-pseudo basin of @xmath2 up to time @xmath8 .
we will end the proof of theorem [ maintheo ] , by applying the following key result which bounds from below the lebesgue measure of the set @xmath9 for any ergodic measure @xmath2 : [ theorem2 ] let @xmath18 be a compact riemannian manifold of finite dimension .
let @xmath46 be anosov with hyperbolic splitting @xmath47 , where @xmath51 and @xmath54 are the stable and unstable sub - bundles respectively .
let @xmath2 be an ergodic measure . then : @xmath278 we notice that the limit at left in equality ( [ eqnl20-b ] ) does not depend on the choice of metric @xmath30 that induces the weak@xmath1 topology in the space @xmath26 of borel probability measures .
in fact , to prove the latter assertion it is enough to argue as in the proof of part ( d ) of theorem [ propositionweaklypseudophysicalmeasures ] in section [ sectionpropertiesweaklypseudophysicalmeasures ] .
so , to prove theorem [ theorem2 ] we choose and fix the following metric in @xmath26 : @xmath279 where @xmath280 is any fixed countable family of real continuous functions @xmath281)$ ] that is dense in @xmath282)$ ] . note that
, according to the metric @xmath30 , the balls are convex . in other words
, if a finite number of probability measures belong to the ball with centre @xmath2 and radius @xmath32 , then any convex combination of those measures also belongs to it .
fix any real value @xmath32 .
the real funcion @xmath285 is continuous because @xmath28 is of class @xmath0 and the sub - bundle @xmath54 is continuous .
thus , from the definition of the weak@xmath1 topology in the space @xmath26 of probability measures , we deduce that there exists @xmath286 such that @xmath287 in particular , for @xmath288 we deduce : @xmath289 since @xmath2 is an ergodic probability measure , we have @xmath290 for @xmath2-a.e .
so , for the fixed value of @xmath255 as above , and for @xmath2-a.e .
@xmath45 , there exists @xmath291 such that @xmath292 for any natural value of @xmath293 , define the set @xmath294 since @xmath295 and @xmath296 , there exists @xmath297 such that @xmath298 in the sequel , we fix such a value of @xmath293 . from the definition of the metrizable weak@xmath1-topology in the space @xmath26 of borel - probability measures ,
it is standard to check that the dirac delta probability @xmath299 depends uniformly continuously on the point @xmath45 .
since the empiric probability @xmath21 is a convex combination of dirac delta measures , and the balls in @xmath26 are convex , we deduce that there exists @xmath300 such that , for any pair of points @xmath301 and for any natural value of @xmath189 , the following assertion holds : for the fixed value of @xmath303 at the beginning , we construct the real numbers @xmath304 ( where @xmath305 is expansivity constant ) , and @xmath251 , as in lemma [ lemmafoliation ] .
we consider any markov partition @xmath175 with diameter smaller than @xmath306 and , for each rectangle @xmath176 , we construct the @xmath0-foliation @xmath234 that satisfies the properties ( a ) to ( d ) of lemma [ lemmafoliation ] . from equality ( [ eqnl31 ] ) , assertion ( [ eqnl35 ] ) , and the triangle property of the metric , we deduce the following assertion for all @xmath307 : @xmath308 @xmath309 recalling equality ( [ eqnl16 ] ) , from the above assertion we deduce that @xmath310 for all @xmath311 . since the rectangle @xmath182 is any piece of the partition @xmath312 that intersects @xmath313 , we deduce the following statement for all @xmath307 : @xmath314 therefore @xmath315 besides , joining assertion ( [ eqnl30 ] ) and inequality ( [ eqnl37 ] ) , we deduce the following property for all @xmath307 : @xmath316 @xmath317 now , for any @xmath307 , let us compute @xmath318 for any rectangle @xmath216 such that @xmath319 . since @xmath320 ,
to compute @xmath318 we will use the fubini decomposition of the lebesgue measure along the local pseudo - unstable @xmath0-foliation @xmath234 . applying part ( d ) of lemma [ lemmafoliation ]
consider the point @xmath241 and the submanifold @xmath321 .
taking the fubini decomposition of @xmath7 we obtain : @xmath322 where @xmath323 is a local @xmath0-diffeomorhism that parameterizes the neighborhood of @xmath176 and trivializes the @xmath0-foliation @xmath234 .
therefore , @xmath324 is continuous and bounded away from zero by a constant , say @xmath325 . since @xmath326 is an open subset of @xmath263 in the topology of this local stable manifold
, we obtain : @xmath327 changing variables @xmath328 in the integral at right , we obtain : @xmath329 @xmath330 since @xmath331 , we can apply inequality at left of part c ) of lemma [ lemmafoliation ] : @xmath332 @xmath333 since @xmath334 and @xmath319 , we can apply inequality ( [ eqnl38 ] ) : @xmath335 @xmath336 @xmath337 from part ( d ) of lemma [ lemmafoliation ] we know that @xmath338 for all @xmath339 , and besides @xmath340 .
thus , we have proved the following inequality for all @xmath307 , and for all @xmath216 such that @xmath319 : @xmath341 joining the above inequality with inequality ( [ eqnl36 ] ) , we deduce , for all @xmath307 : @xmath342 therefore , @xmath343 @xmath344 finally , applying lemma [ lemmaentropia ] , we deduce that @xmath343 @xmath345 so , from equality ( [ eqnmetricentropy ] ) , we conclude @xmath346 ending the proof of theorem [ theorem2 ] .
part b ) of theorem [ maintheo ] , necessary condition : _ if @xmath68 is anosov and if @xmath2 is an invariant measure satisfying pesin s entropy formula , then its ergodic components @xmath65 are weak pseudo - physical for @xmath2-a.e .
_ _ _ so @xmath350 is increasing with @xmath32 .
thus @xmath351 but since @xmath7 is a probability measure , we conclude that @xmath352 applying definition [ definitionweakpseudophysicalmeasure ] , we deduce that @xmath2 is weak pseudo - physical . we have proved that any ergodic measure that satisfies pesin s entropy formula is weak pseudo - physical . now let us consider a non ergodic measure @xmath2 that satisfies pesin s entropy formula . from theorem [ theoremconvexhull ]
we know that its ergodic components @xmath65 also satisfy that formula for @xmath2-a.e .
we conclude that the ergodic components @xmath65 of @xmath2 are weak pseudo - physical for @xmath2-a.e .
@xmath45 , as wanted . | we consider @xmath0 anosov diffeomorphisms on a compact riemannian manifold . we define the weak pseudo - physical measures , which include the physical measures when these latter exist .
we prove that ergodic weak pseudo - physical measures do exist , and that the set of invariant probability measures that satisfy pesin s entropy formula is the weak@xmath1-closed convex hull of the ergodic weak pseudo - physical measures . in brief , we give in the @xmath0-scenario of uniform hyperbolicity , a characterization of pesin s entropy formula in terms of physical - like properties .
2010 _ math . subj . class . : _
primary 37a35 , 37d20 .
secondary 37d35 ; 37a05
. _ keywords and phrases : _ anosov diffeomorphisms ; pesin s entropy formula ; physical measures . |
the chiral and deconfinement phase transitions of nonperaturbative systems are important issues of continuous interests both theoretically and experimentally .
although the mechanism is unknown , the originally chiral symmetric system may undergo chiral phase transition ( cpt ) into a phase with dynamical chiral symmetry breaking ( dcsb ) which explains the origin of constituent - quark masses in qcd and underlies the success of chiral effective field theory @xcite . in the chiral limit , the order parameter of cpt is defined via the fermion propagator @xmath1=\int\frac{\mathrm{d}^dp}{(2\pi)^d}\frac{4b(p^2)}{a^2(p^2)p^2+b^2(p^2)}.\ ] ] the two functions @xmath2 and @xmath3 in the above equation are related to the inverse fermion propagator @xmath4 the deconfinement phase transition is then related to the observation of the free particle and also the corresponding propagator . if the full fermion propagator has no mass singularity in the timelike region , it can never be on mass shell and the free particle can never be observed where the confinement happens @xcite .
accordingly , the appearance of the mass singularity in the system directly implies deconfinement .
so in this way we can learn the deconfinement phase transition from the analytic structure of the fermion propagator .
to indicate dcsb and confinement , it is very suggestive to study some model that reveals the general nonperaturbative features while being simpler .
three - dimensional quantum electrodynamics ( qed@xmath0 ) is just such a model which has many features similar to quantum chromodynamics ( qcd ) , such as dcsb and confinement @xcite .
moreover , its superrenormalization obviate the ultraviolet divergence which is present in qed@xmath5 . due to these reasons
, it can serve as a toy model of qcd . in parallel with its relevance as a tool through which to develop insight into aspects of qcd
, qed@xmath0 is also found to be equivalent to the low - energy effective theories of strongly correlated electronic systems .
recently , qed@xmath0 has been widely studied in graphene @xcite and high - t@xmath6 cuprate superconductors @xcite .
the study of dcsb in qed@xmath0 has been an active subject near 30 years since appelquist _
et al_. found that dcsb vanishes when the flavor of massless fermions reaches a critical number @xmath7 @xcite .
they gain this conclusion by solving the truncated dyson - schwinger equation ( dse ) for the fermion propagator in the chiral limit . later
, extensive analytical and numerical investigations showed that the existence of dcsb in qed@xmath0 remains the same after including higher order corrections to the dse @xcite . on the other hand ,
the achievement in research of the mass singularity and confinement in qed@xmath0 is caused by a paper of p. maris who found that the fermion is confined by the truncated dse for the full fermion and boson propagators at @xmath8 @xcite where chiral symmetry is broken .
this result might imply that the existence of confinement and dcsb depend on the same boundary conditions .
moreover , the authors of ref .
@xcite pointed out that restoration of chiral symmetry and deconfinement are coincident owing to an abrupt change in the analytic properties of the fermion propagator when a nonzero scalar self - energy becomes insupportable .
nevertheless , the above result will be altered when the gauge boson acquires a finite mass @xmath9 through the higgs mechanism @xcite . for a fixed @xmath10 and with the increasing boson mass ,
the fermion chiral condensate falls and diminishes at a critical value @xmath11 ( which , of course , depends on @xmath12 ) and then chiral symmetry restores
. since dcsb and confinement are nonperaturbative phenomena , both of them occur in the low energy region and might disappear with the rise of boson mass .
therefore , it is very interesting to investigate whether or not both phase transitions occur at the same critical point in this case . in this paper , we will adopt the truncated dses for the full propagators to study the behaviors of the mass singularity and the fermion chiral condensate with a range of gauge boson mass and try to answer this question .
the lagrangian for massless qed@xmath0 in a general covariant gauge in euclidean space can be written as @xmath13 where the 4-component spinor @xmath14 is the massless fermion field , @xmath15 is the gauge parameter .
this system has chiral symmetry and the symmetry group is @xmath16 .
the original @xmath16 symmetry reduces to @xmath17 when the massless fermion acquires a nonzero mass due to nonperaturbative effects . just as mentioned in sec .
i , the chiral symmetry is broken by the dynamical generation of the fermion mass ( here @xmath18 ) .
if one adopts the full boson propagator , the results of euclidean - time schwinger function reveal that the fermion propagator has a complex mass singularity and thus corresponds to a nonphysical observable state @xcite which means the appearance of confinement . on the contrary , if the schwinger function exhibits a real mass singularity of the propagator , the fermion is observable and the fermion is not confined @xcite . therefore , we also adopt this method to analyze those nonperaturbative phenomena .
the schwinger function can be written as @xmath19 with @xmath20 .
if there are two complex conjugate mass singularities @xmath21 associated with the fermion propagator , the function will show an oscillating behavior @xmath22 for large ( euclidean ) @xmath23 .
however , the system reveals a stable observable asymptotic state with a mass @xmath24 for the fermion propagator , then @xmath25 by this way , the analysis of mass singularity can be used to determine whether or not the fermion is confined .
since the schwinger function is determined by the fermion propagator and the dses provide us an powerful tool to study it , we shall use the coupled gap equations to calculate this function .
now let us turn to the calculation of @xmath2 and @xmath3 .
these functions can be obtained by solving dses for the fermion propagator , @xmath26 , \label{eq2}\ ] ] where @xmath27 is the full fermion - photon vertex and @xmath28 .
the coupling constant @xmath29 has dimension one and provides us with a mass scale . for simplicity , in this paper temperature ,
mass and momentum are all measured in unit of @xmath30 , namely , we choose a kind of natural units in which @xmath31 .
( [ s0 ] ) and eq .
( [ eq2 ] ) , we obtain the equation satisfied by @xmath2 and @xmath3 @xmath32,\\ b(p^{2})&=&\frac{1}{4}\int\frac{\mathrm{d}^{3}k}{(2\pi)^{3}}tr\left[\gamma_{\sigma}s(k)\gamma_{\nu}(p , k)d_{\sigma\nu}(q)\right].\end{aligned}\ ] ] another involved function @xmath33 is the full gauge boson propagator which is given by@xcite @xmath34+\zeta^2}+\xi\frac{q_\sigma q_\nu}{q^4},\ ] ] where @xmath35 is the vacuum polarization for the gauge boson which is satisfied by the polarization tensor @xmath36\ ] ] and @xmath9 is the gauge boson mass which is acquired though higgs mechanism which happens when the gauge field interacts with a scalar filed in the phase with spontaneous gauge symmetry breaking ( here , we adopt the massive boson propagator to investigate the oscillation behavior of schwinger function in dcsb phase , more details about higgs mechanism in qed@xmath0 can be found in ref .
@xcite ) . using the relation between the vacuum polarization @xmath35 and @xmath37 , @xmath38
we can obtain an equation for @xmath39 which has an ultraviolet divergence .
fortunately , it is present only in the longitudinal part and is proportional to @xmath40 .
this divergence can be removed by the projection operator @xmath41 and then we obtain a finite vacuum polarization@xcite . finally , we choose to work in the landau gauge , since the landau gauge is the most convenient and commonly used one .
once the fermion - boson vertex is known , we immediately obtain the truncated dses for the fermion propagator and then analyze the deconfinement and chiral phase transitions in this higgs model .
the simplest and most commonly used truncated scheme for the dses is the rainbow approximation , @xmath42 since it gives us rainbow diagrams in the fermion dse and ladder diagrams in the bethe - salpeter equation for the fermion - antifermion bound state amplitude . in the framework of this approximation
, the coupled equations for massless fermion and massive boson propagators reduce to the three coupled equations for @xmath2 , @xmath3 and @xmath39 , @xmath43},\\ \label{rfp } b(p^2)&=&\int\frac{\mathrm{d}^{3}k}{(2\pi)^3}\frac{2b(k^2)}{g(k^2)[q^2(1+\pi(q^2))+\zeta^2]},\\ \label{rpi1 } \pi(q^2)&=&\int\frac{\mathrm{d}^{3}k}{(2\pi)^3}\frac{2a(k^2)a(p^2)}{q^2g(k^2)g(p^2)}\times\nonumber\\&&[2k^2 - 4(k\cdot q)-6(k\cdot q)^2/q^2],\end{aligned}\ ] ] with @xmath44 . by application of iterative methods ,
we can obtain @xmath45 and @xmath46 .
after solving the above coupled dses in rainbow approximation by means of the iteration method , we can obtain the three function @xmath52 for the propagator and plot them in fig .
[ fig1 ] .
( left ) and their infrared values ( right ) as functions of the boson mass in dcsb phase.,title="fig:",scaledwidth=23.5% ] ( left ) and their infrared values ( right ) as functions of the boson mass in dcsb phase.,title="fig:",scaledwidth=23.5% ] + ( left ) and their infrared values ( right ) as functions of the boson mass in dcsb phase.,title="fig:",scaledwidth=23.5% ] ( left ) and their infrared values ( right ) as functions of the boson mass in dcsb phase.,title="fig:",scaledwidth=23.5% ] + ( left ) and their infrared values ( right ) as functions of the boson mass in dcsb phase.,title="fig:",scaledwidth=23.5% ] ( left ) and their infrared values ( right ) as functions of the boson mass in dcsb phase.,title="fig:",scaledwidth=23.5% ] + from fig .
[ fig1 ] it can be seen that @xmath2 increases with increasing momenta but almost equal to one at large @xmath53 . in the range of small momenta , it decreases but does not vanish when @xmath54 . both of the other two functions @xmath3 and @xmath39 decrease at large momenta but their rates of decreasing are different .
@xmath3 decreases as rapidly as @xmath55 , while @xmath39 decreases as rapidly as @xmath56 .
in addition , all the three functions are constant in the infrared region .
thus , we can obtain the values of the corresponding functions @xmath45 and @xmath46 at zero momenta , which , as functions of the gauge boson mass @xmath9 , are also shown in fig .
[ fig1 ] . as @xmath9 increases , both @xmath57 and @xmath58 decrease , and @xmath58 vanishes when @xmath9 reaches a critical gauge boson mass @xmath59 , whereas the function @xmath60 rises and diverges at the same critical boson mass @xmath61 . based on eq .
( [ cds ] ) , the critical boson mass can be regarded as the point of chiral phase transition . then , substituting the obtained @xmath62 and @xmath63 into eq .
( [ sf ] ) , we immediately obtain the behavior of the schwinger function with nonzero boson mass which is shown in fig . [ fig2 ] . at small @xmath9
, the schwinger function reveals its typical oscillating behavior which illustrates the conjugate mass singularities like @xmath21 @xmath64 associated with the fermion propagator and thus the free particle can never be observed where the fermion is confined .
as the rise of @xmath9 , the oscillating behavior remains but it vanishes at another critical value @xmath65 and around which both of the propagators do not exhibit any singularity . beyond @xmath66 ,
the function @xmath67\sim - mt$ ] where the stable asymptotic state of the fermion is observable @xmath68 and hence the deconfinement phase transition happens , but the dcsb remains . with the enlargement of @xmath9 , the absolute slope of @xmath67 $ ] decreases and @xmath24 disappears at @xmath61 . for the rainbow approximation.,title="fig:",scaledwidth=45.0% ]
+ vertex with a range of @xmath9.,title="fig:",scaledwidth=45.0% ] + vertex with a range of @xmath9.,title="fig:",scaledwidth=45.0% ] + to validate the difference between @xmath11 and @xmath69 , we also give the behavior of the schwinger function beyond rainbow approximation in fig .
[ fig3 ] . in the bc@xmath70 truncated scheme for dse
, the oscillation of the schwinger function only appears at small @xmath9 , which denotes the existence of confinement , but it disappears at @xmath71 , which exhibits that deconfinement phase transition occurs but here @xmath72 . as the rise of @xmath9 , the schwinger function shows the real mass singularity of the propagator and chiral symmetry gets restored when the boson mass reaches @xmath73 .
the primary goal of this paper is to investigate chiral and deconfinement phase transition by application of an abelian higgs model through a continuum study of the schwinger function .
based on the rainbow approximation of the truncated dses for the fermion propagator and numerical model calculations , we study the behavior of the schwinger function and the fermion chiral condensate .
it is found that , with the rise of the gauge boson mass , the vanishing point ( @xmath69 ) of the oscillation behavior of the schwinger function is apparently less than that of the fermion chiral condensate and each of the propagators _ does not _ reveal any singularity near @xmath69 . to make know the difference between the two critical points , we also work in an improved scheme for the truncated dses and show that the above conclusion remains despite the two critical numerical values alter .
the result indicates that , with the increasing gauge boson mass in the chiral model , the occurrence of de - confinement phase transition is apparently earlier than that of chiral phase transition .
we would like to thank prof . wei - min sun and guo - zhu liu for their helpful discussions . this work was supported by the national natural science foundation of china ( under grant nos .
11105029 , 11275097 and 11205227 ) and the fundamental research funds for the central universities ( under grant no 2242014r30011 ) .
roberts , prog . part .
. phys . * 61 * , 50 ( 2008 ) .
a. bashir , a. raya , i.c .
cloet and c.d .
roberts , phys . rev .
c * 78 * , 055201 ( 2008 ) .
p. maris , phys .
d * 52 * , 6087 ( 1995 ) .
m. r. pennington and d. walsh , phys .
b * 253 * , 246 ( 1991 ) .
burden , j. praschifka , and c.d .
roberts , phys .
d * 46 * , 2695 ( 1992 ) .
feng , b. wang , w.m .
sun , and h.s .
zong , euro .
j. c 73 , 2444 ( 2013 ) .
feng , b. wang , w.m .
sun , and h.s .
zong , phys .
d * 86 * , 105042 ( 2012 ) .
feng , y.q .
zhou , p.l .
yin , and h.s .
zong , phys .
d * 88 * , 125022 ( 2013 ) .
khveshchenko , phys .
lett . * 87 * , 246802 ( 2001 ) .
v.p . gusynin and s.g .
sharapov , phys . rev
. lett . * 95 * , 146801 ( 2005 ) .
drut and t.a .
lahde , phys .
lett . * 102 * , 026802 ( 2009 ) .
n. dorey and n. e. mavromatos , nucl . phys . *
b386 * , 614 ( 1992 ) .
m. franz , z. tesanovic , and o. vafek , phys . rev .
b * 66 * , 054535 ( 2002 ) .
w. rantner and x.g .
wen , phys .
b * 66 * , 144501 ( 2002 )
lee , n. nagaosa , and x.g .
wen , rev .
phys . * 78 * , 17 ( 2006 ) .
t. appelquist , d. nash , and l.c.r .
wijewardhana , phys .
lett . * 60 * , 2575 ( 1988 ) .
d. nash , phys .
lett . * 62 * , 3024 ( 1989 ) .
fischer , r. alkofer , t. dahm , and p. maris , phys .
d * 70 * , 073007 ( 2004 ) .
hofmann , a. raya , and s.s .
madrigal , phys . rev .
d 82 , 096011 ( 2010 ) .
liu and g. cheng , phys .
d * 67 * , 065010(2003 ) .
feng , w.m .
sun , f. hu , and h.s .
zong , inter . j. mod
a * 20*(13 ) , 2753 ( 2005 ) .
feng , y. jiang , w.m .
sun , and h.s .
zong , mod .
a * 27 * , 1250026 ( 2012 ) .
n. brown and m.r .
pennington , phys .
d * 39 * , 2723 ( 1989 ) .
hollenberg , c.d .
roberts , and b.h.j .
mckellar , phys .
c * 46 * , 2057 ( 1992 ) .
roberts and a.g .
williams , prog . part .
* 33 * , 477(1994 ) .
curtis and m.r .
pennington , phys .
d * 42 * , 4165 ( 1990 ) .
m.r . pennington and d. walsh ,
b * 253 * , 246 ( 1991 ) .
kondo and p. maris , phys .
* 74 * , 18 ( 1995 ) .
semenoff , p. suranyi , and l.c.r .
wijewardhana , phys .
d * 50 * , 1060 ( 1994 ) .
j.s . ball and t.w .
chiu , phys .
d * 22 * , 2542 ( 1980 ) . | based on the truncated dyson - schwinger equations for fermion and massive boson propagators in qed@xmath0 , the fermion chiral condensate and the mass singularities of the fermion propagator via the schwinger function are investigated .
it is shown that the critical point of chiral phase transition is apparently different from that of deconfinement phase transition and in nambu phase the fermion is confined only for small gauge boson mass . |
this work was carried out under the visiting researcher@xmath31s program of kyokugen at osaka university and was supported in part by a grand - in - aid for scientific research from the ministry of education , science , sports and culture .
m.h . would like to thank professor k. nonoyama of konan women@xmath31s junior college for information about the synthesis of na@xmath0[cu(pba)]@xmath26h@xmath0o and professor s. yamamoto of okayama university for fruitful discussions .
thanks are also due to the chemical analysis units in riken . 99 f. d. m. haldane : phys . rev .
* 50 * ( 1983 ) 1153 .
l. j. de jongh and a. r. miedema : adv .
* 23 * ( 1974 ) 1 .
m. steiner , j. villan and c. g. windsor : adv .
* 25 * ( 1976 ) 87 .
m. hase , i. terasaki and k. uchinokura : phys .
* 70 * ( 1993 ) 3651 .
e. dagotto and t. m. rice : science * 271 * ( 1996 ) 618 .
a. k. kolezhuk , h .- j .
mikeska and s. yamamoto : phys .
b * 55 * ( 1997 ) r3336 .
s. k. pati , s. ramasesha and d. sen : phys .
b * 55 * ( 1997 ) 8894 .
s. brehmer , h .- j .
mikeska and s. yamamoto : j. phys . : condens
* 9 * ( 1997 ) 3921
. t. kuramoto : cond - mat/9710229 .
m. drillon , j. c. gianduzzo and r. georges : phys . lett .
a * 96 * ( 1983 ) 413 .
m. drillon , e. coronado , r. georges , j. c. gianduzzo and j. curely : phys .
b * 40 * ( 1989 ) 10992 .
m. matsuura , y. okuda , m. morotomi , h. mollymoto and m. date : j. phys .
* 46 * ( 1979 ) 1031 .
y. pei , m. verdaguer , o. kahn , j. sletten and j. p. renard : inorg .
* 26 * ( 1987 ) 138 .
p. j. van koningsbruggen , o. kahn , k. nakatani , y. pei , j. p. renard , m. drillon and p. legoll : inorg . chem . * 29 * ( 1990 ) 3325 .
k. nonoyama , h. ojima and m. nonoyama : inorg .
* 20 * ( 1976 ) 127 . | we report the results of magnetic measurements on a powder sample of nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o ( pba=1,3-propylenebis(oxamato ) ) which is one of the prototypical examples of an @xmath3=1/2 and 1 ferrimagnetic chain .
susceptibility(@xmath4 ) shows a monotonous increase with decreasing temperature ( t ) and reaches a maximum at about 7 k. in the plot of @xmath4@xmath5 versus @xmath6 , the experimental data exhibit a broad minimum and are fit to the @xmath4@xmath5 curve calculated for the ferrimagnetic heisenberg chain composed of @xmath3=1/2 and 1 . from this fit , we have evaluated the nearest - neighbor exchange constant @xmath7=121 k , the g - values of ni@xmath8 and cu@xmath8 , @xmath9=2.22 and @xmath10=2.09 , respectively .
applied external field dependence of @xmath4@xmath5 at low temperatures is reproduced fairly well by the calculation for the same ferrimagnetic model .
extensive studies of one - dimensional systems were prompted by haldane s theoretical work @xcite in 1983 after the initial wave of studies @xcite in the late 1960s and early 1970s .
recently , quantum spin systems with singlet ground states , namely haldane systems @xcite ( linear chain heisenberg antiferromagnets with integer spin values ) , inorganic spin - peierls systems @xcite and even - leg spin ladder systems , @xcite have been studied extensively . in particular , cuprate systems have attracted much attention because of the relation to high @xmath11 superconductors . in regard to the one - dimensional systems with magnetic ground states , an @xmath3=1/2 and 1 ferrimagnetic chain has been theoretically investigated recently , @xcite in addition to some pioneering theoretical works @xcite published in the 1980s . from the low dimensionality and small spin values in this system , we expect a kind of quantum effect .
theoretical studies show some remarkable features as follows : ( 1 ) between two low - lying , gapless and gapped excitation branches , the gapped branch lies higher than that deduced from a conventional spin wave theory . from reliable calculations , @xcite
the gap ( @xmath12/@xmath13 ) has been evaluated to be [email protected] where @xmath13 is the nearest - neighbor exchange constant .
the definition of the hamiltonian will be shown later .
( 2 ) the spin correlation length between sublattice moments is extremely short .
the length is below unit cell length and can not be evaluated with numerical accuracy .
( 3 ) the full magnetization curve up to saturated magnetization is calculated and is obviously different from that for a classical ferrimagnet .
@xcite on the other hand , although some candidates for the ferrimagnetic heisenberg chain composed of spin 1/2 and 1 exist in real bimetallic substances , @xcite only preliminary magnetic measurements and comparisons with numerical calculations were made .
@xcite thus , we investigate precisely the magnetic properties of an alternating ni and cu chain compound nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o ( pba=1,3-propylenebis(oxamato ) ) . in these measurements , we use a deutrated sample because it is of superior quality to a hydrated one , although reason for this remain unclear and we plan to perform neutron scattering measurements on this deutrated sample .
the format used in this letter is as follows : in the next section , we discuss the synthesis and crystal structure of nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o .
we then report the results of magnetic measurements and of the comparison with numerical calculations for @xmath3=1/2 and 1 ferrimagnetic heisenberg chain .
finally , we show the field dependence of @xmath4 times @xmath6 and compare the experimental data with calculated ones for the same model .
powder samples of nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o were synthesized according to the procedure reported in ref . 13 .
na@xmath0[cu(pba)]@xmath26h@xmath0o was prepared from cuso@xmath15 , naoh and 1,3-trimethylenebis(oxamido ) @xcite which was previously synthesized from ethyl oxamate and 1,3-propanediamine .
then , the title compound was obtained by slow diffusion of aqueous solutions ( d@xmath0o @xmath1699.8% ) of na@xmath0[cu(pba)]@xmath26h@xmath0o and ni(clo@xmath15)@xmath0@xmath26h@xmath0o in a u - tube .
chemical analysis showed a slight deviation of h content from the ratio in the ideal deutrated sample , but the molecular weight of this sample was only about 1% smaller than that of nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o .
crystal structure of nicu(pba)(h@xmath0o)@xmath1@xmath22h@xmath0o has not been analyzed , but that of a similar compound mncu(pba)(h@xmath0o)@xmath1@xmath22h@xmath0o where mn replaces ni , has been analyzed . @xcite powder x - ray diffraction patterns of these compounds show that these belong to the same space group .
thus , nicu(pba)(h@xmath0o)@xmath1@xmath22h@xmath0o crystallizes in the orthorhombic system and belongs to the @xmath17 space group .
@xcite as shown in fig . 1 , the structure consists of ordered bimetallic chains along the @xmath18 axis with octahedral ni@xmath8 and square - pyramidal cu@xmath8 ions bridged by oxamato groups . at the apical positions of ni and cu , water molecules are bound .
magnetic measurements were carried out with a squid magnetometer ( quantum design s mpms - xl7l ) at kyokugen in osaka university .
we show in fig .
2 the dc magnetic susceptibility @xmath4(=@xmath19/@xmath20 where @xmath19 and @xmath20 represent magnetization of the sample and the external magnetic field , respectively ) of a powder sample of nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o .
the susceptibility of nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o increases monotonously with decreasing temperature until about 7 k , at which the susceptibility reaches a maximum . below 7 k
, the long - range order probably occurs due to the interchain couplings .
figure 3 shows @xmath4 times @xmath6 of nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o as a function of temperature .
this plot is familiar to chemists but not to physicists .
therefore , we explain this plot in some details . if a magnetic system is paramagnetic , @xmath4@xmath5 is constant over the whole temperature range . if a magnetic system has a dominant ferromagnetic ( antiferromagnetic ) interaction , @xmath4@xmath5 increases ( decreases ) when the temperature is decreased . in fig . 3
, @xmath4@xmath5 decreases when decreasing the temperature from 300 k , implying that antiferromagnetic coupling exists between the nearest neighbor spins , and reaches a rounded minimum at about 70 k. then , @xmath4@xmath5 increases and reaches a maximum at about 10 k , and hereafter , when the temperature is decreased further , it decreases rapidly .
the increase in @xmath4@xmath5 below 70 k implies that this ferrimagnetic system behaves like a ferromagnetic chain at low temperatures .
interchain ( antiferromagnetic ) couplings probably give rise to the steep decrease of @xmath4@xmath5 below 10 k. we compare the experimental data with numerical calculations ( exact diagonalization method up to five unit cells ( ten sites ) ) for the @xmath3=1/2 and 1 ferrimagnetic heisenberg chain .
the hamiltonian of this system in a magnetic field is defined by @xmath21}\\ -g_{s } \mu_{\rm b}h \sum_{i=1}^{l } { { \bf s_{i}}}-g_{s } \mu_{\rm b}h \sum_{i=1}^{l } { { \bf s_{i}}},\ ] ] where * s * and * s * are the @xmath3=1 and @xmath3=1/2 spin operators , respectively , and @xmath22 and @xmath23 the g - values of the @xmath3=1 and @xmath3=1/2 magnetic moments , respectively , and @xmath24 the bohr magneton and @xmath25 the external magnetic field . here , the periodic boundary condition is imposed , so that * s*@xmath26=*s*@xmath27 . the solid line in fig .
3 shows the result of the best fit to the experimental data between 30 k and 150 k. good agreement between experimental and calculated results is achieved between 30 k and 150 k. slight deviation at high temperatures may arise from the error enhancement of @xmath4@xmath5 at high temperatures or the omission of the single ion anisotropy term which exists in ni(@xmath3=1 ) compounds . from this fit
, we obtain the exchange constant @xmath7=121 k , @xmath22(=@xmath9)=2.22 and @xmath23(=@xmath10)=2.09 .
next , we show the magnetic field dependence of @xmath4@xmath5 in fig . 4 .
experimental data of @xmath4@xmath5 at 0.1 t ( open squares ) , 1 t ( open triangles ) and 7 t ( open circles ) are plotted in the upper panel .
experimental data at 0.1 t and 1 t have a similar tendency , but @xmath4@xmath5 at 7 t at low temperatures deviates significantly from the others .
this behavior is reproduced in the calculation shown in the lower panel . here
, the designated @xmath20/@xmath13 figures represent those of @xmath28/@xmath13 with @xmath29=2.0 and @xmath7=121 k , and magnitudes of @xmath4@xmath5 are calculated for the above hamiltonian using @xmath22=2.22 and @xmath23=2.09 .
the low temperature behavior of @xmath4@xmath5 at 7 t implies that the n@xmath30el order tends to be fixed and the ferromagnetic fluctuation of magnetic moments is suppressed by the external field
. in conclusion , magnetic properties of alternating ni and cu chain compound nicu(pba)(d@xmath0o)@xmath1@xmath22d@xmath0o were investigated by magnetic susceptibility measurements . from comparison with a numerical calculation for the ferrimagnetic heisenberg chain composed of @xmath3=1/2 and 1
, we have obtained the values of the exchange constant and the @xmath29-values of ni and cu .
field dependence of @xmath4@xmath5 at low temperatures has been reproduced by similar calculations in magnetic fields . |
acceleration of charge particles in the universe happens on scales from planetary magnetospheres to clusters of galaxies and at energies ranging from nonrelativistic values to @xmath010@xmath1 ev ultra high energy cosmic rays ( uhecrs ) .
the particles are observed directly as cosmic rays ( crs ) , solar energetic particles , or indirectly by their interactions with background matter and electromagnetic fields ( magnetic fields and photons ) , which give rise to heating and ionization of the plasma , and nonthermal radiation extending from long wavelength radio to @xmath0tev gamma - rays . in spite of more than a century of observations , the exact mechanism of acceleration
is still being debated and the detailed model parameters are poorly constrained .
clearly electric fields are involved in any acceleration mechanism .
large scale electric fields have been found to be important in some unusual astrophysical sources such as magnetospheres of neutron stars ( pulsars and perhaps magnetars ) and in so - called double - layers . however , here we are interested in commonly considered mechanisms based on the original fermi process @xcite , which involves scattering of particles by fluctuating electric and magnetic fields ( or plasma turbulence ) or converging flows as in shocks .
the usual approach of determining the acceleration model and its characteristics is to use the forward fitting ( ff ) method , whereby the model particle spectra based on an assumed mechanism and some parametric form of its characteristics are fitted to observations . for radiating sources ,
ff is carried out in two stages , first fitting the photon spectra to an assumed radiation mechanism from a parametrized particle spectrum , then fitting the latter to the acceleration model .
this approach , even though one can never be certain of the uniqueness of the results , has been fairly successful , and for some observations , e.g. , those with poorly determined spatially unresolved spectra , is the best one can do .
but in sources with richer observations one can do better .
in this paper we present a new approach which allows a non - parametric determination of acceleration parameters , mainly their energy dependence , irrespective of some of the details of the acceleration mechanism , directly from the observed radiation or otherwise deduced particle spectra .
this is done by the _ inversion _ of the kinetic differential equations describing the particle acceleration and transport . in our first paper on this subject @xcite
, we applied this technique to inversion of hard x - ray images of solar flares from the reuven ramaty high energy solar spectroscopic imager ( _ rhessi _ ) and determined the energy dependence of the escape time from the acceleration region and from it the energy dependence of the rate of scattering of the particles , presumably due to plasma turbulence , which is related to the pitch angle diffusion coefficient @xmath2 , where @xmath3 is the cosine of the pitch angle . in a more recent paper @xcite
, we have shown that from the same data we can also determine the energy diffusion coefficient @xmath4 , which is related to the momentum diffusion coefficient @xmath5 . in both papers we formulated this in the framework of stochastic acceleration ( sa ) by plasma waves or turbulence , which is same as the original fermi process , nowadays referred to as second - order fermi acceleration process . here
we extend this approach to simultaneous determination of the scattering and acceleration rates , which depend primarily on @xmath2 and @xmath5 , to situations where both sa by turbulence and acceleration by a shock play important roles .
as in previous papers we carry this out in the framework of the so called leaky box model . in the next section we present the kinetic equation describing both acceleration processes , and in [ inv ]
we describe the process of the inversion and the required data for it . in [ sec_snr ] we describe possible application of this method to the acceleration of electrons in supernova remnants ( snrs ) .
interpretation and discussions of the results are shown in [ sec_interp ] and a brief summary
is presented in [ sec_sum ] .
the discussion below is a brief summary of this subject given in a recent review by @xcite describing the conditions under which the so - called leaky - box model is a good approximation . as emphasized in this review , and recognized by the community at large , it is clear now that plasma waves or turbulence play an essential role in the acceleration of charged particles in a variety of magnetized astrophysical and space environments .
turbulence is expected to be produced by large scale flows in most astrophysical situations because of the prevailing large reynolds numbers . once generated on a scale @xmath6 comparable to the size of the source it undergoes dissipationless cascade from large to small spatial scales , or from small wave numbers
@xmath7 up to the dissipation scale given by @xmath8 , generally with a power law energy density distribution @xmath9 .
resonant interactions between particles and small amplitude electromagnetic fluctuations of turbulence cause diffusion of particles in the phase space . for magnetized plasmas this process can be described by the fokker - planck ( fp ) kinetic equation for gyro - phase averaged , four dimensional ( 4-d ) particle distribution function @xmath10 , where @xmath11 is the distance along the magnetic field lines .
this equation involves , in addition to @xmath2 and @xmath5 , a third coefficient @xmath12 , and @xmath3 and are @xmath13 , where @xmath14 is the particle gyro frequency and @xmath15 is the ratio of the turbulent to total magnetic field energy densities ( see e.g. @xcite .
] as well as a source term @xmath16 and energy losses or gains due to interactions of particles with background plasma ( with density @xmath17 , temperature @xmath18 , magnetic field @xmath19 and soft photon energy density @xmath20 ) .
these interactions cause stochastic acceleration , e.g. , @xcite , in which particles systematically gain energy with a rate that is proportional to the square of the wave - to - particle velocity ratio as in the second - order fermi process . also shown in @xcite
, the 4-d differential equation can be reduced to a 3-d equation , when the scattering time @xmath21 is shorter than the dynamic time @xmath22 and the crossing time @xmath23 .
is the particle velocity and in what follows the size @xmath6 refers to the length of the bundle of magnetic lines the particles are tied to . for chaotic fields
this could be much larger than the physical size of the turbulent acceleration region .
] then the momentum distribution is nearly isotropic and one can define the pitch angle averaged quantities , @xmath24 and @xmath25 , and use three pitch angle - averaged transport coefficients @xmath26 ( see @xcite ) to describe spatial and momentum diffusion rates . @xcite and
others , in most subsequent applications of this equation , were interested in acceleration by alfvn waves ( with velocity @xmath27 ) , in which case the diffusion coefficients are related as @xmath28 . limiting their analysis to low magnetization and high energy particles ,
i.e. for @xmath29 they used the inequities @xmath30 to obtain the simplified equation .
however , as was pointed out by @xcite , at low energies and for strong magnetic fields , other plasma waves become more important than the alfvn waves and these inequalities are no longer valid , e.g. , @xcite .
@xcite suggested another approximation for the fp equation for the opposite limit , @xmath31 , in which case the momentum diffusion is the dominant term .
these ideas were further developed by @xcite and summarized in @xcite .
it turns out that if again @xmath32 and @xmath22 , then this situation can be described by the same 3-d equation with slightly different coefficients .
( the proof of this assertion will be presented elsewhere . ) finally a second simplification can be used for both cases if the acceleration region is homogeneous , or if one deals with a spatially unresolved acceleration region where one is interested in spatially integrated equations . in this case
it is convenient to define the 2-d distribution function in terms of the particle energy @xmath33 , @xmath34 $ ] and @xmath35 $ ] , introduce spatially averaged terms @xmath36 and replace the spatial diffusion term by an escape term .
then we obtain the following well known equation , sometimes referred to as the _ leaky box model _
, @xmath37 - { \partial \over \partial e } \left[(a(e ) - \dot e_{\rm l } ) n\right ] \nonumber\\ & - { n \over t_{\rm esc } } + { \dot q}_{\rm inj}\,,\end{aligned}\ ] ] where @xmath38 , @xmath39 and @xmath40 are the direct acceleration and energy loss rates , and @xmath41 and @xmath42 represent the rates of injection and escape of particles in and out of the whole acceleration site . for purely sa , the direct acceleration rate ) are written as @xmath43 , where @xmath44 gives the direct energy gain rate . defining the total energy of the accelerated particles as @xmath45 , it is straightforward to show that integration of the above equation over energy gives @xmath46 @xcite , showing that @xmath47 provides a more accurate representation of the direct energy gain rate than @xmath39 . in what follows
we use the form given in eq .
( [ lbox ] ) which is more convenient for the inversion procedure . ]
@xmath48 where @xmath49 the term @xmath50 is nearly equal to 1 at all @xmath51 ( it has a maximum of @xmath521.3 for @xmath53 ) . , where @xmath54 is used in @xcite and our earlier papers .
] because the acceleration rate in stochastic acceleration is proportional to the square of the velocity ratio @xmath55 , it is often regarded to be too slow to account for production of high - energy particles , especially in comparison to acceleration in a shock ( or a converging flow in general ) .
for a shock with velocity @xmath56 , a particle of velocity @xmath57 upon crossing it gains momentum linearly with velocity ; @xmath58 , and therefore this often is referred to as a first - order fermi process .
there are several misconceptions associated with the above statement .
the first is that the diffusion coefficients , in general , increase with decreasing particle energy so that sa can be very efficient in the acceleration of low energy particles in the background plasma , which is where all acceleration processes must start @xcite .
the second is that shock acceleration is not related to the original process @xcite , and the third is that shock acceleration rate is also second order . in an unmagnetized shock , or in a shock with magnetic field parallel to the shock velocity , acceleration requires an scattering agent to recycle particles repeatedly across the shock .
turbulence is the most likely agent for this .
the acceleration rate then is @xmath59 , where the recycling time @xmath60 @xcite .
thus , _ the shock acceleration rate , @xmath61 , is also a second order mechanism_. as shown by @xcite , for oblique shocks ( @xmath62 ) the acceleration rate also varies as the square of shock velocity , but in this case , specifically for a perpendicular shock ( @xmath63 ) the rate could be much higher . in general then , as emphasized in @xcite , in both sa by turbulence and shock acceleration the rates are proportional to the square of the velocity ratios @xmath64 and @xmath65 , respectively , so that the distinction between them is greatly blurred . in either process
, resonant scattering by turbulence provides rapid isotropization of the particle pitch angle distribution , a necessary prerequisite for efficient acceleration , e.g. , @xcite .
more exactly , in the framework of the leaky box model , the shock acceleration rate can be written as @xmath66 where we have introduced the parameter @xmath67 with @xmath68 the compression ratio and @xmath69 which is a somewhat complicated function of the angle and the ratio of the diffusion coefficients parallel to perpendicular to the magnetic field @xmath70 ; @xcite . for a parallel shock @xmath71 and @xmath72 , where subscript 1 and 2 refer to upstream and downstream region of the shock , respectively .
the usual practice is to assume the bohm limit ; @xmath73 , where @xmath74 is the gyro radius .
in what follows we will use a more accurate relation for @xmath75 obtained from wave particle interactions , as those shown in figure ( 1 ) . for a perpendicular shock
the relation again is simple and from @xcite we obtain @xmath76 ( for @xmath77 ) , which amounts to setting @xmath78 . in applications to astrophysical sources we will be dealing with the scattering and stochastic acceleration times defined as as @xmath79 using these in eqs .
( [ asa ] ) and ( [ ash ] ) we can write shock to sa acceleration rate ratio as @xmath80 where @xmath81 . for _ parallel shocks _ ( @xmath71 )
this ratio becomes @xmath82 .
as pointed out above ( and in @xcite ) , at low energies and for strong magnetic fields , @xmath83 indicating the dominance of sa .
but for high energies and alfvnic turbulence @xmath84 and @xmath85 , where @xmath86 is the alfvn mach number , so shock acceleration dominates at high energies and weakly magnetized plasmas .
figure [ savsshock ] shows a comparison between the sa timescale @xmath87 as defined in eq .
( [ times ] ) and shock acceleration time @xmath88 , based on rates obtained for interactions of electrons with parallel propagating plasma waves @xcite , for two values of the spectral index @xmath89 of the turbulence energy density and two degrees of magnetization described by the plasma parameter @xmath90 , which is equal to the ratio of the electron plasma to gyro frequencies .
as evident at low energies and small values of @xmath91 ( strong magnetization ) sa is the dominant mechanism . for oblique shocks the shock rate will be in general higher by some factor which depends on the angle @xmath92 ; e.g. for a high mach number ( @xmath93 ) perpendicular ( @xmath63 ) shock this factor will be @xmath94 . )
calculated using eq .
( [ times ] ) and shock acceleration time ( denoted here as @xmath95 ) for interactions of electrons with parallel propagating plasma waves with power law spectral distribution for two indicies @xmath96 ( kolmogorov ) and 3 . here
@xmath97 , the ratio of the electron plasma to gyro frequencies , is a measure of the degree of magnetization , and @xmath98 is the characteristic rate for wave - particle interactions .
( see more details in @xcite . )
note that for highly magnetized plasma ( @xmath99 , sa is the dominant mechanism at some energies , and even for a plasma with lower magnetization , sa can not be ignored .
the shock velocity is taken to be @xmath100 km / s compared to the alfvn velocity @xmath101\sim 7000 $ ] and 700 km / s , resulting in @xmath102 and 200 , respectively .
the proportionality constants @xmath103 and @xmath104 are set to unity .
note that the @xmath96 solid curves are multiplied by 100 for clarity . ] in @xcite it was suggested that in presence of a shock the acceleration may be a hybrid process dominated by sa at low energies and shock at high energies . in what follows we will consider the combined processes , which depend on wave - particle interactions , shock compression ratio and background plasma parameters . for a solution of the differential eq .
( [ lbox ] ) we also need the energy dependence of the other terms . for the injected spectrum @xmath105 we will consider a maxwellian distribution at a given temperature @xmath106 , and for the energy loss we include ionization and coulomb losses that dominate at low energies ( and depend on background density @xmath17 and ionic composition ) , and synchrotron and inverse compton losses that dominate at high energies ( and depend on the background magnetic field and photon energy densities ) .
coulomb interactions can also cause energy and pitch angle diffusion which become important at low energies , e.g. , @xcite . as we will see below the last term , namely the escape time ,
is the term that can be obtained most readily from observations , which then allows determination of the other terms .
however , the relation of the escape time to the coefficients of the acceleration mechanism is complicated . as shown in @xcite
, it is related to an integral of spatial diffusion term @xmath107 over the acceleration site .
thus , it also depends on the size @xmath6 of this site or crossing time @xmath108 . for the isotropic case with @xmath109 , one expect the diffusion of the particles across the source to follow a random walk process , which means we can write @xmath110 . in the opposite limit , @xmath111 , the escape time @xmath112 .
combining these two cases in the past we ( @xcite ) have used the approximate expression @xmath113 however , other geometric effects such as those produced by the large scale magnetic fields e.g. chaotic field lines , or strongly converging or diverging field configurations ( see @xcite ) , or deviation from isotropy or a simple spherical homogeneous acceleration site , can make the relation between @xmath114 and other acceleration coefficients more complex .
solution of eq . ( [ lbox ] ) requires knowledge of energy and time dependences of the five coefficients involved in the terms on the right side . in situations
where there exist time resolved observations one needs to solve for the time dependence of the accelerated spectrum . however , if the dynamic time @xmath22 describing the evolution is longer than the characteristic timescales associated with these coefficients ( such as @xmath115 and @xmath87 or energy diffusion time @xmath116 ) , then one can use the steady state assumption and set @xmath117 and modulate the results with the time profile of the dynamic process .
this was the case in our application of the inversion method to solar flares . in the opposite situation of short dynamic time , and in the absence of temporally resolved observations
, one can integrate eq .
( [ lbox ] ) over the dynamic time in which case @xmath118 , because we expect @xmath119 for high energy particles . in this case one is dealing with the values of the coefficients averaged over duration @xmath120 of the process ; e.g average injected spectrum @xmath121 .
as discussed below this will be the case for the application to snrs and cosmic ray electrons ( cres ) .
thus , we need to consider only the energy dependence of the coefficients .
two of these , namely @xmath40 and @xmath122 , depend on background plasma parameters @xmath123 and @xmath20 , and are independent of the acceleration process .
we will assume that we have sufficient information on the background plasma so that we know the values and energy dependences of these two terms .
the other three are related to the characteristics of the acceleration mechanism that we want to determine .
one of these is the energy diffusion coefficient @xmath4 ( related to @xmath5 ) .
the escape time depends on the size @xmath6 of the source and on the spatial diffusion rate ( related @xmath2 ) . the final term , namely the direct acceleration rate has contribution from turbulence , which is related to @xmath4 , and from shocks , which is related to @xmath2 and the characteristics of the shocks described above .
assuming that we know the value of the latter and the size @xmath6 of the acceleration site , we are left with two primary unknowns @xmath2 and @xmath5 , or in terms of more directly unknowns @xmath4 and @xmath114 .
therefore , in order to determine the energy dependences of these two coefficients , we need the variation with energy of two independent observed quantities , as described next . as described in @xcite , one of the two functions that observations can provide is the ( spatially integrated ) energy spectrum of the accelerated particles @xmath124 , which can be deduced from the observed total photon spectrum @xmath125 produced in the acceleration region . , where @xmath126 stands for the radiative cross section .
] if the escape time is finite then the rate of particles escaping will be @xmath127 .
if this spectrum is measured directly , we then can obtain the escape time simply as @xmath128 if we assume that eq .
( [ tesct ] ) is an accurate description of how particles escape , we can then obtain also the scattering time @xmath129 this will then give a measure of the pitch angle diffusion coefficients @xmath2 and as shown by eqs.([ash ] ) and ( [ times ] ) , it will also give the acceleration rate by the shock @xmath130 , assuming we know @xmath103 .
given the above information we are left with only two related unknowns , namely the energy diffusion coefficient @xmath4 or the direct sa rate @xmath131 .
this final unknown can be obtained using our knowledge of the accelerated particle spectra @xmath124 and the escape time @xmath132 by the inversion of the leaky box eq .
( [ lbox ] ) as follows .
the key aspect here is to recognize that this ordinary differential equation is only first - order in the derivative of @xmath4 with respect to @xmath33 , instead of second - order that appears to be the case in its alternate form .
thus , by utilizing the relation between @xmath133 and @xmath4 in eq .
( [ asa ] ) we can rewrite the steady state leaky box equation as @xmath134 + \frac{d}{d e}\left[(\dot{e}_{\rm l}-a_{\rm sh})n\right ] \nonumber\\ = \frac{n}{t_{\rm esc}}-{\dot q}_{\rm inj},\end{aligned}\ ] ] integrating this from @xmath33 to @xmath135 gives @xmath136 \nonumber\\ \times\left(2{\bar \xi}- \frac{d \ln{n}}{d \ln{e } } \right)^{-1 } , \end{aligned}\ ] ] from which we obtain @xmath137 .
thus , all the terms on the right - hand side can in principle be obtained directly from observables .
note that for the time integrated equation under consideration here we must have the equality @xmath138 .
but for relevant energies of @xmath139 only a number of particles in the maxwellian tail contribute and @xmath140 .
if this were not true there would be very few particles accelerated and the case is uninteresting .
thus , in what follows we can neglect the injection term .
however , given the temperature of the background particles this term can be easily calculated and included in the results . finally we define an effective acceleration rate as @xmath141 where @xmath142 , and we have introduced the spectral index of the accelerated particles @xmath143 . at relativistic energies @xmath144 and , as we will see below , typically @xmath145 , so this rate is sum of the shock and ( about two times ) sa rates .
escaping particle are measured directly or by the detection of the radiation they produce outside the acceleration site , which we will call the transport region , where their spectrum is modified due to transport effects .
these effects can be treated by a similar kinetic equation without the diffusion and acceleration terms . if the particles are injected into a finite region and if one can neglect further acceleration and assume that pitch angle scattering quickly isotropizes the particle distribution , then the evolution of particles in the transport region can be described by the leaky box eq .
( [ lbox ] ) which now has only the energy loss and escape terms . instead of a thermal background source term , the spectrum of particles _ injected in the transport region _ is same as those escaping the acceleration site ; @xmath146 in application to the transport of the we are dealing with a long dynamic time so that we can use the steady state equation , which has the formal solution giving the effective spectrum of particles integrated over the transport region @xcite , @xmath147 \,,\end{aligned}\ ] ] where we have defined the energy loss time @xmath148 . of special importance , in general and in particular for the applications described below , is the case when the particles escaping the acceleration site lose all their energy in the transport region .
this is referred to as the _ thick target _ or _ totally cooled _
spectral model , where one sets @xmath149 and get a simpler integral solution @xmath150 first , differentiating this equation we derive the desired expression for the escape time as @xmath151\eta_{\rm eff}^{-1},\end{aligned}\ ] ] where @xmath152 , and we have defined the spectral index @xmath153 .
second , we note that this last integrand is identical to the third term inside the square brackets on the right - hand side of eq .
( [ dee ] ) , so that with the help of this equation we can derive a new simpler relation for the energy diffusion rate as @xmath154 where @xmath155 is the energy loss time scale averaged over the acceleration region .
finally , we define the effective acceleration time ( a combination of shock and sa times ) @xmath156^{-1}\nonumber\\ & = \left[{1\over \tau_{\rm l}^{\rm acc } } + { 1\over \eta_{\rm eff}\tesc } \right]^{-1}.\end{aligned}\ ] ] for pure shock acceleration , the acceleration time @xmath157 and for pure sa , the time @xmath158 . note
that while the escape time depends on only the ratio of effective to acceleration spectra , the acceleration times involve both this ratio and the energy loss time in the acceleration site . in the opposite limit
when particles lose very little of their energy in the transport region , i.e. when @xmath159 , which is called the _ thin target _ model , eq .
( [ neff ] ) simplifies even further to @xmath160 from which we get @xmath161.\end{aligned}\ ] ] for the diffusion coefficient in this case we have to replace the last term inside the first pairs of parenthesis on the right - hand side of eq .
( [ deethick ] ) by @xmath162 .
_ in what follows we will consider only the thick target case .
_ in summary , the above equations show that one can determine the pitch angle and momentum diffusion coefficients in the acceleration region directly from measurements of the particle spectra in the acceleration and transport regions .
as mentioned at the outset , in @xcite we have demonstrated the power of the procedure in application to solar flares .
here we explore the possibility of using the radiative signatures of snrs and observed spectra of cres in the interstellar medium ( ism ) to determine the characteristics of the acceleration mechanism in snrs .
it has been the common belief that snrs are the source of the observed crs ( at least up to the knee at @xmath5210@xmath163 ev ) and recent high energy gamma - ray observations of snrs have enforced this belief considerably .
if this is true then we can get information on the two functions required for our inversion process .
the observed radiative spectrum of snrs from radio to gamma - rays gives the spectrum of the the accelerated particles , @xmath164 , and the observed spectrum of the crs provides information on the spectrum of accelerated particles escaping the snrs , @xmath165 .
although in principle this information is available for both electrons and protons , there are only some preliminary solid observations on the radiative signature of protons in snrs .
therefore , in what follows we will focus on the acceleration of electrons .
however , it should be emphasized that the situation here is not as straightforward as in solar flares where these two functions are determined simultaneously for individual flares . here
we need knowledge of the transport to the earth of the electrons escaping the snrs , and a more important complexity is that , many and a diverse set of snrs , resulting from explosions of different progenitor stars in different environments , contribute to the crs in the ism .
we will address these complexities in the following sections .
many snrs are observed optically and at radio .
the radio radiation produced via the synchrotron mechanism provides the original indication of presence of electrons with energy @xmath166 gev in a magnetic field of @xmath167 1020 @xmath3 g .
appearing in the above equations are equal to one .
] several snrs are detected at x - rays which also are attributed to synchrotron radiation by more energetic electrons , perhaps in a stronger magnetic field .
fermi and hess have detected gev and tev gamma - rays in several snrs . in some cases , for example snr rxj1713.73946 ,
a pure leptonic scenario , whereby the gamma - rays are produced by the synchrotron emitting electrons via the inverse compton ( ic ) scattering of cosmic microwave background ( cmb ) or other soft photons , seems to work @xcite .
while in others , e.g. , snr tycho @xcite , the hadronic scenario , whereby the accelerated protons are responsible for the gamma - rays , fits the data better . in some others , e.g. , snr vela jr
@xcite , both models give acceptable fits . in any case the radio and x - ray emission gives information about the spectrum of the _ accelerated electrons _ which is what we will be concerned with here .
we call this spectrum @xmath168 . in the case of solar flares , where nonthermal electron bremsstrahlung produces the hard x - ray radiation , one can use _ regularized inversion _ procedures to determine the spectrum of the radiating electrons non - parametrically and directly from photon count spectra @xcite .
unfortunately this technique can not be used for snrs
. there has not been much effort in inverting synchrotron and ic spectra to obtain electron spectra non - parametrically .
some time ago , @xcite addressed the inversion of synchrotron spectra and recently @xcite used a matrix inversion method of @xcite to invert the ic spectra and applied it to snr rxj1713.73946 .
but , in general , most of the information on @xmath169 is obtained by ff of the observed photon spectra to parametric electron spectra , with the result that the accelerated electron spectra ( integrated over the acceleration region of snr ) can be described by a power low with a high energy exponential cut off at energy @xmath170 . here and in what follows
we express all particle energies in units of a fiducial energy @xmath171 , which we set equal to 100 gev for numerical purposes .
thus , the spectrum of snr can be written as @xmath172 where @xmath173,\end{aligned}\ ] ] with @xmath174 and @xmath175 . in most cases @xmath176 and @xmath177 tev provide good fits down to energies of @xmath178 gev , e.g. , @xcite . note that as defined above @xmath179 , and @xmath180 is a dimensionless quantity .
the analyses that lead to the above spectra also indicate presence of sufficiently strong magnetic field ( @xmath181 g ) that can come about from amplification of the weaker ism field ( @xmath52 1 @xmath3 g ) by the supernova driven forward shock . in this case
synchrotron losses dominate over ic losses and the radiative loss time in the acceleration site required for our procedure can be written as @xmath182 where @xmath183 and @xmath184 is the thomson cross section . as mentioned above , however
, supernova explosions and snrs may have a broad range of characteristics and parameters of acceleration .
in which case the average snr spectrum contributing to the cres would depend on the distribution of the spectral parameters , say @xmath185 , where @xmath186 stands for @xmath187 and @xmath188 . in this case
the average spectral shape @xmath189 will depend on the shape of the distribution @xmath190 . as we will see below only the value of @xmath187
will be important .
this is related to the power - law indicies of the observed radio spectra which shows a small dispersion ( see @xcite ) .
in addition , as is well known from general theoretical considerations ( @xcite ) , the power - law index of accelerated particle spectra are insensitive to shock characteristics ( e.g. compression ratio ) for high mach number shocks , such as those expected from stellar explosion in the cold ism .
thus , the spectral shape given in eq .
( [ snrspec ] ) seems to be a reasonable approximation .
it should be noted though that explosions and environments of the upper end main sequence stars are considerably different than those of lower mass stars ( see e.g. @xcite ) and could possibly yield different accelerated spectra .
unfortunately there are no observations of remnants of such stars .
this is mainly because they are rarer , which would also mean they contribute less to crs .
in addition , explosions into a hot stellar wind environment may lead to a lower mach number shock and a weaker accelerator . on the other hand , being more powerful explosions
could have an opposite effect , which would enhance their contribution . in the absence of observational evidence about the distribution of characteristics of stellar explosions and snr spectra , in what follows
we will use the spectral form given in eq .
( [ snrspec ] ) for the accelerated spectrum @xmath164 , with the cautionary remark that the above unknown may introduce a significant uncertainty in our final results .
as mentioned above , it is widely believed that snrs are the source of all crs , and we will assume this to be the case for cres . therefore , the spectra of crs are related to those of the particles emitting the snr radiation via the escape time from the snrs .
the escaping particles interact with the galactic background matter and electromagnetic fields producing the galactic diffuse emission from radio to high energy gamma - rays .
these interactions and other processes modify the escaping particle spectra during their transport to where they radiate and to near the earth where they are observed directly .
therefore , crs are expected to have different spectra than snrs with the difference being partially due to the energy dependence of the escape time and partially due to energy losses during their transport in the galaxy .
observations witness these differences . for example , radio spectra of snrs are flatter than those of diffuse radio emission in the ism , and the measured cre spectrum @xmath191 is different than that given in eq .
( [ nsnr ] ) .
the spectral flux of cres has been measured by many instruments with varied results .
but most recent measurements by fermi , hess and pamela have produced a very precise spectrum shown in figure [ crespec ] . as discussed extensively in the literature
these spectra show a well defined deviation from pure power law above 10 gev and hess observations provide a clear evidence of a high energy roll over .
( dots ) as observed by pamela ( 3 lowest energies ) , hess ( 5 highest energies ) and fermi ( from @xcite ) , respectively .
the solid - green curve gives an approximate fit to the observations with its logarithmic derivative shown by the dotted - green curve .
we also present a sample variation of @xmath192 ( from @xcite ) , the energy loss time ( multiplied by @xmath33 ; solid - red ) and its logarithmic derivative ( dotted - red ) showing the transition due to the klein - nishina effect ( see eq .
[ tlossgal ] ) .
the dotted - blue curve gives @xmath193 defined in eq .
( [ tescthick ] ) and used in eq .
( [ tescb ] ) . ]
there has been multiple analyses of this data .
many of these use galprop @xcite or other similar numerical schemes ( e.g. , dragon ) to account for transport effects in the galaxy assuming values for background particle and soft photon densities , large scale magnetic field and a spectrum electromagnetic field fluctuations .. this is usually carried out by fitting the observed cre data to some parametric form of the spectrum of _ the total electrons injected throughout the galaxy _ , which is our function @xmath194 ( eq . [ qtr ] ) .
the results usually consist of a primary power law component with index @xmath11 and a high energy exponential cutoff at @xmath195 so that we have @xmath196 where @xmath197 here
@xmath198 is in units of electrons per unit time and @xmath199 .
different analyses give different explanations for the prominent bump seen around 100 gev .
for example , @xcite attribute this bump to a flux of electrons ( plus positrons ) coming from a nearby pulsar yielding @xmath200 and @xmath201 tev .
@xcite explain the bump with yet another spectral break , a slight flattening above 50 gev and similar values for the other parameters .
@xcite , using the spectrum of diffuse galactic radio emission , obtain @xmath202 but do not have the spectral resolution to see the bump around 100 gev nor do they see the tev cutoff .
we can use the above expression in eqs .
( [ qtr ] ) and ( [ thicktarget ] ) to obtain the acceleration characteristics . as described below this
will be one of the two methods we will use , with @xmath203 and @xmath201 tev .
an alternative and simpler explanation of the bump in the cre spectrum was given in @xcite ( see also @xcite ) , who show that the energy dependence of radiative losses due to combined synchrotron and ic scattering ( by star light , infrared and cmb photons ) can account for this deviation .
this is because at low energies star light is the dominant agent of loss , but at higher energies ic scattering by star light enters the klein - nishina ( kn for short ) regime which suppresses these losses and there is a transition to ic losses to infrared and cmb photons ( which are still in the thomson regime up to energies of a few tev ) and/or synchrotron losses ( depending on the value of the magnetic field ) . for typical values of the relevant quantities in the solar neighborhood
this transition occurs near the bump seen in the cre spectrum .
this means that in this case the radiative loss time that enters eq .
( [ tescthick ] ) does not have the simple thomson regime form @xmath204 , but involves an additional function @xmath192 that slowly varies with energy in the range from 1 gev to 1 tev shown in figure [ crespec ] ( taken from fig . 1 of @xcite ) .
the energy loss time in the transport region can then be written as @xmath205 where @xmath206 here @xmath207 7 @xmath3 g in the solar neighborhood , where @xmath208 is the energy density of all soft photons plus the magnetic field . )
required in this scenario is a power law with spectral index @xmath209 with cutoff at @xmath210 tev . ] in this case the observed cre flux spectrum @xmath211 gives directly the effective spectrum as @xmath212 where @xmath213 is the volume of the galaxy filled with cres , @xmath214 , and @xmath215 is the ( dimensionless ) effective total electron number at @xmath171 .
as described below we will use the above two equation , with the exact observed spectrum for @xmath216 , as a second method .
it should be noted that here , unlike in the previous method , which assumes presence of nearby pulsar , we assume the solar neighborhood is a typical location in the galaxy , e.g. does not contain an unusual large scale fluctuation in density , @xmath19 field or turbulence ( see also the discussion below ) .
we have described two possible methods for inversion of observations to obtain acceleration mechanism characteristics in snrs . in what follows
we discuss how these methods work in practice .
the snr spectrum @xmath217 , and either the deduced injected cre spectrum @xmath218 or the observed cre spectrum @xmath216 provide the energy dependence of the two functions @xmath124 and @xmath165 that we need for our analysis but not their normalization which is required for determining their ratio .
we have already discussed the uncertainty in the spectrum @xmath219 above . here
we describe the uncertainty in the normalizations .
this normalization depends not only on @xmath220 and @xmath221 , but also on the rate of snr formation per unit volume @xmath222 .
given this rate we can determine the averaged density of accelerated electrons in the galaxy and the rate of injection of electrons per unit volume in the ism as @xmath223 and @xmath224 where @xmath225 is the birth time of snrs , and @xmath226 , the volume of the galaxy enclosing all snrs is expected to be or less than @xmath213 .
however this difference does not affect our results . in general , the integrands vary in time and space , but
because the active age of a snr , @xmath227 , is much shorter than other ages , in particular the age of the galaxy , only the snr formation rate averaged over the past @xmath227 years enters these equations .. ] moreover , because electrons in several gev to tev range lose their energy quickly , only the quantities within the finite volume of radius @xmath228 kpc around the solar neighborhood are relevant ( here @xmath229 pc at 100 gev is the scattering mean free path of cres in the ism ) .
, where @xmath230 is the size of the transport region , in this case the thickness of the galactic disk as defined by snrs or crs .
] then the injection rate is determined by the value of the integrand of the above equations averaged over a small volume and short time @xmath231 or nearly for @xmath232 , the current age of the galaxy .
thus , we can write @xmath233 where @xmath234,\end{aligned}\ ] ] and @xmath235 in what follows we suppress the time @xmath236 .
these results assume that @xmath237 is the electron spectrum integrated or averaged over the active life of the snrs . and as stressed above , because the number of accelerated electrons may vary from snr to snr , the normalization constants also stand for averaged quantities .
for example , given the distribution function @xmath238 the integrand in eq .
( [ nacctot ] ) is @xmath239 .
* method a : * in this method we use the deduced injected spectrum as given by eq .
( [ qinjobs ] ) . equating this observed spectrum to that in eq .
( [ qinj2 ] ) we obtain the escape time ( from snrs ) as @xmath240,\end{aligned}\ ] ] with @xmath241 and the effective spectrum as @xmath242 here we have defined @xmath243/x= g(x)/\eta_g$ ] , where @xmath244 . as shown in eqs .
( [ deethick ] ) and ( [ teff ] ) the diffusion coefficient and effective acceleration time depend only on the following combination of terms @xmath245 and , in particular , the effective acceleration time is obtained as @xmath246^{-1}.\ ] ] we can lump all the unknown and poorly known factors that enter in these equations into a single parameter @xmath247}{\tau_{\rm
l,0}^{\rm acc}{\dot q}_{\rm inj , 0}^{\rm tr}},\end{aligned}\ ] ] which then gives @xmath248 \eta_{\rm eff}^{-1},\ ] ] and @xmath249^{-1}.\ ] ] thus , both timescales @xmath250 and @xmath251 can be expressed in units of @xmath252 ( which depends only on the average magnetic field in the acceleration region ) , and their values and the energy dependence of @xmath251 vary with the value of the parameter @xmath253 .
note that in this method the ( more uncertain ) energy loss rate in the ism does not enter into these results .
its effect is included in deducing the injected spectrum from the observed cre spectrum . in other words ,
given the magnetic field in the snr acceleration region around the shock the spectra depend only on @xmath253 ( or @xmath254 ) , which involves the properties of the snrs and the normalization of the deduced injected electrons .
* method b : * alternatively , as mentioned above , we can get the effective spectrum directly from the observed cre spectrum as @xmath255 , in which case instead of eq .
( [ effoveracc1 ] ) we have @xmath256 which when substituted into eqs .
( [ tescthick ] ) and ( [ teff ] ) gives the unknown escape and effective acceleration times as @xmath257\eta_{\rm eff},\end{aligned}\ ] ] with @xmath258 and @xmath259^{-1},\ ] ] where we have defined @xmath260\over n_0^{\rm cr}}\left({b_{\rm snr}\over b_{\rm eff}}\right)^2.\ ] ] these are very similar to the expressions from method a but are more directly related to the observations and now the energy loss time in the galaxy comes into play .
thus , in either method we can combine several poorly understood parameters into essentially one unknown ; namely the constant coefficient @xmath253 or @xmath261 .
the latter fixes the normalization of the ratio of the effective to accelerated spectra and determines the relative importance of the two terms that appear in the expressions for @xmath262 in eq .
( [ teff ] ) . as mentioned above
there is uncertainty associated with values of the spectral indicies and energy cutoffs . in what follows we will set @xmath263 tev and @xmath264 tev but will comment on the effects of the uncertainties after presenting the results
thus , the remaining unknown is the dimensionless factors @xmath253 and @xmath265 . before proceeding further we need to estimate their values . considering the relations between the injection rate deduced from the observations and the observed cre spectrum , it is clear that @xmath266 and that @xmath253 and @xmath261 should have similar values .
below we estimate their values based on method b which is more closely related to the observations .
there are reliable estimates for the values of the magnetic fields entering in the expression for @xmath265 in eq .
( [ rb ] ) ; as stated above @xmath267 g and using the starlight and infrared photon densities and magnetic field values in the galaxy one gets @xmath268 g e.g. , @xcite . also using the observed cre flux ( see fig . [ crespec ] ) of @xmath269 gev@xmath270/(s sr m@xmath270 ) , we get @xmath271 , assuming the poorly known volume of the galaxy that is filled with cres to be @xmath272 @xmath273 .
even less well known are the values of the terms in the square brackets in the numerator of eq .
( [ rb ] ) .
the rate of occurrence of supernovae is believed to be about several per century but what fraction of these produce active ( i.e. cr producing ) remnants is not well known .
observations seem to indicate a smaller rate @xmath274 . the active age of snrs
is estimated to be around @xmath275 to @xmath276 yr , which gives a rough estimate of @xmath277 .
the final factor namely @xmath278 can be estimated from the observed synchrotron and x - ray radiation intensities of individual snrs .
for example , snr rxj1713.73946 has an observed peak flux ( at x - rays ) of @xmath279 ev/(s @xmath280 ) and a low energy spectrum @xmath281 . assuming a distance of 6 kpc , we get a good estimate for the total energy of the synchrotron radiation @xmath282 ergs / s .
this is related to the accelerated particle spectra as @xmath283 where @xmath284 is the synchrotron energy loss rate . for the assumed spectral parameters this gives @xmath285 or @xmath286 .
putting all these together we get @xmath287 .
however this is most likely an overestimation because we have used the observations from a bright snr . the number of accelerated electrons for an average snr ( including possibly a substantial population of weak and undetected ones ) would lower this value considerably .
for example , using the general belief that supernovae inject @xmath288 ergs into the ism and that say 10 percent of this going to crs , with an electron share of one to two percent , we get a number of accelerated electrons smaller by a factor of 10 , or @xmath289 or @xmath290 . considering the large uncertainties about all the above numbers , in what follows we present results for three values of @xmath291 and 0.01 spanning a wide enough range to account for all uncertainties .
figure [ fig - timesa ] shows variation with energy of all time scales obtained by method a ( left ) and method b ( right ) normalized to the value of synchrotron energy loss time at 100 gev in the snr ( @xmath292 yr ) . as evident
the two methods give very similar results but method b results end where the observations of cre spectra become unreliable .
let us first consider * the escape time * which is essentially the ratio of the accelerated spectrum to observed cre spectrum multiplied by the loss time . at energies below @xmath293 tev
it is nearly a power law with index @xmath294 in method a and is @xmath295 in method b , with difference primarily due to the kn effect .
@xmath296 starts to increase steeply at @xmath166 12 tev .
this rise makes the escape of high energy electrons from the snrs more difficult , and is the causes of the steep ( exponential ) decline in the observed cre spectrum . *
the acceleration times * for pure shock or pure sa have similar energy dependences ( with a factor of @xmath297 difference between them ; with sa requiring a longer time or a lower rate ) . at low energies these times are dominated by the second term in eqs .
( [ teffobs ] ) and ( [ teffb ] ) , which makes them proportional to the escape time . had this trend continued to higher energies the acceleration time would have exceeded the energy loss time which would have caused a spectral cutoff when these times would have been equal ( e.g. , at 0.1 , 0.5 , and 3 tev for @xmath265 = 1.0 , 0.1 , and 0.01 , respectively ; and at smaller values by a factor of about 2 for method a ) . since the deduced snr electron spectra
are observed to cut off at higher energies ( 6 tev for rxj 1713.7 ) , the acceleration time must decrease to remain below the energy loss time as seen in both figures . as evident from the discussion in [ model ] and [ inv ] we can also obtain * the scattering time * in the acceleration site . for this purpose
we need some information about the background plasma in the acceleration site .
the first is the size @xmath6 of the region .
we will use the fiducial value of 10 pc ( to include the effects of the chaotic structure of the large scale magnetic field ; see footnote 1 ) which gives us a crossing time @xmath298 yr .
we also need the shock , alfvn and sound velocities .
we shall assume a shock velocity of @xmath275 km / s or @xmath299 , and alfvn velocity of 100 km / s ( for @xmath300 @xmath3 g , @xmath301 @xmath302 ) or @xmath303 , so that the alfvn mach numbers is very large as one needs for efficient shock acceleration .
for such such high mach numbers the compression ratio @xmath304 and @xmath305 .
there are , however , two different ways of obtaining the scattering time .
the first way , which is common for both shock or sa , comes from the relation between the escape and scattering times , which if we assume a random walk process of escape is described by eq .
( [ tesct ] ) and involves the crossing time .
given that @xmath306 30 yr we obtain the first estimate for the scattering time as : @xmath307}= \frac{\tau^2_{\rm cross}}{\tesc } \sim 0.025\ { \rm yr}~\left(\frac{l}{10\ { \rm pc}}\right)^2\left(\frac{4\times 10 ^ 4\ { \rm yr}}{\tesc}\right),\end{aligned}\ ] ] which as expected is much shorter than the crossing time .
here and in what follows the numerical values are calculated for @xmath308 gev and @xmath309 .
the second method of determining @xmath310 comes from the relation between the acceleration and scattering times . for pure * shock acceleration * @xmath311 and as seen in eqs .
( [ ash ] ) and ( [ times ] ) the energy dependence of the scattering and acceleration times should be similar but their relative value depends on the shock velocity , the factor @xmath103 , and for perpendicular shocks on @xmath312 . neglecting the latter for now we get @xmath313}=\zeta\left(\frac{u_{\rm sh}}{c}\right)^2\tau_{\rm ac , sh }
\sim 10\ { \rm yr}\left(\frac{\beta^2_{\rm sh}}{10^{-3}}\right)\left(\frac{\tau_{\rm ac , sh}}{4\times 10 ^ 4\ { \rm yr}}\right),\end{aligned}\ ] ] which is about the crossing time and much larger than the the first estimate of scattering time .
it also has a different energy dependence . as can be seen in figure [ scattimes ] ( left ) , for the spectral indexes ( @xmath314 ) assumed above the first estimate ( black curves ) decreases monotonically with energy while the second ( red curves ) first increases with energy and then declines at higher energies . the difference in energy dependence at low energies comes from the fact that here @xmath315 [ see eqs .
( [ timesa ] ) and ( [ timesa1 ] ) ] making @xmath316}\propto 1/\tsc^{[1 , { \rm sh } ] } ( \propto e^{\alpha_1-s}$ ] , for method a ) .
this difference will be less severe for a steeper snr electron spectra ( i.e. for @xmath187 closer to @xmath11 ) , which is the case in some snrs .
for example , in snr s1993j with radio spectral index of @xmath317 one gets @xmath318 ( @xcite ) .
as shown in figure [ scattimes ] ( right ) using @xmath319 we get similar energy dependence for both estimates ( and both methods ) .
however , as shown above the absolute values of the scattering time deduced from the two curves are different by a large factor : @xmath320}}{\tsc ^{[1]}}=4000\left(\frac{\beta^2_{\rm sh}}{10^{-3}}\right)\left(\frac{10\ { \rm pc}}{l}\right)^2\left({\tau_{\rm sh}\tesc\over 1.6\times 10 ^ 9\ { \rm yr}^2}\right).\ ] ] agreement can be obtained for a lower shock velocity ( @xmath321 km / s ) and/or a larger crossing time ( @xmath322 pc ) .
there is more uncertainty in the first of the above two ways of computing the scattering time ; for example , as mentioned above and in footnote 2 , in a chaotic magnetic field of scale @xmath323 the effective crossing time will be larger by @xmath324 , which will reduce the above discrepancy by the square of this factor .
thus , for concordance we require @xmath325 pc ; ( e.g. , @xmath326 pc for @xmath327 pc ) . as mentioned in connection with eq .
( [ rateratio ] ) , for a perpendicular shock this ratio decreases by the factor @xmath328 expected to be much larger than one so that the required conditions may not be as extreme .
more generally , the validity of the use of the random walk relation between escape and scattering times may also be questionable , so that these results may be telling us that the relation of the escape time to the scattering and crossing times is more complicated than given by the random walk hypothesis .
for example , in a near perpendicular shock , where particles spiral up and down the surface of the shock and escape when they are scattered perpendicular to the shock front , the escape time may be proportional to the diffusion coefficient perpendicular to the magnetic field giving @xmath329}\propto \tesc/\eta$ ] , which could bring the shape and value of the first estimate closer to that of the second estimate . to our knowledge
there has not been much discussion of this aspect of the problem in the literature so that these possibilities require further explorations , which are beyond the scope of this paper .
* stochastic acceleration * by turbulence may be important or even dominant if there is weak or no turbulence in the upstream region , conjectured to be generated by the accelerated particles . in this case
most of the acceleration may happen in the downstream turbulent region with particle escaping into the ism once they cross the shock into the upstream region .
however , this mechanism also faces similar difficulties . here
the energy dependence of the acceleration time @xmath330 ( or energy diffusion time ) is related to the scattering time via the relation between @xmath331 and @xmath2 . in most wave - particle interaction scenarios
these two coefficients have fairly similar energy dependences especially at relativistic energies .
electrons with energies above few gev interact mainly with alfvn or fast mode waves in which case @xmath332 so that @xmath333 and @xmath334 ( see e.g. @xcite )
. thus , we have a second estimate for scattering time for sa as well : @xmath335}=\tau_{\rm ac , sa}\beta_a^2=10^{-2}\ { \rm yr}\left(\frac{\beta^2_a}{10^{-7}}\right)\left(\frac{\tau_{\rm acc , eff}}{10 ^ 5\ { \rm yr}}\right).\end{aligned}\ ] ] as shown by the green lines in figure [ scattimes ] , in this case also the energy dependences of @xmath336}$ ] and @xmath337}$ ] disagree at low energies for @xmath338 ( left ) but they roughly agree at high energies , and , again , the agreement is improved for @xmath319 ( right ) , where both times have almost a flat energy dependence requiring a turbulence spectral index of @xmath339 , which is somewhat greater than the kolmogorov index . moreover , now the relative absolute values are in better agreement for the assumed values of alfvn velocity of 100 km / s and effective size of @xmath340 pc .
we consider acceleration of particles in the framework of the leaky box version of the fokker - planck kinetic equation , which provides an adequate description of the pitch angle averaged and spatially integrated ( over the acceleration region ) energy spectrum of the accelerated particles .
this equation describes sa by turbulence and/or acceleration by a shock , where the leaky box encloses the upstream and downstream turbulent regions of the shock .
turbulence plays a central role in both mechanisms , with the momentum diffusion coefficient @xmath5 determining the rate of energy diffusion and acceleration in the sa model , and with the pitch angle diffusion coefficient @xmath2 determining the spatial diffusion coefficient @xmath341 , hence the rate of acceleration by the shock .
in addition , the energy loss rate , shock compression ratio ( or mach number ) and relative values of the spatial diffusion coefficients parallel and perpendicular to the magnetic field , and in the upstream and downstream regions , also come into play . in the leaky box scenario the coefficients @xmath5 and @xmath2
are represented by the energy diffusion coefficient @xmath4 and the escape time @xmath114 of the particles from the acceleration site .
thus , if we can measure the latter two coefficients we can determine the fundamental wave - particle interaction rates and shed light on the nature of turbulence @xmath342 as demonstrated in @xcite , we can obtain the escape time from the measured spectrum of the accelerated particle @xmath124 and that of the escaping particles @xmath343 .
we further demonstrate ( see @xcite ) that with the inversion of the differential kinetic equation into its integral form , we can obtain the energy diffusion coefficient _ non - parametrically and directly _ from observations of the two spectra and the energy loss rate of the particles in the acceleration region .
@xmath342 we also show that the relations between the two unknowns and observables simplifies considerably if the escaping particles lose all their energy in the transport region outside the acceleration site .
@xmath342 we demonstrate how this procedure can give us the two unknown characteristics of the acceleration mechanism in snrs using the spectrum of the accelerated electrons deduced from radio , x - ray and gamma - ray observations of the snrs and the observed galactic cre spectrum .
@xmath342 expressing all the coefficients or rates in terms of their associated timescales ( e.g. , acceleration and scattering times ) , we show that the unknown time scales can be expressed in units of the relatively well known synchrotron energy loss time in the snr and a single parameter which is a combination of various observable scaling factors , such as rate of formation and length of active period of snrs and other secondary factors .
@xmath342 we employ two different methods of treatment of the observations and show the deduced energy dependence of escape and acceleration times for some reasonable value of the parameters , which in principle can be known given sufficient detailed observations . in method
a we use the spectrum of injected electrons into the ism deduced from the observed cre spectrum ( e.g. , using galprop or other similar models for transport of electrons in the ism ) . in method
b we use the observed cre spectrum directly using a simplified transport dominated and ic losses by starlight which is affected by the kn effects as described in @xcite .
@xmath342 for interpretation of the results , we show that we can obtain scattering time ( @xmath344 ) of particles in the acceleration region using two different relations between it and the above timescales .
the first is from its relation to the escape time , which is mediated by the crossing time ( @xmath345 ) as @xmath346 assuming a random walk situation when @xmath347 .
the second is from its relation to the acceleration times . for shock acceleration scattering and acceleration times
are proportional to each other with proportionality constant being @xmath348 ( plus factors @xmath103 and @xmath349 ) . for pure sa of greater than few gev electrons by alfvn or fast mode waves
there is a similar relation but with proportionality constant of @xmath350 .
@xmath342we find that , for the values of the parameters used in our calculation ( specifically the spectral index @xmath351 ) , the two estimates of the scattering time give very different energy dependences for the scattering time .
this discrepancy largely disappears for @xmath319 .
given the caveats stressed in our discussion this is not an unlikely resolution of the problem .
@xmath342 assuming presence of a sufficient intensity of turbulence both in the upstream and downstream regions of the shock , we expect the acceleration in snrs to be dominated by the shock , because of the prevailing high mach numbers . however , for this scenario we find that the absolute values obtained by the two relations are different by a factor of about 1000 for our fiducial values of @xmath275 km / s for shock velocity and @xmath352 pc for size
. this discrepancy will be smaller for a perpendicular shock .
this lead us to our _ first conclusion that , in addition to a steeper spectrum for accelerated electrons , either these values are off by an order of magnitude , or that the escape time is not related to the crossing time in the simple way one obtains from the random walk scenario .
the latter is an important result and needs further exploration . _
@xmath342 on the other hand , in absence of a sufficient intensity of turbulence in the upstream region , whose presence is only conjectured and not established definitely yet , one can have a pure sa of particles in the turbulent downstream region .
it turns out that in this scenario the absolute values of the two scattering times roughly agree .
_ this , lead us to the second conclusion that in the sa scenario having a steeper accelerated electron spectrum is sufficient and it requires a spectrum of turbulence that is slightly steeper than kolmogorov . _ these are clearly preliminary results , but they demonstrate the power of the inversion method developed here .
a more detailed analysis of the existing data on emission from snrs and transport of the cres can provide better values and forms for the observables required for the inversion , and a more detailed analysis of the inversion , e.g. , including time dependence , can constrain the models further .
these will be addressed in future publications .
but we can conclude that the above results indicate that either the spectrum of injected electrons in the ism deduced from cre and galactic diffuse emissions ( eq . [ qinjobs ] ) is incorrect and/or the simple relation between escape and scattering times used assuming the random walk scenario is incorrect .
the latter is more likely to be the case and is similar to the conclusion we reached applying these techniques to solar flares .
there mirroring of electrons in a converging magnetic field configuration was invoked to resolve a similar discrepancy .
perhaps a complex large scale field geometry can help in snrs as well . on the other hand ,
more consistent results are obtained for a pure stochastic acceleration scenario . | we have developed an inversion method for determination of the characteristics of the acceleration mechanism directly and non - parametrically from observations , in contrast to the usual forward fitting of parametric model variables to observations .
in two recent papers @xcite , we demonstrate the efficacy of this inversion method by its application to acceleration of electrons in solar flares based on stochastic acceleration by turbulence . here
we explore its application for determining the characteristics of shock acceleration in supernova remnants ( snrs ) based on the electron spectra deduced from the observed nonthermal radiation from snrs and the spectrum of the cosmic ray electrons observed near the earth .
these spectra are related by the process of escape of the electrons from snrs and energy loss during their transport in the galaxy .
thus , these observations allow us to determine spectral characteristics of the momentum and pitch angle diffusion coefficients , which play crucial roles in both direct acceleration by turbulence and in high mach number shocks . assuming that the average electron spectrum deduced from a few well known snrs is representative of those in the solar neighborhood we find interesting discrepancies between our deduced forms for these coefficients and those expected from well known wave - particle interactions .
this may indicate that the standard assumptions made in treatment of shock acceleration need revision .
in particular , the escape of particles from snrs may be more complex than generally assumed . |
photon dominated regions are predominantly molecular and atomic regions where the physical and chemical processes are dominated by uv radiation ( cf . hollenbach & tielens , 1997 , 1999 ) .
the molecular clouds in the vicinity of the newly formed stars is heated by the fuv photons in the energy range from about 6 to 13.6ev .
these clouds cool through the atomic and molecular spectral lines , such as [ cii]158@xmath0 , [ oi]63,146@xmath0 , [ ci]609,370@xmath0 and the milli - metric and sub - mm co rotational lines .
plane parallel and spherical models of pdrs have been constructed to understand the physical and chemical characteristics of these regions ( e.g. , kaufman et al .
, 1999 ; kster et al . , 1994 ; le bourlot et al . ,
1993 ; sternberg & dalgarno 1995 ; strzer et al . , 1996 ; tielens & hollenbach 1985 ) .
however there are other important factors which affect the uv absorption and scattering as well as the basic heating and cooling processes in pdrs .
several galaxies such as dwarf galaxies , irregular galaxies , the large and small magellenic clouds have low metallicity ( cf .
wilson 1995 ) .
a radial gradient in metallicity of molecular clouds is found in the milky way and several other nearby galaxies ( e.g. , arimoto et al . , 1996 ) .
these low metallicity systems have much higher [ cii]/co and [ ci]/co line ratios as compared to the galactic value ( e.g. bolatto et al .
, 2000 ; madden 2000 ) .
this suggests that the effects of metallicity should be considered while interpreting the molecular and atomic spectral line observations of these sources . since the important surface coolant of the pdrs , the [ cii]158@xmath1 m emission ,
is strongly affected by the metallicity factor , we study the effects of metallicity in pdrs .
in addition , low metallicity pdr models would also help us to understand the star forming regions in dwarf galaxies which resemble the primordial galaxies .
additionally , observations of edge - on pdr have suggested that the molecular clouds are clumpy , and the uv radiation can penetrate deep inside the clouds ( cf .
stutzki et al . ,
1988 ; boiss 1990 ) .
this suggests that a simple single component model may not be sufficient to explain the observed features .
we also include a mass spectrum of clumps to understand the cooling lines of pdrs from low metallicity galaxies .
low metallicity systems have a lower , dust to gas ratio and heavy elemental abundances as compared to the galactic ism .
this reduction in the amount of dust grains affects , the absorption of uv radiation , heating of gas by photo electric emission from dust , formation of h@xmath2 on the dust grains and the cooling of the gas through atomic and molecular lines .
in addition , the chemistry is also affected by the reduction in dust and heavy elements ( van dishoeck & black 1988 ; lequeux et al . , 1994 ) .
we use the self - consistent spherical pdr model of strzer et al .
( 1996 ) to study the metallicity effects .
we scale the dust dependent parameters and the abundance of heavy elements with the metallicity factor , z , in our pdr calculations .
the variation of the temperature at the surface of the pdr clumps with the incident uv field is plotted in figure [ pdrtemp ] , for a clump of mass , m=1m@xmath3 and density , n=10@xmath4 .
the uv field , @xmath5 , is expressed in units of mean uv field of draine ( 1978 ) .
it is seen from the figure that at high uv fields the temperature is proportional to metallicity , whereas at low uv fields there is no significant change in the surface temperature .
this correlation can be understood analytically by balancing the dominant heating and cooling processes . in the case of pdrs exposed to high uv fields grain photo - electric emission ( pe ) dominates the heating .
the photo electric heating rate given by bakes & tielens ( 1994 ) is @xmath6 where z is the metallicity factor .
the cooling is dominated by fine structure [ oi ] emission , gas - grain collisions and fine - structure [ cii ] emission .
the cooling rate can be written as @xmath7 where @xmath8 is the escape probability , @xmath9 is the transition probability and @xmath10 number of atoms at level @xmath11 and @xmath12 is the corresponding frequency of radiation .
following hollenbach & mckee ( 1979 ) , for the population of the first level of the oi atom , the cooling rate of [ oi]63@xmath1 m can be written as , @xmath13 although eqns .
[ peheat ] and [ oicool ] show a similar dependence with metallicity , the grain heating rate has an additional dependence on metallicity through the charge state of the grains .
the charge state of the grains is expressed as @xmath14 which depends on the availability of electrons for recombination .
the main source of electrons at high uv fields is the ionisation of ci . since at the surface almost all of the ci is ionised , the electron density @xmath15 . with this assumption the heating rate , @xmath16 decreases with decreasing @xmath17 whereas the cooling rate
, @xmath18 remains constant .
the equilibrium temperature obtained by solving eqns .
[ peheat ] and [ oicool ] decreases as the metallicity decreases ( cf .
figure [ equitemp ] ) as seen in our pdr calculations .
in our clumpy model , the molecular cloud is modelled as being composed of many spherical clumps of mass spectrum of the form , @xmath19 we use @xmath20 ( e.g. kramer et al . , 1998 ) .
we assume that the turbulent velocity dispersion of the cloud is larger than the intrinsic line width of each clump .
thus the clumps do not interact radiatively and the total intensity of a spectral line from the cloud can be written as , @xmath21 the beam filling factor of each clump is @xmath22 ) where @xmath23 is the solid angle of the clump of mass @xmath24 and @xmath25 is the beam solid angle .
the fraction of the clumps within the beam is given by @xmath26 . by scaling the mass with @xmath27 ( @xmath28 ) @xmath29 @xmath30 where @xmath31 is the beam filling factor , @xmath32 and @xmath33 .
although the total intensity depends on the scaling constant and the beam filling factors , the line ratio between any two spectral lines depends only on @xmath34 and @xmath35 .
the observations of [ cii ] and [ ci ] emission from low metal galaxies have been modelled by bolatto et al .
( 1999 ) , assuming that the size of the cii region scales inversely with metallicity .
it is also assumed that the size of the ci region remains between two limiting scenarios of , an inverse dependence and no variation with metallicity .
our spherical pdr model calculations based on the model of strzer et al .
( 1996 ) show that the size of the cii layer is indeed inversely proportional to the metallicity factor z. however the size of the ci region shows very weak dependence at low z and a roughly inverse dependence at high z for a typical spherical clump of density , @xmath36 .
the intensity ratio @xmath37158\mu m}$]/@xmath38 , observed in many nearby low metal galaxies , show a power law dependence with metallicity .
this dependence has been predicted using a semi - analytical clumpy model by bolatto et al .
we use our clumpy model explained in section [ clumpymodel ] to study the metallicity dependence of this line ratio .
the observed ratios of nearby galaxies can be well represented by a clumpy model of density @xmath39 , exposed to a uv field of @xmath40 as shown in figure [ clumpyplot ] .
the higher observed ratio for the 30 doradus region can be explained by a similar clumpy model , but exposed to an uv field of @xmath41 .
these results compare well with the results shown by the semi - analytical model of bolatto et al .
this trend suggests that at low metallicities cii is a tracer of molecular hydrogen rather than co. however the observed variation of [ ci]/co line ratio with metallicity is steeper than the model prediction .
this is most likely due to large [ ci ] line intensities predicted by the pdr models and requires further investigation .
arimoto n. , sofue y. , & tsujimoto t. 1996 , pasj , 48 , 275 bakes & tielens 1994 boiss p. 1990 , a&a , 228 , 483 bolatto a.d .
, jackson j.m . , ingalls j.g . 1999 ,
apj , 513 , 275 bolatto a.d . , et al . ,
2000 , apj , 541 , 17 draine b.t .
1978 , apjs , 36 , 595 hollenach & mckee 1979 hollenbach d.j . ,
tielens a.g.g.m . 1999 ,
rev.mod.phys , 71,173 hollenbach d.j . ,
tielens a.g.g.m .
1997 ara&a , 35 , 179 kaufman m.j . , et al . , 1999 , apj , 527 , 795 kster et al . , 1994 , a&a 284 , 545 kramer , c. , et al . , 1998 , a&a , 329 , 249 le bourlot , j. , et al . , 1993 , a&a , 267 , 233 lequeux , j. , et al . , 1994 ,
a&a , 292 , 371 madden s.c .
2000 , newar , 44 , 249 sternberg a. & dalgarno a. 1995 , apjs , 99 , 565 strzer h. , stutzki j. & sternberg a. 1996 , a&a , 310 , 592 stutzki j. , et al .
, 1988 , apj , 332 , 379 tielens a.g.g.m .
, hollenbach d. 1985 , apj , 291 , 722 van dishoeck , e. f. & black , j. h. 1988 , apj , 334 , 771 wilson c. d. 1995 , apj,448 , l97 | several galaxies , such as dwarfs and irregulars as well as outer galactic clouds have low metallicity . at low metallicities
a reduction in the amount of dust and heavy elements plays a significant role on the chemistry as well as the heating and cooling of the gas in the molecular regions , called as photon dominated regions ( pdrs ) .
we present here the effects of reduced metallicity in pdrs and study the important pdr cooling lines ( [ cii ] , [ ci ] and co ) . moreover many observational evidences suggest that molecular clouds are clumpy .
we model the molecular emission from galaxies incorporating a mass spectrum of clumps .
we also compare our results with the semi - analytical results obtained by bolatto et al .
( 1999 ) . |
the impact that active galactic nuclei ( agn ) have on their hosts is one of the most critical components of galaxy formation theory @xcite .
observationally , astronomers find fewer luminous galaxies than expected based on the evolution of a @xmath1cdm universe and frequently attribute this phenomenon to agn quenching of star formation @xcite , ultimately reducing the number of bright galaxies .
some cosmological simulations use the injection of thermal energy or mechanical heating into the center of galaxies as a prescription for agn feedback ( e.g. * ? ? ?
* ; * ? ? ?
* ) , and often derive luminosity functions that match up better with observations . however , the physical processes in agn feedback are far more complex than the introduction of heating processes .
recent computational analyses of jet - driven feedback shows that relativistic jets can actually induce star formation in their host galaxies @xcite .
agn jets have largely been assumed to reduce star formation rates because the jets increase the temperature of the gas they collide with , making collapse under self - gravity more difficult .
interestingly , both recent simulations and observations reveal that jets can actually foster star formation by creating some regions of high density and low temperature embedded in the cocoon surrounding the jet @xcite .
expanding jets create cocoons of turbulent gas surrounding the jet , and the overpressured cocoons form bow shocks and hit clouds of cold gas , increasing density and creating regions of star formation @xcite .
the bow shock generated from jets can also potentially collapse pockets of cold gas to form stars as the cocoon expands out along the disk and compresses it .
for example , both minkowski s object and hanny s voorwerp object are galaxy - scale gas - dominated clouds struck by jets that show high star formation rates @xcite .
other observations from @xcite and @xcite also show extended star formation in galaxies that host agn , while @xcite even finds star formation potentially triggered by an agn at a redshift of @xmath2 .
@xcite also provides a theoretical framework for agn triggered star formation . unfortunately , agn jet - triggered star formation is difficult to observe because agn jets are short - lived on a cosmological time scale ( @xmath3 yr ) , which makes simulations critical to comprehending this phenomenon .
furthermore , the difficulty of simultaneously observing jets and star formation in the host may obscure the frequency of positive jet feedback , as discussed in @xcite
. once a more complex , more accurate relationship between jets and star formation is established , previously unexpected characteristics of stellar populations of galaxies at high redshifts may be explicable . in this paper
, we aim to find the impacts of jets on host galaxy stellar populations that endure substantially longer than the jet itself to further help establish the role of jets in galaxy formation and evolution .
we analyze the hydrodynamic simulations by @xcite of jets in a gas rich , @xmath4 galaxy , and then integrate the orbits of stars formed during the simulation for a gyr to find signatures of past agn - triggered star formation .
agn feedback may also link to the evolution of the host galaxy s morphology .
several studies show spherical and cylindrical symmetry in morphological effects generated from agn interactions @xcite .
observations and arguments from @xcite , @xcite , @xcite , and @xcite show that galaxies of a particular mass at @xmath4 are more dense than galaxies of roughly the same mass at a redshift of @xmath5 , indicating that a mechanism , presumable mergers or agn feedback , is expanding the size of these galaxies over a period of roughly 10 gyr .
similar observations show that the central regions of the galaxies at @xmath4 are similar to galaxies of the same mass today , indicating that the growth in size is occurring in the outer regions , a phenomenon we see in our simulation .
we examine radial velocity distributions in our simulations to examine their impact on the growth in effective size of massive galaxies at @xmath4 . in this respect ,
agn jets may be linked to the ongoing mystery of the origin and cause of hypervelocity stars ( hvs ) in our own galaxy .
analytical arguments have been made supporting the idea that an agn jet could be the mechanism generating high velocity stars that may or may not escape their host galaxy @xcite .
the theory is that the bow shock of the jet can both condense and accelerate pockets of gas that will then form stars which share the host cloud s velocity .
obviously observing such a phenomenon would be difficult .
however , both momentum and energy based arguments can be made supporting the idea , as in @xcite . to add to this theoretical argument
, we analyze stars formed in the simulation that eventually escape the host galaxy .
if jets do , in fact , induce star formation , and some of these stars are hypervelocity stars , then these high velocity stars may be excellent tracers of past jet activity and corresponding induced star formation .
many hvs orbits appear to emanate from the center of our galaxy , largely contradicting the possibility of supernovae and birth kicks as the main generator of hvs @xcite .
we analyze our simulations to examine agn as a possible mechanism for the generation of hvs and escape stars , which in turn could be signatures of agn - triggered star formation .
while the milky way is very different than the high redshift gas - rich galaxies simulated in @xcite and analyzed in this paper , patterns and distributions of escaped stars from @xcite could potentially also be applied to our own galaxy despite the important differences in the respective physical parameters .
several studies indicate possible jet activity in the milky way s past , as in @xcite and @xcite .
the presence of the fermi bubbles further suggests past agn activity roughly 10 myr ago @xcite .
furthermore , present data on hypervelocity b stars may indicate a possible agn outburst @xmath6100 myr ago @xcite . if these time scales are indeed accurate , then the time scales of orbit evolution should suffice to extract signs of past jet activity , jet - driven escaped stars , and perhaps jet - induced star formation . particularly in light of the upcoming gaia space astrometry mission , which should produce the positions and velocities of a billion stars in the milky way , further analysis of the jet s impact on stellar positional and velocity distributions as well as galaxy morphology should provide signatures not just of past jet activity , but also of jet - induced star formation triggering in our own galaxy .
we analyze the four simulations of @xcite examining agn jet activity in a massive , gas - rich disk galaxy with an exponential radial profile with a 5 kpc scale radius , a scale height of 1.5 kpc , and a mass of @xmath7 .
gas in the disk initially has a radial and vertical profile @xmath8 on top of a fractal cube that mimics the clumpy interstellar medium .
the simulation utilizes the ramses adaptive mesh refinement code from @xcite with maximum resolution of grid cells that are 62.5 pc on a side .
gravity was not included due to the short time scales in the simulation compared to the disk evolution time scale .
control simulations without a jet were performed to examine the evolution of the disk to isolate the effects of the jet on star formation and gas dynamics . in total , four simulations are performed : disk only as well as disk and jet for both low and high thresholds for star formation . the simulation uses a star formation model prescribed by @xcite which reproduces the kennicutt
schmidt relation @xcite .
stars are created only in regions where the number density of hydrogen is @xmath9 , with a rate controlled by a fixed star formation efficiency value and the local free - fall time @xmath10 where @xmath11 is the star formation efficiency , @xmath12 is the local free - fall time , and @xmath13 is the star formation threshold . this simulation looks at two cases for star formation , the first with @xmath14 and @xmath15 @xmath16 , and the second with @xmath17 and @xmath18 @xmath16 . in this paper
, we examine the positions , velocities , and time of formation of all the stars formed in each of the four simulations . for this study , we assume a gravitational potential that is generated by an exponential thin disk and a nfw @xcite dark matter halo .
we assume that this potential remains constant during our simulated time range .
since the mass of jet - induced stars is considerably smaller than the mass of the old , pre - existing stellar component and the dark matter halo , the change in the potential will be small .
because the hydrodynamic simulation did not include the rotation of the galaxy , a rotational velocity component must be added to each star to properly model the long - term evolution of the stellar population .
this circular velocity is determined from the gravitational potential , and the rotational velocity resulting from an exponential thin disk is @xcite @xmath19 where @xmath20 and @xmath21 are the modified bessel functions of the first and second kind , and @xmath22 g @xmath23 is the central surface density , calculated with @xmath24 the baryonic mass in the simulation .
the simulations by @xcite prescribe the total baryonic mass in the galaxy , and from this baryonic mass we determine the mass of the dark matter halo using a calibration from @xcite , @xmath25 where @xmath26 is the baryon mass , and @xmath27 is the total mass , equal to the sum of the baryon mass and the dark matter mass .
the simulation begins with @xmath28 , resulting in a dark matter mass , @xmath29 .
the nfw potential from @xcite is : @xmath30 where @xmath31 is the virialized mass , @xmath32 is the scale radius of the dark matter halo , c is the concentration parameter , and @xmath33 we set @xmath31 equal to the dark matter mass @xmath34 , calculated above , and @xmath35 kpc .
we use a mass - concentration relationship from @xcite to determine a reasonable value for @xmath36 given the mass of the dm halo , and determine @xmath37 .
the circular velocity resulting from the nfw potential is therefore @xmath38 to combine the effects of the different potentials , we take advantage of the spherical symmetry and add the two rotational velocities resulting from the disk and from the dark matter together in quadrature as in @xcite . because we have no knowledge of where , in the direction perpendicular to the plane of the disk , each star is in its theoretical orbit , we assume that upon the addition of the circular velocity , each star is at its maximum height above or below the disk . as such , at that point in the star s orbit , the component of the star s velocity perpendicular to the plane of the disk is zero .
therefor , as we break down the centripetal velocity to be added to the star s initial velocity , we break it down only into the components in the plane of the galaxy s disk , the y - z plane .
we then integrate all the orbits for 1 gyr after the simulation ends with a runge - kutta - fehlberg routine .
we examine ten snapshots between 0 and 0.1 gyr and ten snapshots between 0.1 and 1 gyr in order to analyze the evolution of the galactic morphology as well as various velocity and parameter distributions .
we calculate the following parameters for the snapshots from 0 to 1 gyr : the spin parameter @xmath39 , the velocity anisotropy parameter @xmath40 , and the shape parameters @xmath41 , @xmath42 , and @xmath43 .
the spin parameter is calculated using the @xcite definition @xmath44 where @xmath45 is the mass contained within a radius @xmath46 , and @xmath47 is the angular momentum within the same radius .
@xmath48 is the circular velocity as a function of radius @xmath46 , calculated as @xmath49 . as a quantification of the proportion of radial to tangential orbits
, we also calculate the velocity anisotropy parameter @xcite : @xmath50 where the radial velocity dispersion is represented by @xmath51 the tangential velocity dispersion by @xmath52 . a galaxy with a @xmath53 is composed of purely radial orbits , and galaxies with more tangential motion have more negative values . in order to quantify the shape and evolving morphology of the galaxy , we begin with the calculation of the inertial tensor as in @xcite : @xmath54 we then diagonalize the inertia tensor and calculate the eigenvalues and eigenvectors .
we define the parameters @xmath55 , @xmath43 , @xmath36 as the square roots of the eigenvalues ( where @xmath56 ) .
finally we compute the shape parameter @xmath41 , which quantifies how spherical a galaxy is , as @xmath57 . a distribution
that is perfectly spherical will have @xmath58 .
to quantify triaxiality , we compute the shape parameter @xmath59 .
a galaxy with @xmath60 ( 0 ) corresponds to an prolate ( oblate ) halo .
these shape parameters allow the quantification of the morphology of the galaxy , which in turn facilitates analyzing the evolution of the galactic distribution of stars as a direct result of the agn jet .
it should be noted that only stars formed in the hydrodynamic simulation ( 10 myr without jet , plus @xmath61 with active jet ) are taken into account . since the pre - existing stellar population is not affected by the jet activity
, the stellar population properties rather indicate the changes in these properties than actual measurable values .
also the feedback duration will impact the strength of these changes .
furthermore it should be kept in mind that the velocities used in this analysis taken from the hydrodynamic simulation , which by far can not resolve the actual scales of star formation within molecular clouds .
the stellar velocities , as inherited by their parent gas cloud , will certainly depend on the efficiency of momentum transfer onto the molecular gas . in the present paper , we can not properly account for these efficiencies and rather have to interpret the velocities conservatively as upper limits .
we find a substantial difference between the age , velocity , and number distribution of stars created in the simulations with and without the relativistic jet , specifically after the jet has begun .
the simulations with the jet show much larger vertical velocity distributions , larger azimuthal ( rotational ) velocity distributions , larger radial velocity distributions . furthermore , there are other locational signatures of jet activity , such as rings of star formation .
gyr for the simulations without the jet ( _ left _ ) and with the jet ( _ right_).,title="fig : " ] gyr for the simulations without the jet ( _ left _ ) and with the jet ( _ right_).,title="fig : " ] figure [ fig : time_formation_stellar_mass ] shows plots of the mean time of formation of stars within the galaxies with and without the jet .
the disk has been divided into 10,000 grid cells and color coded with the mass - weighted average time of formation of stars in each cell .
a circular cavity with a radius of @xmath62 kpc is carved out of the disk galaxy where the mean time of formation is below @xmath63 myr , well before the jet is started . in this central part of the disk ,
little star formation occurs after 10 myr , when the jet starts .
the jet creates a cocoon and subsequent bowshock that re - pressurize the disk , a process that moves radially outward , creating rings of star formation along the disk .
figure [ fig : phase_time_formation_radius ] is a 2d histogram of the time of formation vs. the star s distance from the center at formation , weighted by stellar mass .
the jet begins at 10 myr , so the data before 10 myr reflects galactic evolution without a jet . the data after the jet begins at 10 myr shows not only that more star formation occurs , but also where and when these stars are forming .
the feature beginning at 10 myr and extending from 2 kpc outward radially represents a ring of star formation that begins when the jet turns on .
this ring then moves outward along the disk from a radius of 2 kpc at 10 myr to 6 kpc at 24 myr .
furthermore , the data from the simulation with the jet shows additional star formation further out along the disk beyond a radius of 6 kpc all the way to a radius of 14 kpc . both the bow shock from the jet and the pressurization of the disk by the cocoon cause this additional star formation .
the region within a radius of @xmath62 kpc where star formation nearly ceases shortly after 10 myr is obvious from this plot .
we find substantial differences in stellar velocity distributions between galaxies with and without jets that endure for 1 gyr after the beginning of the simulation , long after the jet has stopped .
the jets induce larger vertical velocities off the disk as well as larger radial velocities along the disk , along with slower azimuthal velocities in the plane of the disk .
these differences in velocity distributions indicate that more stars formed in galaxies with jets are formed in regions that have been accelerated outward in the radial and vertical directions and in regions with rotational velocities that have been deccelerated , likely because of the pressurization of the disk from the jet s cocoon .
figure [ fig : phase_radius_vertical_velocity ] shows phase plots of vertical velocity versus disk radius weighted by stellar mass for the disk only simulation at 0.3 gyr , the disk and jet simulation at 0.3 gyr , and the disk and jet simulation at 1 gyr . we overplot lines on the phase plots showing the @xmath64 , @xmath65 , @xmath66 ( median ) , @xmath67 , and @xmath68 percentiles of velocity .
the disk only simulation shows very small vertical velocities and small deviations from the average vertical velocity of zero , and also changes very little over the next gyr of evolution .
the jet simulation data , however , shows much higher vertical velocities and larger vertical velocity dispersions , particularly at 0.03 gyr , just 5 myrs of evolution after the end of the jet simulation .
however , the distribution does tighten as the galaxy evolves , particularly at radii @xmath69 kpc for @xmath70 gyr ( lower panel ) .
our results show a jet - induced difference in the azimuthal velocity distributions .
we find that in the galaxy with a jet , the rotational velocity of stars within a radius of 4 kpc is greater than in the galaxy without a jet , while the rotational velocity of stars with radii greater than 4 kpc is slower than in the galaxy without the jet .
figure [ fig : phase_radius_rotational_velocity ] shows phase plots of rotational velocity versus disk radius . the simulation without the jet shows only small deviations from the expected rotation curve .
the jet simulation data , however , shows an initial spreading of the expected rotation curve yielding larger distributions of azimuthal velocity .
however , as the stellar motion within the galaxy evolves , stars within a radius of 4 kpc have a higher rotational velocity than the expected rotation curve , while stars further out that 4 kpc have a significantly slower rotational velocity than the disk only simulation .
this signature also lasts throughout the entire 1gyr evolution .
it is possible that the increase in rotational velocity in smaller radii and the decrease in rotational velocity at larger radii is the result of a larger fraction of elliptical orbits , whose tangential velocities will be greatest at the closest point to the center of the galaxy and smallest at the furthest point form the center of the galaxy .
gyr and are weighted by stellar mass . ]
we also find substantially higher radial velocities in the simulations with the jet .
figure [ fig : phase_radius_radial_velocity ] shows phase plots of radial vtelocity versus disk radius for the disk only simulation at 20 myr , the disk and jet simulation at 20 myr , and the disk and jet simulation at 1 gyr . without the jet , only little to no radial velocity is found for the entire 1 gyr evolution .
the jet simulation data , however , contains stars that are formed with a large radial boost , particularly stars formed between radii of @xmath71 kpc to @xmath72 kpc .
this boost is caused by the bow shock emanating from the jet and spreading through the galaxy .
as the galaxy evolves , however , stars with large positive radial velocities are eventually pulled back toward the center of the galaxy .
the positive bias in radial velocity persists through 20 myr , before these stars are pulled back toward the center of the galaxy . by around 30 myr
, the distribution of radial velocities evolves toward a large negative bias with stars heading back toward the galaxies center . by roughly 0.2 gyr ,
however , the fluctuation has stopped , and instead a distribution of radial velocities far larger than in the corresponding snap shot form the disk only simulation is present . that larger distribution of radial velocities is still present after 1 gyr of evolution .
the analysis of the radial velocities has important ramifications on assessing the possible origin of hvs because of the large distribution of high radial velocities : roughly @xmath73 of stars formed around a radius of 2 kpc have radial velocities @xmath74 km s@xmath75 .
we define escaped stars as those whose positions exceed spherical radii of 30 kpc . no stars from the disk
only simulations escape the galaxy , and no stars from the jet simulation escape the galaxy until the jet begins at 10 myr .
figure [ fig : escape_time_formation_radius ] is a phase plot of time of formation vs. initial disk radius of the escaped stars , and clearly shows the bow shock driven by the jet propagates out through the disk and induces the formation of stars in gas clouds with enough momentum to escape the galaxy .
some of the stars have velocities @xmath74 km s@xmath75 .
we find a decrease in hvs production and a drop in the velocity of those stars after 15 myr , after the bow shock has passed through the most central part of the disk , consistent with the idea that the shock is generating the hvs .
some of the stars escape the galaxy with high vertical velocities , moving almost straight off the disk .
these stars were formed closed to the onset of jet activiy .
other stars escaped the galaxy with high velocities in the plane of the disk , shooting out the edge of the disk .
most of the escaped stars are formed within a radius of 4 kpc , but some are formed at central distances of @xmath76 kpc . of the escaped stars formed 1015 myr in the simulation ,
@xmath77% reach a distance from the center of 30 kpc while still remaining within 15 kpc from the jet axis , indicating that most of their velocity is in the vertical direction .
in fact , @xmath78% of these stars formed 1015 myr in the simulation have initial vertical velocities greater than 200 km s@xmath75 .
of stars formed within 4 kpc distance from the center , about 50% reach a distance of 30 kpc while still within 10 kpc from jet the axis , and roughly 75% reach a distance of 30 kpc while still within a 15 kpc distance from the jet axis , indicating again that most of these stars motion is vertical to the disk .
an analysis of the time of formation versus the star s radius at formation , as shown in figure [ fig : escape_time_formation_radius ] traces very closely the pattern of jet induced formation as demonstrated in figure [ fig : phase_time_formation_radius ] .
this shows that the jet s bow shock not only induces star formation but also generates high - velocity pockets of star formation .
+ [ fig : escape_time_formation_radius ]
to quantify the differing evolution in morphology and velocity distributions , we calculate the sphericity parameter @xmath41 , the triaxiality parameter @xmath42 , the spin parameter @xmath39 , and the velocity anisotropy parameter @xmath40 as a function of time and radius .
we calculate these parameters for the twenty snap shots mentioned in sect .
[ sec : method ] . in agreement with the increased distribution in radial velocity distribution ,
the simulation with the disk and jet has a substantially lower triaxiality for 200300 myr after the jet , meaning that the simulations with the jet have more of a flat , oblate disk shape during that time .
the galaxies with the jet also show lower sphericity for the duration of the 1 gyr evolution after the jet .
this difference in both parameters indicates that the agn jet s bow shock creates a more extended stellar structure along the disk of the galaxy .
this conclusion agrees with observationally motivated ideas about agn feedback ` puffing up ' the size of massive galaxies at @xmath4 discussed by @xcite .
@xcite show that galaxies of a given mass at @xmath4 are more dense than galaxies of similar mass at a redshift of @xmath5 , indicating that some mechanism , presumable mergers or agn feedback , is expanding the size of these galaxies over a period of roughly 10 gyr .
agn jets in our simulation trigger star formation in regions with significant radial velocities acting to expand the effective size of the galaxy , particularly at the outer regions ( figs .
[ fig : time_formation_stellar_mass ] and [ fig : phase_radius_rotational_velocity ] ) .
similar observations show that the central regions of the galaxies at @xmath4 are similar to galaxies of the same mass today , indicating the growth in size is occurring at the outer regions @xcite , a phenomenon we see also in our simulation .
the spin parameter quantifies the coherence of the rotation of the galaxy , and is much higher in the disk only simulations than in the disk and jet simulations .
this is another indicator that the jet disrupts the basic rotation of the galaxy , in agreement with previous analysis of the higher radial and vertical velocity distributions in the simulation with the jet as well as the mostly slower rotational velocity distributions in the simulations with the jet .
the difference in the spin parameter is larger when the calculation includes stars at larger radii , indicating that as an analysis includes more of the outer regions of a galaxy hosting a jet , the less coherent the rotation becomes .
this difference further emphasizes the impact of the jet on the outer regions of the galaxy . as function of time for both the disk only simulation and the disk and jet simulation , splitting the stars into four radial regions ( axial distance).,title="fig : " ] + figure [ fig : beta ] shows the velocity anisotropy parameter @xmath40 over time for both galaxies with and without the jet for four different maximum radii .
the jet causes an extreme difference in the velocity anisotropy parameter @xmath40 , and this agrees with our previous analysis of the differing distributions of vertical , radial , and rotational velocity distributions .
the velocity anisotropy parameter for the galaxy with the jet is far closer to 1 , between 0.2 and 0.8 for most of the maximum radii , reflecting a much higher bias toward radial motion in the host galaxy than in the disk only simulation , in which @xmath40 is between -20 and -1 for the four maximum radii .
there is more fluctuation in the velocity anisotropy early on for the galaxy with the jet , but the difference in @xmath40 persists for the duration of the gyr evolution , making increased velocity anisotropy and bias toward radial motion in the host galaxy a lasting signature of past jet activity .
our analysis identifies several signatures of past agn jet activity .
the first is the ring of star formation created shortly after the jet becomes active that moves outward radially along the disk starting from a radius of 2 kpc and moving out to a radius of 6 kpc , which corresponds to @xmath71 to 4 scale heights of the gaseous disk
. this ring will be one of the most observable signatures to come out of these simulations and in fact may have already been observed in the nearby ( @xmath79 ) radio galaxy cygnus a. @xcite finds a ring of young , blue stars around the center of the galaxy , perpendicular to the direction of the jets .
they contend that the blue light is from young stars and not from scattered quasar radiation .
the energy spectrum more closely indicates a possibly instantaneous starburst less than 10 myr ago .
that time frame would be consistent with the age of agn activity in the galaxy , also on the order of 10 myr .
thus , cygnus a could provide observational evidence for this effect . however
, because jet activity is so brief on a cosmological time scale , the signatures of jet activity that are still observable long after jet activity ends will be the most useful .
this process also creates a lasting gradient in the stellar populations time of formation , with the oldest stars having formed in the center of the galaxy and the average age of stellar populations becoming younger as one moves out along the disk .
this also generates the formation of young stars along the outer edges of the galaxy , far from the center .
in fact , not only do @xcite and @xcite find possible agn - triggered star formation , @xcite finds rings of star formation in galaxies with agn but with no current jet activity , consistent with an inside - out positive feedback model as we see in our simulations .
the gradient in the age of stellar populations from the center of the galaxy outward could be a strong signature of jet - triggered star formation .
furthermore , @xcite finds extended star formation at high radii in galaxies that host agn .
while these galaxies do not host current jet activity , star formation at the outer parts of the galaxy is consistent with our results .
our analysis , also shows much wider distributions for radial , vertical and azimuthal velocities .
these increased distributions all persist up to a gyr after the jet stops .
of course , our simulations only show these signatures for the stars formed during these 25 myr and not for stars formed for the next gyr after , which would effect how observable these effects would be .
these increased distributions also create galaxies with higher velocity anisotropy parameters and lower spin parameters , indicating more random and less coherent orbits within the galaxy , with much higher biases toward radial motions in galaxies that hosted jets in the past .
these characteristics all may endure for a gyr after the jet ends , depending on the stellar mass originating from the jet - induced star formation . in agreement with the increased distribution of radial velocities , in particular the positive bias in radial velocity with the initialization of the jet
, we find that high - velocity stars and escaped stars are formed in the exact same pattern as the previously discussed rings of star formation , as characterized by their radius of formation and time of formation .
therefore , a distribution of escaped stars that matches that pattern could also indicate past jet activity , and perhaps past jet induced star formation . while the milky way is a @xmath80 , barred spiral galaxy , and the galaxies we simulate are high - redshift and gas - rich , it is possible that some of the signatures of agn jet activity in the galaxies we simulate may also be present in the milky way .
these results may also be examined with the data of the upcoming space astrometry mission gaia , which will accurately measure the positions , proper motions , and velocities of a billion stars in the milky way .
these comparisons would have to be between general results , such as a noticeably positive bias in the radial velocity distribution .
@xcite has already measured the velocity anisotropy parameter of some halo stars in the milky way .
as similar analyses are published , comparison with a theoretical approach as described here will yield productive results for the role of agn jets possibly in the milky way and more generally in galaxy formation and evolution .
zd was supported by a centre for cosmological studies balzan fellowship .
vg was supported by the sonderforschungsbereich sfb 881 `` the milky way system '' ( subproject b4 ) of the german research foundation ( dfg ) .
the research of js has been supported at iap by the erc project 267117 ( dark ) hosted by universit pierre et marie curie - paris 6 and at jhu by nsf grant oia-1124403 . | to investigate feedback between relativistic jets emanating from active galactic nuclei ( agn ) and the stellar population of the host galaxy , we analyze the long - term evolution of the galaxy - scale simulations by @xcite of jets in massive , gas - rich galaxies at @xmath0 and of stars formed in the host galaxies .
we find strong , jet - induced differences in the resulting stellar populations of galaxies that host relativistic jets and galaxies that do not , including correlations in stellar locations , velocities , and ages .
jets are found to generate distributions of increased radial and vertical velocities that persist long enough to effectively extend the stellar structure of the host .
the jets cause the formation of bow shocks that move out through the disk , generating rings of star formation within the disk .
the bow shock often accelerates pockets of gas in which stars form , yielding populations of stars with significant radial and vertical velocities , some of which have large enough velocities to escape the galaxy .
these stellar population signatures can serve to identify past jet activity as well as jet - induced star formation . |
non - stationary effects now a days attract much attention and are vital both from fundamental and technological points of view .
first of all , non - stationary characteristics provide more information about the properties of nanoscale systems comparing to the stationary ones .
moreover , modern electronic devices design with particular set of transport parameters requires careful analysis of non - stationary effects , transient processes and time evolution of charge and spin states prepared at the initial time moment @xcite,@xcite,@xcite,@xcite,@xcite,@xcite .
correct analysis of the non - stationary dynamics of @xmath0local@xmath0 magnetic moment and electron occupation numbers of the correlated anderson impurity coupled to reservoir requires the investigation of the kondo correlations influence on the system time evolution .
it is necessary to clarify the question how the relaxation rates of @xmath0local@xmath0 magnetic moment and charge density change with the appearance of the kondo correlations .
one can distinguish two main problems .
the first one is widely discussed in the literature and deals with the kondo correlations decay ( correlations already exist at the initial time moment ) due to the inelastic processes connected with the many - particle interaction , external field and so on @xcite , @xcite,@xcite,@xcite,@xcite .
in such situation the typical rate , when the kondo correlations disappear is usually connected with the inverse decoherence time @xmath1 .
dephasing rate caused by the inelastic electron - electron scattering was analyzed in @xcite .
authors obtained the dependence of spin - flip rate on transferred energy in two limiting cases : the temperature is higher than the kondo temperature and much lower than the kondo temperature .
non - equilibrium decoherence rate induced by the voltage driven current in quantum dot systems was analyzed in @xcite .
the authors have demonstrated that in the regime of large voltage ( higher than the kondo temperature ) tunneling current prevents the development of the kondo correlated singlet state and have found decoherence rate induced by applied voltage .
later , the dependence of typical spin - flip rate on the value of external magnetic field in non - equilibrium case was investigated in @xcite .
the authors demonstrated that inelastic processes associated with the finite current through the dot result in the spin - flip effects with typical rate determined by the renormalized exchange energy .
so , tunneling conductivity and magnetization were found to be universal functions of @xmath2 and @xmath3 , where @xmath4 - ia the applied bias voltage , @xmath5 - external magnetic field and @xmath6 - is the equilibrium kondo temperature .
the decay rate of the kondo correlated state due to photon assisted processes was analyzed in @xcite , @xcite .
it was demonstrated in @xcite that the dot driven out of equilibrium by an ac field is also characterized by universal behavior : the dot s properties depend on the ac field only through the two dimensionless parameters , which are the frequency and the amplitude of the ac perturbation , both divided by @xmath6 .
another problem which deserves careful analysis deals with the investigation of the kondo correlations appearance rate , when coupling to the reservoir is switched on at the initial time moment ( in such situation the kondo correlations and any correlations between the localized and reservoir states are initially absent ) .
so , the present paper is devoted to the investigation of the typical time scales responsible for the kondo correlations formation and the influence of the kondo effect on the system non - stationary characteristics .
we show that in the non - stationary case there exists the only one time scale @xmath7 , responsible for the formation of the kondo correlations , which are initially absent .
we consider non - stationary processes in the system of the single - level impurity coupled to an electronic reservoir with the coulomb interaction of the localized electrons ( see fig.[figure1 ] ) .
the model hamiltonian has the form : @xmath8 index @xmath9 labels continuous spectrum states in the lead , @xmath10- tunneling transfer amplitude between the continuous spectrum states and localized state with the energy @xmath11 .
@xmath10 is considered to be independent on momentum and spin .
operators @xmath12 correspond to the electrons creation / annihilation in the continuous spectrum states @xmath13 .
@xmath14-localized state electron occupation numbers , where operator @xmath15 destroys electron with spin @xmath16 on the energy level @xmath11 .
@xmath17 is the on - site coulomb repulsion for the double occupation of the localized state .
our investigations deal with the low temperature regime when fermi level is well defined and the temperature is much lower than all the typical energy scales in the system .
consequently the distribution function of electrons in the leads ( band electrons ) is close to the fermi step .
we are interested in the system dynamics , when coupling to the reservoir is switched on at the initial time moment .
so , any correlations between localized and reservoir states are not present initially .
let us consider @xmath18 elsewhere .
kinetic equations for the electron occupation numbers operators have the form : @xmath19 previously we analyzed long living magnetic moments time evolution for deep impurities ( @xmath20 ) and demonstrated that for the @xmath0paramagnetic@xmath0 initial conditions ( @xmath21 ) relaxation rate to the stationary state is determined by @xmath22 and in the case of the @xmath0magnetic@xmath0 initial conditions ( @xmath23 ) relaxation rate to the stationary state is determined by @xmath24 , where @xmath25 @xcite .
consequently , the long living @xmath0magnetic@xmath0 moments are present in the system .
@xmath26\cdot e^{-\lambda_1 t } \label{01}\end{aligned}\ ] ] we assumed that the kondo correlations are absent at the initial time moment are not significant , because they evolve much slower , than the magnetic moment relaxes . here arises an important question : what is the typical rate of the kondo correlated state formation , which initially does nt exist in the system
. this rate can be quite different from the characteristic decay rate of the kondo correlation , which were initially present in the system .
another important problem we are interested in is how the kondo correlations appearance influence on the local magnetic moment dynamics .
so , we try to clarify how the kondo correlations reveal in long living magnetic moment relaxation and how these correlations evolve with time .
for simplicity we ll analyze these problems in the case of infinitely large coulomb correlations @xmath27
. then @xmath28-h.c.)-2\cdot\gamma\langle\hat{n}_{1\sigma}\rangle \label{1}\end{aligned}\ ] ] where @xmath29 . the first term describes the corrections to the system dynamics beyond slowly varying amplitudes approximation .
operator equations of motion have the following form : @xmath30}{\partial t}&=&-t_{k}\sum_{k^{'}}(\hat{n}_{1k^{'}-\sigma}-\hat{n}_{k^{'}1-\sigma})\hat{n}_{k\sigma}-\nonumber\\ & -&t_{k}(1-\hat{n}_{1-\sigma})(\hat{n}_{1k\sigma}-\hat{n}_{k1\sigma}),\nonumber\\ \langle\hat{n}_{k\sigma}\rangle\frac{\partial \langle(1-\hat{n}_{1-\sigma})\rangle}{\partial t}&=&-t_{k}\sum_{k^{'}}\langle\hat{n}_{1k^{'}-\sigma}-\hat{n}_{k^{'}1-\sigma}\rangle\langle\hat{n}_{k\sigma}\rangle\nonumber\\ \label{2}\end{aligned}\ ] ] let s consider equation for @xmath31 . after averaging over fast oscillations it can be expressed as : @xmath32 applying sequential iteration procedure one can obtain : @xmath33 let us now analyze the nontrivial logarithmic divergent terms , which appear beyond the decoupling approximation .
@xmath34 here we used the following relation @xmath35 . retaining the most divergent logarithmic terms in the higher order iterations ,
we obtain : @xmath36^{2}+ ..... )\sim\nonumber\\ \sim(\varepsilon_k<<\gamma)\sim-\frac{t_{k}(1-\hat{n}_{1-\sigma})(\hat{n}_{k\sigma})}{\varepsilon_1(1-\frac{\gamma}{\varepsilon_1}\cdot
ln(\varepsilon_k / w)+i\frac{\gamma}{\varepsilon_1})}+ ... \label{6}\end{aligned}\ ] ] introducing @xmath37 in the usual way , one can easily get : @xmath38 finally , substitution of expression ( [ 7 ] ) to eq.([2 ] ) yields @xmath39 let s define @xmath40 . the main contribution to @xmath41 comes from @xmath42 .
so , one can obtain : @xmath43,\nonumber\\ \frac{\partial < n_{1\sigma}>}{\partial t}&=&-2\gamma[<n_{1\sigma}>-(1-<n_{1-\sigma } > ) n_{k\varepsilon}^{\sigma}-\nonumber\\&-&\frac{\gamma}{2\varepsilon}\cdot k^{-\sigma\sigma}(t)],\nonumber\\ \frac{\partial k^{\sigma-\sigma}(t)}{\partial t}&=&-\gamma_{kondo}\cdot[k^{\sigma-\sigma}(t)-k_{0}^{\sigma-\sigma}(t)],\nonumber\\ \frac{\partial k^{-\sigma\sigma}(t)}{\partial t}&=&-\gamma_{kondo}\cdot[k^{-\sigma\sigma}(t)-k_{0}^{-\sigma\sigma}(t)]\label{9}\nonumber\\\end{aligned}\ ] ] where @xmath44 and @xmath45\label{11}\nonumber\\\end{aligned}\ ] ] initial conditions are @xmath46 and @xmath47 . at @xmath48 the kondo correlations and any correlations between the impurity and reservoir states are absent .
so , the characteristic time scale of the kondo correlations formation can be defined as @xmath49 .
this time scale of the kondo correlations formation differs from the typical time of their decay , caused by the inelastic interaction ( voltage driven current , electron - phonon interaction etc . ) .
the appearance of the kondo correlated state is governed by the exchange interaction between localized and conduction electron in reservoir , while the decay of the kondo correlations is determined by inelastic spin - flip processes with the characteristic rate @xmath50 . at low temperature
@xmath51 this rate can be determined as @xmath52 , where @xmath53 is the typical transferred energy value due to inelastic interaction . in the case of voltage
driven current @xmath53 has to be replaced by @xmath4 .
obtained results ( see fig .
[ figure2 ] ) also demonstrate long living magnetic moment relaxation with typical rate @xmath54 up to the time @xmath55 , which is the time of the kondo correlations ( absent at initial time moment ) appearance . for the times @xmath56 electron occupation numbers time evolution @xmath57
demonstrate spin - flip effects , caused by the presence of the kondo correlations .
spin - flip effects lead to the non - monotonic behavior of electron occupation numbers and to the changing of local magnetic moment sign ( see fig . [ figure2 ] ) . but this effect is weak , because @xmath54 strongly exceeds @xmath58 and local magnetic moment nearly approaches to it s stationary value , when the kondo correlations appear .
slow changing of local magnetic moment sign near it s stationary value ( which is equal to zero ) is clearly seen from fig .
[ figure2 ] . with further time increasing local magnetic moment reaches it s stationary zero value for non - magnetic reservoir . at the large time
scales @xmath56 kondo correlations lead to slight decreasing of local magnetic moment relaxation rate . at such time
scales the correlation function @xmath41 is close to it s stationary value .
so , the magnetic moment relaxation rate @xmath59 is replaced by @xmath60 .
we analyzed non - stationary processes of the kondo correlations formation , when coupling to reservoir is switched on at the initial time moment .
it was found out that the typical time scale of the kondo correlations appearance is @xmath49 , which is quite different from decoherence time associated with the inelastic spin - flip processes .
the influence of the kondo effect on the non - stationary dynamics of local magnetic moment and electron occupation numbers of the correlated anderson impurity coupled to reservoir was investigated .
it was demonstrated that for the times @xmath56 electron occupation numbers time evolution is weakly influenced by the spin - flip effects , caused by the appearance of the kondo correlations , because the relaxation rate of local magnetic moment strongly exceeds @xmath58 .
it was also revealed that for the large time scales @xmath56 the kondo correlations lead to slight decreasing of local magnetic moment relaxation rate .
i. bar - joseph , s.a .
gurvitz , _
phys.rev b _ , * 44 * , 3332 , ( 1991 ) .
gurvitz , m.s .
marinov , _
phys.rev a _ , * 40 * , 2166 , ( 1989 ) . p.i .
arseyev , n.s .
maslova , v.n .
mantsevich , _
european physical journal b _ , * 85*(7 ) , 249 , ( 2012 ) .
stafford , n.s .
wingreen , _ phys .
_ , * 76 * , 1916 , ( 1996 ) .
hazelzet , m.r .
wegewijs , t. h. stoof , y.v .
nazarov , _ phys .
b _ , * 63 * , 165313 , ( 2001 ) .
e. cota , r. aguado , g. platero , _ phys .
_ , * 94 * , 107202 , ( 2005 ) .
a.kaminski , l.i.glazman , _ phys . rev .
_ , * 86 * , 2400 , ( 2001 )
. a. rosch , j. kroha , p. wolfle , _ phys .
_ , * 87 * , 156802 , ( 2001 ) .
a. rosch , j. paaske , j. kroha , p. wolfle , _ phys .
_ , * 90 * , 076804 , ( 2003 ) .
j. konig , t. pohjola , h. schoeller , g. schon , _ physica e _ , * 6 * , 371 , ( 2000 ) .
p. nordlander , n.s .
wingreen , y. meir , d.c .
langerth , _ phys .
b _ , * 61 * , 2146 , ( 2000 ) . | we investigated the typical time scales of the kondo correlations formation for the single - state anderson model , when coupling to the reservoir is switched on at the initial time moment .
the influence of the kondo effect appearance on the system non - stationary characteristics was analyzed and discussed . |
in march of 2010 , the _ fermi_-lat collaboration announced a new gev transient in the galactic plane , fgl j2102 + 4542 , that was identified as a nova outburst in the symbiotic binary v407 cygni ( hereafter , v407 cyg ) .
at least seven gev transients located near the galactic plane have been discovered by egret , _
fermi_-lat and agile .
only two have been identified at other wavelengths : v407 cyg , which is the first nova to be detected at gev energies , and j0109 + 6134 , which was likely a background blazar @xcite .
the physical nature of the other five sources is unknown @xcite , and some of these gev transients may represent a new class of gamma - ray emitting objects . the _ fermi_-lat collaboration reported variable gamma - ray emission in the 0.110 gev band from fgl j2102 + 4542 during 2010 march 1026 ( mjd 5526555281 ) @xcite .
its flux in gamma rays , binned on a day - to - day basis , peaked 2010 march 1314 with a flux of @xmath2 above 100 mev @xcite .
the gev gamma - ray activity lasted approximately two weeks .
the initial report of gev emission triggered very energetic radiation imaging telescope array system ( veritas ) observations of the object at very high energy ( vhe ; e @xmath0 100gev ) as part of an ongoing campaign to observe transients detected by _
fermi_-lat .
using multi - wavelength data , it was determined that the new transient was most likely associated with v407 cyg , a binary system consisting of a mira - type pulsating red giant and a white dwarf companion @xcite . a nova outburst from v407 cyg
was detected in the optical waveband on 2010 march 10 @xcite with a magnitude of @xmath3 , while pre - outburst magnitudes from the previous two years of monitoring ranged between magnitude 9 and 12 @xcite .
v407 cyg has been optically monitored for decades and has experienced previous outbursts , but the system had never been observed to be as bright as during the nova ( e.g. , * ? ? ?
* ; * ? ? ?
the onset of the optical outburst corresponds to the first significant detection of the source by the _
fermi_-lat on 2010 march 10 .
novae in red giant / white dwarf systems have been known to produce expanding shocks that can result in x - ray emission ( e.g. , the recurrent nova rs oph , * ? ? ?
* ; * ? ? ?
* ) , and indeed , x - ray emission from v407 cyg was detected after the nova @xcite .
based on the observed x - ray emission from the 2006 nova outburst of rs oph , before the launch of _
fermi_-lat , @xcite suggested that particles could be accelerated in novae up to tev energies , but gamma - ray emission from a nova had never previously been detected . here , we discuss the veritas observations of v407 cyg and their implications for gamma - ray emission from the nova .
we also describe an improved event reconstruction technique for stereo observations by imaging atmospheric cherenkov telescopes ( iacts ) made at large zenith angles ( lza ) .
veritas is a ground - based vhe gamma - ray observatory located at the fred lawrence whipple observatory in southern arizona .
it consists of four iacts sensitive from approximately 100 gev to above 30 tev .
each veritas telescope has a 12 m tessellated reflector with a total area of 110m@xmath4 .
each camera s focal plane contains 499 closed - packed circular photomultiplier tubes , giving a total field - of - view of @xmath5 .
gamma - rays incident onto the upper atmosphere induce a particle cascade , called an air shower , in which some charged particles have sufficient speed to emit cherenkov light .
the direction and energy of the original gamma ray can be reconstructed from images of the cherenkov light recorded by the telescopes .
when observing at small zenith angles ( @xmath6 ) , the array has an energy resolution of @xmath7 at 1 tev and an angular resolution of better than @xmath8 at 1 tev @xcite . for observations at lza , the energy and angular resolution are degraded and the energy threshold is increased .
veritas observed v407 cyg for several nights after the announcement of the _
fermi_-lat detection , during days 916 of the outburst ( 2010 march 1926 ) .
the zenith angle of these observations ranged between @xmath9 and @xmath10 .
the veritas telescopes are regularly operated in a mode called wobble mode , during which the location of the object to be observed is offset from the center of the field of view ( fov ) by 0.5@xmath11 , allowing for simultaneous source and background measurements @xcite .
the offset direction cycles between north , south , east and west for sequential observing segments to reduce systematic effects . after filtering the data for contamination due to poor weather or instrumental problems , 304 minutes of live time remained from the original 335 minutes of observations , see table 1 . to test the improved reconstruction technique discussed in section [ sec : two reconstruction methods ] ,
veritas observations of the crab nebula were also analyzed .
we selected 203 minutes of good time intervals from 17 data segments taken on the crab nebula during 2010 march 1216 ( mjd 5526755271 ) with similar zenith angles ranging from @xmath12 to @xmath13 .
all data were analyzed using the standard analysis package for veritas data @xcite .
the raw data were calibrated and cleaned , and quality selection criteria based on the number of photomultiplier tubes contained in the images and the position of the image in the camera were applied .
the shape and orientation of the gamma - ray images were parametrized by their principal moments @xcite . in order to produce gamma - ray images of the sky ,
it is necessary to reconstruct the putative source location for each shower in the camera plane ( hereafter arrival direction " ) .
when imaging showers with multiple iacts , the arrival direction of a shower is usually found using simple geometric arguments .
the major axes of the images produced by a shower in each iact camera intersect near the location of the arrival direction .
the shower arrival direction is calculated by minimizing the perpendicular distance to each image s semi - major axis , weighted by the size of each image .
this method , here called the standard method , is effective at small zenith angles .
however , at lza , the major axes of the air shower images from an individual gamma - ray event are generally close to parallel .
thus , the uncertainty of the intersecting point increases , resulting in a loss of angular resolution . due to this effect , a reconstruction technique that does not depend on the intersection of the axes is desirable for lza observations .
the displacement method is a direction reconstruction algorithm that is useful for lza observations @xcite . in very general terms
, it consists of calculating the arrival direction using the shape and brightness of a given air shower image .
more specifically , it relies on the determination of the _ disp _ parameter , defined as the angular distance from the image centroid to the arrival direction .
this method was used by several experiments in the past @xcite , with varying ways of calculating _
disp_. the basis of the displacement method is the relationship of the _ disp _ parameter to other image parameters @xcite .
the implementation of the algorithm in veritas is as follows : we estimate _ disp _ as a function of three other image parameters , _ size _ , _ length _ and _ width _
@xcite , using monte carlo simulated gamma - ray showers .
the method results in two different arrival directions , one on each side of a telescope image along the semi - major axis , also known as head - tail ambiguity @xcite .
this ambiguity is eliminated by choosing the cluster of arrival directions closest to one another , one coming from each image .
finally , the arrival direction is estimated independently for each telescope image and an average weighted by _
size _ is taken .
this method proves to be more powerful than the standard method @xcite when reconstructing events with zenith angle larger than @xmath9 .
quantitatively , an improvement of @xmath1430% in detection significance for a source having 1% of the strength of the crab nebula has been observed .
the cosmic ray background rate for iacts is typically more than @xmath15 times the gamma ray rate , so it is important to reduce this background while retaining as many gamma - ray events as possible . by exploiting the differences in the development of gamma ray and cosmic ray
induced showers , the background due to cosmic rays can be reduced significantly , while still retaining a high fraction of gamma - ray like events .
the background reduction is performed by placing standard selection criteria , optimized using monte carlo simulations and real data from the crab nebula on the shower image parameters .
the selection criteria for the size of the telescope images , the mean scaled width and mean scaled length parameters @xcite , the height of maximum cherenkov emission and the angular distance from the anticipated source location to the reconstructed arrival direction of each shower ( @xmath16 ) are given in table [ cuts ] . to perform a background subtraction of the surviving cosmic ray events , an estimation of these background counts
is made using the reflected - region background model @xcite .
events within an angular distance @xmath16 of the anticipated source location are considered on events .
background measurements ( off events ) are taken from regions of the same size and at the same angular distance from the center of the fov .
for this analysis , a minimum of eight background regions was used .
the excess number of events from the anticipated source location is found by subtracting the number of off events ( scaled by the relative exposure , @xmath17 ) from the on events .
statistical significances are calculated using a modified version of eqn .
17 of @xcite to allow for varying number of off - source regions due to nearby star 60 cygni @xcite .
more details about veritas , the calibration procedure and the analysis techniques can be found in @xcite .
analysis of the veritas data did not show a significant detection at the location of v407 cyg .
the results from both event reconstruction methods were used to calculate upper limits on the flux from v407 cyg with the method described by @xcite and the assumption of a gaussian - distributed background .
the upper limits for v407 cyg are calculated at the decorrelation energies of 1.8 tev for the standard method and 1.6 tev for the displacement method and assume that any emission takes the form of a power law with a photon index of -2.5 . the decorrelation energy is the energy at which the dependence of the upper limit calculation on the assumed photon index is minimized .
this energy is found by performing multiple upper limit calculations , with different spectral indices , and determining the region where the resulting upper limit functions intersect .
the energy threshold for the observations of v407 cygni with veritas , defined as the maximum of the product of the assumed spectral shape and the effective area , is 1.2 tev for both methods .
the analysis results for v407 cyg are presented in table [ results ] and figure [ fig : mapfermi ] .
in addition , results from observations of the crab nebula at similar zenith angles are presented in table [ results ] and figure [ fig : mapcrab ] .
the efficiency of the displacement method for event reconstruction can be observed in the increase in both gamma - ray rate and significance for the crab measurements .
the increased sensitivity also results in a reduction of the upper limit for v407 cyg compared to the standard method .
the gev detection of v407 cyg provides evidence for previously unobserved gamma - ray emission from novae in white dwarf / red giant systems . expanding shock waves
have been known to accelerate particles to high energies , and gamma rays are observed from supernova remnants .
the discovery by the _
fermi_-lat team , however , suggests that the same phenomenon occurs in some novae , adding a new class of gamma - ray emitting objects .
the lack of a significant detection in the vhe band suggests that either particles were not accelerated to sufficient energies to produce vhe photons during the v407 cyg outburst or that vhe photons were produced , but then absorbed .
the key to creating gamma rays is the acceleration of sufficiently energetic charged particles . in the case of v407 cyg ,
the expanding matter from the nova collides with the stellar wind from the red giant and causes a shock , which accelerates the particles near the shock to relativistic energies .
a rough estimation for the maximum energy attainable by first order fermi acceleration of a particle at a shock can be found , following the discussion of @xcite . if @xmath18 is the magnetic flux density where a shock proceeds and the shock travels with velocity @xmath19 , the maximum energy of a particle with charge @xmath20 is @xmath21 , where @xmath22 is the time allowed for particle acceleration .
this means that the highest attainable energy is proportional to the magnetic field in which the nova travels , the square of shock speed and the time for acceleration .
the mean magnetic field in the shock can be estimated as @xmath23^\frac{1}{2}$ ] , where @xmath24 is the density of gas molecules with respect to the distance from the center of the red giant and t is the wind temperature @xcite .
we assume a wind temperature of t=700k , corresponding to the temperature of the dust envelope measured by @xcite . @xcite
were able to directly measure the temperature of the red giant wind in a similar symbiotic system , eg and , and found that it can reach @xmath148000 k near the red giant . using
this wind temperature would increase the estimate for maximum particle energy by a factor of three .
@xcite carried out detailed hydrodynamic simulations of the v407 cyg nova with various gas distribution models and could accurately reproduce the x - ray light curve of v407 cygni .
their model for the distribution of gas that best reproduced the light curve included what they call a circumbinary density enhancement , " a region of density exceeding the typical @xmath25 profile of the stellar wind in the binary system , and had a binary separation of 15.5 au @xcite . for the temporal profile of the nova shock velocity
, we used the equation that @xcite found from fits to the broad components of the h@xmath17 spectra they measured beginning at day 2.3 after the outburst ( 13 march 2010 ) and thereafter : @xmath26 . for the velocity between day 0 and day 2 of the outburst
, we can assume two cases that bound the possible velocity profiles : i ) the nova shell experienced free expansion at a constant velocity before day 2 , with the assumption that the mass collected by the nova shell during this period was small ( free expansion model ) or ii ) extrapolate the above equation for the velocity to times before day 2 ( extrapolation model ) .
we then find that at the start of veritas observations of v407 cyg ( day 9 of the outburst ) , @xmath27 tev for the free expansion model and @xmath28 tev for the extrapolation model .
this suggests that particles could have been accelerated to tev energies by the time of the veritas observations . to check the importance of absorption
, we calculated the opacity along the photon path for gamma rays generated at the shock front .
electron - positron pair production via photon - photon collision is the dominant interaction @xcite .
we modeled the red giant spectrum as a blackbody with a temperature of 2500 k and a radius of 500 @xmath29 , found the photon density as a function of position and energy following @xcite , and used the cross section for the photon - photon collision from @xcite .
we found that the opacity for tev photons only becomes significant when the tev emission region is located directly behind the red giant with the system viewed edge on . though this case can not be ruled out , it is statistically unlikely .
in addition , if the suggested orbital parameters of @xcite are accurate , the system is unlikely to have been in such an orientation at the time of the nova .
the upper limits placed by veritas can put some restrictions on the gamma - ray emission mechanism in v407 cyg .
two physical models of gamma - ray production at the shock - front have been suggested @xcite . in the hadronic model , gamma rays
are produced in the decay of @xmath30 particles generated by collisions of high energy protons accelerated in the shock . in the leptonic model , gamma rays
are produced via inverse - compton scattering of infrared photons emitted from the red giant on high energy electrons accelerated in the shock .
the electron threshold energy for production of a gamma - ray photon via inverse - compton scattering off the red giant photons can be estimated as : @xmath31 , where @xmath32 is the electron threshold energy , @xmath33 is the gamma - ray energy and @xmath34 is the energy of the red giant photons .
the electron threshold energy for a 1 tev gamma ray scattering off 0.6 ev photons at the peak of the red giant spectrum is 1.1 tev .
though the above calculation indicates that particles could reach tev energies if continuously accelerated for the full nine days from the initial outburst to the start of the veritas observations , the inverse - compton cooling time would be significantly less than a day ( the time estimated by @xcite for 5 gev electrons ) , meaning electrons that are accelerated in the first few days of the outburst would not likely retain sufficient energy to produce vhe photons by the time of the veritas observations .
vhe emission near the time of the veritas observations would therefore require freshly accelerated particles , however , recently accelerated particles would likely not have enough time to reach tev energies .
if electrons did reach tev energies , they would be approaching the klein - nishina regime , where the cross section for inverse - compton scattering would be reduced , resulting in a longer cooling time .
however , electrons with tev energies would be well above the exponential cutoff , @xmath35 gev , of the electron spectrum in the best - fit leptonic model of @xcite .
these two factors imply that the veritas upper limits place no new constraints on leptonic models .
for hadronic models , the _ fermi_-lat data alone provide relatively poor constraints on the extension of the proton spectrum to high energies .
in particular , the cutoff energy @xmath36 is not well bounded from above if the spectral index is steep .
the veritas data can be used to improve the constraints on the hadronic model parameters . to do so
, we re - fit the hadronic model used by @xcite to the _
fermi_-lat points with the addition of the veritas upper limit .
figure [ fig : spectrum ] shows the _
fermi_-lat data @xcite and the veritas flux upper limit compared to the best fitting hadronic model .
the gamma - ray spectrum is calculated via the method of @xcite , assuming a cosmic proton spectrum of the following form : @xmath37 ( protons gev@xmath38 ) , where ( @xmath39 are kinetic energy , cut - off energy , and mass of the proton and @xmath40 is the power law index ) .
figure [ fig : revisedmap ] shows a confidence region map for the parameters of the hadronic model using both the _ fermi_-lat and veritas data .
the gamma - ray spectrum was modeled as described above and fitted to the _
fermi_-lat data by varying @xmath40 and @xmath41 .
the spectrum was then compared to the veritas upper limit , and a @xmath42 value for the veritas data point alone was calculated . specifically , we calculated the model flux in the veritas energy band and compared this to the flux upper limit determined via the displacement method .
this @xmath42 value was then added to the @xmath42 calculated for the _
fermi_-lat data .
the confidence levels were then calculated for the two parameters of interest , @xmath40 and @xmath41 .
as can be seen from figure [ fig : revisedmap ] , the veritas observations place greater restrictions on the model proton spectral index for high cutoff energies .
the 90% confidence limits are @xmath43 5 tev ( comparable to @xmath44 tev , calculated above for the extrapolation model ) and @xmath45 0.01 tev ( much lower than @xmath46 ) .
it is possible that the peak energy of the particles produced by the shock could be reduced if the magnetic field is weaker than estimated above .
@xcite argue for a larger binary separation of 2025 au , based on the presence of lithium burning in the mira , and evidence that the white dwarf in the system is massive . using this larger separation distance
would lead to weaker magnetic fields .
the limits placed by the veritas observations are near the threshold for the observations , so it is also possible that simply not enough particles were accelerated to high enough energies to produce a significant detection by veritas .
_ acknowledgments .
_ we would like to thank pierre jean of the _ fermi_-lat team for providing the contour data from the _
fermi_-lat results and useful discussion .
this research is supported by grants from the u.s .
department of energy office of science , the u.s . national science foundation and the smithsonian institution , by nserc in canada , by science foundation ireland ( sfi 10/rfp / ast2748 ) and by stfc in the u.k .
we acknowledge the excellent work of the technical support staff at the fred lawrence whipple observatory and at the collaborating institutions in the construction and operation of the instrument .
abdo , a. a. , et al .
, 2010 , science , 329 , 817 acciari , v. a. , et al .
( veritas collaboration ) , 2008 , apj , 679 , 1427 aharonian , f. , et al .
, 2004 , a&a , 421 , 529 aharonian , f. , et al .
( hegra collaboration ) 2001 , a&a , 370 , 112 bassani , l. et al . , 2010 , atel .
@xmath472498 bassani , l. et al . , 2010 ,
@xmath472506 bode , m. et al .
2006 , apj , 652 , 629 elik , . , 2008 , phd thesis at the university of california , los angeles chaty , s. , 2010 , arxiv:1005.0100 cheung , c. c. et al . , 2010 ,
@xmath472487 cogan , p. , et al .
, 2007 , in proc .
30th international cosmic ray conference , mrida , 3 , 1385 [ arxiv:0709.4233 ] daum , a. et al .
, 1997 , astrop .
8 , 1 domingo - santamara , e. et al . ,
2005 , icrc proceedings , 5 , 363 - 366 espey , b. r. & crowley , c. , 2008 , rs ophiuchi ( 2006 ) and the recurrent nova phenomenon , asp conference series , 401 fomin , v. p. , et al . , 1994 , astropart .
, 2 , 137 gould , r. j. & schrder , g. p. , 1966
letters 16 , 252 hays , e. a. , & fermi lat collaboration , 2010 , bulletin of the american astronomical society , 41 , 670 helene , o. , 1983 _ nucl .
instr . and meth .
_ 212 , 319 helene , o. , 1990 _ nucl .
instr . and meth .
_ 300 , 132 hillas , a. m. , 1985 _ 19th int .
cosmic ray conf . , la jolla _ 3 , 445 hillas , a. m. , 1996 _ space science rev _ 75 , 17 hofmann , w. et al .
, 1999 , astropart .
122 , 135 - 143 kamae , t. , karlsson , n. , mizuno , t. , abe , t. & koi , t. , 2006 , apj , 647 , 692 kertzman , m. , sembroski , g. m. , 1994 _ nucl .
instr . and meth .
, _ 343 , 629 kolotilov , e. a. et al . , 1998 , astronomy letters , 24 , 451 kranich ,
d. & stark , l. s. , for the hegra collaboration , 2003 , icrc proceedings krawczynski , h. , carter - lewis , d. a. , duke , c. , holder , j. , maier , g. , le bohec , s. , & sembroski , g. , 2006 , astropart .
physics , 25 , 380 lessard , r. w. et al .
, 2001 , astropart .
15 , 1 longair , m. s. , 2011 , _ high energy astrophysics _ ( _ third edition _ ) , cambridge , cambridge , uk : cambridge university press li , t .-
, & ma , y .- q . , 1983 , apj , 272 , l317 munari , u. , margoni , r. , & stagni , r. , 1990 , mnras , 242 , 653 munari , u. et al . , 2011 , mnras , 410 , l52 nelson , t. et al . , 2012 , apj , 748 , 43 k. nishiyama , f. kabashima , international astronomical union central bureau for astronomical telegrams , reported by h. maehara , no .
2199 ( 2010 ) .
unez , p. d. et al .
, 2011 , apj , 731 , 105 orlando , s. & drake , j. j. 2011 , mon . not .
r. astron .
soc . 000 , 1 - 10 rolke , w. a. , lpez , a. m. , & conrad , j. , 2005 , _ nucl . instr . and meth .
_ , 551 , 493 entrk , g. d. et al . , 2011 , arxiv:1109.6044 sokoloski , j. l. , et al . , 2006 ,
nature 442 , 276 sabatini , s. , et al . , 2010 , apj 712 , l10 tatischeff , v. , & hernanz , m. , 2007 , apj , 663 , l101 tavani , m. , mukherjee , r. et al .
, 1997 , apj 479 , l109 vandenbrouke , j. , et al . , 2010 , apj 718 , l166 statistical error , and arrows indicate 2@xmath48 upper limit .
the rightmost arrows show the 99% confidence level ( 3@xmath48 ) veritas upper limit calculated using the displacement method ( at 1.6 tev ) and the standard method ( at 1.8 tev ) for event reconstruction ( see section [ sec : results ] ) .
the fitting curve was constructed with the method of @xcite with the parameters ( @xmath40 , log@xmath36)=(2.15 , 1.5 ) ( see section [ sec : discussion ] ) . ] | we report on very high energy ( e @xmath0 100 gev ) gamma - ray observations of v407 cygni , a symbiotic binary that underwent a nova outburst producing 0.110 gev gamma rays during 2010 march 1026 .
observations were made with the very energetic radiation imaging telescope array system during 2010 march 1926 at relatively large zenith angles , due to the position of v407 cyg .
an improved reconstruction technique for large zenith angle observations is presented and used to analyze the data .
we do not detect v407 cygni and place a differential upper limit on the flux at 1.6 tev of @xmath1 ( at the 95% confidence level ) .
when considered jointly with data from _ fermi_-lat , this result places limits on the acceleration of very high energy particles in the nova . |
there have been much interest in the study of spin models with orbital degeneracy @xcite due to experimental progress related to many transition - metal and rare - earth compounds such as lamno@xmath1 , ceb@xmath2 , and perovskite lattice , as in kcuf@xmath1@xcite .
those systems involve orbital degree of freedom in addition to spin ones .
almost three decades ago , kugel and khomskii@xcite had pointed out the possibility of orbital excitations in these systems . as a model system
, it exhibits some fascinating physical features which is lack without orbital degree of freedom . the isotropic case of spin system with orbital degeneracy
was shown to have an enlarged su(4 ) symmetry @xcite , and one dimensional model is known to exactly solvable @xcite .
materials related to spin - orbital systems in one dimension include quasi - one - dimensional tetrahis - dimethylamino - ethylene ( tdae)-c@xmath3@xcite , artificial quantum dot arrays@xcite , and degenerate chains in na@xmath4ti@xmath4sb@xmath4o and na@xmath4v@xmath4o@xmath5 compounds @xcite .
it is therefore worthwhile to systematically study the features of one dimensional model .
theoretical studies@xcite has found the strong interplay of orbital and spin degrees of freedom in the excitations spectra .
it has been noticed that the presence of orbital may results in various interesting magnetic properties . applying a conventional magnetic field ,
the spin orbital chain with su(4 ) symmetry is shown to reduce to a model with orbital su(2 ) symmetry@xcite in the ground state .
recently , we showed that the magnetization process becomes more complicated if taking account of the contribution from orbital sector @xcite .
we have explained that the competition between spin and orbital degree of freedom leads to an orbital anti - polarization phase .
however , the external field we introduced in ref@xcite is not the most general one for su(4 ) systems . in this paper
, we study a su(4 ) spin - orbital chain in the presence of a generalized external field on the basis of its bethe ansatz solution .
our paper is organized as follows . in sec .
[ sec : betheansatz ] , we introduce the bethe - ansatz solution and the zeeman like term which is going to be added to the original su(4 ) hamiltonian . in sec . [
sec : gsd ] , we give some useful remarks on the quantum number configurations for the ground state in the presence of external field that is characterized by three parameters .
we also simply demonstrate the thermodynamic limit of the bethe - ansatz equation and briefly present the dress energy description of ground state in the presence of external field . in sec .
[ sec : su3preserved ] , we study the magnetization properties of a hamiltonian in the regime with one - parameter symmetry breaking . in sec .
[ sec : su2preserved ] , we study both magnetization and the phase diagram in the regimes with two - parameter symmetry breaking . various phases and the quantum phase transitions
are obtained by both numerical calculation and analytical formulation . concerning to various phases we present detailed explanation in terms of group theory .
[ sec : discussion ] gives a brief summery .
we start from the following hamiltonian @xmath6 .
\label{eq : hamiltonian}\ ] ] where @xmath7 and @xmath8 denote respectively spin and orbital operators at site @xmath9 , both are generators of su(2 ) group characterizing the spin and orbital degree of freedom of outer shell electrons in some transitional metal oxides at the insulating regime .
the coupling constant is set to unit for simplicity .
it has been pointed that the above hamiltonian possesses an enlarged su(4 ) symmetry@xcite rather than su(2)@xmath10su(2 ) symmetry .
the four states that carry out the fundamental representation of su(4 ) group is denoted by @xmath11 these bases are labelled by the eigenvalues of @xmath12 and @xmath13 , _ i.e. _ , @xmath14 . as the @xmath15(4 ) lie algebra is of rank 3 , there exists third generator @xmath16 which possesses simultaneous eigenvalue together with @xmath12 and @xmath13 . for convenience
, we denote this new generator by @xmath17 hereafter . in the terminology of group theory
, however , the quadruplet can also be labelled by the weight vectors which is defined by eigenvalues of @xmath18 that constitute the cartan subalgebra of @xmath19 lie algebra , here we adopt the chevalley basis because the physical quantities can be conveniently expressed in this basis .
the eigenvalues of @xmath20 as well as that of @xmath21 are given in table i , the relation between these two basis reads@xcite , @xmath22 [ table : dgfgf ] .the eigenvalue of @xmath12 , @xmath13 , @xmath17 and @xmath23-component of @xmath24 for the four basis states ( eq . [ eq : basis ] ) . [ cols="^,^,^,^,^,^,^ " , ] the present model ( [ eq : hamiltonian ] ) has been solved by bethe - ansatz method@xcite , its energy spectrum is given by @xmath25 where the @xmath26 s are solutions of the following coupled transcendental equations @xmath27 where @xmath28 .
the @xmath29 and @xmath30 are rapidities related to the three generators of the cartan subalgebra of the @xmath15(4 ) lie algebra .
the quantum numbers @xmath31 specify a state in which there are @xmath32 number of sites in the state @xmath33 , @xmath34 in @xmath35 , @xmath36 in @xmath37 , and @xmath38 in @xmath39 .
hence the @xmath23-component of total spin , orbital and @xmath17 are obtained as @xmath40 , @xmath41 , and @xmath42 . in present su(4 ) model
, a three - parameter external field @xmath43 can be introduced to write out a most general zeeman - like energy .
@xmath44 for more clear physics implication , we re - choose the parameters to write the effective magnetization @xmath45 , @xmath46 where @xmath47 , @xmath48 , @xmath49 are generalized land @xmath0 factors for @xmath12 , @xmath13 and @xmath17 respectively .
( [ eq : magddgdg ] ) can be expressed in terms of the number of rapidities , @xmath50 because the zeeman - like term commutes with the su(4 ) hamiltonian ( [ eq : hamiltonian ] ) , the energy spectrum in the presence of external field are simply related to the energy spectrum in the absence of external field , @xmath51 where @xmath52 is determined by eqs .
( [ eq : bae ] ) .
obviously , the application of the external field with different magnitude just brings about various level crossings . in terms of @xmath53 ,
the magnetization ( [ eq : magddgdg ] ) becomes @xmath54 which breaks su(4 ) symmetry down to various lower symmetries depending on the distinct regions in the parameter space .
based on the bethe ansatz solution of the model , we first give the quantum number description of the ground state , which is useful for numerical method .
we also give the dress energy description for the ground state and propose the conditions to determine quantum phase transitions , which is useful for analytic study .
it has been known @xcite that the ground state of the hamiltonian ( [ eq : hamiltonian ] ) is a su(4 ) singlet for the case of @xmath55 .
the configuration of the quantum number for ground state @xmath31 ( @xmath56 ; @xmath57 ; @xmath58 ) are consecutive integers ( or half integers ) arranging symmetrically around the zero . in the presence of magnetic field
, however , the zeeman term brings about level crossings and the state with @xmath59 is no longer the ground state .
therefore the numbers @xmath60 for the lowest energy state are related to the magnitude of the applied external field . in order to solve the bethe ansatz equation numerically ,
we need to determine the possible configuration of quantum numbers for give values @xmath61 , @xmath62 and @xmath38 . the property of young tableau requires that max@xmath63 , max@xmath64 , max@xmath65 , and @xmath66 for a given @xmath67 .
then one is able to analyze the change of energy level for each state , which determines the true ground state for a given external field .
one can also calculates the magnetization by eq .
( [ eq : mag ] ) . in the thermodynamic limit
, the energy ( [ eq : tenergy ] ) is expressed in terms of densities of the rapidities , @xmath68d\lambda \nonumber \\ & & + ( g_s - g_t)h\int_{-\mu_0}^{\mu_0 } \omega(\mu)d\mu \nonumber\\ & & + ( g_t - g_u)h\int_{-\nu_0}^{\nu_0}\tau(\nu)d\nu \label{eq : energy}\end{aligned}\ ] ] these densities satisfy the following coupled integral equations @xmath69 where @xmath70 and @xmath71 and @xmath72 are determined by @xmath73 it is more convenient to introduce dress energy@xcite .
the iteration of eq.([eq : density ] ) gives rise to @xmath74 where @xmath75 , and @xmath76 are the dress energies in @xmath26 , @xmath77 and @xmath30 sectors respectively .
it is worthwhile to point out that the dress energy is also the thermal potentials at zero temperature , i.e. , @xmath78 , @xmath79 , and @xmath80 .
the thermal bethe - ansatz equation in the zero - temperature limit turns to eq.([eq : thermodeq ] ) . in terms of dress energy , the energy ( [ eq : energy ] )
is simplified to @xmath81 apparently , the ground state is a quasi - dirac sea where the states of negative dress energy , @xmath82 , @xmath83 , @xmath84 , are fully occupied .
the fermi points of the three rapidities are determined by @xmath85 the system will be magnetized if the applied field enhance the dress energy , because it makes the corresponding fermi points decline .
the quantum phase transition occurs when any of the fermi points shrinks to zero . as a result ,
the critical values of the external field are solved by @xmath86 these conditions together with eq.([eq : thermodeq ] ) enable one to calculate those critical values .
the application of external field makes the su(4 ) symmetry break down to various regimes with different residual symmetry . in this section , we shall discuss the simplest cases of single parameter hierarchy .
there are three special directions in the weight space of su(4 ) . if the external field is supplied along those directions , i.e. , either @xmath87 , @xmath88 and @xmath89 in eq .
( [ eq : zeemdgkgds ] ) is not vanished , a partial breaking of a su(2 ) to u(1 ) will take place .
let us consider them respectively .
if @xmath90 , the zeeman interaction ( [ eq : mago3 ] ) becomes @xmath91 the occurrence of operator @xmath92 makes the hamiltonian noncommutable with @xmath93 .
thus a su(2 ) subgroup generated by @xmath92 , @xmath94 @xmath95 is broken down to u(1 ) . analyzing the level crossing from eq .
( [ eq : tenergy ] ) , we shown the magnetization curve in fig .
[ suheis_figure_magn67 ] . because the terms @xmath96 and @xmath97 in the first two equations of eqs .
( [ eq : thermodeq ] ) are non - positive , the external field can not enhance the two dress energies @xmath98 and @xmath99 , the fermi points in both @xmath26 and @xmath77 sectors are fixed . on the contrary
, the zeeman term has a positive contribution to @xmath100 , and its two fermi points will decline when the external field increases . although it is a su(4 ) singlet labelled by young tableau @xmath101 $ ] in the absence of external field , the ground state possess a residual su(3)@xmath10u(1 ) symmetry in the presence of the aforementioned one parameter external field at small magnitude , which corresponds to phase iv labelled by four - row young tableau . in this regime
are there still three type of rapidities that solve the bethe - ansatz equation .
the u(1 ) is generated by @xmath92 , while the su(3 ) is generated by the following eight operators @xmath102 there exists a critical field when those two fermi points shrink to zero , the rapidity @xmath30 disappears in the bethe - ansatz equation .
thus a quantum phase transition occurs at the critical field which separate two phases , we call phase iv and phase iii . the magnetization process can be clearly illustrated by the evolution of young tableau , ( 18,5)(3,0 ) ( -1 , 3.35)(0,1)0.7 ( -1.2,3.35)(1,0)0.4 ( -1 , 2.15)(0,1)0.7 ( -1.2,2.85)(1,0)0.4 ( -1 , 1.85)(0,-1)0.7(-1.2,1.15)(1,0)0.4 ( -1 , 0.65)(0,-1)0.7 ( -1.2,0.65)(1,0)0.4 ( 0,1)(1,0)1 ( 0,2)(1,0)1 ( 0,3)(1,0)1 ( 0,0)(1,0)1 ( 3,0)(1,0)1 ( 0,4)(1,0)1 ( 3,4)(1,0)1 ( 3,1)(1,0)1 ( 3,2)(1,0)1 ( 3,3)(1,0)1 ( 1.1,0)(1,0)0.1 ( 1.4,0)(1,0)0.1 ( 1.7,0)(1,0)0.1 ( 2.0,0)(1,0)0.1 ( 2.3,0)(1,0)0.1 ( 2.6,0)(1,0)0.1 ( 2.9,0)(1,0)0.1 ( 1.1,1)(1,0)0.1 ( 1.4,1)(1,0)0.1 ( 1.7,1)(1,0)0.1 ( 2.0,1)(1,0)0.1 ( 2.3,1)(1,0)0.1 ( 2.6,1)(1,0)0.1 ( 2.9,1)(1,0)0.1 ( 1.1,2)(1,0)0.1 ( 1.4,2)(1,0)0.1 ( 1.7,2)(1,0)0.1 ( 2.0,2)(1,0)0.1 ( 2.3,2)(1,0)0.1 ( 2.6,2)(1,0)0.1 ( 2.9,2)(1,0)0.1 ( 1.1,3)(1,0)0.1 ( 1.4,3)(1,0)0.1 ( 1.7,3)(1,0)0.1 ( 2.0,3)(1,0)0.1 ( 2.3,3)(1,0)0.1 ( 2.6,3)(1,0)0.1 ( 2.9,3)(1,0)0.1 ( 1.1,4)(1,0)0.1 ( 1.4,4)(1,0)0.1 ( 1.7,4)(1,0)0.1 ( 2.0,4)(1,0)0.1 ( 2.3,4)(1,0)0.1 ( 2.6,4)(1,0)0.1 ( 2.9,4)(1,0)0.1 ( 3,1)(1,0)1 ( 3,2)(1,0)1 ( 3,3)(1,0)1 ( 4,0)(0,1)4 ( 3,0)(0,1)4 ( 1,0)(0,1)4 ( 0,0)(0,1)4 ( 4.5,3.5)(1,0)2 ( 7,0)(0,1)4 ( 8,0)(0,1)4 ( 9.5,0)(0,1)4 ( 10.5,0)(0,1)4 ( 11.5,1)(0,1)3 ( 13,1)(0,1)3 ( 14,1)(0,1)3 ( 7,0)(1,0)1 ( 9.5,0)(1,0)1 ( 9.5,1)(1,0)2 ( 13,1)(1,0)1 ( 7,4)(1,0)1 ( 9.5,4)(1,0)2 ( 13,4)(1,0)1 ( 7,1)(1,0)1 ( 7,2)(1,0)1 ( 7,3)(1,0)1 ( 9.5,2)(1,0)2 ( 9.5,3)(1,0)2 ( 13,2)(1,0)1 ( 13,3)(1,0)1 ( 8.1,4)(1,0)0.1 ( 8.4,4)(1,0)0.1 ( 8.7,4)(1,0)0.1 ( 9.0,4)(1,0)0.1 ( 9.3,4)(1,0)0.1 ( 8.1,3)(1,0)0.1 ( 8.4,3)(1,0)0.1 ( 8.7,3)(1,0)0.1 ( 9.0,3)(1,0)0.1 ( 9.3,3)(1,0)0.1 ( 8.1,2)(1,0)0.1 ( 8.4,2)(1,0)0.1 ( 8.7,2)(1,0)0.1 ( 9.0,2)(1,0)0.1 ( 9.3,2)(1,0)0.1 ( 8.1,1)(1,0)0.1 ( 8.4,1)(1,0)0.1 ( 8.7,1)(1,0)0.1 ( 9.0,1)(1,0)0.1 ( 9.3,1)(1,0)0.1 ( 8.1,0)(1,0)0.1 ( 8.4,0)(1,0)0.1 ( 8.7,0)(1,0)0.1 ( 9.0,0)(1,0)0.1 ( 9.3,0)(1,0)0.1 ( 11.6,4)(1,0)0.1 ( 11.9,4)(1,0)0.1 ( 12.2,4)(1,0)0.1 ( 12.5,4)(1,0)0.1 ( 12.8,4)(1,0)0.1 ( 11.6,3)(1,0)0.1 ( 11.9,3)(1,0)0.1 ( 12.2,3)(1,0)0.1 ( 12.5,3)(1,0)0.1 ( 12.8,3)(1,0)0.1 ( 11.6,2)(1,0)0.1 ( 11.9,2)(1,0)0.1 ( 12.2,2)(1,0)0.1 ( 12.5,2)(1,0)0.1 ( 12.8,2)(1,0)0.1 ( 11.6,1)(1,0)0.1 ( 11.9,1)(1,0)0.1 ( 12.2,1)(1,0)0.1 ( 12.5,1)(1,0)0.1 ( 12.8,1)(1,0)0.1 ( 14.5,3.5)(1,0)2 ( 17,1)(0,1)3(18,1)(0,1)3 ( 20.5,1)(0,1)3 ( 21.5,1)(0,1)3 ( 17,1)(1,0)1 ( 20.5,1)(1,0)1 ( 17,4)(1,0)1 ( 20.5,4)(1,0)1 ( 17,2)(1,0)1 ( 17,3)(1,0)1 ( 20.5,2)(1,0)1 ( 20.5,3)(1,0)1 ( 18.1,4)(1,0)0.1 ( 18.4,4)(1,0)0.1 ( 18.7,4)(1,0)0.1 ( 19.0,4)(1,0)0.1 ( 19.3,4)(1,0)0.1 ( 19.6,4)(1,0)0.1 ( 19.9,4)(1,0)0.1 ( 20.2,4)(1,0)0.1 ( 18.1,3)(1,0)0.1 ( 18.4,3)(1,0)0.1 ( 18.7,3)(1,0)0.1 ( 19.0,3)(1,0)0.1 ( 19.3,3)(1,0)0.1 ( 19.6,3)(1,0)0.1 ( 19.9,3)(1,0)0.1 ( 20.2,3)(1,0)0.1 ( 18.1,2)(1,0)0.1 ( 18.4,2)(1,0)0.1 ( 18.7,2)(1,0)0.1 ( 19.0,2)(1,0)0.1 ( 19.3,2)(1,0)0.1 ( 19.6,2)(1,0)0.1 ( 19.9,2)(1,0)0.1 ( 20.2,2)(1,0)0.1 ( 18.1,1)(1,0)0.1 ( 18.4,1)(1,0)0.1 ( 18.7,1)(1,0)0.1 ( 19.0,1)(1,0)0.1 ( 19.3,1)(1,0)0.1 ( 19.6,1)(1,0)0.1 ( 19.9,1)(1,0)0.1 ( 20.2,1)(1,0)0.1 which shows the evolution of young tableau in magnetization process ,
i.e. , from su(4 ) singlet to su(3)@xmath10u(1 ) states and then to a su(3 ) singlet .
physically , we have @xmath103 at zero external field due to there are a quart of the total sites being respectively in the states @xmath33 , @xmath35 , @xmath104 and @xmath39 .
turning on the external field leads to the spin - orbital flipping , @xmath105 , @xmath106 , which result in non - vanishing magnetization .
the su(3)@xmath10u(1 ) symmetry makes the above three flipping processes favor to occur simultaneously .
when the external field succeeds a critical value , it goes into phase iii where all the states @xmath39 have been flipped over . in this phase the @xmath23-component of total spin and total orbital keep positive constant @xmath107 while that of @xmath17 keeps a negative constant @xmath108 .
consequently , the magnetization reaches a saturate value @xmath109 and the ground state becomes the su(3 ) singlet regardless of the magnitude of the external field in phase iii . applying the external field along the direction of the second simple root of @xmath19
lie algebra , we will have a symmetry breaking from su(4 ) to su(2)@xmath10u(1)@xmath10su(2 ) for ground state .
this is realized by the choice of land g factors @xmath110 , which makes the magnetization to be @xmath111 those two su(2 ) are generated respectively by @xmath112 as @xmath97 in the second equation of ( [ eq : thermodeq ] ) is positive but both @xmath96 and @xmath113 in the first and third equation are negative , the external field makes the fermi points in @xmath77 sector to shrink .
the critical value of the external field when quantum phase transition occurs is determined by @xmath114 .
this critical point separates two different phases , we call phase iv and phase ii . the magnetization curve is shown in fig .
[ suheis_figure_magn67 ] .
it can also be illustrated by the evolution of the young tableau , ( 18,5)(3,0 ) ( -1 , 3.35)(0,1)0.7 ( -1.2,3.35)(1,0)0.4 ( -1 , 2.15)(0,1)0.7 ( -1.2,2.85)(1,0)0.4 ( -1 , 1.85)(0,-1)0.7(-1.2,1.15)(1,0)0.4 ( -1 , 0.65)(0,-1)0.7 ( -1.2,0.65)(1,0)0.4 ( 0,1)(1,0)1 ( 0,2)(1,0)1 ( 0,3)(1,0)1 ( 0,0)(1,0)1 ( 3,0)(1,0)1 ( 0,4)(1,0)1 ( 3,4)(1,0)1 ( 3,1)(1,0)1 ( 3,2)(1,0)1 ( 3,3)(1,0)1 ( 1.1,0)(1,0)0.1 ( 1.4,0)(1,0)0.1 ( 1.7,0)(1,0)0.1 ( 2.0,0)(1,0)0.1 ( 2.3,0)(1,0)0.1 ( 2.6,0)(1,0)0.1 ( 2.9,0)(1,0)0.1 ( 1.1,1)(1,0)0.1 ( 1.4,1)(1,0)0.1 ( 1.7,1)(1,0)0.1 ( 2.0,1)(1,0)0.1 ( 2.3,1)(1,0)0.1 ( 2.6,1)(1,0)0.1 ( 2.9,1)(1,0)0.1 ( 1.1,2)(1,0)0.1 ( 1.4,2)(1,0)0.1 ( 1.7,2)(1,0)0.1 ( 2.0,2)(1,0)0.1 ( 2.3,2)(1,0)0.1 ( 2.6,2)(1,0)0.1 ( 2.9,2)(1,0)0.1 ( 1.1,3)(1,0)0.1 ( 1.4,3)(1,0)0.1 ( 1.7,3)(1,0)0.1 ( 2.0,3)(1,0)0.1 ( 2.3,3)(1,0)0.1 ( 2.6,3)(1,0)0.1 ( 2.9,3)(1,0)0.1 ( 1.1,4)(1,0)0.1 ( 1.4,4)(1,0)0.1 ( 1.7,4)(1,0)0.1 ( 2.0,4)(1,0)0.1 ( 2.3,4)(1,0)0.1 ( 2.6,4)(1,0)0.1 ( 2.9,4)(1,0)0.1 ( 3,1)(1,0)1 ( 3,2)(1,0)1 ( 3,3)(1,0)1 ( 4,0)(0,1)4 ( 3,0)(0,1)4 ( 1,0)(0,1)4 ( 0,0)(0,1)4 ( 4.5,3.5)(1,0)2 ( 7,0)(0,1)4 ( 8,0)(0,1)4 ( 9.5,0)(0,1)4 ( 10.5,0)(0,1)4 ( 11.5,2)(0,1)2 ( 13,2)(0,1)2 ( 14,2)(0,1)2 ( 7,0)(1,0)1 ( 9.5,0)(1,0)1 ( 9.5,1)(1,0)1 ( 7,4)(1,0)1 ( 9.5,4)(1,0)2 ( 13,4)(1,0)1 ( 7,1)(1,0)1 ( 7,2)(1,0)1 ( 7,3)(1,0)1 ( 9.5,2)(1,0)2 ( 9.5,3)(1,0)2 ( 13,2)(1,0)1 ( 13,3)(1,0)1 ( 8.1,4)(1,0)0.1 ( 8.4,4)(1,0)0.1 ( 8.7,4)(1,0)0.1 ( 9.0,4)(1,0)0.1 ( 9.3,4)(1,0)0.1 ( 8.1,3)(1,0)0.1 ( 8.4,3)(1,0)0.1 ( 8.7,3)(1,0)0.1 ( 9.0,3)(1,0)0.1 ( 9.3,3)(1,0)0.1 ( 8.1,2)(1,0)0.1 ( 8.4,2)(1,0)0.1 ( 8.7,2)(1,0)0.1 ( 9.0,2)(1,0)0.1 ( 9.3,2)(1,0)0.1 ( 8.1,1)(1,0)0.1 ( 8.4,1)(1,0)0.1 ( 8.7,1)(1,0)0.1 ( 9.0,1)(1,0)0.1 ( 9.3,1)(1,0)0.1 ( 8.1,0)(1,0)0.1 ( 8.4,0)(1,0)0.1 ( 8.7,0)(1,0)0.1 ( 9.0,0)(1,0)0.1 ( 9.3,0)(1,0)0.1 ( 11.6,4)(1,0)0.1 ( 11.9,4)(1,0)0.1 ( 12.2,4)(1,0)0.1 ( 12.5,4)(1,0)0.1 ( 12.8,4)(1,0)0.1 ( 11.6,3)(1,0)0.1 ( 11.9,3)(1,0)0.1 ( 12.2,3)(1,0)0.1 ( 12.5,3)(1,0)0.1 ( 12.8,3)(1,0)0.1 ( 11.6,2)(1,0)0.1 ( 11.9,2)(1,0)0.1 ( 12.2,2)(1,0)0.1 ( 12.5,2)(1,0)0.1 ( 12.8,2)(1,0)0.1 ( 14.5,3.5)(1,0)2 ( 17,2)(0,1)2(18,2)(0,1)2 ( 20.5,2)(0,1)2 ( 21.5,2)(0,1)2 ( 17,4)(1,0)1 ( 20.5,4)(1,0)1 ( 17,2)(1,0)1 ( 17,3)(1,0)1 ( 20.5,2)(1,0)1 ( 20.5,3)(1,0)1 ( 18.1,4)(1,0)0.1 ( 18.4,4)(1,0)0.1 ( 18.7,4)(1,0)0.1 ( 19.0,4)(1,0)0.1 ( 19.3,4)(1,0)0.1 ( 19.6,4)(1,0)0.1 ( 19.9,4)(1,0)0.1 ( 20.2,4)(1,0)0.1 ( 18.1,3)(1,0)0.1 ( 18.4,3)(1,0)0.1 ( 18.7,3)(1,0)0.1 ( 19.0,3)(1,0)0.1 ( 19.3,3)(1,0)0.1 ( 19.6,3)(1,0)0.1 ( 19.9,3)(1,0)0.1 ( 20.2,3)(1,0)0.1 ( 18.1,2)(1,0)0.1 ( 18.4,2)(1,0)0.1 ( 18.7,2)(1,0)0.1 ( 19.0,2)(1,0)0.1 ( 19.3,2)(1,0)0.1 ( 19.6,2)(1,0)0.1 ( 19.9,2)(1,0)0.1 ( 20.2,2)(1,0)0.1 the spin - orbital flipping process caused by the applied external field has several characteristics . during the flipping process , @xmath104 and @xmath115 flips simultaneously into @xmath35 and @xmath116 pairs , which makes four - row young tableau reduce to the two - row young tableau when across the critical field .
apparently , the eigenvalues of both @xmath13 and @xmath17 do not change during the magnetization process .
only @xmath12 contributes to the magnetization @xmath117 .
the total spin is completely polarized ( i.e. , @xmath118 is saturated ) once the phase iv transits to phase ii .
the phase that yamashita _ et al _ @xcite discussed agrees with this special case . if @xmath119 and @xmath120 , the magnetization becomes @xmath121 which implies that the external field was applied along the first simple root of @xmath19 lie algebra .
this gives rise to a symmetry breaking down to u(1)@xmath10su(3 ) for ground state .
for the sake of saving space , we omitted the operators that generate those symmetry .
because this parameter choice implies that @xmath96 in @xmath26 sector positive but the @xmath0 factor terms in both @xmath77 and @xmath30 sectors are non - positive , the quantum phase transition is only related to @xmath26 sector .
the critical value is determined by @xmath122 .
this critical point separates the system into two phases , phase iv and phase i. the magnetization process is shown in fig.[suheis_figure_magn67 ] .
it is helpful to illustrated this process by the evolution of young tableau , ( 18,5)(3,0 ) ( -1 , 3.35)(0,1)0.7 ( -1.2,3.35)(1,0)0.4 ( -1 , 2.15)(0,1)0.7 ( -1.2,2.85)(1,0)0.4 ( -1 , 1.85)(0,-1)0.7(-1.2,1.15)(1,0)0.4 ( -1 , 0.65)(0,-1)0.7 ( -1.2,0.65)(1,0)0.4 ( 0,1)(1,0)1 ( 0,2)(1,0)1 ( 0,3)(1,0)1 ( 0,0)(1,0)1 ( 3,0)(1,0)1 ( 0,4)(1,0)1 ( 3,4)(1,0)1 ( 3,1)(1,0)1 ( 3,2)(1,0)1 ( 3,3)(1,0)1 ( 1.1,0)(1,0)0.1 ( 1.4,0)(1,0)0.1 ( 1.7,0)(1,0)0.1 ( 2.0,0)(1,0)0.1 ( 2.3,0)(1,0)0.1 ( 2.6,0)(1,0)0.1 ( 2.9,1)(1,0)0.1 ( 1.1,1)(1,0)0.1 ( 1.4,1)(1,0)0.1 ( 1.7,1)(1,0)0.1 ( 2.0,1)(1,0)0.1 ( 2.3,1)(1,0)0.1 ( 2.6,1)(1,0)0.1 ( 2.9,1)(1,0)0.1 ( 1.1,2)(1,0)0.1 ( 1.4,2)(1,0)0.1 ( 1.7,2)(1,0)0.1 ( 2.0,2)(1,0)0.1 ( 2.3,2)(1,0)0.1 ( 2.6,2)(1,0)0.1 ( 2.9,2)(1,0)0.1 ( 1.1,3)(1,0)0.1 ( 1.4,3)(1,0)0.1 ( 1.7,3)(1,0)0.1 ( 2.0,3)(1,0)0.1 ( 2.3,3)(1,0)0.1 ( 2.6,3)(1,0)0.1 ( 2.9,3)(1,0)0.1 ( 1.1,4)(1,0)0.1 ( 1.4,4)(1,0)0.1 ( 1.7,4)(1,0)0.1 ( 2.0,4)(1,0)0.1 ( 2.3,4)(1,0)0.1 ( 2.6,4)(1,0)0.1 ( 2.9,4)(1,0)0.1 ( 3,1)(1,0)1 ( 3,2)(1,0)1 ( 3,3)(1,0)1 ( 4,0)(0,1)4 ( 3,0)(0,1)4 ( 1,0)(0,1)4 ( 0,0)(0,1)4 ( 4.5,3.5)(1,0)2 ( 7,0)(0,1)4 ( 8,0)(0,1)4 ( 9.5,0)(0,1)4 ( 10.5,0)(0,1)4 ( 11.5,3)(0,1)1 ( 13,3)(0,1)1 ( 14,3)(0,1)1 ( 7,0)(1,0)1 ( 9.5,0)(1,0)1 ( 9.5,3)(1,0)2 ( 13,3)(1,0)1 ( 7,4)(1,0)1 ( 9.5,4)(1,0)2 ( 13,4)(1,0)1 ( 7,1)(1,0)1 ( 7,2)(1,0)1 ( 7,3)(1,0)1 ( 9.5,2)(1,0)1 ( 9.5,1)(1,0)1 ( 8.1,4)(1,0)0.1 ( 8.4,4)(1,0)0.1 ( 8.7,4)(1,0)0.1 ( 9.0,4)(1,0)0.1 ( 9.3,4)(1,0)0.1 ( 8.1,3)(1,0)0.1 ( 8.4,3)(1,0)0.1 ( 8.7,3)(1,0)0.1 ( 9.0,3)(1,0)0.1 ( 9.3,3)(1,0)0.1 ( 8.1,2)(1,0)0.1 ( 8.4,2)(1,0)0.1 ( 8.7,2)(1,0)0.1 ( 9.0,2)(1,0)0.1 ( 9.3,2)(1,0)0.1 ( 8.1,1)(1,0)0.1 ( 8.4,1)(1,0)0.1 ( 8.7,1)(1,0)0.1 ( 9.0,1)(1,0)0.1 ( 9.3,1)(1,0)0.1 ( 8.1,0)(1,0)0.1 ( 8.4,0)(1,0)0.1 ( 8.7,0)(1,0)0.1 ( 9.0,0)(1,0)0.1 ( 9.3,0)(1,0)0.1 ( 11.6,4)(1,0)0.1 ( 11.9,4)(1,0)0.1 ( 12.2,4)(1,0)0.1 ( 12.5,4)(1,0)0.1 ( 12.8,4)(1,0)0.1 ( 11.6,3)(1,0)0.1 ( 11.9,3)(1,0)0.1 ( 12.2,3)(1,0)0.1 ( 12.5,3)(1,0)0.1 ( 12.8,3)(1,0)0.1 ( 14.5,3.5)(1,0)2 ( 17,3)(0,1)1(18,3)(0,1)1 ( 20.5,3)(0,1)1 ( 21.5,3)(0,1)1 ( 17,3)(1,0)1 ( 20.5,3)(1,0)1 ( 17,4)(1,0)1 ( 20.5,4)(1,0)1 ( 18.1,4)(1,0)0.1 ( 18.4,4)(1,0)0.1 ( 18.7,4)(1,0)0.1 ( 19.0,4)(1,0)0.1 ( 19.3,4)(1,0)0.1 ( 19.6,4)(1,0)0.1 ( 19.9,4)(1,0)0.1 ( 20.2,4)(1,0)0.1 ( 18.1,3)(1,0)0.1 ( 18.4,3)(1,0)0.1 ( 18.7,3)(1,0)0.1 ( 19.0,3)(1,0)0.1 ( 19.3,3)(1,0)0.1 ( 19.6,3)(1,0)0.1 ( 19.9,3)(1,0)0.1 ( 20.2,3)(1,0)0.1 from fig.[suheis_figure_magn67 ] , we see that the spin , orbital as well as @xmath17 are polarized simultaneously versus external field .
the above young tableau indicates that the states @xmath123 simultaneously because they carries out a su(3 ) representation and the system possess su(3 ) symmetry .
after the system is fully polarized , the magnetization reach the maximum value @xmath124 .
then the residual symmetry of the ground state is only of u(1 ) .
in the previous section , we discussed the simplest case where merely one su(2 ) subgroup symmetry is broken . in the following
, we will consider the regimes with two - parameter symmetry breaking , which involves more su(2 ) subgroup . under the restriction @xmath125 , the magnetization ( [ eq : mag ] ) becomes ,
@xmath126 which indicates that the residual symmetry of ground state is u(1)@xmath10 u(1 ) @xmath10 su(2 ) , in which the su(2 ) is generated by @xmath127 and @xmath128 .
the magnetization curves for different @xmath48 is plotted in fig .
[ suheis_figure_magn3 ] , and the phase diagram in terms of @xmath129 versus @xmath130 is given in fig .
[ suheis_figure_phase3 ] . on the one hand , because the eigenvalue of @xmath13 equals that of @xmath17 for both states @xmath33 and @xmath35 , the flipping process occurred in phase ii contributes to both magnetization of @xmath13 and @xmath17 equivalently . on the other hand ,
the flipping from @xmath104 and @xmath39 occurs simultaneously in the phase iv due to the su(2 ) symmetry . as a results ,
the magnetization of @xmath13 and @xmath17 are expected to be the same in the whole process , which can be seen from our numerical calculation in fig .
[ suheis_figure_magn3 ] . for @xmath131 , there exists three distinct phases , denoted by iv , ii and i respectively according to the number of the rows of young tableau .
the su(2 ) symmetry makes the states @xmath104 and @xmath115 flip simultaneously when the external field increases .
this makes the four - row young tableau turn to two - row young tableau directly , hence the phase iii labelled by three - row young tableau will not take place . the boundary between phase iv and phase ii
is determined from @xmath132 and @xmath133 together , @xmath134 where @xmath135 can be computed from eqs .
( [ eq : thermodeq ] ) numerically . for sufficient large external field ,
all @xmath136 , @xmath137 and @xmath138 are frozen to the @xmath23-direction , which brings about the occurrence of phase i. the boundary between phase i and phase ii is determined by @xmath139 , i.e. , @xmath140 this can also be derived from the competition between the states related to the young tableau @xmath141 $ ] and [ @xmath67 ] .
the asymptotic behavior at large @xmath130 is @xmath143 , which implies that the phase i will never occur as long as @xmath144 . the magnetization process in the region @xmath145 can be illustrated by the following evolution of young tableau , ( 18,5)(3,0 ) ( -1 , 3.35)(0,1)0.7 ( -1.2,3.35)(1,0)0.4 ( -1 , 2.15)(0,1)0.7 ( -1.2,2.85)(1,0)0.4 ( -1 , 1.85)(0,-1)0.7(-1.2,1.15)(1,0)0.4 ( -1 , 0.65)(0,-1)0.7 ( -1.2,0.65)(1,0)0.4 ( 0,1)(1,0)1 ( 0,2)(1,0)1 ( 0,3)(1,0)1 ( 0,0)(1,0)1 ( 0,4)(1,0)1 ( 2.5,0)(1,0)1 ( 2.5,4)(1,0)1 ( 1.1,0)(1,0)0.1 ( 1.4,0)(1,0)0.1 ( 1.7,0)(1,0)0.1 ( 2.0,0)(1,0)0.1 ( 2.3,0)(1,0)0.1 ( 2.6,0)(1,0)0.1 ( 2.9,0)(1,0)0.1 ( 1.1,1)(1,0)0.1 ( 1.4,1)(1,0)0.1 ( 1.7,1)(1,0)0.1 ( 2.0,1)(1,0)0.1 ( 2.3,1)(1,0)0.1 ( 2.6,1)(1,0)0.1 ( 2.9,1)(1,0)0.1 ( 1.1,2)(1,0)0.1 ( 1.4,2)(1,0)0.1 ( 1.7,2)(1,0)0.1 ( 2.0,2)(1,0)0.1 ( 2.3,2)(1,0)0.1 ( 2.6,2)(1,0)0.1 ( 2.9,2)(1,0)0.1 ( 1.1,3)(1,0)0.1 ( 1.4,3)(1,0)0.1 ( 1.7,3)(1,0)0.1 ( 2.0,3)(1,0)0.1 ( 2.3,3)(1,0)0.1 ( 2.6,3)(1,0)0.1 ( 2.9,3)(1,0)0.1 ( 1.1,4)(1,0)0.1 ( 1.4,4)(1,0)0.1 ( 1.7,4)(1,0)0.1 ( 2.0,4)(1,0)0.1 ( 2.3,4)(1,0)0.1 ( 2.6,4)(1,0)0.1 ( 2.9,4)(1,0)0.1 ( 2.5,1)(1,0)1 ( 2.5,2)(1,0)1 ( 2.5,3)(1,0)1 ( 3.5,0)(0,1)4 ( 2.5,0)(0,1)4 ( 1,0)(0,1)4 ( 0,0)(0,1)4 ( 4,3.5)(1,0)1 ( 5.5,0)(0,1)4 ( 6.5,0)(0,1)4 ( 8,0)(0,1)4 ( 9,2)(0,1)2 ( 10.5,2)(0,1)2 ( 11.5,3)(0,1)1 ( 13,3)(0,1)1 ( 5.5,0)(1,0)1 ( 8,2)(1,0)1 ( 10.5,3)(1,0)1 ( 5.5,4)(1,0)1 ( 8,4)(1,0)1 ( 10.5,4)(1,0)1 ( 5.5,1)(1,0)1 ( 5.5,2)(1,0)1 ( 5.5,3)(1,0)1 ( 8,3)(1,0)1 ( 6.6,4)(1,0)0.1 ( 6.9,4)(1,0)0.1 ( 7.2,4)(1,0)0.1 ( 7.5,4)(1,0)0.1 ( 7.8,4)(1,0)0.1 ( 6.6,3)(1,0)0.1 ( 6.9,3)(1,0)0.1 ( 7.2,3)(1,0)0.1 ( 7.5,3)(1,0)0.1 ( 7.8,3)(1,0)0.1 ( 6.6,2)(1,0)0.1 ( 6.9,2)(1,0)0.1 ( 7.2,2)(1,0)0.1 ( 7.5,2)(1,0)0.1 ( 7.8,2)(1,0)0.1 ( 6.6,1)(1,0)0.1 ( 6.9,1)(1,0)0.1 ( 7.2,1)(1,0)0.1 ( 7.5,1)(1,0)0.1 ( 7.8,1)(1,0)0.1 ( 6.6,0)(1,0)0.1 ( 6.9,0)(1,0)0.1 ( 7.2,0)(1,0)0.1 ( 7.5,0)(1,0)0.1 ( 7.8,0)(1,0)0.1 ( 9.1,4)(1,0)0.1 ( 9.4,4)(1,0)0.1 ( 9.7,4)(1,0)0.1 ( 10,4)(1,0)0.1 ( 10.3,4)(1,0)0.1 ( 9.1,3)(1,0)0.1 ( 9.4,3)(1,0)0.1 ( 9.7,3)(1,0)0.1 ( 10,3)(1,0)0.1 ( 10.3,3)(1,0)0.1 ( 9.1,2)(1,0)0.1 ( 9.4,2)(1,0)0.1 ( 9.7,2)(1,0)0.1 ( 10,2)(1,0)0.1 ( 10.3,2)(1,0)0.1 ( 11.6,3)(1,0)0.1 ( 11.9,3)(1,0)0.1 ( 12.2,3)(1,0)0.1 ( 12.5,3)(1,0)0.1 ( 12.8,3)(1,0)0.1 ( 11.6,4)(1,0)0.1 ( 11.9,4)(1,0)0.1 ( 12.2,4)(1,0)0.1 ( 12.5,4)(1,0)0.1 ( 12.8,4)(1,0)0.1 ( 13.5,3.5)(1,0)1 ( 15,2)(0,1)2 ( 16,2)(0,1)2 ( 17.5,2)(0,1)2 ( 18.5,3)(0,1)1 ( 20,3)(0,1)1 ( 15,2)(1,0)1 ( 17.5,3)(1,0)1 ( 15,4)(1,0)1 ( 17.5,4)(1,0)1 ( 15,3)(1,0)1 ( 16.1,4)(1,0)0.1 ( 16.4,4)(1,0)0.1 ( 16.7,4)(1,0)0.1 ( 17,4)(1,0)0.1 ( 17.3,4)(1,0)0.1 ( 16.1,3)(1,0)0.1 ( 16.4,3)(1,0)0.1 ( 16.7,3)(1,0)0.1 ( 17,3)(1,0)0.1 ( 17.3,3)(1,0)0.1 ( 16.1,2)(1,0)0.1 ( 16.4,2)(1,0)0.1 ( 16.7,2)(1,0)0.1 ( 17,2)(1,0)0.1 ( 17.3,2)(1,0)0.1 ( 18.6,3)(1,0)0.1 ( 18.9,3)(1,0)0.1 ( 19.2,3)(1,0)0.1 ( 19.5,3)(1,0)0.1 ( 19.8,3)(1,0)0.1 ( 18.6,4)(1,0)0.1 ( 18.9,4)(1,0)0.1 ( 19.2,4)(1,0)0.1 ( 19.5,4)(1,0)0.1 ( 19.8,4)(1,0)0.1 ( 20.5,3.5)(1,0)1 ( 22,3)(0,1)1 ( 23,3)(0,1)1 ( 24.5,3)(0,1)1 ( 25.5,3)(0,1)1 ( 22,3)(1,0)1 ( 22,4)(1,0)1 ( 24.5,3)(1,0)1 ( 24.5,4)(1,0)1 ( 23.1,3)(1,0)0.1 ( 23.4,3)(1,0)0.1 ( 23.7,3)(1,0)0.1 ( 24.0,3)(1,0)0.1 ( 24.3,3)(1,0)0.1 ( 23.1,4)(1,0)0.1 ( 23.4,4)(1,0)0.1 ( 23.7,4)(1,0)0.1 ( 24.0,4)(1,0)0.1 ( 24.3,4)(1,0)0.1 the boundary between phase iv and phase i is determined by @xmath139 , @xmath132 and @xmath146 together , @xmath147 the common solution of eqs .
( [ eq : phase3_line1]-[eq : phase3 line3 ] ) gives @xmath148 and @xmath149 , which is a three - phase - coexist point .
if the external field along the direction of the second simple root is quenched but those along the other directions are kept , we will have symmetry breaking down to u(1)@xmath10 su(2 ) @xmath10 u(1 ) .
such kind of symmetry breaking is caused by zeeman term of the following magnetization @xmath150 the magnetization curves with different @xmath49 is plotted in fig .
[ suheis_figure_magn4 ] , and the phase diagram in terms of @xmath151 versus @xmath130 is given in fig .
[ suheis_figure_phase4 ] . for @xmath152 ,
there exists three phases denoted by iv , iii and i respectively .
the boundary between phases iv and iii is determined from @xmath153 , which can be solved numerically . for sufficient large external field
, the magnetization is saturated reaching phase i. this phase transition occurs at @xmath154 thus if @xmath155 , the phase i will never occur regardless of the magnitude of external field .
so in the region @xmath156 , the magnetization process can be illustrated by young tableau ( 18,5)(3,0 ) ( -1 , 3.35)(0,1)0.7 ( -1.2,3.35)(1,0)0.4 ( -1 , 2.15)(0,1)0.7 ( -1.2,2.85)(1,0)0.4 ( -1 , 1.85)(0,-1)0.7(-1.2,1.15)(1,0)0.4 ( -1 , 0.65)(0,-1)0.7 ( -1.2,0.65)(1,0)0.4 ( 0,1)(1,0)1 ( 0,2)(1,0)1 ( 0,3)(1,0)1 ( 0,0)(1,0)1 ( 0,4)(1,0)1 ( 2.5,0)(1,0)1 ( 2.5,4)(1,0)1 ( 1.1,0)(1,0)0.1 ( 1.4,0)(1,0)0.1 ( 1.7,0)(1,0)0.1 ( 2.0,0)(1,0)0.1 ( 2.3,0)(1,0)0.1 ( 2.6,0)(1,0)0.1 ( 2.9,0)(1,0)0.1 ( 1.1,1)(1,0)0.1 ( 1.4,1)(1,0)0.1 ( 1.7,1)(1,0)0.1 ( 2.0,1)(1,0)0.1 ( 2.3,1)(1,0)0.1 ( 2.6,1)(1,0)0.1 ( 2.9,1)(1,0)0.1 ( 1.1,2)(1,0)0.1 ( 1.4,2)(1,0)0.1 ( 1.7,2)(1,0)0.1 ( 2.0,2)(1,0)0.1 ( 2.3,2)(1,0)0.1 ( 2.6,2)(1,0)0.1 ( 2.9,2)(1,0)0.1 ( 1.1,3)(1,0)0.1 ( 1.4,3)(1,0)0.1 ( 1.7,3)(1,0)0.1 ( 2.0,3)(1,0)0.1 ( 2.3,3)(1,0)0.1 ( 2.6,3)(1,0)0.1 ( 2.9,3)(1,0)0.1 ( 1.1,4)(1,0)0.1 ( 1.4,4)(1,0)0.1 ( 1.7,4)(1,0)0.1 ( 2.0,4)(1,0)0.1 ( 2.3,4)(1,0)0.1 ( 2.6,4)(1,0)0.1 ( 2.9,4)(1,0)0.1 ( 2.5,1)(1,0)1 ( 2.5,2)(1,0)1 ( 2.5,3)(1,0)1 ( 3.5,0)(0,1)4 ( 2.5,0)(0,1)4 ( 1,0)(0,1)4 ( 0,0)(0,1)4 ( 4,3.5)(1,0)1 ( 5.5,0)(0,1)4 ( 6.5,0)(0,1)4 ( 8,0)(0,1)4 ( 9,1)(0,1)3 ( 10.5,1)(0,1)3 ( 11.5,3)(0,1)1 ( 13,3)(0,1)1 ( 5.5,0)(1,0)1 ( 8,2)(1,0)1 ( 10.5,3)(1,0)1 ( 5.5,4)(1,0)1 ( 8,4)(1,0)1 ( 10.5,4)(1,0)1 ( 5.5,1)(1,0)1 ( 5.5,2)(1,0)1 ( 5.5,3)(1,0)1 ( 8,3)(1,0)1 ( 6.6,4)(1,0)0.1 ( 6.9,4)(1,0)0.1 ( 7.2,4)(1,0)0.1 ( 7.5,4)(1,0)0.1 ( 7.8,4)(1,0)0.1 ( 6.6,3)(1,0)0.1 ( 6.9,3)(1,0)0.1 ( 7.2,3)(1,0)0.1 ( 7.5,3)(1,0)0.1 ( 7.8,3)(1,0)0.1 ( 6.6,2)(1,0)0.1 ( 6.9,2)(1,0)0.1 ( 7.2,2)(1,0)0.1 ( 7.5,2)(1,0)0.1 ( 7.8,2)(1,0)0.1 ( 6.6,1)(1,0)0.1 ( 6.9,1)(1,0)0.1 ( 7.2,1)(1,0)0.1 ( 7.5,1)(1,0)0.1 ( 7.8,1)(1,0)0.1 ( 6.6,0)(1,0)0.1 ( 6.9,0)(1,0)0.1 ( 7.2,0)(1,0)0.1 ( 7.5,0)(1,0)0.1 ( 7.8,0)(1,0)0.1 ( 9.1,4)(1,0)0.1 ( 9.4,4)(1,0)0.1 ( 9.7,4)(1,0)0.1 ( 10,4)(1,0)0.1 ( 10.3,4)(1,0)0.1 ( 9.1,3)(1,0)0.1 ( 9.4,3)(1,0)0.1 ( 9.7,3)(1,0)0.1 ( 10,3)(1,0)0.1 ( 10.3,3)(1,0)0.1 ( 9.1,2)(1,0)0.1 ( 9.4,2)(1,0)0.1 ( 9.7,2)(1,0)0.1 ( 10,2)(1,0)0.1 ( 10.3,2)(1,0)0.1 ( 9.1,1)(1,0)0.1 ( 9.4,1)(1,0)0.1 ( 9.7,1)(1,0)0.1 ( 10,1)(1,0)0.1 ( 10.3,1)(1,0)0.1 ( 8,1)(1,0)1 ( 11.6,3)(1,0)0.1 ( 11.9,3)(1,0)0.1 ( 12.2,3)(1,0)0.1 ( 12.5,3)(1,0)0.1 ( 12.8,3)(1,0)0.1 ( 11.6,4)(1,0)0.1 ( 11.9,4)(1,0)0.1 ( 12.2,4)(1,0)0.1 ( 12.5,4)(1,0)0.1 ( 12.8,4)(1,0)0.1 ( 13.5,3.5)(1,0)1 ( 15,1)(0,1)3 ( 16,1)(0,1)3 ( 17.5,1)(0,1)3 ( 18.5,3)(0,1)1 ( 20,3)(0,1)1 ( 15,1)(1,0)1 ( 17.5,3)(1,0)1 ( 15,4)(1,0)1 ( 17.5,4)(1,0)1 ( 15,3)(1,0)1 ( 15,2)(1,0)1 ( 16.1,3)(1,0)0.1 ( 16.4,3)(1,0)0.1 ( 16.7,3)(1,0)0.1 ( 17,3)(1,0)0.1 ( 17.3,3)(1,0)0.1 ( 16.1,2)(1,0)0.1 ( 16.4,2)(1,0)0.1 ( 16.7,2)(1,0)0.1 ( 17,2)(1,0)0.1 ( 17.3,2)(1,0)0.1 ( 16.1,4)(1,0)0.1 ( 16.4,4)(1,0)0.1 ( 16.7,4)(1,0)0.1 ( 17,4)(1,0)0.1 ( 17.3,4)(1,0)0.1 ( 16.1,1)(1,0)0.1 ( 16.4,1)(1,0)0.1 ( 16.7,1)(1,0)0.1 ( 17,1)(1,0)0.1 ( 17.3,1)(1,0)0.1 ( 18.6,4)(1,0)0.1 ( 18.9,4)(1,0)0.1 ( 19.2,4)(1,0)0.1 ( 19.5,4)(1,0)0.1 ( 19.8,4)(1,0)0.1 ( 18.6,3)(1,0)0.1 ( 18.9,3)(1,0)0.1 ( 19.2,3)(1,0)0.1 ( 19.5,3)(1,0)0.1 ( 19.8,3)(1,0)0.1 ( 20.5,3.5)(1,0)1 ( 22,3)(0,1)1 ( 23,3)(0,1)1 ( 24.5,3)(0,1)1 ( 25.5,3)(0,1)1 ( 22,3)(1,0)1 ( 22,4)(1,0)1 ( 24.5,3)(1,0)1 ( 24.5,4)(1,0)1 ( 23.1,3)(1,0)0.1 ( 23.4,3)(1,0)0.1 ( 23.7,3)(1,0)0.1 ( 24.0,3)(1,0)0.1 ( 24.3,3)(1,0)0.1 ( 23.1,4)(1,0)0.1 ( 23.4,4)(1,0)0.1 ( 23.7,4)(1,0)0.1 ( 24.0,4)(1,0)0.1 ( 24.3,4)(1,0)0.1 in phase iii , the length of the second row and that of the third row in their corresponding young tableau is always equal due to the su(2 ) symmetry .
thus the probability of pure spin - flipping latexmath:[$|\overline\downarrow\rangle \rightarrow latexmath:[$|\overline\downarrow\rangle \rightarrow latexmath:[$|\overline\downarrow\rangle \rightarrow of @xmath12 and @xmath13 .
these properties result in the same magnetization curves of @xmath12 and @xmath13 shown in fig .
[ suheis_figure_magn4 ] .
flipping over the state @xmath158 which has positive eigenvalue of @xmath17 brings about a negative magnetization of @xmath17 in phase iv . the phase i will never occur for @xmath155 , as is similar to the case of su(3)@xmath10u(1 )
actually , it recovers the case of residual su(3)@xmath10u(1 ) symmetry at @xmath159 , if @xmath160 , we can see from fig .
[ suheis_figure_phase4 ] that the phase iv transits into phase i directly along with the increase of external field .
the boundary which separates these two phase is determined by @xmath161 obviously , the point at @xmath162 and @xmath163 is a three - phase - coexist point .
the magnetization properties in the region of @xmath160 is similar to the case of u(1)@xmath10su(3 ) , particulary , a larger u(1)@xmath10su(3 ) symmetry remains for @xmath164 .
the third case of two - parameter symmetry breaking is produced by the zeeman term with the restriction to the the land g factor @xmath165 .
the magnetization now reads @xmath166 which breaks the su(4 ) symmetry down to su(2)@xmath10 u(1 ) @xmath10 u(1 ) .
the magnetization curves for different @xmath129 is plotted in fig .
[ suheis_figure_magn5 ] , and phase diagram in terms of @xmath129 versus @xmath130 is given in fig .
[ suheis_figure_phase5 ] . in the region of @xmath167 ,
there exist three phases denoted by iv , iii , and ii respectively .
the final state is characterized by two rows young tableau at sufficient large external field due to the su(2 ) symmetry .
the boundary separates phase iii and ii is determined by @xmath132 , which reads @xmath168 obviously if @xmath169 , the phase ii will never occur regardless the magnitude of external field , and the magnetization process is similar to the case of su(3)@xmath10u(1 ) . in the region @xmath145
, the magnetization process can be illustrated by the following young tableau , ( 18,5)(3,0 ) ( -1 , 3.35)(0,1)0.7 ( -1.2,3.35)(1,0)0.4 ( -1 , 2.15)(0,1)0.7 ( -1.2,2.85)(1,0)0.4 ( -1 , 1.85)(0,-1)0.7(-1.2,1.15)(1,0)0.4 ( -1 , 0.65)(0,-1)0.7 ( -1.2,0.65)(1,0)0.4 ( 0,1)(1,0)1 ( 0,2)(1,0)1 ( 0,3)(1,0)1 ( 0,0)(1,0)1 ( 0,4)(1,0)1 ( 2.5,0)(1,0)1 ( 2.5,4)(1,0)1 ( 1.1,0)(1,0)0.1 ( 1.4,0)(1,0)0.1 ( 1.7,0)(1,0)0.1 ( 2.0,0)(1,0)0.1 ( 2.3,0)(1,0)0.1 ( 2.6,0)(1,0)0.1 ( 2.9,0)(1,0)0.1 ( 1.1,1)(1,0)0.1 ( 1.4,1)(1,0)0.1 ( 1.7,1)(1,0)0.1 ( 2.0,1)(1,0)0.1 ( 2.3,1)(1,0)0.1 ( 2.6,1)(1,0)0.1 ( 2.9,1)(1,0)0.1 ( 1.1,2)(1,0)0.1 ( 1.4,2)(1,0)0.1 ( 1.7,2)(1,0)0.1 ( 2.0,2)(1,0)0.1 ( 2.3,2)(1,0)0.1 ( 2.6,2)(1,0)0.1 ( 2.9,2)(1,0)0.1 ( 1.1,3)(1,0)0.1 ( 1.4,3)(1,0)0.1 ( 1.7,3)(1,0)0.1 ( 2.0,3)(1,0)0.1 ( 2.3,3)(1,0)0.1 ( 2.6,3)(1,0)0.1 ( 2.9,3)(1,0)0.1 ( 1.1,4)(1,0)0.1 ( 1.4,4)(1,0)0.1 ( 1.7,4)(1,0)0.1 ( 2.0,4)(1,0)0.1 ( 2.3,4)(1,0)0.1 ( 2.6,4)(1,0)0.1 ( 2.9,4)(1,0)0.1 ( 2.5,1)(1,0)1 ( 2.5,2)(1,0)1 ( 2.5,3)(1,0)1 ( 3.5,0)(0,1)4 ( 2.5,0)(0,1)4 ( 1,0)(0,1)4 ( 0,0)(0,1)4 ( 4,3.5)(1,0)1 ( 5.5,0)(0,1)4 ( 6.5,0)(0,1)4 ( 8,0)(0,1)4 ( 9,1)(0,1)3 ( 10.5,1)(0,1)3 ( 11.5,2)(0,1)2 ( 13,2)(0,1)2 ( 10.5,2)(1,0)1 ( 5.5,0)(1,0)1 ( 8,2)(1,0)1 ( 10.5,3)(1,0)1 ( 5.5,4)(1,0)1 ( 8,4)(1,0)1 ( 10.5,4)(1,0)1 ( 5.5,1)(1,0)1 ( 5.5,2)(1,0)1 ( 5.5,3)(1,0)1 ( 8,3)(1,0)1 ( 6.6,4)(1,0)0.1 ( 6.9,4)(1,0)0.1 ( 7.2,4)(1,0)0.1 ( 7.5,4)(1,0)0.1 ( 7.8,4)(1,0)0.1 ( 6.6,3)(1,0)0.1 ( 6.9,3)(1,0)0.1 ( 7.2,3)(1,0)0.1 ( 7.5,3)(1,0)0.1 ( 7.8,3)(1,0)0.1 ( 6.6,2)(1,0)0.1 ( 6.9,2)(1,0)0.1 ( 7.2,2)(1,0)0.1 ( 7.5,2)(1,0)0.1 ( 7.8,2)(1,0)0.1 ( 6.6,1)(1,0)0.1 ( 6.9,1)(1,0)0.1 ( 7.2,1)(1,0)0.1 ( 7.5,1)(1,0)0.1 ( 7.8,1)(1,0)0.1 ( 6.6,0)(1,0)0.1 ( 6.9,0)(1,0)0.1 ( 7.2,0)(1,0)0.1 ( 7.5,0)(1,0)0.1 ( 7.8,0)(1,0)0.1 ( 9.1,4)(1,0)0.1 ( 9.4,4)(1,0)0.1 ( 9.7,4)(1,0)0.1 ( 10,4)(1,0)0.1 ( 10.3,4)(1,0)0.1 ( 9.1,3)(1,0)0.1 ( 9.4,3)(1,0)0.1 ( 9.7,3)(1,0)0.1 ( 10,3)(1,0)0.1 ( 10.3,3)(1,0)0.1 ( 9.1,2)(1,0)0.1 ( 9.4,2)(1,0)0.1 ( 9.7,2)(1,0)0.1 ( 10,2)(1,0)0.1 ( 10.3,2)(1,0)0.1 ( 9.1,1)(1,0)0.1 ( 9.4,1)(1,0)0.1 ( 9.7,1)(1,0)0.1 ( 10,1)(1,0)0.1 ( 10.3,1)(1,0)0.1 ( 8,1)(1,0)1 ( 11.6,3)(1,0)0.1 ( 11.9,3)(1,0)0.1 ( 12.2,3)(1,0)0.1 ( 12.5,3)(1,0)0.1 ( 12.8,3)(1,0)0.1 ( 11.6,4)(1,0)0.1 ( 11.9,4)(1,0)0.1 ( 12.2,4)(1,0)0.1 ( 12.5,4)(1,0)0.1 ( 12.8,4)(1,0)0.1 ( 11.6,2)(1,0)0.1 ( 11.9,2)(1,0)0.1 ( 12.2,2)(1,0)0.1 ( 12.5,2)(1,0)0.1 ( 12.8,2)(1,0)0.1 ( 13.5,3.5)(1,0)1 ( 15,1)(0,1)3 ( 16,1)(0,1)3 ( 17.5,1)(0,1)3 ( 18.5,2)(0,1)2 ( 20,2)(0,1)2 ( 17.5,2)(1,0)1 ( 15,1)(1,0)1 ( 17.5,3)(1,0)1 ( 15,4)(1,0)1 ( 17.5,4)(1,0)1 ( 15,3)(1,0)1 ( 15,2)(1,0)1 ( 16.1,3)(1,0)0.1 ( 16.4,3)(1,0)0.1 ( 16.7,3)(1,0)0.1 ( 17,3)(1,0)0.1 ( 17.3,3)(1,0)0.1 ( 16.1,2)(1,0)0.1 ( 16.4,2)(1,0)0.1 ( 16.7,2)(1,0)0.1 ( 17,2)(1,0)0.1 ( 17.3,2)(1,0)0.1 ( 16.1,4)(1,0)0.1 ( 16.4,4)(1,0)0.1 ( 16.7,4)(1,0)0.1 ( 17,4)(1,0)0.1 ( 17.3,4)(1,0)0.1 ( 16.1,1)(1,0)0.1 ( 16.4,1)(1,0)0.1 ( 16.7,1)(1,0)0.1 ( 17,1)(1,0)0.1 ( 17.3,1)(1,0)0.1 ( 18.6,4)(1,0)0.1 ( 18.9,4)(1,0)0.1 ( 19.2,4)(1,0)0.1 ( 19.5,4)(1,0)0.1 ( 19.8,4)(1,0)0.1 ( 18.6,3)(1,0)0.1 ( 18.9,3)(1,0)0.1 ( 19.2,3)(1,0)0.1 ( 19.5,3)(1,0)0.1 ( 19.8,3)(1,0)0.1 ( 18.6,2)(1,0)0.1 ( 18.9,2)(1,0)0.1 ( 19.2,2)(1,0)0.1 ( 19.5,2)(1,0)0.1 ( 19.8,2)(1,0)0.1 ( 20.5,3.5)(1,0)1 ( 22,2)(0,1)2 ( 23,2)(0,1)2 ( 24.5,2)(0,1)2 ( 25.5,2)(0,1)2 ( 22,2)(1,0)1 ( 22,4)(1,0)1 ( 24.5,2)(1,0)1 ( 24.5,4)(1,0)1 ( 22,3)(1,0)1 ( 24.5,3)(1,0)1 ( 23.1,3)(1,0)0.1 ( 23.4,3)(1,0)0.1 ( 23.7,3)(1,0)0.1 ( 24.0,3)(1,0)0.1 ( 24.3,3)(1,0)0.1 ( 23.1,2)(1,0)0.1 ( 23.4,2)(1,0)0.1 ( 23.7,2)(1,0)0.1 ( 24.0,2)(1,0)0.1 ( 24.3,2)(1,0)0.1 ( 23.1,4)(1,0)0.1 ( 23.4,4)(1,0)0.1 ( 23.7,4)(1,0)0.1 ( 24.0,4)(1,0)0.1 ( 24.3,4)(1,0)0.1 fig . [ suheis_figure_magn5 ] shows that that the magnetization process of @xmath12 , @xmath13 and @xmath17 are quite different . in phase iv ,
the flippings from @xmath39 to other three states possess positive contribution to @xmath12 and @xmath13 , but negative contribution to @xmath17 . in phase iii
, @xmath12 undergoes polarization continually , while @xmath17 undergoes polarization but @xmath13 undergoes anti - polarization . since in the final phase ,
both eigenvalues are zero and the spin magnetization are saturated , this phase will not change for any further increase of the external field .
the point of @xmath170 , @xmath171 is a three phase coexistence point .
it can be seen that there exist two phases merely for @xmath155 , and phase iv transits into phase ii directly .
actually , the residual symmetry of su(2)@xmath10u(1)@xmath10su(2 ) is restored when @xmath172 .
in above , we studied the magnetization properties of a su(4 ) spin - orbital chain in the presence of a generalized external field that has three parameters due to the cartan subalgebra of @xmath19 lie algebra has three generators .
those three parameters are re - chosen so that to relate them with the spin @xmath136 , orbital @xmath137 and their _ product _ @xmath138 .
we called them three land g factors .
then all possible symmetry breaking and the corresponding magnetization process induced by that external field are studied respectively . the ground state phase diagram caused by the competition of quantum fluctuation and zeeman - like effect
are studied by solving the bethe - ansatz equations numerically .
the phase transition boundaries derived by studying the dress energy equations analytically .
the features of various phases and transitions between them are explained in detailed in terms of group theory analysis . | the ground state phases of a one - dimensional su(4 ) spin - orbital hamiltonian in a generalized external field are studied on the basis of bethe - ansatz solution . introducing three land @xmath0 factors for spin , orbital and their products in the su(4 ) zeeman term ,
we discuss systematically the various symmetry breaking .
the magnetization versus external field are obtained by solving bethe - ansatz equations numerically .
the phase diagrams corresponding to distinct residual symmetries are given by means of both numerical and analytical methods . |
total and differential cross sections for the reaction @xmath1 have been measured with high accuracy at pion energies above 100 mev .
currently , experimental efforts in this energy range are aimed at measuring spin dependent observables .
the situation at pion energies below 100 mev is less complete , especially for energies of a few tens of mev@xcite .
the disagreement between the experimental data for the total cross section in some cases exceeds 20% , much larger than the quoted uncertainties .
the experimental data show a deviation from smooth behavior in the total cross section of the reaction @xmath2 close to @xmath3 ( @xmath4)@xcite . more recent measurements of the partial total cross section for the inverse reaction at saturne have verified this effect@xcite . in a recent experiment of the dubna - gatchina group on pion absorption on carbon@xcite a dip in the energy dependence of the quasi - deuteron component of absorption near @xmath5=28 mev has been observed .
the measurements of pion absorption on the deuteron@xcite performed by the same group also indicated possible structure at a pion energy of 30 mev .
a more recent experiment at lampf@xcite did not observe any dip on carbon , but these data suggest a change in the angular distribution of protons at pion energies near 25 mev .
some possible alternative explanations of the structure in this energy range are given in ref.@xcite .
one of them attributed the structure to the excitation of a diproton resonance in the @xmath6 state .
however , these experiments are not accurate enough to reach a conclusion concerning the structure under the discussion .
a satisfactory solution fitting the available database has not been found in partial wave analysis , suggesting underlying systematic problems with the data base .
the experiment described here was intended to resolve the discrepancies in existing data and to verify the observed structure in the excitation function of pion absorption on the deuteron at low pion energies .
the differential and total cross sections of the @xmath7 reaction have been measured with fine steps in incident pion energy .
the experiment was performed at the low energy pion channel ( lep ) of the clinton p. anderson meson physics facility ( lampf ) .
positive pions with energies of 21 , 23 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 33 , 35 , 37 , 39 , 41 and 45 mev were used .
the pion beam had an average intensity of a few 10@xmath8/sec , allowing the beam particles to be counted while not overloading the bgo detectors .
the pion fraction varied from 78% for 45 mev beam to 33% for 21 mev beam .
the momentum bite of the lep channel was set to 1% for the highest energies and increased to 4% for the lowest ones in order to increase the pion flux .
the target was composed of cd@xmath9 with a cross sectional area of @xmath10 and an areal density of 0.469 g/@xmath11 .
it was attached to a thin paper pipe and placed in the center of the bgo ball .
the supporting pipe was aligned along the beam axis .
the diameter of the pipe was big enough to keep its walls out of the beam .
a 0.25-mm - thick plastic scintillator s1 with a cross section of @xmath12 was located just before the target .
downstream of the target a rectangular array of nine csi scintillators da1da9 was preceded by a 10-mm - thick plastic scintillator , s2 .
da1da9 and s2 could be used for the detection of the most forward going reaction products as well as for a determination of beam composition . in this experiment only the central detector , da5 , of the array was used . a coincidence between s1 and the central detector , da5 , of the downstream array
was used as a beam monitor .
every 1000th beam event ( event triggered by a @xmath13 coincidence ) was read out for further analysis to determine the pion fraction which was needed for absolute normalization of the cross sections .
a large solid angle detector , the lampf bgo ball , was used to detect the reaction products in this study .
detailed information on the bgo ball can be found in ref . .
the bgo ball consists of 30 phoswich detectors .
the detectors of the array were of pentagonal and hexagonal shape and tightly packed to form a truncated icosahedron of 32 sides .
two of the 32 sides are opened for the beam entry and exit .
the detectors were distributed about an inner radius of 6.1 cm from the center of the array to the center of each crystal face , and were arranged in six groups centered at laboratory scattering angles of @xmath14 , and @xmath15 .
each detector had a solid angle of about @xmath16 sr and was supported in a 0.5-mm - thick electro - formed nickel can which had a 0.05-mm - thick entrance window .
each detector consisted a 3-mm - thick ne102 plastic scintillator optically coupled to the front of a 5.6-cm - thick bismuth germanate ( bgo ) crystal , with a 7.62-cm - diameter photomultiplier tube on the back . since the decay constant of the bgo scintillator is much longer than that of the plastic scintillator ( 250 ns vs 1.5 ns ) , the anode signal was time sliced to provide both @xmath17 ( fast ) and @xmath18 ( slow ) signals for charged particle identification ( pions , protons , deuterons , etc . ) , and for identification of neutrons and gamma rays .
the crystals were thick enough to stop up to 185-mev protons and 90-mev pions .
the time resolution of the detectors was about 1 ns , sufficient to eliminate events with hits from different beam bursts ( the lampf beam has a 5-ns microstructure ) .
the light output of bgo scintillator depends significantly on the temperature of bgo material@xcite . to minimize fluctuations in temperature of the bgo
, a tent - like structure was built to isolate the bgo ball from its surroundings .
the event trigger consisted of a coincidence between the target detector , s1 , and at least one bgo crystal in anti - coincidence with da5 .
the raw data for each event contains information about the energy deposited in the plastic and the bgo for all detectors and timing information with respect to the beam counter . in the first step of the analysis a time gate
was applied to all of the signals to remove accidentals from further analysis .
the next step was the determination and application of the constants used for unmixing the @xmath19 information from the phoswich detectors .
the signal from each phoswich detector was integrated in two different adc channels with different time gates , 50 and 250 ns long .
the @xmath20 and @xmath18 information was separated using : @xmath21 @xmath22 where @xmath23 and @xmath24 represent the raw data from the adc s with the short and long gates , respectively , and @xmath25 , @xmath26 are the coefficients for conversion of raw adc data to energy in mev .
the relative fraction of a long signal in a short gate and vice versa are given by the mixing parameters , @xmath27 and @xmath28 , respectively .
the @xmath29 and offsets , @xmath30 , were determined using a least square fit for the events along @xmath17 and @xmath18 axes , where only one component of a signal ( short or long ) exists .
these are produced by charge particles which stop in the plastic and leave a @xmath17 with no @xmath18 signal , or by neutral particles which interact in the bgo and leave a @xmath18 with no @xmath17 signal .
an example of a @xmath31 distribution from a phoswich is shown in fig .
one can clearly see regions corresponding to protons and pions .
the outlined region along the @xmath18 axis is due to neutral particles and the outlined region along the @xmath17 axis is due to short range charged particles stopped in the plastic scintillator .
an initial energy calibration of the bgo ball ( @xmath25 and @xmath26 ) was made by using two proton coincidences from the @xmath32 reaction .
an alternate method used elastically scattered pions from @xmath33 .
pions from this reaction give a strong peak ( fig .
[ ede ] ) with well defined energy for each ring of detectors in the bgo ball .
this eliminated the strong kinematic energy dependence in the former procedure .
use of the cd@xmath9 as a target and a single hit trigger made this possible .
the @xmath17 and @xmath18 gains were fitted continuously during data taking and off - line data analysis .
this continuous stabilization was important because of instability in the bgo ball detectors .
the parameters @xmath29 and @xmath30 also require continuous stabilization because both the gain and the decay constant of the bgo scintillator is temperature dependent .
we used the following stabilization algorithm .
for sequential subsets of raw events the complete set of parameters was fitted .
the weighted average of old and new values of the parameters was used for the next subset of events .
this procedure was completely automated .
[ pp ] ( described below ) shows the energy resolution and accuracy of this calibration procedure .
the width of the two proton total energy peak from the @xmath34 reaction , summed over entire ball is @xmath35 mev .
after the system has been calibrated one can easily determine the event multiplicity , identify detected particles and measure their energies .
we selected events using following criteria : * the multiplicity must be equal 2 ; * both particles must be protons ; and * the opening angle between the two protons and their total energy must satisfy the kinematics of the reaction @xmath36 . fig .
[ pp ] is the energy spectrum for two proton events .
the events are summed over all kinematically allowed combinations of bgo ball detectors .
the narrow peak at the higher energy corresponds to the reaction @xmath36 .
the lower energy broad peak with a long tail results from pion absorption on carbon .
one observes that there is practically no background under the @xmath36 reaction peak .
measurements with no target but with the supporting pipe in its normal position and with a pure carbon target show that the background under the peak does not exceed 0.1% so no background subtraction was necessary . because of the low beam flux in this experiment it was possible to directly count the beam particles . pulse height and timing
information from s1 , s2 and da5 detectors was used to determine the pion fraction .
first , a cut was placed on the time between s1 and s2 to eliminate accidental coincidences between particles from different beam bursts .
another cut was applied to a two dimensional @xmath37 distribution ( pulse height from s2 and da5 ) to eliminate positrons . the resulting pulse height spectrum from s2
is shown in fig .
the left peak corresponds to muons , the right one to pions .
two modified moyal functions:@xcite @xmath38 one for pions and one for muons , were fitted to the energy loss distribution . here
@xmath18 is the energy loss in s2 and @xmath39 are adjusted parameters .
the pion distribution was integrated to obtain the pion fraction . at the two lowest energies ( 21 mev and 23 mev ) some pions loose all their energy in s2 and
do not hit da5 and the corresponding pion peak has a different , and more complicated shape . in this case
only the muon peak was fitted and the pion fraction was determined by subtracting the muon fraction from the integral number of beam particles . a monte carlo simulation of beam particles passing through the setup based on the geant code@xcite has been done .
the corrections for energy losses , straggling , multiple scattering and decay were calculated from this simulation .
the resulting correction to the total number of beam pions is 7% for the highest beam energy and 44% for the lowest one . for beam energies of 21 and 23 mev
this correction accounts for pions , which passed through the target but did not hit da5 and were lost from the beam trigger .
this simulation also was used for calculation of the pion energy in the center of a target .
the uncertainty in the number of beam pions , mainly due to fitting errors , is 3 5% , with the contribution from the statistical error less than 1% .
the efficiency was calculated using a monte carlo code which incorporates the actual geometry of the experimental setup .
the efficiency accounts for the effective solid angle ( geometrical solid angle reduced by the requirement of both protons detection ) , reaction losses , and dead time .
the effective solid angles for all kinematically allowed pairs of the detectors have been calculated .
the missing solid angle due to the nickel cans surrounding the crystals is about 6% .
reaction loss correction due to the interaction of protons with the bgo were obtained using data from ref .
the typical value of this correction is 10% .
finally , the data were corrected for dead time .
a typical value of the dead time in this experiment was 5% .
we estimate the total uncertainty in the efficiency of the setup is about 5% .
the differential cross section has been calculated for the selected events at each specific scattering angle .
the data for the backward angles were translated to the forward hemisphere since the angular distribution in the center of mass frame is symmetrical around @xmath40 due to indistinguishability of the two protons .
the resulting differential cross sections are presented in table [ tdsig ] . to obtain the total error
, we added the statistical and normalization errors in quadrature .
an overall systematic uncertainty of about 5% in the efficiency calculation is not included in the total error .
this affects the global normalization of all 90 differential cross section points and does not change their relative value .
we fitted the cross sections using a simple parameterization given by the legendre polynomial series : @xmath41 at each energy we used only the statistical errors of the cross sections but not the normalization error because it does not change the relative shape of the angular distribution .
the total cross sections were determined by : @xmath42 using the parameterization of eq .
( [ leg ] ) .
the integration over @xmath43 instead of @xmath44 is necessary because there are two indistinguishable protons in the final state .
the fitted coefficients , total cross sections , and @xmath45 are presented in table [ tsgt ] .
errors for @xmath46 and @xmath47 are obtained from the fit . for the total cross section errors we added the normalization errors in quadrature .
overall systematic uncertainty of 5% is not included for the same reason as for differential cross section .
the differential cross sections measured in this experiment are plotted in fig .
solid lines show the best fit results with eq .
( [ leg ] ) .
the results listed in table [ tsgt ] show that only s- and p- pionic waves are important in this energy region . the absence of a need for higher partial waves agrees with previous measurements@xcite .
fixed angle excitation functions are presented in fig .
[ dsigr ] .
[ a2a0 ] shows the behavior of the angular distribution as a function of energy by plotting the ratio of the coefficients @xmath47 to @xmath46 .
the shape of the angular distribution changes smoothly with pion energy .
no structure , within the accuracy of the experiment , is observed in the differential cross section as a function of energy .
the total cross section is presented in fig .
[ sgt ] along with the rest of the world s data .
the current data follows the general trends .
the biggest disagreement in the total cross section data , which existed around @xmath48 mev is resolved by the new data .
we compared our cross sections with the predictions of the recent partial wave analysis of ref .
the new vpi solution , sm95 , that includes the current data is compared with the previous sp94 results which do not include these data .
table [ tab3 ] and figs .
[ dsigr],[sgt ] show the level of agreement between or data and the phase shift fits .
the normalization factor averaged over all 15 energies are 1.04 and 1.01 for sp94 and sm95 solutions respectively .
the values of the normalization factors for both solutions show that our estimate of systematic uncertainties is reasonable .
table [ tab4 ] presents a comparison of our data with other partial wave analysis .
the recent vpi results , which include our data set , give better agreement than the previous results by the queen mary college group ( 9 256 mev ) @xcite and by the hiroshima university group ( 6 256 mev ) @xcite . the good agreement between the predictions of sp94 and recent sm95 vpi solutions with the experimental data
support the absence of any anomalous behavior of the cross sections in this energy region within the accuracy of the present experimental results .
we have presented new data on the total and differential cross section for the reaction @xmath2 for pion energies from 18 mev to 44 mev . in this experiment ,
90 experimental data points for the differential cross section and 15 points for the total cross section have been measured with the same systematic uncertainty .
the number of the experimental data points in this energy range is almost doubled by these results .
the data generally follow the trends of previous results and are in a good agreement with the recent sm95 solution of the vpi partial wave analysis . the previous disagreement in total cross section data
is resolved by this experiment .
we do not observe any structure either in total cross section or in differential cross section .
the authors wish to acknowledge valuable discussions with r. a. arndt , r. a. giannelli , r. d. ransome , and b. g. ritchie .
this work was supported in part by russian fundamental research foundation grant 93 - 02 - 3995 and by the u.s .
department of energy .
i. i. strakovsky and e. a. pasyuk , in : _
`` experimental program of the investigations at inst .
moscow meson factory '' _ ( proc . of 4th
all union seminar on experimental program of investigations at inr moscow meson factory , 1985 ) moscow , russia : institute for nuclear research press , 1986 , p. 241 .
j. saudinos , b. bonin , j - cl .
duchazeaubeneix , j - cl .
faivre , m. garson , d. legrand , m. rouger , l. antonuk , a. boudard , j. arvieux , and g. gaillard , in : _ proc . of int .
workshop on nucl . phys . in gev region , kek , nov . 1315 , 1984 _ , kek report 84 - 20 , 1985 , p. 228 ; nouvelles de saturne * 10 * , 25 ( 1985 ) .
k. akimov , m. g. gornov , m. i. gostkin , yu .
b. gurov , i. i. haysak , s. p. kruglov , s. i. merzlyakov , k. o. oganesyan , e. a. pasyuk , s. yu . porokhovoy , a. i. rudenko , a. v. shiskov , and i. i. strakovsky , nucl .
* a451 * , 433 ( 1992 ) .
s. i. gogolev , m. g. gornov , m. i. gostkin , yu .
b. gurov , i. i. haysak , s. p. kruglov , s. i. merzlyakov , k. o. oganesyan , e. a. pasyuk , s. yu . porokhovoy , a. i. rudenko , and i. i. strakovsky , phys .
* b300 * , 24 ( 1993 ) .
r. d. ransome , v. p. cupps , s. dawson , r. w.fergerson , a. green , c. l. morris , j. a. mcgill , j. r. comfort , b. g. ritchie , j. tinsley , j. d. zumbro , r. a. loveman , p. c. gugelot , d. l. watson , and c. fred moore , phys .
64 * , 372 ( 1990 ) ; phys . rev . *
c 42 * , 1500 ( 1990 ) .
r. d. ransome , in _ pion - nucleus physics : future directions and new facilities at lampf _ ,
proceedings of the los alamos conference on pion - nucleus physics , aip conf .
proc . * 163 * , edited by r. j. peterson and d. d. strottman ( aip , new york , 1988 ) p. 270 .
a. zucchatti , p. levi sandri , c. schaerf , a. a. cowley , j. v. pilcher , l. mazzaschi , m. anghinolfi , n. bianchi , l. casano , p. corvisiero , a. dangelo , e. de sancctis , g. gervino , b. girolami , v. lucherini , v. mokeev , v. muccifora , p. picozza , e. polli , a. r. reolon , g. ricco , m. ripani , p. rossi , m. sanzone , and m. taiuti , nucl .
instr.&meth . in phys
* a321 * , 219 ( 1992 ) .
b. g. ritchie , r. d. edge , d. j. malbrough , b. m. preedom , f. e. bertrand , e. e. gross , f. e. obenshain , j. r. wu , m. blecher , k. gotow , r. l. burman , r. carlini , m. e. hamm , m. j. leitch , and m. a. moinester , phys .
* c 24 * , 552 ( 1981 ) . | a study of the reaction @xmath0 has been performed in the energy range of 18 44 mev .
total cross sections and differential cross sections at six angles have been measured at 15 energies with an energy increment of 1 2 mev .
this is the most systematic data set in this energy range .
no structure in the energy dependence of the cross section has been observed within the accuracy of this experiment . |
kanbur et al ( 2002 ) , hendry et al ( 1999 ) , tanvir et al ( 2004 ) introduced the use of principal component analysis ( pca ) in studying cepheid light curves .
they showed that a major advantage of such an approach over the traditional fourier method is that it is much more efficient : an adequate fourier description requires , at best , a fourth order fit or 9 parameters , whilst a pca analysis requires only 3 or 4 parameters with as much as 81@xmath0 of the variation in light curve structure being explained by the first parameter .
later , leonard et al ( 2003 ) used the pca approach to create cepheid light curve templates to estimate periods and mean magnitudes for hst observed cepheids .
the purpose of this paper is to apply the pca technique to the study of rr lyrae light curves .
the mathematical formulation and error characteristics of pca are given in k02 and will only be summarized here .
the data used in this study were kindly supplied by kovacs ( 2002 private communication ) and used in kovacs and walker ( 2001 , hereafter kw ) .
these data consist of 383 rrab stars with well observed v band light curves in 20 different globular clusters .
kw performed a fourier fit to these data , which , in some cases , is of order 15 .
details concerning the data can be found in kw .
the data we work with in this paper is this fourier fit to the magnitudes and we assume that the fourier parameters published by kw are an accurate fit to the actual light curves .
we start with the data in the form used in kw : a list of the mean magnitude , period and fourier parameters for the v band light curve .
the light curve can thus be reconstructed using an expression of the form @xmath1 where @xmath2 is the mean magnitude , @xmath3 , @xmath4 the period , @xmath5 the fourier parameters given in kw .
these light curves are then rephased so that maximum light occurs at phase 0 and then rewritten as @xmath6 the @xmath7 are the light curve characteristics entering into the pca analysis ( k02 ) .
we then solve equation ( 4 ) of k02 , either after , or before removing an average term from the fourier coefficients in equation ( 2 ) .
with pca , the light curve is written as a sum of `` elementary '' light curves , @xmath8 where @xmath9 is the magnitude at time t , @xmath10 etc .
are the pca coefficients and the @xmath11 are the elementary light curves at phase or time t. these elementary light curves are not a priori given , but are estimated from the dataset in question .
each star has associated with it a set of coefficients @xmath12 and these can be plotted against period just as the fourier parameters in equation ( 1 ) are plotted against period .
we also note that the pca results are achieved as a result of the analysis of the _ entire _ dataset of 383 stars whereas the fourier method produces results for stars individually .
this feature of pca is particularly useful when performing an ensemble analysis of large numbers of stars obtained from projects such as ogle , macho and gaia .
solving equation ( 4 ) of k02 yields the principal component scores and the amount of variation carried by each component .
what we mean by this is the following : if we carry out an @xmath13 order pca fit , then pca will assume that all the variation in the dataset is described by @xmath14 components and simply scale the variation carried by each component accordingly .
table 1 shows this `` amount of variation '' quantity with and without the average term removed .
we see that in the case when we do not remove the average term the first pc explains as much as @xmath15 of the variation in the light curve structure . in the case
when we do remove the average term from the fourier coefficients , the first pca coefficient explains as much 81 percent of the variation in light curve structure . in either case , the first four components explain more than @xmath16 of the variation .
figures 1 and 2 show some representative light curves from our rrab dataset . in each panel of these two figures , the solid line is the fourier decomposition of order 15 ( that is 31 parameters ) used by kw , whilst the dashed line is a pca generated light curve of order 14 ( that is 15 parameters ) .
straightforward light curves such as the one given in the bottom and top left panels of figures 1 and 2 respectively are easily reproduced by our method .
the top left panel of figure 1 provides an example of an rrab light curve with a dip and sharp rise at a phase around 0.8 .
this is well reproduced by pca .
it could be argued that pca does not do as well as fourier in mimicking this feature , for example , in the bottom right panel of figure 2 .
however , the difference in the peak magnitudes at a phase of around 0.8 is of the order of 0.02mags .
it is also important to remember that the pca method is an ensemble method and analyzes all stars in a dataset simultaneously .
with fourier , it is possible to tailor a decomposition to one particular star .
this difference can be seen either as a positive or negative point about either technique . given this
, we contend that pca does remarkably well in describing the full light curve morphology of rrab stars . on the other hand , the fourier curve in the bottom left panel of figure 2 at this phase is not as smooth as the pca curve .
in fact the pca curves do not change much after about 8 pca parameters .
even though table 1 implies that the higher order pca eigenvalues are small , we feel justified in carrying out such a high order pca fit because its only after about 8 pca components that the fitted light curve assumes a stable shape .
the left panel of figure 3 displays an eighth order pca fit ( 9 parameters , dashed line ) and a fourth order fourier fit ( 9 parameters , solid line ) .
the fourier curve still has some numerical wiggles whilst the pca curve is smoother .
in addition , the two curves disagree at maximum light .
the right panel of figure 3 shows , for the same star , the same order pca curve as the left panel and an eighth order fourier fit ( 17 parameters ) .
now the two light curves agree very well .
note that in portraying the pca and fourier fits of reduced order in this figure , we simply truncated the original representations to the required level .
we suggest that figures 1 - 3 and table 1 provide strong evidence that pca is an _
efficient _ way to describe rrab light curve structure without compromising on what light curve features are captured by this description . figures 4 - 6 display plots of the first three pc scores plotted against log period for our sample .
the errors associated with these pca scores are discussed in section 4 of k02 and given in equation 6 of that section .
the orthogonal nature of these scores may well provide insight into the physical processes causing observable features in the light curve structure .
a detailed study of these plots , in conjunction with theoretical models , is left for a future paper .
figure 7 graphs v band amplitude against the first pca coefficient ( after averaging ) . we see a very tight correlation .
since table 1 implies that pca1 explains about @xmath17 of the variation in light curve structure , figure 6 shows that the amplitude is a good descriptor of rrab light curve shape , at least for the data considered in this paper .
although the fourier amplitudes are also correlated with amplitude , with pca , we can quantify , very easily , the amount of variation described by each pca component .
this has implications for both modeling and observation . on the modeling side , a computer code that can reproduce the observed amplitude at the correct period ,
will also do a good job of reproducing the light curve structure . on the observational side
, this provides insight into why we can use the amplitude , rather than a full blown pca or fourier analysis , to study the _ general _ trends of light curve structure .
this is why comparing theoretical and observational rrab light curves on period - amplitude diagrams works reasonably well , though we caution that a careful analysis should consider the finer details of light curve structure .
figures 6 and 7 display plots of the first two pca coefficients and fourier amplitudes , respectively , for our data , plotted against each other . whilst @xmath18 and @xmath19 are correlated with each other ,
@xmath20 and @xmath21 are not , by construction .
a similar situation would occur had we plotted @xmath18 or @xmath19 against @xmath22 .
this is another advantage of pca analysis of variable star light curves : the different pca components are orthogonal to each other .
a practical advantage of this feature is outlined in the next section .
a major goal of stellar pulsation studies is to find formulae linking global stellar parameters such as luminosity or metallicity to structural light curve properties .
if we are interested in the @xmath23 band magnitude , then we can write , @xmath24 where , since we do not know the function @xmath25 , we try to estimate it empirically .
two different approaches to quantifying light curve structure will , in general , yield different formulations of the function @xmath25 , but if there does exist a true underlying function @xmath25 , then both methods should give similar answers for @xmath26 , given the same input data . with a fourier based method ,
the function @xmath25 is related to the fourier amplitudes and phases , @xmath27 , usually with a linear relation . with a pca approach ,
we use the pca scores plotted in figures 2 - 4 . hence a pca relation , though also linear , will be different .
the nature of pca implies that the error structure in such formulae will be simpler and we quantify this below .
both formulations should , of course , give similar numbers for the final estimated value of the physical parameter in question , in this case , @xmath26 .
kw used the fourier method and found relations of the form , @xmath28 and , @xmath29 we note that these relations were obtained through an iterative procedure whereby outliers were removed and the relations re - fitted ( kovacs 2004 ) . in this paper , we use the pca method , but also , we use the entire dataset c mentioned in kw , consisting of 383 stars , and fit the relations just once .
we do not remove any outliers .
this may be why we obtain slightly different versions of the fit using fourier parameters than that published in kw .
for ease of comparison , we include in table 2 results obtained using both pca and fourier parameters .
this table gives the name for the relation , the independent variables considered and coefficients together with their standard errors .
the value of chi - squared in the table is defined as @xmath30 where @xmath31 is the fitted value of @xmath26 and @xmath32 are the number of stars and parameters respectively in the fit .
an examination of this table strongly suggests that * \1 ) similar relations to equations ( 4 ) and ( 5 ) between @xmath26 and the pca coefficients exist . *
\2 ) we can use an f test ( weisberg 1980 ) to test for the significance of adding a second and then a third pca parameter to the regression .
the f statistic we use is @xmath33 where @xmath34 are the residual sum of squares under the null and alternate ( nh and ah ) hypothesis respectively .
similarly , @xmath35 and @xmath36 are the degrees of freedom under these two hypotheses . for this problem ,
the null hypothesis is that the model with the smaller number of parameters is sufficient whilst the alternative hypothesis is that the model with the greater number of parameters is required . under the assumption of normality of errors , equation
( 7 ) is distributed as an @xmath37 , ( weisberg , 1980 , p. 88 ) . applying
this @xmath38 test implies firstly , that adding the first parameter @xmath20 is a significant addition to @xmath39 and secondly , that adding a second and third parameter , @xmath21 and @xmath40 are also highly significant with a p value less than 0.0004 . in the case of fourier parameters , adding the @xmath18 parameter to @xmath39 is highly significant and adding the @xmath22 parameter to this is also highly significant .
however , a formula involving ( @xmath41 ) has a p value of 0.0058 and a formula involving all 3 fourier amplitudes and @xmath39 is not a significant addition to a formula involving @xmath42 . * \3 ) the standard deviation of the fits given in the last column is generally slightly higher for the pca case , when considering similar numbers of parameters .
this is perhaps caused by the fact that the different pca components carry orthogonal sets of information . *
\4 ) the errors on the coefficients in the pca fits are always significantly smaller .
this is an important point when we evaluate the errors on the final fitted value of the absolute magnitude . *
\5 ) if we write the absolute magnitude as a function of parameters , @xmath43 , @xmath44 then the error on the absolute magnitude is given by , @xmath45 as table 2 indicates , @xmath46 is always smaller when the @xmath47 are pca coefficients rather than fourier amplitudes .
figure 8 and 9 portray graphs of @xmath20 vs @xmath21 and @xmath18 verses @xmath19 respectively .
we note that @xmath48 .
table 3 presents sample correlation and covariance coefficients between the period and pca parameters and period and fourier parameters .
table 3 , and figures 6 and 7 demonstrate that the correlation coefficient amongst any pair of pca coefficients is smaller than between any pair of fourier coefficients .
hence the error on the fitted value of @xmath26 , @xmath49 , _ has _ to be smaller when using a pca based formula .
we can use table 3 and equation ( 9 ) to formally calculate the error on @xmath50 .
table 4 presents these results .
the label in the top row of this table ( p1 , f1 , etc .
, ) refers to the appropriate relation in table 2 .
we see clearly that the pca formulae do better than their fourier counterparts with a similar number of parameters . when we consider the @xmath51 and @xmath52 variables ,
then the `` error advantage '' using a pca based method is a factor of two .
this occurs not just because the pca coefficients are orthogonal to each other , but also because the errors on the coefficients in a pca based formula are significantly smaller than in the fourier case .
figure 10 displays a plot of the predicted absolute magnitudes obtained using a two parameter ( @xmath53 ) fourier fit and the three parameter ( @xmath54 ) pca fit .
the two approaches are displaced from each other because we do not consider the constants in this study . disregarding this
, it can be seen that the slope of this plot is 1 : hence the two methods produce similar relative absolute magnitudes .
we have shown that the method of pca can be used to study rr lyrae light curves .
it has distinct advantages over a fourier approach because * \a ) it is a more efficient way to characterize structure since fewer parameters are needed .
a typical fourier fit requires 17 parameters whereas a pca fit may only need 9 . *
\b ) using the pca approach , we see clearly why the amplitude is a good descriptor of rrab light curve shape . * \c ) the different pca components are orthogonal to each other whereas the fourier amplitudes are highly correlated with each other .
this leads to relations linking light curve structure to absolute magnitude using pca having coefficients with smaller errors and leading to more accurate estimates of absolute magnitudes .
this can reduce the formal error , in some cases , by a factor of 2 . in the present formulation of our pca approach ,
the input data is a fourier analysis .
if these input data , that is the fourier decompositions , contain significant observational errors , the error bars on the resulting principal components will be larger . neither the pca or fourier approach can compensate fully for noisy data . in this sense , the sensitivity of pca to noisy data should be similar to fourier , though the fact that pca is an ensemble approach in which we initially remove an average term does guard against individual points having too much undue influence .
as an example , table 4 of kw gives 17 outliers ( in terms of their fourier parameters ) , which kw removed in their analysis relating absolute magnitude to fourier parameters .
we do _ not _ remove these outliers , yet , in terms of the final fitted magnitudes presented in figure 10 , pca and fourier produce very similar results .
further , even with the inclusion of these 17 stars , the pca method still produces pca coefficients with smaller errors as given in tables 2 and 3 .
kanbur et al ( 2002 ) discuss in detail the nice error properties of the pca method as applied to variable stars and give a recipe with which to calculate errors on pca coefficients .
their figure 2 , albeit for cepheids , displays error bars on these coefficients .
we see that even with noisy data , the progression of pca parameters with period is preserved , though of course , the error bars on the pca coefficients are larger .
ngeow et al ( 2003 ) developed a simulated annealing method which can reduce numerical wiggles in fourier decomposition of sparse data .
ngeow et al ( 2003 ) give specific examples of how such an approach improves fourier techinues using ogle lmc cepheids .
a similar result will hold true for rr lyraes . hence this annealing technique couple with a principal component analysis should prove very useful when dealing with noisy rr lyrae data and will be treated in detail in a subsequent paper .
our pca results are based on a sample of 383 stars in globular clusters .
how transferable are our results and how can our results be used to obtain pc coefficients for a new rr lyrae light curve which appears to be normal ( ie no signs of blazhko effects etc . ) ?
our results are transferable to the extent that the original 383 stars are a good representation of the entire population of rrab stars , including variation in metallicity and differences between field and cluster variables .
given this caveat , we suggest two methods to reproduce the light curve of a new rrab star .
firstly , it is straightforward to include the new star in the pca analysis with the existing dataset .
this is our recommended approach and preserves the `` ensemble analysis '' property of our pca method .
our second method will be the subject of future paper but briefly it is this .
we fit the progression of the pca coefficients with period , such as given in figures 4 and 5 , with simple polynomial functions . as an aside
, we remark that figure 4 contains significant scatter , perhaps associated with metallicity , so that it would be best to include metallicity in such polynomial fits . for a new star , we then guess its period and read off , for that period , the value of the pca coefficients . equation ( 3 ) then allows us to generate the light curve .
we iterate this until a specified error criterion is satisfied .
we can then use existing formulae relating absolte magnitude to light curve structure as defined by pca .
this pca template approach has been used , with considerable success , in analysing hst cepheid data ( leonard et al 2003 ) .
we note from table 2 that the chi square on the fitted relations are similar for pca and fourier .
does this mean that despite the smaller formal errors with pca , both methods ability to predict rrab absolte magnitudes is limited by the intrinsic properties of rrab stars themselves ?
to some extent this is true .
jurcsik et al ( 2004 ) , in analysing accurate data for 100 rrab stars in m3 , show that for some 16 stars , amongst which there exist some pairs whose absolute mean magnitudes differ by about 0.05 mags ( the accuracy of the photometry is about 0.02mags ) , the fourier parameters and periods are very similar .
that is , an empirical method relating absolute magnitude to period and fourier parameters in one waveband could not distinguish between these stars .
since , as jurcsik et al ( 2004 ) point out , their data contains a small range of both mass and metallicity , temperature is the only other variable , it may be the case that multiwavelength information is needed .
it is worthwhile to investigate how pca fares with this dataset .
here we give an outline that suggests that pca can be more efficient at extracting information from the light curve .
for the sixteen stars which had differing absolute magnitudes but very similar fourier parameters , we can perform the following procedure : for every pair , @xmath55 , we calculate @xmath56 @xmath57 @xmath58 @xmath59 and @xmath60 where @xmath61 are the fourier amplitudes and @xmath62 are the pca coefficients and @xmath63 are the mean magnitudes . in the above
, we always take the absolute value of the differences .
we need to take fractional changes because the fourier amplitudes and pca coefficients have different ranges .
we now plot diff3 against diff1 and diff2 .
this is presented in figure 11 , where the open squares are diff1 and the closed squares are diff2 .
we see that with pca ( closed squares ) , the differences between light curve structure parameters are greater than with fourier ( open squares )
. this could imply that pca can be more efficient though the limitations associated with using a single waveband are still present
. a more rigorous , quantitative discussion of this , in a fisher information sense , will be given in a future paper .
in other future work we plan to investigate the applicability of this method to light curve structure - metallicity relations , rrc stars and a comparison of observed and theoretical light curves using pca .
.percentage of variation explained by pc components [ cols="<,^,^,^,^,^,^",options="header " , ]
smk thanks geza kovacs for stimulating discussions and for kindly supplying the rrab dataset .
smk thanks d. iono for help writing the pca and least squares program and c. ngeow for help with latex .
hm thanks fcrao for providing a summer internship in 2001 when part of this work was completed .
we also thank the referee for constructive comments .
hendry , m. a. , tanvir , n. a. , kanbur , s. m. , 1999 , asp conf .
series , 167 , p. 192
kanbur , s. , iono , d. , tanvir , n. & hendry , m. , 2002 , mnras , 329 , 126 jurcsik , j. , benko , j. , bakos , g. , szeidl , b. , szabo , r. , 2004 , apjl , 597 , 49 kovacs , g. , walker , a. r. , 2001 , a&a , 371 , 579 kovacs , g. , 2004 , private communication leonard , d. , kanbur , s.,m . , ngeow , c. , tanvir , n. , 2003 , apj , 594 , 247 ngeow , c. , kanbur , s. m. , nikolaev , s. , tanvir , n. , hendry m. , 2003 , 586 , 959 tanvir , n. r. , hendry , m. a. , kanbur , s. m. , 2004 , in preparation weisberg , s. , 1980 , _ applied linear regression _ , john wiley & sons , @xmath64 ed . | in this paper , we analyze the structure of rrab star light curves using principal component analysis .
we find this is a very efficient way to describe many aspects of rrab light curve structure : in many cases , a principal component fit with 9 parameters can describe a rrab light curve including bumps whereas a 17 parameter fourier fit is needed . as a consequence we show show statistically why the amplitude is also a good summary of the structure of these rr lyrae light curves .
we also use our analysis to derive an empirical relation relating absolute magnitude to light curve structure . in comparing this formula to those derived from exactly the same dataset but using fourier parameters , we find that the principal component analysis approach has distinct advantages .
these advantages are , firstly , that the errors on the coefficients multiplying the fitted parameters in such formulae are much smaller , and secondly , that the correlation between the principal components is significantly smaller than the correlation between fourier amplitudes .
these two factors lead to reduced formal errors , in some cases estimated to be a factor of 2 , on the eventual fitted value of the absolute magnitude .
this technique will prove very useful in the analysis of data from existing and new large scale survey projects concerning variable stars .
rr lyraes stars : fundamental parameters |
at the present time more than @xmath8 exo - planets are known which have been mostly discovered by the indirect method of the variation of the radial velocity induced by planets orbiting around the central star .
it can be shown that young planets can be detected with adaptive optics ( ao ) imaging techniques at @xmath21-@xmath6 meter telescopes if they are sufficiently distant from their parent star .
since it is reasonable to assume a coevality between the parent star and the planet , we can retrieve from atmospherical models ( burrows et al .
1997 , baraffe et al . 2003 ) that planets having a mass in the range @xmath22-@xmath6 m@xmath4 and orbiting around young late - type stars ( @xmath6-@xmath23 myr ) have a typical brightness contrast with respect to the parent star of the order of @xmath24-@xmath25 , corresponding to a magnitude difference of @xmath26 of @xmath3-@xmath14 mag .
this means that a planet can be detected at a few tens of aus from parent stars having distances @xmath27 @xmath1 pc .
first results on the search for young and massive planets are reported by neuhuser et al .
( 1997 ) using 4 m class telescopes and macintosh et al .
( 2001 ) and keisler et al .
( 2003 ) using 10 m class telescopes . recently
( chauvin et al . 2004 ) , a companion of @xmath12 @xmath3@xmath28@xmath29m@xmath4 was detected at around @xmath1 au from the central star 2m1207 , a very late type star ( m8 ) in twa hydrae ( @xmath12 @xmath6 myr ) .
this object is , at the present time , the most interesting giant planet candidate ever detected and further confirmation of the orbital motion is needed . from the point of view of the scientific motivations related to this kind of search
it is certainly an important challenge to image directly a planet around another star , since it can provide a direct evidence for the existence of such a planet . at the present time
only a few exo - planets have been detected with the photometric technique called _
transit _ ( hd209458b - charbonneau et al .
2000 , ogle - tr-113 ( konacki et al . , 2004 , bouchy et al . 2004 ) , ogle - tr-56 ( torres et al . , 2004 ) , ogle - tr-132 ( bouchy et al . 2004 ) , ogle - tr-111 ( pont et al . ,
2004 ) and trres-1 ( alonso et al . , 2004 ) ) . moreover , direct imaging is complementary to other detection methods like the radial velocity variation technique .
the latter is sensitive to objects orbiting at close distances from the parent star ( the radial velocity induced by the presence of a companion is indeed proportional to the inverse of the square root of the orbital distance ) while the ao imaging presented here would be sensitive to planets orbiting at much larger distances .
according to statistical results retrieved by radial velocity studies ( marcy et al .
2003 ) , most of the known exo - planets orbit at distances smaller than @xmath19 au and the semi - major axes of exo - planets found so far are not larger than @xmath12@xmath30 au .
since the radial velocity technique is heavily biased towards short separations , the derived major - axis distribution is certainly not representative .
not much is known at the present about potential planets that could exist at distances larger than about @xmath30 au .
we note that the formation of exo - planets is quite unlikely at distances of a few tens of au . at these large distances
the density of the planetesimals is low and the gravity from the central star is weak .
this makes the process of planet formation inefficient and slow ( thommes et al .
, 1999 , 2002 ) . however , different models predict the transfer ( jump ) or migration of planets to distances larger than @xmath3-@xmath6 au from the central stars . among these models ,
we recall the gravitational scattering model which claims that in multi - planetary systems , due to gravitational interactions between the planets , it can happen that the lowest mass objects are ejected on hyperbolic trajectories and some planets can move on stable orbits of a few tens of au from the parent star .
such models were proposed by weidenschilling @xmath31 marzari ( 1996 ) and rasio @xmath31 ford ( 1996 ) and revised more recently by papaloizou & terquem ( 2001 ) and ford et al .
an alternative mechanism that could induce an outward migration is the planet - disc interaction ( veras & armitage , 2004 ) .
it was proven that stable orbits of a few tens of au can be attained by a few jupiter mass planet under the effect of torques induced by the disc ( i.e at early stages of the star formation ) if the system has photoevaporation wind .
numerical simulations ( veras & armitage , 2004 ) showed that , under the assumption of a typical solar - mass photoevaporation wind of the order of @xmath19-@xmath3@xmath32@xmath33 m@xmath34yr@xmath35 , an outward migration of a planet with few jupiter masses can take place with a frequency of the order of @xmath36 @xmath2 ( with respect to a sample of a few hundreds of simulations ) .
it is , therefore , interesting to find out if young and massive planets are present up to a few tens of au from the central star . for these reasons , we planned and performed an extended survey aiming to search for massive exo - planets orbiting around nearby and young stars using one of the best available instruments at the present time for high - contrast ao imaging observations : naco at the very large telescope ( vlt ) ( see section [ naco ] ) .
the plan of the paper is the following . in section [ targets ]
the criteria used for the selection of the targets are presented . in section [ naco ]
we describe the observational strategy and we provide technical information on the observations . in section [ proc ]
we present our data processing , in section [ planets ] we discuss an analysis of planet - like features identified in deep images and in section [ sensit ] we give a detection limit estimate for the observed targets . in section [ disc ]
we discuss our results and finally , in section [ conc ] the conclusions of our study are presented .
the properties of planets that we are searching for are : * ( a ) * the contrast @xmath26 between the planet and the central star should be @xmath3-@xmath14 mag * ( b ) * and the planets should orbit at distances @xmath37 @xmath38 @xmath19 au from the central star . the planet intrinsic luminosity , retrieved by evolutionary models ( burrows et al . 1997 , baraffe et al .
2003 ) , depends on the age of the system and the mass of the planet .
the magnitude of a planet depends also on the wavelength used for the observations . due to the fact that planets strongly fade with age we selected stars as young as possible . following the evolutionary models
, we define our selection ranges as follows : age @xmath12 @xmath6-@xmath23 myr , mass of the planets @xmath12 @xmath22-@xmath6 m@xmath4 and wavelengths in the @xmath39 and @xmath40 bands .
obviously the nearer the target is from the observer , the smaller is the distance from the central star that we can reach and probe .
therefore , we gave priority to nearby objects .
we note that most ( @xmath12 @xmath41 @xmath2 ) of our targets have distances @xmath42 @xmath43 pc and only @xmath22 objects are located between @xmath1 and @xmath44 pc . in order to optimize the contrast conditions , we gave priority to late - type stars ( spectral type k - m ) .
only one target ( hd @xmath45 ) has a spectral type g5iv .
table [ tab1 ] and [ tab2 ] summarizes the properties of the selected targets and the theoretical magnitude of planets that might orbit around them .
the magnitude of the stars in the @xmath39 and @xmath40 bands are retrieved from the magnitude in v band and the spectral type ( simbad astronomical database[multiblock footnote omitted ] ) .
table [ tab1 ] gives also the references from which the targets were selected .
a relatively large fraction of targets belong to the tw hydrae local associations , @xmath11 pictoris moving groups and tucana / horologium association ( zuckerman et al .
2001a , 2001b ; song et al .
2003 , barrado y navascues et al . 1999 ,
barrado y navascues 2001 ) .
the age of tw hydrae is estimated to be @xmath12 @xmath6 myr ( stauffer et al .
1995 ) , the age of @xmath11 pictoris is @xmath12 @xmath13 myr ( barrado y navascues et al .
1999 ) and the tucana / horologium age is @xmath12 @xmath46-@xmath47 myr ( zuckerman & webb 2000 ) .
several other targets , members of young stellar kinematic groups , were selected from montes et al .
( 2001 ) .
we want to emphasize that the ages given in table [ tab1 ] are often approximate values , particularly for those stars which are not members of associations or moving groups . for sufficiently young low - mass stars ( @xmath42 @xmath46-@xmath1 myr )
their location in theoretical hertzsprung russel diagrams can be used for an age estimate . for older stars ( @xmath38 @xmath1 myr ) the isochrones are often indistinguishable from stars on the zams
( see for example barrado y navascues 1998 ) and therefore other age calibrators have to be employed like the stellar magnetic activity , the lithium abundance , the stellar rotation rate and the x - ray flux .
in addition to these photometric signatures , the galactic space velocity is often employed for an age estimate and for defining young moving groups .
young stars are indeed characterized by low dispersion of the galactic space - velocity components ( u , v , w ) .
a few criteria exist in the literature defining precise ranges for these components ( eggen 1996 , jeffries 1995 ) .
these same criteria are employed by montes et al .
( 2001 ) .
we avoided the selection of binary systems with separations smaller than @xmath12 @xmath29@xmath48 because the probability that a planet can find a stable orbit around one of the binary component is rather small .
the binary systems included in our survey are : twa @xmath3 ( @xmath29@xmath48 ) , twa @xmath21 ( @xmath49@xmath48 ) , twa @xmath50 ( @xmath30@xmath48 ) , gj @xmath51 ( @xmath52 ) and bd @xmath53 @xmath54 ( @xmath55 ) .
the observations were performed with the ao system naco , consisting of the infrared camera conica ( lenzen et al .
1998 ) assisted by the adaptive optics facilities naos ( rousset et al .
2000 ) placed at the focus of the unit `` yepun '' of the eso / vlt @xmath21 m telescope .
we observed @xmath46 targets during three observing runs : 16 - 19 february 2003 , 20 - 23 july 2003 and 7 - 9 january 2004 .
two targets ( gj @xmath56 and gj @xmath57 ) were observed in service mode .
table [ tab3 ] summarizes the observational instrumental parameters . due to the rather bright central star ,
our observational strategy was the following .
first , we observed each target in a narrow - band filter ( table [ tab3 ] ) for a few minutes in order to obtain a non - saturated point spread function ( psf ) .
we obtained a strehl ratio ( sr ) in the range @xmath58-@xmath59 over the all runs .
second , we obtained deep and saturated ( @xmath12 @xmath60-@xmath46 min ) @xmath40 or @xmath39 broad - band observations of each target ( table [ tab3 ] ) in order to increase the detectability on the psf wings .
we used naco at a pixel scale of @xmath61 mas yielding a fov of @xmath62 .
the optical wavefront sensor was locked on the target star yielding a typical resolution of @xmath12@xmath1 mas in @xmath63 band . for the deep observations , we selected the exposure time ( dit in table [ tab3 ] ) of a single frame to obtain a number of counts ( for the central star ) larger than @xmath12 @xmath64 times the saturation limit .
the deep images are saturated over a central diameter having a size of @xmath65-@xmath66 . for each target , a data set of dithered frames was obtained in order to correct for sky emission and the pixel to pixel variations in sensitivity . in a few cases ( e.g. targets characterized by a smaller brightness ) we used a broad - band filter with a neutral density filter ( nd short - see table [ tab3 ] ) instead of a narrow - band filter to obtain a non - saturated psf . for hd @xmath67 ,
due to its high brightness , we used a different observational strategy based on the absorption properties of a planet atmosphere .
this target was observed with the intermediate - band filters ( ib @xmath68 and ib @xmath69 ) centered out and in the methane ( ch@xmath70 ) absorption band ( @xmath12 @xmath69 @xmath71 m ) .
we expect that planet features would appear in the continuum ( ib @xmath68 ) but not in the absorption band ( ib @xmath69 ) . due to the higher strehl ratio ( sr )
reachable in the @xmath63 band compared to the @xmath39 band , we gave priority to the former band in order to optimize the probability to detect a low - mass companion .
we chose the @xmath39 band only for the oldest stars ( @xmath8-@xmath23 myr ) , i.e. for planets having a theoretical colour ( @xmath39-@xmath72 ) @xmath73 @xmath74 @xmath75 mag .
we note that , according to the cond models ( baraffe et al .
2003 ) ( see fig.[fig1 ] ) and for planet masses of @xmath22-@xmath6 m@xmath4 , the @xmath39 band is comparable to the k band for an age minor than @xmath8 myr . for targets with an age of @xmath8-@xmath23 myr the gain of the h band with respect to the k band is significant .
the procedure used to reduce our data set includes iraf and eclipse packages .
each frame was cleaned by a bad - pixel mask and divided by the flat field .
in addition , the sky background was eliminated in each frame .
the final deep image ( shift and add of all the frames ) was obtained using a cross - correlation routine in eclipse .
we underline that the shift and add had to be done using a cross - correlation method since our psfs are saturated in the central zone .
alternative methods , based on the gaussian fit to the peak of a psf fail in such a case . in order to save observational time
, we did not observe any reference standard star but we filtered the final deep images from the low spatial frequencies using two different methods . in this way the large - scale features of the saturated psf were removed .
in the first method , we convolved the deep image of the scientific target with a two - dimensional gaussian function having a width roughly equivalent to the fwhm of the psf of the star .
then we subtracted the result from the original deep image . only features identified outside the central saturated region ( @xmath12 @xmath65-@xmath66 )
are reliable in all cases .
we note that , if the fwhm of the gaussian function is much smaller than the fwhm of the psf of the star , the filtering is not efficient .
this means that large - scale features of the saturated psf are not sufficiently eliminated and the potential planet - like features close to the star are not visible .
if , on the contrary , the fwhm of the gaussian function is too large , the filtering eliminates also the potential planet - like features close to the star i.e. the structures having small spatial scale . in the second method , we selected , for each point of the deep image , an annulus . calculating the average of these frames we obtain the estimate of the background of the deep image inside an annulus around the centre of our target .
we finally subtracted the background from the original image .
the annulus is characterized by an inner radius of half of the psf @xmath76 and an outer radius @xmath12@xmath22 larger than the internal radius .
the central part of the psf is excluded in the calculation of the background because it introduces an offset in the average estimation . reducing the size of the inner radius
, we increase the filtering of the low spatial frequencies . increasing the outer radius the estimation of the background
is affected by features belonging to regions far away from the central part of the psf .
we define an outer radius equal to @xmath22 times the inner radius in order to estimate a background on a limited region around the central part of the psf .
we note that the final detection limit only weakly depends on the values of the inner and outer radius . repeating the filtering with several combinations of inner and outer radii the final detection limit shows differences within @xmath58 mag .
the filtering was applied in a region of @xmath12 @xmath30@xmath48@xmath32@xmath30@xmath48 centred on each target .
although the two methods have differences , they are based on the same principle aiming to eliminate the large - scale features of the saturated psfs . in section [ sensit ]
the similarities / differences of the two methods are discussed with respect to the results that we obtained .
after data reduction and filtering of all the observations as described in sec.[proc ] , a visual inspection of all the deep images was performed and a few point - like structures were identified in some of the images : v@xmath77 nor ( @xmath63 band ) , sao @xmath78 ( @xmath39 band ) , gj @xmath57 ( @xmath39 band ) , bd @xmath79 @xmath80 ( @xmath39 band ) , bd @xmath81 @xmath82 ( @xmath39 band ) , gj @xmath83 ( @xmath39 band ) and hd @xmath84 ( @xmath39 band ) .
all the other targets did not reveal any point - like sources around them .
we discuss here the nature of these objects trying to distinguish between artifacts ( ghost images ) and faint point sources at low galactic latitude .
we have verified that all the targets observed in h band show three point - like features at the same position at about @xmath19@xmath48 from the star .
fig.[bd_filter ] shows deep images of two of these targets : bd @xmath79 @xmath80 and bd @xmath81 @xmath82 .
three point - like features are visible ( indicated by arrows and letters a , b and c ) at the same relative angular separation from the star .
these features are probably ghost images produced by some imperfection of the broad - band filter or static aberration residuals that depend on the wavelength .
this could justify the fact that they are visible only in @xmath39 band . in only two deep images ( v@xmath77 nor and sao @xmath78 - fig.[sao_cand ] ) we observed clear point - like sources different from those created by the @xmath39 filter that we have just described .
fig.[sao_cand]a shows the deep image of sao @xmath78 ( galactic latitude = @xmath85 ) .
an object at around @xmath36-@xmath46 au from the star is visible .
the estimated magnitude of this object is @xmath12@xmath86 mag in @xmath39 band . due to the low galactic latitude of sao @xmath78 , the probability that this feature is a background star is not negligible .
fig.[sao_cand]b shows the deep image of v343 nor ( galactic latitude = @xmath87 ) .
several sources were detected around this star .
the estimated magnitude of all the sources is in the range @xmath14-@xmath60 mag .
we note that a @xmath3m@xmath4 planet orbiting around this @xmath13 myr old star would have @xmath12 @xmath88 mag in @xmath72 band . besides this
, we observe that the galactic latitude of v@xmath77 nor is relatively low ( @xmath89 ) and several among these point - like sources appear a few arcsec away from the central star .
thus the probability that these sources are background stars is quite high .
we can not totally exclude that the faint sources around sao @xmath78 and v@xmath77 nor are gravitationally bound massive planets .
for this reason , a second epoch observation is necessary in order to clarify the nature of the detected point - like sources .
knowing that the proper motions ( hipparcos catalog , perryman et al .
1997 ) of sao @xmath78 are ( @xmath90@xmath91 -202 mas / yr and @xmath92@xmath91 -270 mas / yr ) and those of v@xmath77 nor are ( @xmath90@xmath91 -@xmath93 @xmath28 @xmath94
mas / yr and @xmath92@xmath91 -@xmath95 @xmath28 @xmath96 mas / yr ) we deduce that after one year a re - observation could throw light on the nature of these objects .
a forthcoming run is planned to definitely verify if these objects are indeed background stars .
we remind the reader that an estimation of the probability that a point source having a faint magnitude comparable to ours in a field of view of @xmath29@xmath48 is a background star was done for a galactic latitude of @xmath12 + @xmath97 ( brandner et al .
a probability of @xmath41@xmath2 was estimated in that case .
our point sources are placed at a lower galactic latitudes ( @xmath87 and @xmath85 ) , and the point - sources are placed at around @xmath29@xmath48 from the central star ( in the sao @xmath78 case , fig.[sao_cand ] ) and at distances larger than @xmath29@xmath48 ( in the v@xmath77 nor case , fig.[sao_cand ] ) .
we conclude , that the probability that such sources are background stars is larger than @xmath41@xmath2 .
we conclude , thus , that no very promising massive - planet candidate was identified in our survey .
to estimate quantitatively the detection limit of our deep images as a function of the angular separation from the central star , we used the following procedure . along a radial direction we calculated the standard deviation of the intensity over a box of @xmath98@xmath32@xmath98 pixels ( with typical values of @xmath98 of @xmath15-@xmath30 pixels which corresponds to the fwhm of the non - saturated psfs ) and one pixel step .
we then averaged over all radial directions ( @xmath99@xmath100 ) and finally we calculated the contrast @xmath26 at @xmath3@xmath101 with respect to the peak of the non - saturated psf of the star . since the central star is saturated in the broad - band images we had to use the fluxes measured in the non - saturated narrow - band filters ( see table [ tab3 ] ) to calculate what peak flux our saturated broad - band images would have if they were not saturated .
the non - saturated psf had to be multiplied by a constant factor ( @xmath102 ) to take into account the different transmission and width of the filters .
if @xmath103 is the filter used to obtain the saturated psf and @xmath104 the one used to obtain a non - saturated psf , @xmath102 is given by : @xmath105 where @xmath106 is the filter transmission coefficient , @xmath76 is the full width half maximum of the filters and dit is the integration time of each single frame ( table [ tab3 ] ) . in order to illustrate that our detection limit is reasonable , we scaled the non - saturated psf of one of our targets ( v343 nor ) to a flux value corresponding to a @xmath3m@xmath4 planet ( representative of the typical planets we are interested in ) and we placed this artificial planet in the deep image at different angular separations from the star .
fig.[mascia ] ( a)-(b ) shows the filtered deep images of the same target processed with the two filtering methods described above ( section [ proc ] ) .
this `` artificial planet '' is placed at @xmath107 and @xmath108 from the central star .
fig.[mascia ] ( a)-(b ) shows that a separation of @xmath108 is roughly the minimum angular separation at which a @xmath3m@xmath4 planet can be detected .
figures [ det_lim1]-[det_lim4 ] show the detection limits as a function of angular separation for @xmath0 objects of our survey for which deep images have been obtained .
we note that , in one case ( gj @xmath109 ) , a technical problem did not allow to observe the non - saturated psf and this prevented us to calculate the detection limit . in the case of the binary gj @xmath51 , we could calculate only the detection limit for gj @xmath51 a. the separation of @xmath12 @xmath52 between the two components was slightly too large to include both stars in the same frame in all the dithered positions . for all targets , the detection limit is reliable starting for angular separations @xmath38 @xmath65 . for angular separations @xmath42
@xmath65 the psf of the deep images is usually saturated .
fig.[det_lim4 ] shows the detection limit of the two targets ( hd @xmath67 and bd -17@xmath100 @xmath54 ) that we observed through the more narrow `` methane band filters '' ( see section [ naco ] ) .
since the optical quality of these intermediate - band filters is worse compared to the broad- and narrow - band filters , the final deep images appear dominated by structures characterized by high spatial frequencies .
for this reason , the filtering is less efficient for these two deep images ( the dashed , dotted and full thin lines are quite similar - fig.[det_lim4 ] ) . from fig.[det_lim1 ]
, [ det_lim2 ] , [ det_lim3 ] we can see that : * the detection limit of the filtered deep images is very similar for the two different filtering methods .
this indicates that the achieved gain from the applied filtering is practically method independent . * the filtering of large spatial structures ( low spatial frequencies )
has a variable efficiency .
the angular separation range over which the filtering shows its maximum efficiency is @xmath110-@xmath111 from the central star . in this region the gain in contrast @xmath26 with respect to the non - filtered image ranges from @xmath112 to @xmath29 mag .
* the detection limit is not the same for all the stars .
this is not surprising because it depends on several parameters : the age , the magnitude of the central star , the wavelength used for the observations , the distance of the star from the observer and also on the exposure time . in one case ( the brown dwarf twa @xmath3b ) , the detection limit is rather low because we did not saturate the psf since the central star is relatively faint . to calculate the typical performances of naco at subarcsecond distances , we averaged the contrast @xmath26 obtained at @xmath110 and @xmath107 for all the observed targets having a total integration time within the range @xmath60-@xmath46 minutes .
we find , in the @xmath63 band , a @xmath26 @xmath91 @xmath50 mag at @xmath110 and a @xmath26 @xmath91 @xmath113 mag at @xmath107 .
we find , in @xmath39 band , a @xmath26 @xmath91 @xmath114 mag at @xmath110 and a @xmath26 @xmath91 @xmath115 mag at @xmath107 .
we recall that @xmath110 corresponds , for our sample , to projected separations of @xmath29-@xmath116 au , @xmath107 to projected separations of @xmath15-@xmath41 au .
luhman & jayawardhana ( 2002 ) , observed @xmath36 targets with nicmos ( keck ii ) and found a @xmath26 @xmath12 @xmath6-@xmath117 mag at @xmath107 in the @xmath39 band with an integration time of @xmath12 @xmath118 minutes .
this illustrates nicely that naco is a well performing instrument for high - contrast imaging observations .
the detection limit estimates also permit us to calculate an upper limit of the projected separation between a star and a potential planet , as follows .
we consider a typical @xmath3m@xmath4 planet , we calculate the theoretical value ( retrieved by the models ) of the star / planet contrast , we then place this value on the ordinates of the detection limit curve ( see for example fig.[det_lim1 ] ) and finally we retrieve ( on the abscissae ) the distance beyond which we can affirm that such a planet should not exist ( because it was not detected ) .
table [ tab4 ] shows , for each target , such upper limits for the star / exo - planet distance .
table [ tab5 ] shows the frequency at which we would be able to detect a @xmath3m@xmath4 planet at different distances from the parent star . by analysing these results from a statistical point of view
, we can calculate ( fig.[prob ] ) the cumulative distribution of the upper limit of the distance star / exo - planet for our sample of targets .
the thin line represents the cumulative distribution obtained for @xmath3m@xmath4 planets .
we find that ( fig.[prob ] ) , with respect to our sample of targets , in @xmath119 of cases ( median value ) , there are not planets at distances larger than @xmath5 au and , in @xmath120 of cases , there are no planets at distances larger than @xmath10 au .
we underline that we are considering a @xmath3m@xmath4 planet just as a reference , but the same calculation could be done for more and/or less massive planets in the range @xmath22-@xmath6 m@xmath4 . in the case of more massive planets , the cumulative distribution should have a median value ( @xmath119 ) associated with an angular separation smaller than @xmath5 au . in the case of less massive planets , the cumulative distribution should have a median value ( 50 @xmath2 ) associated with an angular separation larger than @xmath5 au . to better appreciate the sensitivity with respect to the mass of the planets we also calculated the cumulative distribution for @xmath6m@xmath4 planets ( fig.[prob ] - bold line ) .
we find that , in @xmath119 of the cases ( median value ) there are no planets at distances larger than @xmath7 au and in @xmath120 of the cases there are not planets at distances larger than @xmath9 au .
table [ tab4 ] and table [ tab5 ] show the upper limit of the projected star / planet separation and statistical results also for @xmath6m@xmath4 planets .
the numbers given in these tables illustrate that it is relatively unlikely that a massive planet exists beyond a certain distance from the star .
for example the fact that no @xmath3m@xmath4 planet was found for @xmath118 of our targets beyond @xmath60 au ( see table[tab4 ] ) means that the likelood for the existance of such planets beyond this distance is @xmath12 @xmath121 @xmath2 i.e. @xmath12 @xmath122@xmath2 . this last estimation is useful for the reader to have an idea of how much the size of the sample affects the statistical results .
we recall that we are using the term _ probability _ in a less conventional way . a precise ( statistically speaking ) definition of _ probability _ should need some positive detection that we obviously do not have .
which are the implications of these results with respect to models such as those cited in section [ intr ] ( gravitational scattering model and outward migration ) , explaining the transfer of planets to large distances ( some tens of au ) from the central star ?
our results indicate that the frequency of planet jumps or outward migration is quite low .
more precisely , it seems difficult for @xmath6m@xmath4 planets to find a stable orbit at distances larger than @xmath88 au , and that @xmath3m@xmath4 planets at distances larger than @xmath123 au , at least in the range of early ages that we considered .
we note that the statistical analysis presented in fig.[prob ] related to the @xmath6m@xmath4 planets is particularly interesting because , due to a steeper cumulative distribution concentrated at small projected separations from the central star , it permits us to better constrain the probability of location of planets .
we also recall that such massive planets are rarely detected at distances smaller than @xmath30 au using radial velocity measurements .
as described in the introduction , the radial velocity technique is sensitive only to planets orbiting at closer distances ( @xmath73 @xmath30 au ) from the central star .
it is interesting to analyse what the direct imaging technique reveals about @xmath6m@xmath4 planets at distances larger than @xmath30 au .
looking at our results , we can state that , for most of the targets ( @xmath12 @xmath124 of cases ) , there are no @xmath6m@xmath4 planets at distances larger than @xmath12 @xmath88 au .
we find that the probability to find planets equal or more massive than @xmath6m@xmath4 at large distances is low as well as the probability to find them at close distances ( @xmath42 @xmath30 au ) as appears from radial velocity surveys ( marcy et al .
massive planets are expected to form beyond the so called _
line _ distance ( @xmath12 @xmath3 au for the solar - like stars au for typical solar - like star .
this distance is calculated for simple thermal equilibrium and is estimated equal to @xmath19-@xmath29 au for late type stars ( black 1980 ) ] ) ( lissauer , 1993 ) and from this distance they might migrate inward .
our study shows that , at least for late spectral type ( k and m ) stars , if such massive ( @xmath12 @xmath6m@xmath4 ) planets form beyond the snow line , we can envisage the following possibilities : * ( 1 ) * massive planets form and can be found in the narrow range of distances _
snow line _ - @xmath88 au from the central star , * ( 2 ) * at distances larger than @xmath88 au only planets having masses smaller than @xmath6m@xmath4 can form and can be found , * ( 3 ) * if any @xmath6m@xmath4 planet would have formed beyond the _
snow line _ , the formation and inward migration necessarily happen on a time scale @xmath42 @xmath6myr .
to test the hypotheses ( 1 ) and ( 2 ) one should improve the contrast at small angular separations . to test the hypothesis ( 3 )
one should select a sample of stars younger than ours .
unfortunately , stars having this property are placed preferably at distances @xmath38 @xmath8 pc from the observer .
even future ground - based planet finders will be able to detect planets only at larger separations ( a few tens of au ) from the central star if the targets are placed at such distances ( @xmath38 @xmath8 pc ) .
for this reason , it would be probably more efficient and appropriate to observe them from space . in this way
, one could by - pass the limitations imposed by the atmosphere on the attainable angular resolution and to probe the same projected separations ( as those analysed in this paper ) from these distant stars .
the resolution reachable in space by the hubble space telescope ( hst ) is not good enough to probe such small physical distances . in a previous study ( brandner et al .
2000 ) , a set of pre - main sequence t tauri stars in the chamaleon t ( @xmath19-@xmath3 myr at @xmath125 pc ) and scorpius - centaurus ob ( @xmath3-@xmath14 myr at @xmath125 pc ) associations were observed .
results showed that at @xmath46 au it would be possible to detect only more massive planets ( @xmath12 @xmath60m@xmath4 ) than those we would detect with naco ( this paper ) .
we should thus wait for a new generation of space telescope to probe such close distances at so early ages .
other interesting issues can be retrieved by our analysis concerning the core - accretion model .
our results provide interesting information concerning mechanisms to explain planet formation .
we are considering here the core - accretion model originally proposed by pollack ( 1996 ) since it provides an explanation for the formation of all gaseous as well as solid planets .
it claims that a growing solid core , after reaching a critical mass , accretes a massive atmosphere . according to our results , this model could work , assuming that massive planets ( @xmath38 @xmath6m@xmath4 ) form in a narrow @xmath3-@xmath14 au distance from the central star in a time scale of @xmath12@xmath25 yr ( typical time scale of disk lifetime ) .
one of the critical aspects of the core - accretion model has always been that it can not explain the presence of planets observed with the radial velocity technique at close distances from the central star .
the time necessary for the formation plus the migration should be longer than the typical disk lifetime ( @xmath12@xmath25 yr , haisch et al .
2001 ) . the modified version of the pollack model , recently proposed by alibert et al .
( 2004 ) , seems to solve this problem .
using a detailed description of the composition ( gas plus solid ) of the disk this model reduces the time necessary for a planet to form and migrate by a factor of @xmath6 . with such a modification ,
the process of formation and migration might need a time shorter than the typical disk lifetime .
following the predictions of alibert s model , it would be possible that we did not detect massive planets simply because they formed at ages ( @xmath42 @xmath6myr ) earlier than those of our selected sample .
this means that our results are consistent with the core - accretion model in the alibert formulation .
this model seems thus promising due to the fact that it matches with both observational issues coming from radial velocity measurements and also from direct imaging observations .
we finally mention that our results indicate that a sort of _ exo - planet desert _ for massive ( @xmath12 @xmath6m@xmath4 ) planets is observed at distances larger than @xmath14 au from the stars of our survey .
this desert seems to extend to quite large distances from the central star .
a recent survey ( mccarthy @xmath31 zuckerman , 2004 ) carried out with the near - infrared camera nirc and a coronograph at keck did not detect any @xmath3-@xmath13 m@xmath4 exo - planet in the range @xmath126-@xmath127 au around @xmath12 @xmath128 stars having a mean age of @xmath127 myr .
we would like to mention the relevant improvement in terms of sensitivity reached by our study ( using naco at the vlt ) with respect to previous published results .
a recent analysis of massive planet detection ( neuhuser et al . 2003 ) , carried out with the mpe speckle camera at the eso s @xmath129 m ntt telescope , put an upper limit of @xmath16 au on the distance between the star and @xmath6m@xmath4 exo - planets in the case of targets belonging to @xmath11 pictoris ( @xmath12@xmath13 myr ) .
results from our study show that we can exclude any @xmath6m@xmath4 planets at distances larger than @xmath14 au ( a physical distance @xmath15 times smaller ) for targets belonging to the same moving group .
our results show a better sensititity also with respect to previous observations made with hst ( neuhuser et al .
estimates provided by this study give an upper limit distance between the star and @xmath3m@xmath4 planets of @xmath12 @xmath1 au .
stars observed by neuhuser et al .
( 2002 ) belong to a @xmath12@xmath29myr old systems ( well younger than our sample - @xmath6-@xmath23 myr ) .
models predict a mean difference in h and k band of the order of at least @xmath22 mag between the age of stars in neuhuser et al .
study and ours . we thus deduce that the same instrument ( hst ) would have an upper limit for the star / planet distance quite a bit larger than @xmath1 au for @xmath3m@xmath4 planets in the @xmath6-@xmath23 myr age range
. how can we improve our results to better constrain planet - formation models ?
it is evident that an increase of the achievable contrast at smaller angular separations shifts the median value of the cumulative distributions to smaller angular separations ( fig.[prob ] ) .
this means that the upper limit of the projected star / exo - planet separation decreases .
however , we note that , using the observational strategy employed in our survey , we can hardly improve the sensitivity at subarcsecond distances using a @xmath21 meter class telescope as illustrated in fig.[sdi_pot ] .
this figure shows the detection limit obtained for one of our targets for an integration time of @xmath3 , @xmath6 and @xmath130 minutes .
the deep image was filtered from low spatial frequencies before we calculated the detection limit .
we observe that , beyond @xmath19@xmath48 , the reachable contrast increases with increasing the integration time ( @xmath12 @xmath19 mag in @xmath14 min ) . at distances smaller than @xmath19@xmath48 , on the contrary , a longer integration time does not increase the contrast because we are strongly limited by the speckle noise . the differential imaging technique proposed by racine et al .
( 1999 ) is a rather efficient method to reduce the speckle noise at small angular separations from the central star .
images taken simultaneously at slightly different wavelengths are subtracted from each other . since the speckle noise of both images is nearly the same/@xmath131 factor . ] , the speckle noise is eliminated in the resulting image providing one can sufficiently limit static aberrations .
the differential technique finds a useful application in detecting cool methane - rich faint objects .
rosenthal et al .
( 1996 ) first proposed observations of the same star done simultaneously in two different wavelengths which are located _ `` on '' _ and _ `` off '' _ the methane ( ch@xmath70 ) absorption band .
if one subtracts the image taken in the absorption band from the one taken outside the absorption band , planet / brown dwarf features appear in the continuum .
the absorption at @xmath132 @xmath71 m , in the ch@xmath70 band , is a clear signature of any object having temperature t @xmath27 @xmath133 k ( giant exo - planets and t - type brown dwarfs ) .
we can thus increase the sensitivity ( contrast ) at subarcsecond distances using the recently implemented naco / sdi ( simultaneous differential imager ) instrument .
this instrument employs the differential technique in @xmath39 band .
preliminary results ( close et al . 2004 , lenzen et al . 2004 ) indicate a @xmath26 @xmath12 @xmath134 mag at @xmath110 from the central star . considering our averaged estimate of @xmath26 @xmath91 @xmath114 mag in the @xmath39 band ( see section [ sensit ] )
, we should gain @xmath12 @xmath29 mag in contrast using naco / sdi at @xmath110 ( see fig.[sdi_pot ] ) .
a further and fundamental improvement in contrast should be attained by future ground - based `` planet finders '' , i.e. instruments supported by ao systems providing a better strehl ratio .
the contrast reachable by such an instrument at @xmath110 should be @xmath26 @xmath12 @xmath135 mag in @xmath136-@xmath39 bands ( see fig.[sdi_pot ] ) .
we refer in particular to the top level requirements of cheops , a second generation instrument for the vlt conceived for planet detection ( gratton et al .
we think that the principal goals of these instruments ( naco / sdi and ground - based planet finders ) in the context of the detection of massive and young exo - planets and the understanding of mechanisms responsible for exo - planet formation are the following : * ( a ) * they will permit us to provide answers to the hypothesis ( 1 ) and ( 2 ) ( see section [ dist ] ) . *
( b ) * they will be rather appropriate to do studies similar to the one proposed in this paper but applied to a sample of targets characterized by an earlier spectral type ( i.e. solar - like type ) in the same range of ages .
this is due to the better contrast reachable at subarcsecond distances ( see fig.[sdi_pot ] ) . *
( c ) * they will be appropriate instruments to detect even fainter exo - planets ( i.e. in the @xmath19-@xmath3 m@xmath4 range ) at subarcsecond angular separations from the central star .
in this paper we have presented results of a survey carried out with naco at the vlt aiming to detect massive exo - planets ( @xmath22-@xmath6 m@xmath4 ) orbiting around @xmath46 young ( @xmath6-@xmath23 myr ) and nearby ( @xmath98 @xmath42 @xmath44 pc with @xmath41 @xmath2 of them in the first @xmath43 pc ) stars . no promising candidate was identified .
all faint sources found in the vicinity of the central star are presumably background stars .
a detailed estimate of the detection limit in @xmath0 of the observed targets was presented .
this permitted us to calculate the typical contrast ( @xmath26 ) between a planet and a central star at @xmath110 and @xmath19@xmath48 provided by naco for a typical integration time of @xmath60-@xmath46 min ( see section [ sensit ] ) , to present a statistical analysis of the upper limit for the star / planet separation and also to discuss our results in the framework of the mechanisms that are supposed to explain the formation of exo - planets .
our most relevant result is that , in our sample of targets , in @xmath8@xmath2 of cases , no @xmath6m@xmath4 ( and more massive ) exo - planets was detected at distances larger than @xmath9 au and no @xmath3m@xmath4 ( and more massive ) exo - planets was detected at distances larger than @xmath10 au .
if one considers the median values of the cumulative distributions we have that , in @xmath1@xmath2 of cases , no @xmath6m@xmath4 ( and more massive ) exo - planets should exist at distances larger than @xmath7 au and no @xmath3m@xmath4 ( and more massive ) exo - planets should exist at distances larger than @xmath5 au . we emphasize that , these statistical estimates were done for the projected star / exo - planet distance .
we are not considering here the possibility that planets might not be visible in part of their orbit .
our study indicates that massive planets ( mass @xmath38 @xmath3m@xmath4 ) are rare at distances larger than @xmath30 au as well as at distances smaller than @xmath30 au as indicated by radial velocity estimates ( see section [ dist]).we underline that for the category of massive , warm and self - luminous planets that we looked for in our survey , it is not possible to calculate a reliable probability distribution for detecting planets .
the reason is that , to do this , one should need to know the planet frequency distribution _
a posteriori_. this is obviously impossible to do if no - planets have already been detected . for this category of planets , one can only define some merit functions ( depending on several parameters such as the age , the distance from the observer , the magnitude of the central star , the mass of the planet ) to define some criteria to select targets for which the detection should be particularly favourable . for this reason
, we think that it is fundamental to report results ( even of non - detections ) for the category of exo - planets that we are searching for and to try to get as homogeneous analysis as possible of the results in order to improve the statistics of the upper limit for star / planet distance .
* gj 803 ( au mic ) : * recent estimates of emission at optical wavelengths ( liu et al .
2004 , kalas et al . 2004 ) of a circumstellar disk surrounding gj 803 ( au mic ) were published .
the disk is detected between @xmath12 @xmath1 and @xmath137 au corresponding to around @xmath3@xmath48 - @xmath138@xmath48 ( d @xmath12 @xmath6 pc ) .
the inner region was occulted by the presence of a coronograph so no information is provided on the disk structure at distances smaller than @xmath3@xmath48 .
the field of view of our observations is much more smaller ( @xmath12 @xmath29@xmath48 in radius around the central star ) than that used for the disk detection . within @xmath12 @xmath1
au , the typical time scale of dust is shorter than the stellar age so it was suggested ( kalas et al .
2004 ) that grains observed here must be continually replenished by the collisional erosion of much larger objects ( like comets and asteroids ) . in the inner region of the disk ,
the presence of planetisimals might originate planets by accretion .
kalas et al .
( 2004 ) suggest that in the @xmath29 - @xmath46 au region , the presence of gas giant planets should be possible .
our results indicate that no planets equal or more massive than @xmath3m@xmath4 are detected at distances larger than @xmath15 au and no planets equal or more massive than @xmath6m@xmath4 are detected at distances larger than @xmath29 au . besides this
, we can not exclude that some planets with mass smaller than @xmath3m@xmath4 exist at such close angular separations .
we have to wait for observations at higher contrast to have a more precise answer to this question . also it was suggested ( liu et al .
2004 ) that the spatially localized enhancements and deficits found in the gj 803 disk at distances larger than @xmath14 au from the star might be signposts of ongoing planet formation .
our observations can not infer any information at this large distance from the star .
we thank the eso staff for their support during the observations .
we also acknowledge i. baraffe and a. burrows for providing informations on their models and w. brandner and a. raga for useful discussions .
cccccccccc hip 2729 & 00 34 51.2 & -61 54 58 & k5v & 49.5 & 9.56 & tuc .
& 35 & -55 & ( 5),(6),(7 ) + hd 17925 & 02 52 32.1 & -12 46 11 & k1 & 10.4 & 6.0 & local ass . & 50 & -58 & ( 8),(12 ) + gj 179 & 04 52 05.7 & 06 28 35 & m3.5 & 12.1 & 11.96 & & 200 & -23 & ( 2 ) + gj 182 & 04 59 34.8 & 01 47 01 & m0.5 & 26.7 & 10.1 & & 35 & -24 & ( 2),(10 ) + hip 23309 & 05 00 47.1 & -57 15 26 & k7v & 26.3 & 10.09 & beta pic .
& 12 & -37 & ( 3),(10),(13 ) + gj 207.1 & 05 33 44.8 & 01 56 43 & m2.5 & 16.8 & 11.5 & & 50 & -2 & ( 1),(2 ) + ao men & 06 18 28.2 & -72 02 42 & k5v & 38.5 & 10.09 & beta pic .
& 12 & -28 & ( 3),(10),(13 ) + bd + 2@xmath100 1729 & 07 39 23.0 & 02 01 01 & k7 & 9.8 & 9.82 & local ass .
& 100 & 12 & ( 2 ) + lq hya & 09 32 25.6 & -11 11 05 & k0v & 18.3 & 7.80 & & 50 & 28 & ( 1),(8 ) + twa 6a & 10 18 28.8 & -31 50 02 & k7 & 77.0 & 12.0 & tw hya .
& 10 & 21 & ( 4),(5 ) + bd 1@xmath100 2447 & 10 28 55.5 & 00 50 28 & m2 & 7.2 & 9.63 & local ass .
& 100 & 47 & ( 1),(2 ) + twa 5b ( 2@xmath48 ) & 11 31 55.4 & -34 36 27 & m8.5 & 50.0 & - & tw hya . & 10 & 25 & ( 4),(5 ) + twa 8a ( 13@xmath48 ) & 11 32 41.5 & -26 51 55 & m2 & 21.0 & ( r=11.2 ) & tw hya . & 10 & 33 & ( 4),(5 ) + twa 8b & 11 32 41.5 & -26 51 55 & m5 & 21.0 & ( r=13.85)&tw hya . &
10 & 33 & ( 4),(5 ) + twa 9a@xmath139 ( 9@xmath48 ) & 11 48 24.2 & -37 28 49 & m1 & 64.0@xmath140 & 11.3 & tw hya .
& 10 & 24 & ( 4),(5 ) + twa 9b & 11 48 24.2 & -37 28 49 & m1 & 64.0@xmath140 & ( r=12.98 ) & tw hya . & 10 & 24 & ( 4),(5 ) + sao 252852 & 14 42 28.1 & -64 58 43 & k5v & 16.4 & 8.50 & & 100 & -4 & ( 2),(9 ) + v343 nor & 15 38 57.6 & -57 42 27 & k0v & 39.8 & 8.14 & beta pic .
& 12 & -2 & ( 3),(10),(13 ) + v2306 oph & 16 30 18.1 & -12 39 45 & m3.5 & 4.4 & 10.12 & local ass .
& 100 & 24 & ( 2 ) + hd155555 ab@xmath141 & 17 17 25.5 & -66 57 02 & g5iv & 31.4 & 7.20 & beta pic .
& -16 & 24 & ( 3),(10),(13 ) + hd 155555 c & 17 17 25.5 & -66 57 00 & m4.5 & 31.4 & 12.71 & beta pic .
& 12 & -16 & ( 3),(10),(13 ) + cd -64@xmath100 1208 & 18 45 36.9 & -64 51 48 & m0 & 29.2 & 9.20 & beta pic .
& 12 & -24 & ( 3),(10),(13 ) + pz tel & 18 53 05.9 & -50 10 50 & k0 & 49.6 & 8.42 & beta pic./tuc .
& 12 & -21 & ( 3 ) + gj 799a ( 2@xmath142 ) & 20 41 51.1 & -32 26 07 & m4.5 & 10.2 & 11.0 & beta pic .
& 12 & -36 & ( 3),(10),(13 ) + gj 799b & 20 41 51.1 & -32 26 09 & m4.5 & 10.2 & 11.0 & beta pic .
& 12 & -36 & ( 3),(10),(13 ) + gj 803 ( au mic ) & 20 45 09.3 & -31 20 24 & m0 & 9.9 & 8.61 & beta pic .
& 12 & -37 & ( 3),(10),(13 ) + bd -17@xmath100 6128 ( 2@xmath143 ) & 20 56 02.7 & -17 10 54 & k7 & 47.7 & 10.60 & beta pic .
& 12 & -35 & ( 3),(10),(13 ) + gj 813 & 20 57 25.4 & 22 21 45 & m2 & 13.6 & 12.0 & & 200 & -15 & ( 2 ) + gj 890 & 23 08 19.5 & -15 24 35 & m0 & 21.8 & 10.84 & & 100 & -63 & ( 1),(2),(11 ) + hd 221503 & 23 32 49.4 & -16 50 44 & k5 & 13.9 & 8.60 & local ass . &
@xmath38 100 - 200 & -69 & ( 1),(2),(9 ) + [ tab1 ] cccccccc hip 2729 & 5.96 & 6.13 & 18.60 & 18.75 & 12.64 & 12.62 & -0.02 + hd 17925 & 3.50 & 3.60 & 16.28 & 15.99 & 12.78 & 12.39 & -0.4 + gj 179 & 7.77 & 8.01 & 20.24 & 18.31 & 12.47 & 10.30 & -2.17 + gj 182 & 6.49 & 6.66 & 17.25 & 17.40 & 10.76 & 10.74 & -0.02 + hip 23309 & 6.37 & 6.54 & 15.19 & 15.59 & 8.82 & 9.05 & 0.23 + gj 207.1 & 7.20 & 7.41 & 17.33 & 17.04 & 10.13 & 9.63 & -0.29 + ao men & 7.80 & 7.97 & 15.98 & 16.38 & 8.18 & 8.41 & 0.23 + bd + 2@xmath100 1729 & 6.61 & 6.76 & 18.11 & 16.98 & 11.50 & 10.22 & -1.28 + lq hya & 5.97 & 6.07 & 17.51 & 17.23 & 11.54 & 11.16 & -0.38 + twa 6a@xmath139 & 7.97 & 8.17 & 17.35 & 17.65 & 9.38 & 9.48 & 0.15 + bd 1@xmath100 2447 & 5.58 & 5.78 & 17.44 & 16.31 & 11.86 & 10.53 & -1.33 + twa 5b ( 2@xmath48 ) & 11.5 & 11.95 & 16.41 & 16.71 & 4.91 & 4.76 & -0.15 + twa 8a ( 13@xmath48 ) & 7.44 & 7.72 & 14.53 & 14.83 & 7.09 & 7.11 & 0.14 + twa 8b & 9.01 & 9.36 & 14.53 & 14.83 & 5.47 & 5.52 & 0.00 + twa 9a ( 9@xmath48 ) & 7.68 & 7.95 & 16.95 & 17.25 & 9.27 & 9.30 & 0.1 + twa 9b & 9.14 & 9.34 & 16.95 & 17.25 & 7.81 & 7.91 & 0.1 + sao 252852 & 5.68 & 5.85 & 19.22 & 18.09 & 13.54 & 12.24 & -1.3 + v343 nor & 6.31 & 6.41 & 16.05 & 16.46 & 9.74 & 10.05 & 0.31 + v2306 oph & 5.93 & 6.17 & 16.30 & 15.17 & 10.37 & 9.00 & -1.37 + hd155555 ab & 5.62 & 5.7 & 15.54 & 15.94 & 9.92 & 10.24 & 0.32 + hd 155555 c & 7.25 & 7.51 & 15.54 & 15.94 & 8.29 & 8.43 & 0.14 + cd -64@xmath100 1208 & 5.73 & 5.88 & 15.38 & 15.78 & 9.65 & 9.90 & 0.25 + pz tel & 6.69 & 6.79 & 16.54 & 16.94 & 9.85 & 10.15 & 0.3 + gj 799a ( 2@xmath142 ) & 5.54 & 5.8 & 13.10 & 13.50 & 7.56 & 7.70 & 0.14 + gj 799b & 5.54 & 5.8 & 13.10 & 13.50 & 7.56 & 7.70 & 0.14 + gj 803 ( au mic ) & 5.00 & 5.15 & 13.03 & 13.44 & 8.03 & 8.29 & 0.26 + bd -17@xmath100 6128 ( 2@xmath143 ) & 7.13 & 7.28 & 16.45 & 16.85 & 9.32 & 9.57 & 0.25 + gj 813 & 7.95 & 8.15 & 20.50 & 18.57 & 12.55 & 10.42 & -2.13 + gj 890 & 7.37 & 7.52 & 19.84 & 18.71 & 12.47 & 11.19 & -1.13 + hd 221503 & 5.78 & 5.95 & @xmath38 20.55 & @xmath38 18.62 & @xmath3814.77 & @xmath3812.67 & @xmath38 -2.10 + [ tab2 ] cccccc hip 2729 & nb 2.12 & @xmath40 & 1.30 & @xmath12 28 & 23/07/2003 + hd 17925 & ib 2.09 & ib 2.24 & 1.70 & @xmath12 23 & 22/07/2003 + gj 179 & nb 1.64 & @xmath39 & 1.74 & @xmath12 36 & 26/12/2003 + gj 182 & nb 2.12 & @xmath40 & 1.30 & @xmath12 35 & 26/12/2003 + hip 23309 & nb 2.12 & @xmath40 & 2.06 & @xmath12 20 & 23/07/2003 + gj 207.1 & @xmath40/nd short & @xmath40 & 9.00 & @xmath12 30 & 08/01/2004 + ao men & nb 2.12 & @xmath40 & 2.50 & @xmath12 20 & 17/02/2003 + bd + 2@xmath100 1729 & nb 1.64 & @xmath39 & 1.00 & @xmath12 17 & 19/02/2003 + lq hya & nb 2.12 & @xmath40 & 0.80 & @xmath12 23 & 09/01/2004 + twa 6a & nb 2.12 & @xmath40 & 5.00 & @xmath12 14 & 09/01/2004 + bd 1@xmath100 2447 & @xmath39/nd short & @xmath39 & 0.36 & @xmath12 21 & 09/01/2004 + twa 5b ( 2@xmath48 ) & nb 2.12 & @xmath40 & 5.00 & @xmath12 31 & 17/02/2003 + twa 8a ( 13@xmath48 ) & nb 2.12 & @xmath40 & 4.20 & @xmath12 19 & 17/02/2003 + twa 8b & @xmath40/nd short & @xmath40 & 19.00 & @xmath12 22 & 17/02/2003 + twa 9a ( 9@xmath48 ) & @xmath40/nd short & @xmath40 & 21.00 & @xmath12 25 & 18/02/2003 + twa 9b & @xmath40/nd short & @xmath40 & 21.00 & @xmath12 25 & 18/02/2003 + sao 252852 & nb 1.64 & @xmath39 & 0.40 & @xmath12 24&21/07/2003 + v343 nor & nb 2.12 & @xmath40 & 1.90 & @xmath12 22 & 19/02/2003 + v2306 oph & nb 2.12 & @xmath40 & 1.30 & @xmath12 25 & 19/02/2003 + hd155555 ab & nb 2.12 & @xmath40 & 1.20 & @xmath12 8 & 18/02/2003 + hd 155555 c & nb 2.12 & @xmath40 & 5.40 & @xmath12 38 & 20/07/2003 + cd -64@xmath100 1208 & nb 2.12 & @xmath40 & 1.10 & @xmath12 24 & 22/07/2003 + pz tel & nb 2.12 & @xmath40 & 1.56 & @xmath12 30 & 23/07/2003 + gj 799a ( 2@xmath142 ) & nb 2.12 & @xmath40 & 0.58 & @xmath12 23 & 21/07/2003 + gj 799b & nb 2.12 & @xmath40 & 0.58 & @xmath12 23 & 21/07/2003 + gj 803 ( au mic ) & ks / nd short & @xmath40 & 0.36 & @xmath12 30 & 21/07/2003 + bd
-17@xmath100 6128 ( 2@xmath143 ) & nb 2.12 & @xmath40 & 2.50 & @xmath12 27 & 21/07/2003 + gj 813 & nb 1.64 & @xmath39 & 0.70 & @xmath12 28 & 23/07/2003 + gj 890 & nb 2.12 & @xmath40 & 3.00 & @xmath12 24 & 21 - 22/07/2003 + hd 221503 & nb 1.64 & @xmath39 & 0.70 & @xmath12 22 & 22/07/2003 + [ tab3 ] ccccc hip 2729 & 35 & 49.5 & 64 & 36 + gj 179 & 200 & 12.1 & 9 & 5 + gj 182 & 35 & 26.7&19 & 11 + hip 23309 & 12 & 26.3&11 & 7 + ao men & 12 & 38.5&13 & 9 + bd + 2@xmath100 1729 & 100 & 9.8 & 8 & 4 + lq hya & 50 & 18.3&22 & 8 +
twa 6a & 10 & 77.0&52 & 25 + bd 1@xmath100 2447 & 100 & 7.2 & 6 & 3 + twa 5b & 10 & 50.0 & 10 & 7 + twa 8a & 10 & 21.0&7 & 3 + twa 8b & 10 & 21.0&7 & 2 + twa 9a & 10 & 64.0&38 & 19 + twa 9b & 10 & 64.0&22 & 13 + sao 252852 & 100 & 16.4&19 & 10 + v343 nor & 12 & 39.8&24 & 12 + v2306 oph & 100 & 4.4&4 & 1 + hd 155555 ab & 12&31.4&22 & 11 + hd 155555 c & 12 & 31.4&13 & 8 + cd -64@xmath100 1208 & 12 & 29.2&15 & 9 +
pz tel & 12 & 49.6&26 & 15 + gj 799 a & 12 & 10.2&4 & 2 + gj 799 b & 12 & 10.2 & 6 & 2 + gj 803 ( au mic ) & 12 & 9.9&4 & 2 + bd -17@xmath100 6128 & 12 &
47.7&25 & 14 + gj 813 & 200 & 13.6&8 & 4 + gj 890 & 100 & 21.8 & 26 & 14 + hd 221503 & @xmath38 100 - 200 & 13.9 & 17 & 9 + [ tab4 ] cc | cc 10 cases & 4 @xmath42 @xmath144 @xmath145 10 & 9 cases & 1 @xmath42 @xmath144 @xmath145 5 + 8 cases & 10 @xmath42 @xmath144 @xmath145 20 & 8 cases & 5 @xmath42 @xmath144 @xmath145 10 + 8 cases & 20 @xmath42 @xmath144 @xmath145 40 & 9 cases & 10 @xmath42 @xmath144 @xmath145 20 + 2 cases & 40 @xmath42 @xmath144 @xmath145 65 & 2 cases & 20 @xmath42 @xmath144 @xmath145 36 + [ tab5 ] | we report on a survey devoted for the search of exo - planets around young and nearby stars carried out with naco at the vlt . the detection limit for @xmath0 among the best available targets vs. the angular separation from the star
is presented .
the non - detection of any planetary mass companion in our survey is used to derive , for the first time , the frequency of the upper limit of the projected separation planet - stars . in particular
, we find that in @xmath1@xmath2 of cases , no @xmath3m@xmath4 planet ( or more massive ) has been detected at projected separations larger than @xmath5 au and no @xmath6m@xmath4 planet ( or more massive ) has been detected at projected separations larger than @xmath7 au . in @xmath8@xmath2 of cases , these values increase to @xmath9 au and @xmath10 au respectively .
the excellent sensitivity reached by our study leads to a much lower upper limit of the projected planet - star separation compared with previous studies .
for example , for the @xmath11 pictoris group , ( @xmath12@xmath13 myr ) , we did not detect any @xmath6m@xmath4 planet at distances larger than @xmath14 au . a previous study carried out with @xmath15 m class telescopes put an upper limit for @xmath6m@xmath4 planets at @xmath12@xmath16 au . for our closest target ( v@xmath17 oph - d = @xmath18 pc )
it is shown that it would be possible to detect a @xmath6m@xmath4 planet at a minimum projected separation from the star of @xmath19 au and a @xmath3m@xmath4 planet at a minimum projected separation of @xmath20 au .
our results are discussed with respect to mechanisms explaining planet formation and migration and to forthcoming observational strategies and future planet finder observations from the ground . |
the _ spitzer space telescope ( sst ) _ , with its very high mid - infrared sensitivity , has revealed a new type of low - mass protostars called very low luminosity objects ( vellos ; e.g. , young et al .
2004 ) . defining the internal luminosity of a source , , to be the total luminosity of the central protostar and circumstellar disk ( if present ) ,
a vello is defined to be an object embedded within a dense core that meets the criterion @xmath3 0.1 ( di francesco et al . 2007 ) .
vellos are primarily being discovered in cores previously classified as starless based on observations with the _ infrared astronomical satellite ( iras ) _ and the _ infrared space observatory ( iso)_. the three most well - studied vellos to date are iram 04191 ( andr et al . 1999 ; dunham et al .
2006 ) , l1014 ( young et al . 2004
; crapsi et al . 2005 ; bourke et al . 2005 ; huard et al . 2006 ) , and l1521f ( bourke et al . 2006 ) , which show very different properties in molecular line observations despite their similar internal luminosities .
iram 04191 ( andr et al . 1999 ; belloche et al . 2002 ) and l1521f ( crapsi et al . 2004 ) show evidence for infall whereas l1014 does not ( crapsi et al .
iram 04191 is associated with a well - collimated outflow ( andr et al .
1999 ) ; the other two are not , although at least l1014 and possibly l1521f feature weak , compact outflows ( bourke et al . 2005 ; bourke et al .
2006 ) . the discovery of vellos with _
spitzer _ has put into question the picture of low - mass star formation as a continuous process of constant mass accretion at the standard rate of @xmath4 @xmath5 ( shu , adams , & lizano 1987 ) through a single evolutionary sequence , the well - established class system progressing from class 0 to iii ( myers & lada 1993 , andr et al .
this standard accretion rate predicts a much higher luminosity than observed for vellos ; vellos must feature some combination of a very low central mass and a very low accretion rate ( e.g. dunham et al .
if the accretion continues at the current low rate to the very small central mass , it might not make a star .
however , the accretion rate is not necessarily constant .
for instance , fu orionis ( fu ori ) objects undergo outbursts ( bell et al .
1995 and references therein ) .
studies ( vorobyov & basu 2005 and references therein ) for the nature of the fu ori variables suggest accretion bursts from the disk to the central star by the thermal instability of the disk
. therefore , two potential explanations for the very low luminosities of vellos are 1 ) proto - brown dwarfs , and 2 ) objects in a quiescent phase of the episodic accretion process .
the former can be discriminated from the latter with studies of the chemistry since they involve vastly different thermal histories , which is crucial to the chemical evolution .
the thermal history is especially important in interactions between gas and ice ; ice evaporation and gas freeze - out from and onto grain surfaces , respectively , depend on the dust temperature ( lee et al .
proto - brown dwarfs , with their very low masses , will never experience a hot phase , whereas the outbursts of a cycle of episodic accretion , a short time period when the majority of the mass is dumped onto the central protostar , involve significant warming of the surrounding dust .
the quiescent states between outbursts feature much colder dust temperatures . as a result
, envelopes of proto - brown dwarfs will be similar to starless cores in their chemical distributions , while objects in a quiescent state of episodic accretion will show different chemical distributions from starless cores or normal , embedded class 0/i objects .
iram 04191 may be undergoing episodic accretion since it features a strong outflow which predicts a higher accretion rate by two orders of magnitude than inferred from the internal luminosity of the source ( andr et al .
1999 ; dunham et al . 2006 ) .
furthermore , the n@xmath0h@xmath1 emission , observed with the plateau de bure interferometer ( pdbi ) and the iram 30 m telescope , shows a hole in the center of the envelope ( belloche & andr 2004 ) .
in general , n@xmath0h@xmath1 emission tends to peak towards the center of starless cores ( lee et al .
2003 ) , but be deficient from the centers of class 0/i sources due to destruction by co as it evaporates ( lee et al .
iram 04191 shows moderate co depletion ( crapsi et al .
2004 ) ; belloche & andr ( 2004 ) suggest that freeze - out of n@xmath0 in the high - density , inner envelope might result in the observed n@xmath0h@xmath1 hole
. however , if freeze - out of n@xmath0 is significant enough to explain this hole , there should be significantly more deuteration and depletion of co than observed ( crapsi et al . 2004 ) , similar to that seen in prestellar cores ( lee et al .
2003 ) . in this study
, we model the chemical evolution in the process of episodic accretion to provide a possible explanation of the chemical distributions of co and n@xmath0h@xmath1 in gas and to predict observable consequences in the co@xmath0 ice feature in vellos such as iram 04191 that show strong evidence for undergoing such a process .
we use the chemo - dynamical model developed by lee et al .
this model calculates the chemical evolution of a model core evolving from the prestellar stage through the embedded protostellar stages .
the dynamical evolution is described by combining a sequence of bonnor - ebert spheres ( bonnor 1956 , ebert 1955 ) with the inside - out collapse model ( shu 1977 ) , where the accretion rate from the envelope onto the star+disk system is constant .
the model also includes the first hydrostatic core ( fhsc ) stage , which results from the first gravitational collapse of a dense molecular core and lasts until the core temperature reaches 2000 k and the dissociation of molecular hydrogen causes the second collapse ( boss 1993 , masunaga et al .
the radius of the fhsc is about 5 au , and the disk is not yet well developed at this stage . as a result
, the accretion luminosity arising from accretion onto the fhsc is not significant .
the evolution of the central luminosity follows that adopted by young & evans ( 2005 ) , who incorporated the evolution of an unresolved disk and stellar photosphere into the central luminosity .
this luminosity is proportional to the accretion luminosity arising from envelope accretion following the inside - out collapse model ( see young & evans 2005 for details ) . in this study
, we assume that the accretion from the envelope to the disk is constant , but the accretion from the disk to the star is episodic .
thus , the envelope density structure is identical to that of the inside - out collapse model , but the internal luminosity is changed by the episodic accretion process .
the initial mass of the model core , with an effective sound speed of 0.27 and outer radius of 6200 au , is 1 ( equation 1 of young & evans 2005 ) .
the episodic accretion is assumed to begin after the fhsc phase , which lasts for @xmath6 years .
the dust temperature profile at each time step is calculated with the one dimensional continuum radiative transfer code dusty ( ivezi et al . 1999 ) , using as input the density profile from the inside - out collapse model . the input internal radiation field at each time step is calculated by the model of young & evans ( 2005 ) with their standard parameters of @xmath7 , @xmath8 , and @xmath9 ( see young & evans 2005 for a detailed explanation of these parameters ) combined with our prescription for episodic accretion . in this prescription ,
accretion bursts 10 times more massive than normal ( @xmath10 m@xmath11 yr@xmath12 ) occur during @xmath13 years every @xmath14 years .
for the @xmath15 years between the accretion burst events , the accretion from the disk to the star is decreased by a factor of 100 .
vorobyov & basu ( 2005 ) have showed that @xmath16 m@xmath11 is accreted from the disk to the central source with the accretion rate of @xmath17 m@xmath11 yr@xmath12 during a burst , and the duration of the intervening quiescent accretion phase is @xmath18 years for @xmath19 m@xmath11 yr@xmath12 . according to their calculations and for the accretion rates in the burst and quiescent phases in this model , which are about an order of magnitude smaller than the rates of vorobyov & basu ( 2005 )
, we adopt the timescales of @xmath13 and @xmath14 years in the two phases , respectively , and these timescales are consistent with the estimated from hl tauri by lin et al .
( 1994 ) .
we calculate the radiative transfer independently at each time step since the dust temperature reaches a new equilibrium essentially instantly compared to the length of our time step . for example , one silicate grain with a radius of 0.1 and a temperature of 100 k will cool down to 10 k within one second if the heating source suddenly disappears ( draine & anderson 1985 ) .
following young & evans ( 2005 ) , the envelope is heated externally by the black - draine interstellar radiation field ( isrf ) ( evans et al .
2001 ) attenuated by a visual extinction of 0.5 mag .
also following young & evans , we used the dust model of ossenkopf & henning ( 1994 ) appropriate for thin ice mantles after @xmath20 years of coagulation at a gas density of @xmath21 @xmath22 ( oh5 dust ) .
1a shows the evolution of the total internal luminosity in the model , and fig .
1b shows the temperature evolution of a few infalling grid points . as seen in this figure , the temperature that an infalling gas parcel experiences peaks at the time step of each massive accretion and is lowered significantly during the quiescent states . betwen 4 and 5.5@xmath23 years ,
the luminosity and temperature increase steeply with time even in the quiescent phase because of the transition from the fhsc to class 0 , which occurs from 2@xmath23 years until 5.5@xmath23 years .
this transition timescale is the same as that used in lee et al .
( 2004 ) , but much longer than the 100 years assumed by young & evans ( 2005 ) .
the evolution of the spectral energy distribution ( sed ) around @xmath24 years is presented in fig .
an accretion burst occurred at @xmath25 years , after which a quiescent state continues until the next burst @xmath14 years later .
the accretion luminosity powered in the burst phase at @xmath25 years is about 5 , whereas the accretion luminosities at 5.0 and 5.4 @xmath23 years are about 0.004 and 0.02 , respectively .
the model sed at @xmath26 years fits well the observed sed of iram 04191 .
the total luminosity in the model is about 0.09 , where the luminosity arising from heating by the isrf ( @xmath27 ) is larger than the internal luminosity ( @xmath28 ) .
this result is not very consistent with that of dunham et al .
( 2006 ) , who found that the internal luminosity must be greater than 0.04 .
however , the density profile and isrf in their model are significantly different from ours , making it hard to directly compare the two results .
the density and dust temperature profiles are shown in fig .
3 at the same time steps shown for the seds in fig .
the infall radius is 0.014 pc ( 2800 au ) when the accretion burst occurs at @xmath25 .
the density at the infall radius is about @xmath29 @xmath22 .
the density profile has the power law index of @xmath30 and 2.0 inside and outside the infall radius , respectively . except for the accretion burst phase
, the dust temperature is below the co evaporation temperature ( @xmath31 k ) at almost all radii .
in prestellar cores , which are very evolved starless cores , carbon - bearing molecules such as cs and co are readily frozen - out onto grain surfaces ( bergin & langer 1997 , aikawa et al .
2001 , lee et al .
2003 , lee et al .
in contrast , nitrogen - bearing molecules such as n@xmath0h@xmath1 and nh@xmath32 are considered good tracers of ( column ) density in cold , dense cores since n@xmath0 , the parent molecule of n@xmath0h@xmath1 and nh@xmath32 , has a long formation timescale such that it becomes abundant later in the core evolution .
therefore , the comparison between co and n@xmath0h@xmath1 abundances is a good indicator of the physical properties of a dense core .
the low binding energy of n@xmath0 on grain surfaces has also been suggested as an explanation for the long - lived molecules in the cold regions ( lee et al .
however , berg et al . ( 2005 ) showed that the binding energy of n@xmath0 on grain surfaces is not different from that of co. the high abundance of n@xmath0h@xmath1 in the cold , dense regions may be due to the replenishment of n@xmath0 through the interactions between n@xmath0h@xmath1 and grains since n@xmath0h@xmath1 recombines with electrons on the grain surfaces .
the significant depletion of co and electrons , the principal destroyers of n@xmath0h@xmath1 in the gas phase , is another explanation for the high abundance of n@xmath0h@xmath1 in these cold , dense regions .
we updated the chemical network used in lee et al .
( 2004 ) to include recent results on the n@xmath0h@xmath1 chemistry , including a new binding energy of n@xmath0 identical to that of co ( berg et al . 2005 ) and new rates of dissociative recombination of n@xmath0h@xmath1 with electrons ( geppert et al . 2004 ) .
the initial abundances of other species are the same as those in table 3 of lee et al .
bare silicate grains were assumed for binding energies of molecules .
4 presents the evolution of the co abundance in gas and ice at the time steps @xmath13 years before , during , and 3 and @xmath33 years after the accretion burst at @xmath25 years since the model sed at @xmath26 years fits the observations of iram 04191 reasonably well .
@xmath13 years before the accretion burst , co is frozen - out onto grain surfaces and depleted from the gas at most radii except around the co evaporation radius , inside which the dust temperature is higher than the co evaporation temperature , developed by the previous accretion burst at @xmath34 years . during the burst phase , the co evaporation radius jumps to a much larger radius due to the significant increase in accretion luminosity .
once the burst phase ends and the core makes the transition back to the quiescent phase , the accretion luminosity , and therefore the temperature , decreases greatly .
co is then gradually frozen back out of the gas and onto the grain surfaces .
the evolution of the n@xmath0h@xmath1 abundance is shown in fig .
before the accretion burst at @xmath25 years , the n@xmath0h@xmath1 abundance peaks at a rather large radius and decreases inward .
( the sharp drop of the n@xmath0h@xmath1 abundance at the edge of the model core results from the dissociative recombination of n@xmath0h@xmath1 with electrons . ) at radii of 0.002 pc ( log r@xmath35 ) and 0.0008 pc ( log r@xmath36 ) , the bumps of n@xmath0h@xmath1 are caused by the dips of co at these radii . the n@xmath0h@xmath1 abundance decreases at radii smaller than 0.0008 pc even though co is depleted here .
n@xmath0h@xmath1 forms mainly by the reaction between n@xmath0 and h@xmath37 , which is destroyed by co. as mentioned above , the evaporation temperature of n@xmath0 is the same as that of co , i.e. , @xmath3820 k. therefore , at the radii smaller than 0.0008 pc where the dust temperature is lower than @xmath31 k , and the density is greater than @xmath21 @xmath22 , the significant freeze - out of n@xmath0 causes the low abundance of n@xmath0h@xmath1 . during the burst phase , the n@xmath0h@xmath1 abundance drops significantly at radii smaller than the co evaporation radius since co destroys n@xmath0h@xmath1 . in the quiescent states 3 , 5 , and @xmath39 years after the burst , n@xmath0h@xmath1 slowly increases at small radii as co is frozen - out onto the grain surfaces .
most vellos were classified as starless core before _ spitzer _ observations , and they showed wide variation in chemical and physical properties .
for instance , iram 04191 ( andr et al . 1999 ) and l673 - 7 ( m. dunham 2007 , private communication ) have well - collimated molecular outflows , but no molecular outflow has been detected in l1521f even though its _ spitzer _
irac images show a compact bipolar scattered emission nebula ( bourke et al .
l1014 features a very compact outflow detected with the submillimeter array ( bourke 2005 ) . in chemical properties ,
iram 04191 and l1014 are similar with moderate co depletion but different in their n@xmath0h@xmath1 emission , which peaks at the center of l1014 but shows a hole at the center of iram04191 .
l1521f , which was believed to be a highly evolved starless core ( crapsi et al .
2004 ) , shows significant co depletion and centrally peaked n@xmath0h@xmath1 emission . before the _ spitzer _ observations , this variation in chemical properties
was believed to arise from an as - of - yet incompletely understood evolutionary sequence of starless cores .
the discovery of the vellos has cast doubt on our understanding of the continuous and unique sequence of star formation consisting of the evolution of a starless core into a prestellar core followed by the formation of a class 0 object that then evolves through the class system .
the sources without strong outflows might be in the final accretion phase of substellar objects or in the fhsc stage ( omukai et al .
on the other hand , vellos with strong , well - collimated outflows such as iram 04191 are hard to understand since the outflow suggests a much higher accretion luminosity than the observed internal luminosity of the source .
a possible explanation for such objects is episodic accretion . in this picture , vellos are objects in quiescent phases of a cycle of episodic mass accretion .
the strong outflows observed can be generated during accretion bursts that occurred prior to the current quiescent phase .
these two processes predict vastly different thermal histories , which can be discriminated by their different predictions for the chemical evolution of the core .
observations probing these different predictions are challenging , however , since high spatial resolution is required to resolve the regions most affected by the different thermal histories .
in addition , observations of ice features require high sensitivities due to the very low luminosities of the central source . as seen in the previous section , the relation between co and n@xmath0h@xmath1 abundances in an object featuring episodic accretion is different from that of starless cores or hot corinos , where co and n@xmath0h@xmath1 have an anti - correlation . in the quiescent phase after an outburst , both co and n@xmath0h@xmath1 have not had enough time to reach chemical equilibrium . as a result
, co can be less depleted and n@xmath0h@xmath1 more depleted compared to a core forming a proto - brown dwarf with a similar central luminosity to iram 04191 .
co in ice can also be a discriminator since co ice features from vellos in a quiescent phase of episodic accretion will be weaker than those in proto - brown dwarfs .
additionally , the co@xmath0 ice features can also be very important tracers of the thermal history of vellos .
the co@xmath0 15.2 @xmath2 bending mode has been detected toward massive star forming regions ( gerakines et al .
1999 ) , low mass star forming cores ( pontoppidan et al . 2007 ) and even dark molecular clouds ( knez et al . 2005 ) .
the formation process of the co@xmath0 ice is controversial , but it must form on grain surfaces since the abundance of the co@xmath0 gas in dark clouds is too low ( @xmath40 , bergin et al .
1995 ) to explain the observed co@xmath0 ice features .
co@xmath0 ices are found both in water- and co - rich grain surfaces .
therefore , the co@xmath0 ice likely forms on grain surfaces by the reactions between co frozen - out from gas and oh or o photodissociated from water or co in ice .
the photodissociation of water or co on grain surfaces , especially in low mas star forming cores , can occur by the cosmic ray induced uv photons . as a result
, the timescale of the formation of co@xmath0 on grain surfaces must be long ( @xmath41 years in the density of @xmath20 @xmath22 , pontoppidan et al .
2007 . )
the evaporation temperature of the co@xmath0 ice is about 35 k , higher and lower than the co ( 20 k ) and h@xmath0o ( 100 k ) evaporation temperatures , respectively . in outburst phases of episodic accretion ,
the dust temperature can rise above the co@xmath0 evaporation temperature at some inner radii .
there is also a range of radii where the dust temperature is between the co and co@xmath0 evaporation temperatures , so co@xmath0 remains on grain surfaces while co evaporates . in this range
, co@xmath0 will produce the double - peaked feature indicative of pure co@xmath0 ice ( pontoppidan et al .
2007 ) . in the following quiescent phase
, co will be gradually frozen - out as seen above , and thus , combined with the long formation timescale of co@xmath0 , the solid co@xmath0 abundance profile developed in the outburst phase will not change greatly . as a result
, we expect that objects in quiescent stages of episodic accretion will have a weaker solid co@xmath0 feature ( if co@xmath0 evaporates up to a large radius ) but a stronger pure co@xmath0 feature ( if the range of co@xmath0 evaporation is not large ) compared to substellar objects . fig .
6 shows the temperature profile during an outburst with marks for the co and co@xmath0 ice mixture .
however , if the vellos are proto - brown dwarfs , which never raise the dust temperature above the co and co@xmath0 evaporation temperatures in the majority of their envelopes , then the solid co@xmath0 feature will be strong as seen toward background sources without a strong pure co@xmath0 component ( knez et al .
therefore , the co@xmath0 ice feature as well as the abundance distribution of co and n@xmath0h@xmath1 in gas can be good tracers of the formation processes of vellos .
aikawa , y. , ohashi , n. , inutsuka , s .-
i . , herbst , e. & takakuwa , s. 2001 , apj , 552 , 639 andr , p. , ward - thompson , d. , & barsony , m. 1993 , apj , 406 , 122 andr , p. , motte , f. , & bacmann , a. 1999 .
apj , 513 , l57 bergin , e.a . ,
langer , w.d .
, & goldsmith , p.f .
1995 , apj , 441 , 222 bell , k.r . ,
lin , d.n.c . , hartmann , l.w .
, & kenyon , s.j .
1995 , apj , 444 , 376 belloche , a. , andr , p. , despois , d. , & blinder , s. 2002 , a&a , 393 , 927 belloche , a. , & andr , p. 2004 , a&a , l35 bergin , e.a . & langer , w.d .
1997 , apj , 486 , 316 bonnor , w.b . 1956 ,
mnras , 116 , 351 boss , a.p .
1993 , apj , 410 , 157 bourke , t.l .
, crapsi , a. , myers , p.c . ,
evans , n.j .
ii , wilner , d.j . , huard , t.l . ,
jrgensen , j.k . , & young , c.h .
2005 , apj , 633 , l129 bourke , t.l .
2006 , apj , 649 , l37 crapsi , a. , caselli , p. , walmsley , c.m . ,
tafalla , m. , lee , c.w . ,
bourke , t.l .
, & myers , p.c .
2004 , a&a , 420 , 957 crapsi , a. , et al .
2005 , a&a , 439 , 1023 di francesco , j. , evans . n.j .
, ii , caselli , p. , myers , p.c . , shirley , y. , aikawa , y. , & tafalla , m. 2007 , in protostars and planets v , ed .
b. reipurth , d. jewitt , & k. keil ( tucson : univ .
arizona press ) draine , b.t . & anderson , n. 1985 , 292 , 494 dunham , m.m .
2006 , apj , 651 , 945 ebert , r. 1955 , z. astrophys .
, 37 , 217 evans , n.j.ii , rawlings , j.m.c . ,
shirley , y. , & mundy , l.g . 2001 , apj , 557 , 193 geppert , w. d. , thomas , r. , semaniak , j. , ehlerding , a. , millar , t. , @xmath42sterdahl , f. , af ugglas , m. , djuri@xmath43 , n. , pa@xmath44l , a. , & larsson , m. 2004 , apj , 609 , 459 gerakines , p.a . , whittet , d.c.b .
, ehrenfreund , p. , boogert , a.c.a . ,
tielens , a.g.g.m .
, schutte , w.a . , chiar , j.e .
, v an dishoeck , e.f . ,
prusti , t. , helmich , f.p .
, de graauw , t. 1999 , apj , 522,357 huard , t.l . ,
et al . 2006 , apj , 640 , 391 ivezi , . , nenkova , m. , & elitzur , m. 1999 , astro - ph/9910475 knez , c. , boogert , a.c.a . ,
pontoppidan , k.m . , kessler - silacci , j. , van dishoeck , e.f . ,
evans , n.j .
ii , augereau , j .- c . ,
blake , g.a . , & lahuis , f. 2005 , apj , 635 , l145 lee , j .- e . ,
evans , n.j.,ii , shirley , y.l . , & tatematsu , k. 2003 , apj , 583 , 789 lee , j .- e . , bergin , e.a .
, & evans , n.j.ii 2004 , apj , 617 , 360 lin , d.n.c . , hayashi , m. , bell , k.r .
, & ohashi , n. 1994 , apj , 435 , 821 masunaga , h. , miyama , s.m .
, & inutsuka , s .- i .
1998 , apj , 495 , 346 berg k.i .
, van broekhuizen , f. , fraser , h.j . , bisschop , s.e .
, van dishoeck , e.f . , & schlemmer , s. 2005 , apj , 621 , 33 myers , p.c . & ladd , e.f . , 1993 ,
apj , 413 , l47 omukai , k .
2007 , pasj , 59 , 589 ossenkopf v. & henning th .
, 1994 , a&a , 291 , 943 pontoppidan , k.m , boogert , a.c.a . ,
fraser , h.j . ,
van dishoeck , e.f . ,
blake , g.a . ,
lahuis , f. , berg , k.i . , evans , n.j.,ii , & salyk , c. 2007 , apj , in press shu , f.h .
1977 , apj , 214 , 488 shu , f.h . , adams , f.c . , & lizano , s. 1987 , ara&a , 25 , 23 vorobyov , e.i . & basu , s. 2005 , apjl , 633 , 137 young , c.h .
, et al . 2004 , apjs , 154 , 396 young , c.h .
, & evans , n.j .
2005 , apj , 627 , 293 | a new type of object called very low luminosity objects ( vellos ) " has been discovered by the spitzer space telescope .
vellos might be substellar objects forming by accretion . however , some vellos are associated with strong outflows , indicating the previous existence of massive accretion . the thermal history , which significantly affects the chemistry , between substellar objects with a continuous low accretion rate and objects in a quiescent phase after massive accretion ( outburst ) must be greatly different .
in this study , the chemical evolution has been calculated in an episodic accretion model to show that co and n@xmath0h@xmath1 have a relation different from starless cores or class 0/i objects .
furthermore , the co@xmath0 ice feature at 15.2 @xmath2 m will be a good tracer of the thermal process in vellos . |
the hadronization mechanism is an important but still unsolved problem up to now due to its nonperturbative nature .
it is recognized that the hadronization mechanism is universal in all kinds of high energy reactions , _
e.g_. , @xmath2 annihilation , and hadron(nuclear)-hadron(nuclear ) collisions . among these reactions , @xmath2 annihilation at high energies , especially at the @xmath0 factory in the future , is best for studying the hadronization mechanism , since all the final hadrons come from primary ones , all of which are hadronization results . the @xmath3 process is generally divided into four phases ( see fig .
[ hdpro ] ) . 1 . in the electro - weak phase , @xmath4 pair converts into a primary quark pair @xmath5 via virtual photon or @xmath6 .
this phase is described by the electro - weak theory .
2 . the perturbative phase describes the radiation of gluons off the primary quarks , and the subsequent parton cascade due to gluon splitting into quarks and gluons , and the gluon radiation of secondary quarks .
it is believed that perturbative qcd can describe this phase quantitatively .
3 . in the hadronization phase ,
the quarks and gluons interact among themselves and excite the vacuum in order to dress themselves into hadrons , that is , the confinement is ` realized ' .
since this process belongs to the unsolved nonperturbative qcd , investigations employing various models will shed light on understanding this process .
4 . in the fourth phase , unstable hadrons decay .
this phase is usually described by using experimental data .
this paper focuses on step 3 .
the main method is to compare results of various hadronization models with the data , and the future @xmath0 factory is very suitable for this purpose .
the popular hadronization models at the market , lund string fragmentation model(lsfm)@xcite , webber cluster fragmentation model(wcfm)@xcite , and quark combination model(qcm ) , succeed in explaining a lot of experimental data in @xmath7 and @xmath8 processes by adjusting corresponding parameters .
recently , the baryon to meson ratio@xcite and constituent quark number scaling of elliptic flow @xmath9@xcite are measured at rhic experiments , which do not favor the fragmentation model , while the qcm can explain these phenomena [email protected] was first proposed by annisovich and bjorken _
et al_. @xcite .
it was famous for its simple picture and its successful prediction of the percentage of vector mesons .
one of its great merits is that it treats the baryon and meson production in an uniform scheme , so it describes the baryon production naturally .
however , for a long period qcm is regarded as being ruled out , because the prediction for baryon - antibaryon(@xmath1 ) rapidity correlation in @xmath2 annihilation by cerny s monte carlo program , which was alleged to be based on qcm , has great discrepancies with the experimental results of tasso collaboration@xcite . but early in 1987 , the @xmath1 phase space correlation from the naive qcm scheme was analyzed , which showed that there should not be such an inconsistency qualitatively . in the meantime
, quark production rule and quark combination rule(the so - called ` shandong quark combination model(sdqcm ) ' ) were developed .
then a serials of quantitative results obtained by sdqcm@xcite confirmed that sdqcm can naturally explain the @xmath1 short range rapidity correlation together with baryon to meson ratio when the multi - parton fragmentation is included . in order to understand the hadronization phenomena especially in heavy ion collisions ,
it is necessary to study the hadronization mechanism in detail in @xmath2 annihilation once @xmath0 factory is available . on the other hand ,
the hadronization model serves as a bridge between the perturbative qcd and experiments , so it is a very important tool for studying , _
e.g. _ , lhc physics .
the properties of the light hadrons have been studied in our previous works . here
we focus on investigating the production of heavy hadrons(_e.g_. , @xmath10 , _
etc_. ) by lsfm and sdqcm .
this paper is organized as follows : in section [ ii ] , we give a brief introduction to the popular hadronization models , _
i.e. _ , lund string fragmentation model , webber cluster fragmentation model and quark combination model .
some numerical results are given in section [ iii ] .
finally , a short summary and outlook close the study .
string fragmentation model , first proposed by artru and mennesser in 1974@xcite , has been developed by the theory physics group of lund university since 1978 , and corresponding monte - carlo programs(_e.g_. , jetset , pythia , _
etc_. ) are written . by
now pythia is one of the most widely used generators describing the high - energy collisions .
the hadronization of @xmath5 color - singlet is the simplest case for string fragmentation .
lattice qcd supports a linear confinement potential between color charges,_i.e .
_ , the energy stored in the color dipole field increases linearly with the separation between them .
the assumption of linear confinement is the starting point for the string model . as the @xmath11 @xmath12 move away in the opposite direction , the kinetic energy of the system changes into the potential energy of the color string ( or color flux tube ) . when the potential increases to a certain extent the string will break by the production of a new quark pair , and the production possibility is given by the quantum mechanical tunnelling@xcite @xmath13 where @xmath14 is the transverse mass of @xmath15 quark , and @xmath16 , the string constant , denotes the potential energy per unit length .
considering the assumption of no transverse excitation of the string , the @xmath17 is locally compensated between @xmath18 and @xmath19 . from eq .
( [ eq : tunneleffect ] ) , one sees clearly that the probability of the different flavors is about @xmath20 .
so @xmath21 quarks are not expected to be produced through fragmentation , but only in the perturbative qcd procedure . when the new quark pair @xmath22 are excited out in vacuum , the @xmath5 color string splits into @xmath23 and @xmath24 two color - singlets .
if the invariant mass of either of the singlet system is large enough , further break will occur .
the process ends when all string pieces become exactly hadrons on their mass - shells .
this picture describes meson production naturally other than the production of baryons .
more complex diquark and popcorn mechanisms have to be introduced to explain the baryon production .
one prominent feature of lsfm is that energy momentum and flavour are conserved at each step of the fragmentation process . in @xmath7 process at high energies ,
multi - parton states will be produced at the end of perturbative phase , so that the multi - parton fragmentation has to be taken into account .
the corresponding procedures have been developed in ref .
the fragmentation picture of lund model is shown in fig .
[ figld ] .
the cluster fragmentation model has been proposed by wolfram@xcite in 1980 .
webber cluster fragmentation model(wcfm)@xcite is the well - known example .
it has three parts : * the formation of color - singlet cluster .
gluons split into quark pairs after parton shower : @xmath25 .
adjacent quark and anti - quark from different gluons combine to form a color - singlet .
* cluster with large invariant mass fragments into smaller ones . *
little cluster decays into primary hadrons._i.e .
_ , @xmath26 .
note that only two - body decay or fragmentation is adopted in the webber cluster fragmentation model .
the corresponding hadronization picture is local , universal and simple .
its fragmentation picture is shown in fig .
[ figwebber ] . in the framework of sdqcm , quark production rule(qpr ) , quark combination rule ( qcr )
are adopted to describe the hadronization in a color singlet system , and then a simple longitudinal phase space approximation ( lpsa ) is used to obtain the momentum distribution for primary hadrons in its own system .
finally this hadronization scheme is extended to the multi - parton states .
here we briefly introduce the sdqcm and list the relevant equations(for detail , see ref .
@xcite ) . in a color singlet system formed by @xmath5 , @xmath27 pairs of quarks can be produced by vacuum excitation via strong interaction .
we assume that @xmath27 satisfies poisson distribution : @xmath28 where @xmath29 is the average number of those quark pairs . according to qpr
, @xmath29 is given by @xmath30 where @xmath31 is the invariant mass of the system , @xmath32 is a free parameter , @xmath33 is the average mass of newborn quarks , and @xmath34 and @xmath35 are the masses of endpoint quark and anti - quark .
thus we have @xmath27 pairs of quarks according to eqs .
( [ fir ] ) , ( [ fir1 ] ) ( containing one primary quark pair ) .
when describing how quarks and antiquarks form hadrons , we find that all kinds of hadronization models satisfy the near correlation in rapidity more or less .
since there is no deep understanding of the significance and the role of this , the near rapidity correlation has not been used sufficiently . in ref .
@xcite , we have shown that the nearest correlation in rapidity is in agreement with the fundamental requirements of qcd , and determines qcr completely .
the rule guarantees that the combination of quarks across more than two rapidity gaps never emerges and that @xmath27 quarks and @xmath27 antiquarks are exactly exhausted , which guarantees the unitarity ( see ref .
@xcite ) . given that the quarks and antiquarks are stochastically arranged in rapidity space , each order can occur with the same probability .
then the probability distribution for @xmath27 quarks and @xmath27 antiquarks to combine into @xmath36 mesons , @xmath37 baryons and @xmath37 anti - baryons according to qcr is given by @xmath38 the average numbers of primary mesons @xmath39 and baryons @xmath40 are @xmath41 approximately , in the combination for @xmath42 , @xmath39 and baryons @xmath40 can be well parameterized as linear functions of quark number @xmath27 , @xmath43 where @xmath44 and @xmath45 . but for @xmath46 , one has @xmath47 so that , the production ratio of baryon to meson is obtained from eqs .
( [ w2 ] ) and ( [ w3 ] ) @xmath48 we see that sdqcm treats meson and baryon formation uniformly , and there is no extra @xmath49 mechanism and free parameters for the baryon production . here
the @xmath50 ratio is completely determined at a certain @xmath27 , unlike in lsfm it is completely uncertain and has to be adjusted by free parameters , i.e , the ratio of diquark to quark @xmath51 . in order to give the momentum distribution of primary hadrons , each phenomenological model must have some inputs .
for example , in lsfm , they use a symmetric longitudinal fragmentation function @xmath52 where @xmath53 and @xmath54 are two free parameters ( and @xmath53 is flavor dependent ) . in this paper , in order to give the momentum distribution of primary hadrons produced according to qpr and qcr , we simply adopt the widely used lpsa which is equivalent to the constant distribution of rapidity .
hence a primary hadron @xmath55 is uniformly distributed in the rapidity axis , and then its rapidity can be written as @xmath56 where @xmath57 is a random number ; @xmath58 and @xmath59 are two arguments , and can be determined by energy - momentum conservation in such a color singlet system @xmath60 where @xmath61 and @xmath62 denote the energy and the longitudinal momentum of the @xmath55th primary hadron respectively , obtained by @xmath63 where @xmath64 is given by @xmath65 where @xmath66 is the mass of the @xmath55th primary hadron , and @xmath67 obeys the distribution @xmath68 in this paper , we set @xmath69 .
( [ last ] ) is just what lsfm uses .
note that lpsa or the constant rapidity distribution is rather naive , but it is convenient for us to study the correlations without introducing many complicating parameters . at the end of parton showering ,
a final multi - parton state will start to hadronize . to connect the final multi - parton state with sdqcm
, we adopt a simple treatment assumed in wcfm , _
i.e. _ before hadronization , each gluon at last splits into a @xmath70 pair , the @xmath18 and @xmath71 carry one half of the gluon momentum , and each of them forms a color singlet with their counterpart antiquark and quark in their neighbourhood , respectively .
now take the three parton state @xmath72 as an example to illustrate the hadronization of a multi - parton state .
denote the 4-momenta for @xmath11 , @xmath12 , @xmath73 as @xmath74 before hadronization , the gluon splits into a @xmath70 pair and the @xmath18 and @xmath71 carry one half of the gluon momentum , and the @xmath72 system forms two color singlet subsystems @xmath75 and @xmath24 .
the invariant masses of the subsystems are @xmath76 as was commonly argued by sjstrand and khoze recently@xcite , the confinement effects should lead to a subdivision of the full @xmath5 system into color singlet subsystems with screened interactions between these subsystems @xmath75 and @xmath24 .
hence sdqcm can be applied independently to each color singlet subsystem,_i.e .
_ , we can apply the equations in the former two subsections to each subsystem , and obtain the momentum distribution for the primary hadrons in their own center - of - mass system .
then after lorentz transformation , the momentum distribution of the primary hadrons in laboratory frame is given .
this treatment can be extended to a general multi - parton state , for example , see fig .
[ qcm1a ] . obviously , when the emitting gluon is soft or collinear with the direction of @xmath11 or @xmath12 , @xmath72 can not be distinguished from @xmath5 and @xmath77 or @xmath78 is too small for hadronization . for the avoidance of these situations ,
a cut - off mass @xmath79 has to be introduced . here
@xmath79 is a free parameter in the perturbative phase .
its value and energy dependence is theoretically uncertain .
the physical assumption is that @xmath79 is independent of energy@xcite .
in this paper , we study the hadronization process at @xmath0 factory by lsfm and sdqcm .
we focus on investigating the properties of the heavy hadrons , _
e.g_. , the ratio of baryon to meson and the baryon - antibaryon correlations . for lsfm
, we use the default parameter values in pythia , while for sdqcm , we adopt the parameters used in ref .
@xcite which fit data quite well .
the final hadron multiplicity at high energy @xmath2 reactions is always an important topic .
we list the predictions of final hadron multiplicities at @xmath0 factory(@xmath80 ) by lsfm and sdqcm in table i. the results of lsfm are obtained by running pythia6.4 with the default parameters .
the experimental data are from ref .
it is found that the predictions of lsfm and those of sdqcm are consistent with most of the experimental data . however , the multiplicity of the heavy baryons , _
eg_. , @xmath81 , @xmath82 , are still not measured at @xmath0 energy , and the corresponding theoretical predictions obtained by lsfm and sdqcm are quite different . therefore it is important to measure their production rates at @xmath0 pole in order to discriminate different hadronization mechanisms . at the @xmath0 energy
, we study the relation between the integrated luminosity and the producing number of the heavy baryons .
the corresponding results are displayed in fig.[lumz0 ] .
it is obvious that for the lep i , the production number for @xmath81 predicted by lsfm(sdqcm ) is about @xmath83 , and that for @xmath84 is several tens(hundreds ) . in order to study the heavy baryon production mechanism with enough precision
, the integrated luminosity should be increased as large as possible .
for example , if the integrated luminosity of @xmath0 factory can reach about @xmath85 , the produced number of @xmath84 may reach several thousands(ten thousands ) according to the prediction of lsfm(sdqcm ) , which makes it possible to study the properties of the heavy baryons and to test the hadronization models with higher precision .
[ multiplicities ] .results for average hadron multiplicities at @xmath0 factory .
the experimental data are from ref .
[ cols="<,^,^,^",options="header " , ]
in the @xmath86 process , the final particles have no relations with the structure of the initial particles , so it is convenient to study the hadronization mechanism . above all
, such studies help to investigate the multi - production processes in other high energy reactions(_e.g .
_ , at rhic energy ) . in this paper , we report the predictions especially related to heavy hadrons , for example , baryon to meson ratio and @xmath1 flavor correlations obtained by lsfm and sdqcm in the hadronic @xmath0 decay .
these results show that the future experiments at @xmath0 factory to test hadronization mechanism , especially rare hadronized events , such as the doubly heavy baryon production , are significant .
though many studies have been made for the @xmath7 process , still a lot of unclear but important topics are worth studying , such as the color connection effects@xcite .
the color connection among final partons plays a crucial role in the surface between pqcd phase and the hadronization one(see fig .
[ hdpro ] ) .
different choices of the color connections of the final multi - parton system leads to different hadronization results .
recently , the doubly heavy baryon(_e.g .
_ , @xmath87 ) production has been investigated in different reactions@xcite . in @xmath2 annihilation
, the doubly heavy baryon production reveals a special kind of color connection of the final parton system @xmath88(@xmath89 ) .
this kind of color connection has not been considered in the popular hadronization models . in ref .
@xcite , the hadronization effects of this kind of color connection are investigated in an extremely limited case of @xmath90 process at @xmath37 factory energies . there
_ cq _ and @xmath91 form diquark - antidiquark pair , which is treated as a color singlet string and fragments into hadrons in the framework of lsfm .
however due to the phase space limitation , the hadronization effect induced by diquark pair fragmentation is small .
fortunately , in the future @xmath0 factory , a large number of events such as @xmath92 will be produced@xcite , so that the correlated physical observables can be measured with higher statistics .
as a result , the hadronization mechanism and the related non - perturbative nature can be well studied at @xmath0 pole energy .
this work is supported in part by nsfc , natural science foundation of shandong province(zr2009am001 ) , and doctoral science foundation of university of jinan(b0527 ) .
the authors would like to thank prof .
s. y. li for his kindly helpful discussions .
s. s. adler _ et al .
_ [ phenix collaboration ] , phys .
rev.lett . * 91 * , 182301 ( 2003 ) ; j. adams _ et al . _ [ star collaboration ] , phys .
lett . , 052302 ( 2004 ) ; j. adams _ et al . _ [ star collaboration ] , phys .
lett . , 122301 ( 2005 ) .
r. c. hwa and c. b. yang , phys .
c * 67 * , 034902 ( 2003 ) .
f. l. shao , q. b. xie and q. wang , phys .
c * 71 * , 044903 ( 2005 ) ; t. yao , q. b. xie and f. l. shao , chinese physics c * 32 * , 356 ( 2008 ) ; t. yao , w. zhou and q. b. xie , phys .
c * 78 * , 064911 ( 2008 ) .
qu - bing xie , in proceedings of the xixth international symposium on multi - particle dynamics 1988 , edited by d. schiff and j. tran thanh van ( world scientific , singapore 1988 ) , p. 369 , and the references therein .
f. l. shao , q. b. xie , s. y. li and q. wang , phys .
d * 69 * , 054007 ( 2004 ) ; y. jin , q. b. xie , and s. y. li , high ener .
phys . and nucl .
phys . * 27 * , 282 ( 2003)(in chinese ) ; s. y. li , f. l. shao , q. b. xie and q. wang , phys .
d * 65 * , 077503 ( 2002 ) ; f. l. shao and q. b. xie , high ener .
phys . and nucl
. phys . * 25 * , 710 ( 2001)(in chinese ) ; q. wang , g. gustafson , y. jin , and q. b. xie , phys .
d * 64 * , 012006 ( 2001 ) ; q. wang , g. gustafson , and q. b. xie , phys .
d * 62 * , 054004 ( 2000 ) ; s. y. li , z. g. si , q. b. xie and q. wang , phys .
b * 458 * , 370 ( 1999 ) ; z. g. si , and q. b. xie , high energy phys .
* 22 * , 307 ( 1998)(in chinese ) ; z. g. si , q. wang and q. b. xie , phys .
b * 401 * , 107 ( 1997 ) ; l. l. tian , q. b. xie and z. g. si , high energy phys . nucl .
* 17 * , 147 ( 1993 ) ; l. l. tian , q. b. xie and z. g. si , phys . rev .
d * 49 * , 4517 ( 1994 ) .
j. p. ma and z. g. si , phys .
b * 568 * , 135 ( 2003 ) ; s. y. li , z. g. si and z. j. yang , phys .
b * 648 * , 284 ( 2007 ) ; c. h. chang , j. p. ma , c. f. qiao and x. g. wu , j. phys .
g * 34 * , 845 ( 2007 ) ; b. aubert _ et al .
_ [ babar collaboration ] , phys .
d * 74 * , 011103 ( 2006 ) . | we briefly review the hadronization pictures adopted in the lund string fragmentation model(lsfm ) , webber cluster fragmentation model(wcfm ) and quark combination model(qcm ) , respectively .
predictions of hadron multiplicity , baryon to meson ratios and baryon - antibaryon flavor correlations , especially related to heavy hadrons at @xmath0 factory obtained by lsfm and qcm are reported . + * keywords : * @xmath0 factory ; hadronization model ; baryon to meson ratio ; @xmath1 correlation + * pacs : * 12.15.ji , 13.38.dg , 24.10.lx |
topological insulators ( ti ) are a new class of materials whose surfaces / edges host bound spin - polarized dirac fermions with high mobility . these states , in a time reversal invariant system
are protected against perturbation and non - magnetic disorder .
@xcite well - known examples of materials with such properties include bi@xmath0te@xmath1 , bi@xmath0se@xmath1 , sb@xmath0te@xmath1 , bi@xmath3sb@xmath4 alloy , and the cdte / hgte / cdte quantum well heterostructure .
bi@xmath0te@xmath1 , bi@xmath0se@xmath1 , sb@xmath0te@xmath1 , and bi@xmath3sb@xmath4 alloy belong to the class of 3d topological insulators ( 3d ti ) @xcite and host bound states on their surface .
these states are characterized by a linear hamiltonian and form a band - gap closing dirac cone on each surface at the @xmath5 point . in this work , methods that alter the surface state dispersion through proximity induced ferromagnetism , superconductivity , and
finite - size effects are described .
alteration of the surface dispersion is driven in part by the need to utilize these materials as field effect transistors ( fets ) that uses the highly mobile zero - gap surface or edge states as the conducting channel .
the zero band gap surface states though prohibit an easy turn - off mechanism and therefore inducing a band gap is essential for successful transistor operation .
a key idea to open a band gap is destruction of time reversal symmetry @xcite which is responsible for protecting the surface states .
time reversal symmetry , which in principle , can be destroyed in the presence of an external perpendicular magnetic field @xcite and appears to be a straightforward approach in a standard experimental set - up , may not be applied easily to a miniaturized device on account of electromagnetic interference with the surrounding circuitry .
alternatively , the surface can be coated with a ferromagnet with an out - of - plane magnetization component to replicate this behaviour .
it is now well established that doping with cu or pb can introduce superconducting states in bi@xmath0se@xmath1 . additionally , through the proximity effect at the interface between a superconductor ( sc ) and topological insulator
, the superconductor s wave functions can penetrate the surface of a topological insulator and induce superconductivity .
this induced superconductivity , by virtue of its intrinsic energy gap between the fermi - level and the superconducting ground state offers a possible way to open a band - gap in a topological insulator .
a modified version of the bogoliubov - de gennes ( bdg ) hamiltonian for a 3d - ti and an _ s_-wave superconductor is used to compute the dispersion relationship of a ti - sc heterostructure .
the paper is structured as follows : in section [ meta ] , the 4-band k.p hamiltonian to compute the surface and edge state energy dispersion is presented .
dispersion relations obtained through them are utilized later to investigate finite - size effects on surface and edge states of 3d and 2d tis respectively .
the 4-band hamiltonian is modified via an exchange - interaction term to model proximity - induced ferromagnetism on the surface of a 3d ti .
the ti - sc heterostructure is represented by a bdg hamiltonian and dispersion when the superconductor is an _
s_-wave or _
p_-wave type is worked out analytically .
spin matrices using the basis states of the 4-band hamiltonian are set up to compute spin - polarization of surface states .
numerical calculations based on the hamiltonians created for several ti systems in section [ meta ] are collected in section [ res ] .
the paper concludes by summarizing key results in section [ conc ] .
surface and edge states in topological insulators are characterized by a linear dispersion and dirac cones .
they further depend on dimensions and growth conditions of the structure that host them .
low - energy continuum models for 3d and 2d tis used in deriving results included in this paper are described in this section .
the dispersion relations of bi@xmath0te@xmath1 , bi@xmath0se@xmath1 , and sb@xmath0te@xmath1 films are computed using a 4-band k.p hamiltonian . the 4-band hamiltonian @xcite is constructed ( eq . [ eqn1 ] ) in terms of the four lowest low - lying states @xmath6 , @xmath7 , @xmath8 , and @xmath9 .
additional warping effects @xcite that involve the @xmath10 term are omitted in this low - energy effective hamiltonian .
@xmath11 where @xmath12 , @xmath13 and @xmath14 . for bi@xmath0te@xmath1 and bi@xmath0se@xmath1 ,
the relevant parameters are summarized in table .
[ table1 ] .
[ table1 ] lcc parameters & bi@xmath0te@xmath1 & bi@xmath0se@xmath1 + m@xmath15 ( ev ) & 0.30 & 0.28 + a@xmath16 ( ev ) & 2.26 & 2.2 + a@xmath0 ( ev ) & 2.87 & 4.1 + b@xmath16 ( ev @xmath17 ) & 10 & 10 + b@xmath0 ( ev @xmath17 ) & 57.38 & 56.6 + c ( ev ) & -0.18 & -0.0068 + d@xmath16 ( ev @xmath17 ) & 6.55 & 1.3 + d@xmath0 ( ev @xmath17 ) & 29.68 & 19.6 + the dispersion for a 20.0 @xmath18 bi@xmath0se@xmath1 thick film which is approximately twenty quintuple - layers is shown in fig .
the dirac cone is formed at an energy equal to 0.029 @xmath19 confirming that it is indeed a mid - gap state .
the bulk band - gap of bi@xmath0se@xmath1 is approximately 0.32 @xmath19 at the @xmath20 point .
in contrast to the thick - film dispersion , the band profile of a 3.0 @xmath18 ( approximately three quintuple - layers ) bi@xmath0se@xmath1 film has a finite band gap . in the case of a thin - film , the two surface states hybridize .
the hybridization occurs because each state has a definite localization or penetration length .
when the penetration length is comparable to film thickness , the opposite spin - resolved bands of the two surfaces will mix . since bands with identical quantum numbers can not cross , a gap(fig .
[ fig1]b ) opens up at the @xmath20 point and the dispersion changes to dirac - hyperbolas .
it must be stated here , that though a thin film on account of intrinsic hybridization , which is a function of dimension , is not a preferred choice to manipulate the band gap .
the band gap control must be performed externally as described in the latter part of this paper .
two variants of a topological insulator film are considered .
the first is a free - standing symmetric thin - film of bi@xmath0se@xmath1 .
two exactly degenerate dirac cones are formed for a symmetric film .
the second thin film considered is assumed to grow on a substrate which induces an asymmetry between the two surfaces .
the asymmetry can be explained by assuming that the top surface is exposed to vacuum while the bottom surface is fixed to a substrate .
inversion symmetry therefore does not hold @xcite for such a ti film .
this asymmetry is simulated by adding a small symmetry - breaking potential to the hamiltonian along the confinement direction .
it can also be viewed as a simple way of creating two chemically distinct surfaces .
the dirac cones , one on each surface , are therefore no longer degenerate .
they are now split as a function of the applied symmetry - breaking potential and positioned at distinct energies .
this splitting is similar to a rashba - split @xcite in presence of structural inversion asymmetry ( sia ) .
the asymmetry of the surfaces , in this case , causes the sia .
when the conditions of ultra - thin film and asymmetry are simultaneously fulfilled , it is observed that the two surface bands offer another instance of rashba - type splitting . to explain this phenomenon , first a low thickness film is considered .
the two degenerate massless dirac cones in this case would mix and open a finite band - gap as explained above .
each dirac cone can now be represented as a massive spin - degenerate dirac hyperbola .
@xcite if an additional asymmetry is imposed through a potential , the spin degeneracy of each dirac hyperbola is broken and the bands split along @xmath21 in opposite directions .
the degeneracy is only maintained at the @xmath20 point .
the conventional rashba - type splitting follows exactly the same pattern , for example in an asymmetric gaas quantum well .
@xcite a hallmark of the surface bands of a 3d - topological insulator is their intrinsic complete spin - polarization and locking of the spin perpendicular to the momentum . within the framework of the 4-band k.p hamiltonian ,
expectation value of the three spin - polarized vectors is computed .
the operators for the three spin - polarizations are given by eq .
[ eqn2 ] and eq .
[ eqn3 ] in the pauli representation @xmath22 @xmath23 .
@xmath24 @xmath25 the above matrices are written under the basis set ordered as @xmath6 , @xmath7 , @xmath8 , and @xmath9 .
the expectation value for each spin - polarization operator is calculated in the usual way in eq .
( [ eqn4 ] ) .
@xmath26 where @xmath27 the spin - polarization shown in fig .
[ fig2 ] is computed by choosing a @xmath28-vector of magnitude 0.01 @xmath29 at various polar angles.for each such vector , the spin - polarization obtained was perpendicular to momentum with a zero out - of - plane component .
the red - arrows in fig .
[ fig2 ] denote the direction of spin - polarization and is always tangential to the circular contour traced out by the @xmath28-vector .
point for a 20.0 @xmath30 thick bi@xmath0se@xmath1 film .
the spin is locked to momentum ( which is a radial vector on the circle ) shown by the tangential lines on the plot . ]
2d - tis which exhibit edge states are also highly mobile and suitable for nanoscale applications .
@xcite one example of edge states in a 2d - ti is a cdte / hgte / cdte nanoribbon obtained by quantizing one of the edges of the corresponding quantum well
. it would be useful to compare length scales up to which an edge and surface state can be sustained in a 3d and 2d topological insulator system . while an eight - band k.p hamiltonian describes the full set of six valence ( including spin split - off ) and two conduction bands and their mutual interaction through the off - diagonal terms , it is sufficient to focus on bands that exclusively take part in the inversion process @xcite .
this interaction of bands is governed by the coupling of conduction and valence states , represented through a linear term as shown in eq . . @xmath31 where @xmath32 and @xmath33 @xmath34 describes band bending and 2__m@xmath15 _ _ =
-__e__@xmath35 corresponds to energy gap between bands and is negative in the inverted order bands .
this hamiltonian is written in the basis of the lowest quantum well sub - bands @xmath36 , @xmath37 , @xmath38 , and @xmath39 . here
@xmath40 stands for the two kramers partners .
the sign of the gap parameter _ m
_ determines if it is a trivial insulator ( _ m _ @xmath41 0 ) or a topological insulator ( _ m _ @xmath42 0 ) .
experimentally , _
m@xmath15 _ is tuned by changing the quantum well width .
the parameters _ a , b , c , d , and m@xmath15 _ are geometry dependent . for numerical calculations performed in this paper ,
the parameters were set to : @xmath43 , @xmath44 , @xmath45 , @xmath46 , and @xmath47 . before a complete hamiltonian for a ti - superconductor heterostructure
can be written , the conventional bcs description @xcite of an _
s_-wave superconductor must be examined .
the bcs hamiltonian in its simplest form can be written by starting with a hamiltonian ( eq . [ bcs3 ] ) that describes a many - fermion system with a spin - independent interaction potential .
such a hamiltonian in general form is expressed as : @xmath48 by restricting to paired fermions with zero total momentum and opposite spin , the bcs hamiltonian can be written as @xmath49 the spectrum of this hamiltonian when solved using the bogoliubov transformation @xcite yields a band structure @xcite with a gap in the spectrum . for studying proximity effect between a superconductor @xcite and a topological insulator , the 4-band k.p model and the bcs hamiltonian
is used in conjunction .
the fundamental assumption of the bcs hamiltonian is the formation of cooper pairs which are paired electrons with zero momentum and spin .
superconductivity which is induced on the ti side of the ti - sc heterostructure must therefore agree to this principle . at this point
it is worth mentioning again that the four orbitals participating in the electronic bonding process are @xmath6 , @xmath7 , @xmath8 , and @xmath9 .
the composite hamiltonian for the ti - sc structure similar to the bdg hamiltonian can now be written as @xmath50 in eq .
[ bdgprimer ] , @xmath51 denotes the chemical potential and _ t _ is the time reversal operator .
h@xmath52 is the composite hamiltonian and h@xmath53 represents the intrinsic 4-band k.p hamiltonian . in the composite hamiltonian , eq .
[ bdgprimer ] , @xmath54 is the pair potential where @xmath55 is the order parameter and @xmath56 characterizes the strength of the attractive interaction potential as established in the bcs formulation . for the _ s_-wave superconductor considered here , the pair - potential , @xmath57 is isotropic and just a number . the analytic representation of pair - potential changes to a @xmath58 dependent quantity if _
p _ or _ d_-type superconductors are considered .
the pair potential @xmath59 denotes a correlated pair of electrons with opposite momentum and spin . for the case of a ti , which is turned in to a superconductor ,
the orbitals with opposite spin and momentum are paired .
the two sets of orbitals in the 4-band ti hamiltonian are therefore coupled by two pair potentials .
the full ti - sc hamiltonian _
h@xmath52 _ in the basis set @xmath6 , @xmath7 , @xmath8 , @xmath9 , @xmath60 , @xmath61 , @xmath62 , and @xmath63 can be now written as @xmath64 where @xmath12 , @xmath13 and @xmath14 . in the above hamiltonian
, the fermi - level @xmath51 has been set to zero .
also , the pair potential @xmath65 is assumed for the zero - momentum fermion pair @xmath66 .
for the corresponding , @xmath67 , the pair potential is @xmath68 .
@xcite a more accurate order parameter can be obtained through a self - consistent calculation .
for instance the order parameter @xmath65 , is given as @xmath69 in eq .
[ delta1 ] , @xmath70 and @xmath71 refer to the wave function components in the 8 @xmath72 1 column vector that correspond to @xmath6 and @xmath62 .
details of the self - consistent calculation of the order parameter is discussed in section [ res ] . as a concrete demonstration of band - gap alteration , an _
s_-wave superconductor which is characterized by an isotropic energy - gap and conveniently represented as @xmath73 is considered .
the surface states can be modeled through a dirac hamiltonian written as @xmath74 . adding a pair potential such that fermions with opposite spins and momentum are coupled , the hamiltonian takes a simple form @xmath75 for an s - type superconductor , the pair potential is independent of the @xmath76 , the hamiltonian given in eq .
( [ red_ham ] ) can now be solved to obtain a dispersion relationship of the form @xmath77 for a p - type superconductor , the pair potential is generally expressed as @xmath78 .
the corresponding dispersion gives the following eigenvalues @xmath79 it is easy to see that at the @xmath5 point the pair potential is equal to zero and band gap closing is preserved .
this is qualitatively different from the _ s_-wave case discussed earlier .
@xcite applied magnetic field represents a significant way to open a band gap in the spectrum near @xmath80 .
the @xmath80 degeneracy is protected by time reversal symmetry which can be destroyed by a magnetic field .
the applied magnetic field is incorporated in the hamiltonian through the standard peierls substitution @xmath81 where @xmath82 is the magnetic vector potential . the simplest way to add a magnetic field is layer the surface of a ti with a ferromagnet . the ferromagnetic proximity exchange field polarized along an arbitrary direction
@xcite can be written in the most general way as @xmath83 .
to illustrate the effect of a magnetic field , the spectrum of a ti slab under a @xmath84 field directed parallel and perpendicular to the surface is computed . for a parallel orientation ,
the states remain gap - less but are shifted along the in - plane direction perpendicular to the field .
the surface states around each dirac cone are approximately described as @xmath85 this gives a linear dispersion relationship of the form @xmath86 .
a @xmath87 field perpendicular to the surface ( 0,0,@xmath84 ) can be represented using a landau gauge @xmath88 . using the standard peierls transformation ,
the simple dirac hamiltonian is transformed to @xmath89 the eigen values are landau levels given as @xmath90 . a simpler dirac hamiltonian which captures the phenomenlogical behaviour of surface electrons in presence of the exchange field
is shown in eq .
[ dferro ] .
@xmath91 however , for a surface parallel to the magnetic field , the same gauge can now be written as @xmath92 . using the same peierls transformation ,
the @xmath93 vector is now transformed to @xmath94 .
the @xmath95 vector remains unchanged . for an x - y surface , _ z
_ is a constant which means that @xmath96 is a constant too and commutes with @xmath93 .
the new set of eigen values are therefore written as @xmath97 the only difference is that dirac cones now appear at @xmath98 .
in other words , the eigen energies of the system are independent of the choice of gauge .
two structures are considered to understand the geometric dependence of surface and edge states .
3d topological insulators with surface states are modeled as thin films of finite thickness along the _ z _ direction while 2d topological insulators exhibiting edge states are nano - ribbons with two dimensional confinement along _ y _ and _ z _ directions .
it must be pointed out here that a nano - ribbon is constructed out of an inverted quantum well @xcite to allow topological edge states .
2d tis are modeled using the bhz hamiltonian described in section [ meta ] . for a bi@xmath0se@xmath1 film
, it is observed that surface states are preserved for thickness as low as 9.0 @xmath30 while the corresponding edge states in a cdte / hgte / cdte 2d - ti nanoribbon requires a thickness of at least 100 @xmath2 .
a shows a nanoribbon 100.0 @xmath18 wide and 8.0 @xmath18 thick .
this structure shows a dirac crossing and topological edge state .
[ fig3]b is 50.0 @xmath18 wide and 8.0 @xmath18 thick ribbon with a finite band gap . ]
this suggests that much smaller samples that retain their conducting surface states can be fabricated with 3d tis than in the case of a cdte / hgte / cdte 2d - ti .
the physical interpretation of these length scales can be sought in the delocalization length or penetration depth of these states .
the penetration depth is usually expressed as @xmath99 , where _ m _ is the bulk band gap and @xmath100 is the fermi - velocity . by just noting the charge excitation gap , it is obvious that large difference in penetration depth can be easily explained .
the bulk excitation gap of bi@xmath0se@xmath1 is much higher than that of hgte , which is a zero band gap compound .
another result that supports this observation is the film thickness at which a gap opens up in a sb@xmath0te@xmath1 compared to bi@xmath0se@xmath1 . as a numerical example , the band gap opening in a 5.0 @xmath18 thin - film is 0.0084 @xmath19 and 0.0629 @xmath19 in bi@xmath0se@xmath1 and sb@xmath0te@xmath1 respectively . assuming comparable fermi velocities , sb@xmath0te@xmath1 which has a smaller bulk band - gap at @xmath20 ,
the surface states would be more delocalized than bi@xmath0se@xmath1 . as a result of greater delocalization , complete hybridization occurs and gives a larger band gap opening .
the dependence of band - gap for various bi@xmath0se@xmath1 film - thickness is plotted in fig .
[ fig4 ] .
se@xmath1 film thickness .
band gap opens because the two surfaces hybridize . ] for a symmetric thin film with identical surfaces , the two dirac cones representing the surface states are degenerate .
asymmetry due to two different surfaces though can exist on account of inequivalent surface termination or presence of a substrate .
this naturally occurring asymmetry @xcite can be reproduced by applying an electric field along the confinement direction .
an example of a real arrangement of atoms in a bi@xmath0te@xmath1 thin film with bi and tellurium surface termination is shown in fig .
[ fig5 ] . under ideal conditions ( ignoring impurity effects ) , a dipole is formed between the two surfaces which then host a single dirac cone as shown in fig .
the separation of the dirac cones on each surface depends on the asymmetry induced by the potential .
te@xmath1 thin film with two different surfaces .
the two surfaces have bi and te termination thus making them chemically inequivalent ] se@xmath1 thin film with built - in asymmetry .
asymmetry in this film was artificially introduced by using a small potential along the confinement direction .
the two dirac cones are now separated in energy ] a more interesting dispersion relationship arises in the presence of an asymmetric potential in an ultra - thin film .
this situation can be broken in to two parts and results obtained above can then be combined . for an ultra - thin film
whose surfaces have hybridized , two massive dirac hyperbolas are created .
each hyperbola is spin - degenerate as shown in fig .
if film asymmetry impresses an electrostatic potential breaking inversion symmetry , the two dirac hyperbolas will now split in to four sets of dirac hyperbolas ( fig .
[ fig7 ] ) .
the spin degeneracy of the dirac - hyperbolas is only maintained at the @xmath20 because it happens to be one of the time - reversal - invariant - momenta ( trim ) points .
@xcite se@xmath1 film with asymmetry .
the two dirac hyperbolas from the ultra - thin film in presence of asymmetry are now spin - split .
they form four copies , two from conduction and valence band and maintain degeneracy only at the @xmath20 point ] for a symmetric free - standing film , the spin - up and spin - down components of the in - plane spin - polarization are exactly anti - parallel .
interestingly , if spin - polarization is measured in an asymmetric 3d - ti film , the two anti - parallel spin components are no longer of same strength .
the amplitude of each spin - component is now a function of the asymmetry expressed as an electric potential .
the two surfaces of a 3d - ti therefore have unequal spin - amplitudes .
[ fig8 ] was obtained by computing the spin polarization for @xmath101 .
the _ k_-vector is chosen close to the @xmath5 point to satisfactorily represent the dispersion in the dirac regime . the surface band dispersion of a 3d ti is first presented followed by a self - consistent calculation of the order parameter . a 50.0 @xmath2 bi@xmath0se@xmath1 film was layered with an _ s_-wave superconductor . the superconductor is assumed to extend up to 25.0 @xmath2 .
the remaining half of the slab is pristine bi@xmath0se@xmath1 and possesses regular 3d ti properties .
s_-wave superconductor is assumed to have the material properties of aluminium @xcite whose order parameter(@xmath102 in eq .
[ bdg_full ] ) at _ t _ = 0 k is roughly equal to 0.34 @xmath103 .
band dispersion of the surface states for the bi@xmath0se@xmath1 film coupled to the superconductor is shown in fig .
[ fig9]a . since the superconductor extends only until half of the structure , the second surface still shows a dirac cone while the top surface has an open band gap .
this is shown in the zoomed out fig .
additionally , in this calculation , the chemical potential @xmath51 and the second order parameters @xmath104 has been set to zero .
bi@xmath0se@xmath1 film when coated with an _
s_-wave superconductor .
[ fig9]a shows the overall band dispersion while fig .
[ fig9]b displays the energy dispersion around the dirac cones .
the surface with no superconductor penetration has a ti surface state . ] to compute the spatial profile of the order parameter self - consistently , @xmath105 is initially set to the bulk superconducting gap value for aluminium . with this value ,
the ti - sc hamiltonian(@xmath106)is diagonalized and a new order parameter @xmath107 using eq .
[ delta1]is produced . in eq .
[ delta1 ] , the function @xmath108 is set equal to the bulk superconducting energy gap .
@xmath105 in the ti - sc hamiltonian(@xmath106 ) is set to @xmath107 and the process is iterated until convergence is achieved .
the spatial profile shows that @xmath105 is significantly suppressed at 25.0 @xmath2 mark which is the ti - sc interface and drops to zero beyond that as expected since the superconductor extends only until the halfway mark .
the interface therefore strongly distorts the superconductor wave function as demonstrated through this self - consistent calculation .
the oscillations seen in the order parameter is due to finite discretization along the quantized _ z_-direction . for calculations presented here ,
a matrix of the size 2008 @xmath72 2008 was diagonalized .
two spatial profiles are included for @xmath51 = 0 and @xmath51 = 0.2 @xmath19 . as evident ,
the chemical potential does not significantly impact the order parameter .
calculated self - consistently at two different chemical potentials is plotted as a function of the quantized _ z_-axis .
the bi@xmath0se@xmath1 film is 50.0 @xmath2 in thickness and superconductor is assumed to extend until 25.0 @xmath18 . ]
an exchange field operational in a ferromagnet and with a component along the normal ( the normal is chosen to be along the _
z_-axis)to the ti film surface opens a band gap .
a relatively thick slab of bi@xmath0se@xmath1 ( 10.0 @xmath2 in present calculation ) which ensures a decoupled top and bottom surface in presence of a ferromagnetic exchange field ( @xmath109 ) shows a band gap opening .
a gap equal to twice the exchange energy appears and the bands acquire a parabolic character .
the linear dispersion , though is preserved away from the @xmath5 point and the perpendicular spin - momentum locking , is observed again . for thin slabs which allow hybridization of the two surfaces ,
the dirac bands change to dirac hyperbolas as mentioned elsewhere in this paper . in presence of an exchange field , which breaks time reversal symmetry , and near _ k _ = 0 the bands split in two states with spin up energies @xmath110 and two spin down states @xmath111 . for an exchange energy @xmath112 , the gap induced in a thick ti slab as shown in fig .
[ fig11]a is equal to 0.0585 @xmath19 .
this split is roughly twice the chosen exchange energy .
the corresponding split ( fig .
[ fig11]b ) for a thin slab ( 3.0 @xmath2 ) is computed to be 0.0510 @xmath19 . _
z_-directed exchange field .
the exchange field gives different band gap openings for a 3.0 @xmath2 film ( left ) with hybridized surfaces in contrast to decoupled surfaces of a thicker 10.0 @xmath2 film . ] finally , as a demonstration of the assertion made and analytically derived in eq .
( [ bfield_par ] ) , the axis of magnetization for the ferromagnet is assumed to lie along the _ y_-axis .
the magnetic field equal to 10.0 tesla , in landau gauge is represented as @xmath113 where _ z _ is the set of discretized coordinates along the confined axis .
this choice of magnetic vector potential can be easily verified through the standard @xmath87 = @xmath114 relation to represent a @xmath87 field along the _ y_-axis .
further , it was also assumed that the ferromagnetic proximity effect extends only for the first three layers of the film .
two dirac cones are formed as shown in fig .
[ fig12 ] .
one of the dirac cones which corresponds to the surface unaffected by the ferromagnet is at the @xmath115 point .
ti slabs with an in - plane magnetic field ( along _ y_-axis ) of 10.0 tesla .
the dirac cone shifted along _
k@xmath3_-axis comes from the ferromagnet coated surface . ]
the second dirac cone which is from the top surface coated with a ferromagnet has shifted along the _ k@xmath3_-axis to -0.01403 nm@xmath116 .
the energies of the two dirac cones are equal to 0.02914 @xmath117 .
the valence and conduction band component are clearly identified in blue and red colour respectively in fig .
[ fig12 ] .
it is therefore imperative to have an out - of - plane magnetization for a finite band gap and destruction of topological states .
the results presented above demonstrate the finite - sized effects on 2d and 3d topological insulator states .
3d ti films of thickness as low as 6.0 @xmath2 were found to support surface states .
in contrast , a 2d ti cdte / hgte / cdte nanoribbon must be atleast 100.0 @xmath2 in thickness to prevent hybridization of edge states and open a band gap .
the role of asymmetry in a 3d - ti was found to gproduce dirac hyperbolas which for a hybridized thin film is split similar to a rashba splitting .
it was also shown that influence of an external magnetic field can be emulated by layering the surface of a 3d ti with a ferromagnet .
the important condition to note is that the axis of magnetization must have a component perpendicular to the surface .
a ferromagnet whose magnetic axis lies in the plane of the surface will shift the dirac cone and no band gap opening is observed . finally , coating a 3d - ti with an _
s_-wave superconductor leads to a band gap on account of the well - known proximity effect .
the band - gap opening depends on the strength of the order parameter which was computed self - consistently in the calculations shown .
one of us ( ps ) wishes to thank chandan setty from the dept . of physics at purdue university for several illuminating discussions on superconductors .
we also acknowledge support from intel corp . during the early stages of this work . | bi@xmath0te@xmath1 and bi@xmath0se@xmath1 are well known 3d - topological insulators .
films made of these materials exhibit metal - like surface states with a dirac dispersion and possess high mobility .
the high mobility metal - like surface states can serve as channel material for ti - based field effect transistors .
while such a transistor offers superior terminal characteristics , they suffer from an inherent zero band gap problem .
the absence of a band gap for the surface states prevents an easy turn - off mechanism . in this work , techniques that can be employed to easily open a band gap for the ti surface states
is introduced .
two approaches are described : 1 ) coating the surface states with a ferromagnet which has a controllable magnetization axis .
the magnetization strength of the ferromagnet is incorporated as an exchange interaction term in the hamiltonian .
2 ) an _ s_-wave superconductor , because of the proximity effect , when coupled to a 3d - ti opens a band gap on the surface .
this ti - superconductor heterostructure is modeled using the bogoliubov - de gennes hamiltonian .
a comparison demonstrating the finite size effects on surface states of a 3d - ti and edge states of a cdte / hgte / cdte - based 2d - ti is also presented .
3d - ti nanostructures can be reduced to dimensions as low as 10.0 @xmath2 in contrast to 2d - ti structures which require a thickness of at least 100.0 @xmath2 .
all calculations are performed using the continuum four - band k.p hamiltonian . |
envelopes of neutron stars ( nss ) are divided in the inner and outer envelopes .
properties of the inner envelopes are outlined in a companion paper @xcite . here
we focus on the outer ones , at typical densities @xmath6 and temperatures @xmath7 k. these envelopes are relatively thin , but they affect significantly ns evolution . in particular , they provide thermal insulation of stellar interiors .
bulk properties of the ns envelopes ( pressure @xmath8 , internal energy @xmath9 ) are determined mainly by degenerate electron gas .
the electrons are relativistic at @xmath10 , where @xmath11 , @xmath12 is the fermi wave number , @xmath13 is the electron mass , @xmath14 is the electron number density , @xmath15 , @xmath16 and @xmath17 are the ion charge and mass numbers , respectively .
degeneracy is strong at @xmath18 , where @xmath19 and @xmath20 is the lorentz factor .
the ionic component of the plasma can form either strongly coupled coulomb liquid or solid , classical or quantum .
the classical ion coupling parameter is @xmath21 , where @xmath22 is the ion - sphere radius ( @xmath23 ) .
the ions form a crystal if @xmath24 exceeds some critical value @xmath25 ( see below ) .
the quantization of ionic motion becomes important at @xmath26 , where @xmath27 is the ion plasma temperature , and @xmath28 is the ion mass .
the ionic contribution determines specific heat @xmath29 , unless @xmath26 .
the ions and electrons are coupled together through the electron response ( screening ) .
the inverse screening ( thomas fermi ) length is @xmath30 where @xmath31 is the electron chemical potential . if @xmath32 , then @xmath33 and @xmath34 , where @xmath35 is the fine structure constant .
note that @xmath36 at @xmath37 ; therefore the screening effects do not vanish even at high densities .
thermodynamic properties are altered by the magnetic field @xmath38 in the case where the landau quantization of transverse electron motion is important .
the field is called _ strongly quantizing _ when it sets all electrons on the ground landau level .
this occurs at @xmath39 and @xmath40 ( see , e.g. , ref .
@xcite ) , where @xmath41 and @xmath42 g. _ weakly quantizing _ fields ( @xmath43 ) do not significantly alter @xmath8 and @xmath9 but cause oscillations of @xmath29 , other second - order quantities , and electron transport coefficients with increasing density .
characteristic domains in the @xmath44@xmath45 diagram are shown in figure [ fig - domains ] .
the short - dashed lines on the left panel indicate the region of partial ionization .
the dot - dashed lines correspond to @xmath46 ( upper lines ) and @xmath47 ( liquid / solid phase transition ) .
long - dashed lines on the right panel separate three @xmath44@xmath45 regions where the magnetic field is strongly quantizing ( to the left of @xmath48 and considerably below @xmath49 ) , weakly quantizing ( to the right of @xmath48 at @xmath50 ) , or classical ( much above @xmath49 ) .
10.2 cm the dotted lines in fig .
[ fig - domains ] show profiles @xmath51 in the envelope of a `` canonical '' ns with mass @xmath52 and radius 10 km , and with an effective surface temperature @xmath53 k or @xmath54 k ( the values of @xmath55 are marked near these curves ) .
typical temperatures of isolated nss are believed to lie in the hatched region between these two lines . in a magnetized ns ,
@xmath51 depends on strength as well as direction of the magnetic field .
therefore on the right panel we show _ two _ dotted curves for each value of @xmath56 : the lower curve of each pair corresponds to the heat propagation along the field lines ( @xmath57 ,
i.e. , near the magnetic poles ) and the upper one to the transverse propagation ( @xmath58 , near the magnetic equator ) .
thermodynamic functions of the electron gas at arbitrary degeneracy are expressed through the well known fermi - dirac integrals . for astrophysical use ,
it is convenient to employ analytic fitting formulae for these functions presented , e.g. , in ref .
@xcite .
nonideal ( exchange and correlation ) corrections for nonrelativistic electrons at finite temperature have been calculated and parameterized in ref .
@xcite . for the relativistic electrons at low @xmath45 ,
an analytic expansion of the exchange corrections is given , e.g. , in ref .
@xcite ( in this case , the correlation corrections are negligible )
. a smooth interpolation between these two cases has been constructed in ref .
@xcite . for the ionic component at @xmath59 , the main contribution to the thermodynamic functions comes from the ion correlations .
strongly coupled one - component coulomb plasmas ( ocp ) of ions in the _ uniform _ electron background have been studied by many authors .
the thermodynamic functions of the classical ocp liquid have been calculated @xcite at @xmath60 and parameterized @xcite for @xmath61 .
the latter parameterization ensures accuracy @xmath62 ( per particle ) . for the classical coulomb crystal , accurate numerical results and fitting formulae to the free energy @xmath63 ( with anharmonic corrections taken into account )
have been presented in refs .
a comparison of the latter results with the fit @xcite for the liquid yields the liquid solid phase transition at @xmath64 ( curiously , this value could be derived in ref .
@xcite , but it was first noticed in ref .
@xcite ) . in the solid phase
, quantum effects may be important at the considered temperatures .
these effects can be most easily taken into account in the approximation of a harmonic coulomb crystal .
a convenient analytic approximation is provided by a model @xcite which treats two phonon modes as debye modes and the third one as an einstein mode .
numerical calculations @xcite of the harmonic - lattice contributions to @xmath63 , @xmath9 , @xmath29 , and the mean - square ion displacement are reproduced by this model at arbitrary @xmath65 within several percent , which is sufficient for many applications .
8 cm finally , there is a contribution from ion - electron ( @xmath66 ) correlations , which can be treated as polarization of the jellium " of the degenerate electrons . in a coulomb liquid , this contribution has been evaluated in refs .
@xcite using a perturbation theory which involves the electron dielectric function and the static structure factor of ions .
this approximation is justified at @xmath67 ( which is the case at @xmath68 ) . at smaller densities
the @xmath66 contribution in the coulomb liquid has been calculated in ref .
@xcite using the hnc technique ( heavy dots in fig .
[ fig - ie ] ) . in a coulomb
solid , the @xmath66 contribution has been evaluated in ref .
@xcite using the perturbation theory @xcite and a model lattice structure factor . these results , though approximate ,
indicate that the polarization corrections may be rather important .
in particular , @xmath25 may be shifted by @xmath69% when the @xmath66 corrections in the liquid and solid phases are taken into account .
the numerical results have been fitted @xcite by analytic functions of @xmath70 and @xmath24 .
the approximation @xcite for the ocp liquid is shown in fig .
[ fig - ie ] by solid lines .
since the @xmath66 correction is not dominant , the accuracy of the fitting formulae ( @xmath71% ) is sufficient for most of astrophysical applications .
an alternative pad approximation , proposed in ref .
@xcite for the electron - ion fluid , is also accurate at @xmath72 for small and large @xmath24 , but it is less accurate at intermediate @xmath24 and inapplicable at @xmath73 ( dashed lines ) . as mentioned above , quantizing magnetic fields significantly affect the equation of state . in the fully ionized dense plasma ,
the main effects are described by the model of ideal electron gas , studied by many authors ( e.g. , @xcite ) .
heat conduction in ns envelopes is provided mainly by the electrons .
the thermal conductivity tensor @xmath74 is determined mainly by the electron - ion scattering .
calculation of @xmath74 should take into account specific features of the coulomb plasmas in ns envelopes , quite different from the terrestrial liquid and solid metals : ( i ) in the liquid phase , an incipient long - range order emerges at @xmath75 ; ( ii ) in the solid phase , there are usually many brillouin zones within the fermi surface ; ( iii ) in the solid phase , the approximation of one - phonon scattering fails near the melting , which necessitates the use of a more general expression for the structure factor of ions .
these features have been taken into account in ref .
@xcite , where analytic approximations to the electron transport coefficients have also been derived .
magnetic field hampers electron transport across the field lines : transverse thermal and electrical conductivities are reduced by orders of magnitude in typical ns envelopes .
strongly quantizing field significantly changes also longitudinal conductivities ; weakly quantizing field causes de haas van alphen oscillations .
these effects have been outlined , e.g. , in @xcite .
a unified treatment of the electron transport coefficients in the domains of classical , weakly quantizing , and strongly quantizing magnetic field has been developed in @xcite .
a fortran code which implements this treatment is available at ` http://www.ioffe.rssi.ru/astro/conduct/ ` . using this code and solving the thermal diffusion equation ,
we have calculated the @xmath51 profiles shown in fig .
[ fig - domains ] .
ayp is grateful to the theoretical astrophysics group at ecole normale suprieure de lyon for generous hospitality and financial support .
the work of ayp and dgy has been supported in part by intas ( grant 96 - 542 ) and rfbr ( grant 99 - 02 - 18099 ) .
ayp and dgy thank the organizing committee for support provided to attend pnp10 .
99 , this volume , in the equation of state in astrophysics , ed .
g. chabrier , e. schatzman , cambridge univ . ,
cambridge , uk ( 1994 ) 214 , phys .
e * 58 * ( 1998 ) 4941 , phys . rep . * 149 * ( 1987 ) 91 , astron .
, * 314 * ( 1996 ) 1024 , phys . rev .
e * * ( 2000 ) accepted , physica b * 228 * ( 1996 ) 21 , j. chem .
* 111 * ( 1999 ) 6538 , phys .
a * 42 * ( 1990 ) 4972 , phys . rev .
e * 47 * , 4330 ( 1993 ) .
, astrophys .
j. * 414 * ( 1993 ) 695 , ph.d .
thesis , ioffe phys .- tech .
institute , st .
petersburg ( 2000 ) , phys .
rev . a * 14 * ( 1976 ) 816 , astrophys . space phys
* 7 * ( 1989 ) 311 , astron .
( 2000 ) accepted ; astro - ph/0008399 , j. phys .
c * 15 * ( 1982 ) 6233 , astron .
* 346 * ( 1999 ) 345 , astron
. astrophys .
* 351 * ( 1999 ) 787 | outer envelopes of neutron stars consist mostly of fully ionized , strongly coupled coulomb plasmas characterized by typical densities @xmath0@xmath1 and temperatures @xmath2@xmath3 k. many neutron stars possess magnetic fields @xmath4@xmath5 g. recent theoretical advances allow one to calculate thermodynamic functions and electron transport coefficients for such plasmas with an accuracy required for theoretical interpretation of observations . |
a conventional boundary layer theory of fluid flow used for free convective description assumes zero velocity at leading edge of a heated plate .
more advanced theories of self - similarity also accept this same boundary condition @xcite , @xcite , @xcite .
however experimental visualization definitely shows that in the vicinity of edge the fluid motion exists sb , @xcite , @xcite .
it is obvious from the point of view of the mass conservation law . in the mentioned convection descriptions the continuity equation
is not taken into account that diminishes the number of necessary variables .
for example the pressure is excluded by cross differentiation of navier - stokes equation component .
the consequence of zero value of boundary layer thickness at the leading edge of the plate yields in infinite value of heat transfer coefficient which is in contradiction with the physical fact that the plate do not transfer a heat at the starting point of the phenomenon .
the whole picture of the phenomenon is well known : the profiles of velocity and temperature in normal direction to a vertical plate is reproduced by theoretical concepts of prandtl and self-similarity.while the evolution of profiles along tangent coordinate do not look as given by visualisation of isotherms ( see e.g. gdp ) .
it is obvious that isotherms dependance on vertical coordinate @xmath1 significantly differs from power low depandance @xmath3 of boundary layer theories . in this article
we develop the model of convective heat transfer taking into account nonzero fluid motion at the vicinity of the starting edge .
our model is based on explicit form of solution of the basic fundamental equations ( navier - stokes and fourier - kirchhoff ) as a power series in dependant variables .
the mass conservation law in integral form is used to formulate a boundary condition that links initial and final edges of the fluid flow .
we consider a two - dimensional free convective fluid flow in @xmath4 plane generated by vertical isothermal plate of height @xmath2 placed in udisturbed surrounding .
the algorithm of solution construction is following .
first we expand the basic fields , velocity and temperature in power serious of horizontal variable @xmath0 , it substitution into the basic system gives a system of ordinary differential equations in @xmath1 variable .
such system is generally infinite therefore we should cut the expansion at some power .
the form of such cutting defines a model .
the minimal number of term in the modeling is determined by the physical conditions of velocity and temperature profiles . from the scale analysis of the equations we neglect the horizontal ( normal to the surface of the plate ) component velocity .
the minimum number of therms is chosen as three : the parabolic part guarantee a maximum of velocity existence while the third therm account gives us change of sign of the velocity derivative .
the temperature behavior in the same order of approximation is defined by the basic system of equations . the first term in such expansion is linear in @xmath0 , that account boundary condition on the plate ( isothermic one ) .
the coefficient , noted as @xmath5 satisfy an ordinary differential equation of the fourth order .
it means that we need four boundary condition in @xmath1 variable .
the differential links of other coefficients with @xmath6 add two constants of integrations hence a necessity of two extra conditions .
these conditions are derived from conservation laws in integral form .
the solution of the basic system , however , need one more constant choice .
this constant characterize linear term of velocity expansion and evaluated by means of extra boundary condition . in the second section we present basic system in dimensional and dimensionless forms . by means of cross - differentiation
we eliminate the pressure therm and next neglect the horizontal velocity that results in two partial differential equations for temperature and vertical component of velocity . in the third section
we expand both velocity and temperature fields into taylor series in @xmath0 and derive ordinary differential equations for the coefficients by direct substitution into basic system .
the minimal ( cubic ) version is obtained disconnecting the infinite system of equations by the special constraint .
the fourth and fives sections are devoted to boundary condition formulations and its explicit form in therms of the coefficient functions of basic fields .
it is important to stress that the set of boundary conditions and conservation laws determine all necessary parameters including the grasshof anf rayleigh numbers in the stationary regime under consideration .
the last section contains the solution @xmath5 in explicit form and results of its numerical analysis .
the solution parameters values as the function of the plate height @xmath2 and parameters whivh enter the grasshof number @xmath7 estimation are given in the table form , which allows to fix a narrow domain of the scale parameter @xmath8 being the characteristic linear dimension of the flow at the starting level .
let us consider a two dimensional stationary flow of incompressible fluid in the gravity field .
the flow is generated by a convective heat transfer from solid plate to the fluid .
the plate is isothermal and vertical . in the cartesian
coordinates @xmath0 ( horizontal and orthogonal to the palte)@xmath9 ( vertical and tangent to the palte ) the navier - stokes ( ns ) system of equations have the form @xcite.:@xmath10@xmath11 in the above equations the pressure terms are divided in two parts @xmath12 .
the first of them is the hydrostatic one that is equal to mass force @xmath13 , where : @xmath14 is the density of a liquid at the nondisturbed area where the temperature is @xmath15 .
the second one is the extra pressure denoted by @xmath16the part of gravity force @xmath17 arises from dependence of the extra density on temperature , @xmath18 is a coefficient of thermal expansion of the fluid . in the case of gases
@xmath19the last terms of the above equations represents the friction forces with the kinematic coefficient of viscosity @xmath20 the mass continuity equation in the conditions of natural convection of incompressible fluid in the steady state @xcite has the form : @xmath21 the temperature dynamics is described by the stationary fourier - kirchhoff ( fk ) equation : @xmath22 where @xmath23 and @xmath24 are the components of the fluid velocity @xmath25 , @xmath26 - temperature and @xmath27 - pressure disturbances correspondingly and @xmath28 is the thermal diffusivity . from the point of clarity of further transformations
we use the same scale @xmath29 along both variables @xmath0 and @xmath1 .
we will return to the eventual difference between characteristic scales in different directions while the solution analysis to be provided . after introducing variables : @xmath30@xmath31 we obtain in boussinesq approximation ( in all terms besides of buoyancy one we put @xmath32 ) .@xmath33@xmath34 and fk equation is written as @xmath35 where @xmath36 @xmath8 is a characteristic linear dimension and @xmath37 is characteristic velocity : @xmath38 then @xmath39 , @xmath40 and @xmath41is the grashof number , which after plugging ( @xmath42 takes the form:@xmath43 after cross differentiation of equations ( @xmath44 and ( @xmath45 ) we have : @xmath46 = \ ] ] @xmath47 \label{ns1 + 2}\ ] ] the fk equation rescales as @xmath48 and @xmath49 where @xmath50 next we would formulate the problem of free convection around the heated vertical isothermal plate @xmath51 @xmath52 , dropping the primes . in this case
we assume the angle between the plate and a stream line is small that means a possibility to neglect the horizontal component of velocity of fluid , denoting the vertical component as @xmath53 . in this paper
we restrict ourselves by the assumption that @xmath54 and @xmath55 , that yields @xmath56 = 0 , \label{ns - a}\ ] ] @xmath57
the aim of this paper is the theory application to the standard example of a finite vertical plate . having only two basic functions we consider the power series expansions of the velocity and temperature in cartesian coordinates : @xmath58 @xmath59 according to standard boundary conditions on the plate we assume that the both functions tend to zero when @xmath60 ,
so we choose for a calculation the variable that has the zero value for nondimentional temperature ( @xmath61 ) .
it means that the value of @xmath62 outside of the convective flow tends to @xmath63 substituting expressions ( @xmath64 and ( @xmath65 into the equations ( @xmath66 we take into account the linear independance of monomials @xmath67 that gives a system of coupled nonlinear equations for the coefficients @xmath68 @xmath69 , .... and @xmath70 , @xmath71 , @xmath72 such system is infinite hence for a practical use we need to choose appropriate scheme of closed formulation for finite number of variables .
the formulation should be based on physical assumptions for a concrete conditions .
we would like to restrict ourselves by the fourth order approximation for both variables that means we neglect higher order terms starting from fifth one .
the area of the approximations validity is defined by the comparison of terms in expantions ( @xmath64 and ( @xmath73 as it will be clear from further analysis we should consider the functions : @xmath74 @xmath75 @xmath70 and @xmath76 as variables of the first order , while @xmath77 and @xmath78 to be the second one . from the relations that appear after substitution of ( @xmath64 and ( @xmath79 into ( @xmath80 and ( @xmath81 it follows that @xmath82 and @xmath83 finally from both equations ( @xmath84 ( @xmath81 we obtain the system of equations for the coefficients @xmath85 , @xmath86 , @xmath69 and @xmath87@xmath88 @xmath89 @xmath90 @xmath91 the first two ( @xmath92 ( @xmath93 arise from fk equation and the rest of them are from the ns one .
the system of equations is closed if @xmath94 .
it means that the number of equations and the number of unknown functions is the same . in the first approximation
the velocity and temperature are expressed as : @xmath95 from ( @xmath96 ) one has @xmath97 from ( @xmath98 ) it follows that @xmath99 hence ( @xmath100 ) goes to:@xmath101@xmath102 the equation ( @xmath103 ) reads : @xmath104 this results in @xmath105 the form of the equation ( @xmath106 ) indicates that for unique solution one needs four boundary conditions for given parameters @xmath107 and @xmath108 .
apart from such conditions we should also have values for @xmath109 and @xmath110 .
so for expicit determination of @xmath53 and @xmath111 we need eight conditions .
looking for the boundary conditions let us apply conservation laws of mass , momentum and energy , applying the laws to a control volume @xmath112 ( see fog.1 ) .
the first one is the conservation of mass in two dimensions that in steady state looks as : @xmath113 where : @xmath114 is the sum of all lateral surfaces @xmath115 ( fig.1 ) .
the mass conservation law in the integral form ( @xmath116 ) is formulated by a division of the surface @xmath117 to two the lower @xmath118 and upper @xmath119 boundaries only . according to our main assumption about two - dimensionality of the stream we neglect
a dependence of variables on @xmath120 coordinate and @xmath121 $ ] .
hence the condition of total mass conservation looks as follows : @xmath122 where the flow from below @xmath123 is approximately the product of density at temperature @xmath124 and velocity of the incoming flow in the interval @xmath125.$ ] we follow the idea of the velocity field continuity at @xmath126 , hence @xmath127
. for the left side in approximations mentioned above one has ( @xmath128 @xmath129 @xmath130 ) : @xmath131 @xmath132 and outcoming flow @xmath119 is expressed similarily : @xmath133 @xmath134 the mass conservation law yields @xmath135 the next condition is connected with the conservation of energy in a control volume @xmath112 ( area @xmath114 with unit width see fig.1 ) arises from fk equation ( @xmath136 ) by integration over the volume.@xmath137 the left side of the energy conservation equation ( @xmath138 ) is transformed similar applying the identity @xmath139(@xmath140 @xmath141 and ( @xmath142 ) . according to our assumptions we left only flows accross @xmath143 and basing on homogenity of the problem with respect to the coordinate @xmath120
we have : @xmath144 to link the incoming fluid temperature @xmath124 @xmath145 with the solution at @xmath126 and the outgoing fluid ( see @xmath146 ) we put @xmath147 that results in : @xmath148 @xmath149 where @xmath150 .
the equation ( @xmath151 is the ordinary differential equation of the fourth order , therefore its solution needs four constants of integration .
these constants depend on two parameters @xmath107 and @xmath108 , which enter the coefficients of the eq.(@xmath151 .
the function @xmath5 defines the rest functions @xmath152 and @xmath85 via above relations .
it means that we have six constants determining the solution of problem and we need also six corresponding boundary conditions .
the temperature values in the vicinity of the boundary edge point and taken as value -1 ( temperature of incoming from the bottom flow ) . in dimensional
form the interval of consideration has the characteristic length @xmath8 which we identify with a parameter we used when dimensional variables where itroduced ( @xmath61 ) .
let us remind that scale @xmath8 is connected with special ( local , horizontal ) grashof number @xmath153 ( @xmath154).the total height of the plate is denoted @xmath2 for a stationary process an edge conditions may be considered as initial one for a cauchy problem . having a power series approximation of such conditions we choose the coefficients of the series using weierstrass theorem .
it means that we equalize the coefficients to scalar product of intial conditions and orthonormal polynomials on the interval @xmath155 .$ ] in our case the temperature profile @xmath156 represents this condition , while the function is constant @xmath157 on the interval @xmath158 $ ] in nondimensional variables . in the approximation of the thirdpower
orthogonal polynomials we have : @xmath159 because nondimentional temperature of the fluid at the lower half plane , according to above , is @xmath160where the polynomialas are defined as : @xmath161 @xmath162 the normalization for @xmath163 , @xmath164and orthogonality condition @xmath165 give the link between constants : @xmath166 . , which plugging into @xmath167 results in @xmath168 = @xmath169finally @xmath170 substituting the result into @xmath171 gives two equations @xmath172 , @xmath173 , which solving and projecting @xmath174 , @xmath175 @xmath176.yield boundary conditions for the coefficients for temperature expansion : @xmath177 plotting the temperature approximation at the level @xmath178 @xmath179 approximation ] is given by the fig.2 .
let us recall that @xmath180 ( see eq .
( @xmath181))@xmath182 therefore @xmath183 ( @xmath184 ) we will consider as boundary condition for @xmath185 the temperature gradient values @xmath186 on the plate decrease when @xmath1 grows.at the leading edge we pose the condition @xmath187 because the plate lose the contact with the fluid .
it gives third boundary condition ( @xmath146 ) @xmath188
the phenomenon of free convective heat transfer from isothermal vertical plate ( @xmath189 ) imply that temperature gradient on the plate is negative ( @xmath190 ) and decrease along @xmath1 ( @xmath191 ) .
it is also known that velocity profile has maximum at the distance @xmath192 @xmath193 .
the extrema for the curve is defined by derivative of @xmath53 as a function of @xmath194 hence the relation @xmath195 indicates that for @xmath196 , @xmath197 and @xmath198 we have two extremal points @xmath199{\frac{\alpha^{2}}{9\beta^{2}}-\frac{% \gamma}{3\beta}}\text { \ \ \ \ and \ \ \ \ } x_{0}(y)=-\frac{\alpha } { 3\beta}+% \sqrt[2]{\frac{\alpha^{2}}{9\beta^{2}}-\frac{\gamma}{3\beta } } \label{xm0}\ ] ] if @xmath200 notations are chosen to mark maximum position point as @xmath192 while the minimum one is @xmath201 .
in the exeptional case of @xmath202the expression simplifies @xmath203 which is positive for @xmath204the second extremum do not exist now ( see fig.3 ) .
[ fig : fig-3 ] there is a possibility to choose the value @xmath205 considering the @xmath206 as a conditional boundary of the upward stream.we define hence @xmath207 . at the starting horizontal edge of the vertical plate
the vertical velocity component of incoming flow ( @xmath146 ) varies slow so we assume that @xmath208 hence @xmath209 the extrema of the velocity profile ( @xmath210 ) after account of ( @xmath211 ) and ( @xmath130 ) is transformed as , for maximum : @xmath212{\frac{c_{2}{}^{2}}{\left ( \frac{g_{r}}{2}% c\left ( y\right ) \right ) ^{2}}+\frac{2\gamma}{g_{r}c\left ( y\right ) } } $ ] and minimum one : @xmath213{\frac{% c_{2}{}^{2}}{\left ( \frac{g_{r}}{2}c\left ( y\right ) \right ) ^{2}}+\frac{% 2\gamma}{g_{r}c\left ( y\right ) } } $ ] .
the following identity @xmath214 holds for : @xmath215 suppose there exists a level @xmath216 at which @xmath217 where @xmath218 denotes the boundary layer thickness analog . the equation ( @xmath219 ) is solved with respect to @xmath220 that gives : @xmath221 as function of the problem parameters . then plugging ( @xmath222 ) for the expression for the @xmath223 yields @xmath224
let us return to the expression for the temperature ( @xmath146 ) with neglecting the last term in temperature ( the possibility of such assumption will be explained below ) on the level @xmath225 and substitute ( @xmath222 ) and ( @xmath226 ) into it equalizing to the temperature of surraunding @xmath227 .
@xmath228 we have : @xmath229 where : @xmath230 from the equation ( @xmath106 ) after plugging @xmath108 ( @xmath231 ) and taking into account @xmath232 ( @xmath233 ) we have @xmath234 the equation was studied recently @xcite where the solution was given by@xmath235 @xmath236+\exp[-\frac{sy}{2}](b_{1}\cos[\frac{\sqrt{3}}{2}sy]% + b_{2}\sin[\frac{\sqrt{3}}{2}sy ] ) , \label{c(y)}\ ] ] where @xmath237{2\pr g_{r } } \label{s}\ ] ] is expressed via @xmath238 = @xmath239we have also boundary conditions : ( @xmath240 ) @xmath241 solution of the system results in a rather big expression for @xmath242as function of @xmath243which we skip in theis text , going to following approximtion .
the explicit form of the equation ( @xmath244 shows that the three last terms have exponential behaviour as function of @xmath245it means that there are three different domains of the fluid flow structure .
the first is the starting one where all terms are significant .
the leading edge is characterized by two first terms and the medium domain is described by the only first one .
we choose the parameter @xmath216 such that it belongs to that medium range .
in such conditions @xmath246 @xmath247 where @xmath248 plugging @xmath243 in the form of ( @xmath249 into the table of boundary conditions gives @xmath250 a_1=--b_1+,@xmath251b_1=-b_2 + -,@xmath252b_2=.@xmath253 let us consider the natural approxmation @xmath254 .
after substitution of the expression for @xmath255 into @xmath256and next @xmath256 into @xmath257 we have approximate formulas : @xmath250a_0=-,@xmath251 a_1=,@xmath251b_1=-,@xmath251b_2@xmath258-++.@xmath258@xmath259 it defines the expression for @xmath70 as the function of parameters @xmath260 ( @xmath261 , the plate height @xmath2 and the new one ( @xmath262 @xmath263 the velocity profile at the level @xmath225 is defined by ( @xmath264 and the parameters values ( @xmath265 : @xmath266
the mass conservation equation ( @xmath267 ) after substitution of @xmath268 @xmath269 @xmath270 @xmath271 , @xmath272 and denoting @xmath273 has the form : @xmath274 the only real solution of the equation ( @xmath275 ) value that have physical sense is @xmath276 now we can return to the energy conservation equation ( @xmath277 plugging the boundary conditions for the domain restricted by the plate on interval ( @xmath278 ) .
it simpifies the expression for the integral along the plate surface ( heat transfer from the plate on this interval ) .
consequently we change @xmath279 to @xmath223 and neglect the integrant oscilations at vicinity of @xmath126 .
@xmath280 we estimate the heat flux integral from the plate as @xmath281 and take into account the expressions for parameters @xmath282that yields : @xmath283 . as
further considerations show , the value of @xmath225 mayby chosen as close to the plate height @xmath2 .
chosing @xmath285 and plugging the values of the parameters @xmath286 @xmath287and ( @xmath288 ) @xmath289{2ra}=\allowbreak 4.\,\allowbreak5782 $ ] into the table of the function @xmath5 coefficients gives [ cols= " < " , ] substitution of the table values into ( @xmath290 ) we have the expression which allows to plot the function @xmath5 .
in the same approximation the typical velocity profile @xmath291 ( @xmath292 ) at the the stability interval ( y @xmath293 : $ ] ) @xmath294 .
substitution of the grashof number @xmath295= @xmath296 gives @xmath297 that is represented by the plot . .
] in the same condition the temperature profile @xmath298 is defined by the expression ( @xmath146 ) and results in the plot ] to understand the phenomenon it is useful to return to dimensional picture . as a main space scale
it is choosen the parameter @xmath8 ( @xmath61 ) which is connected with the grashof number by ( @xmath154 ) @xmath299{\frac{1}{% bg\phi } \nu^{2}g_{r}}$ ] . where @xmath300 .
for the air example and the temperature @xmath301 @xmath302 , @xmath303 , @xmath304 the viscosity coeficient @xmath305 = @xmath306 the coeficient of thermal expansion @xmath307 and for conditions of our model ( @xmath285 ) @xmath308 we estimate @xmath8 as : @xmath309{\frac{1}{\frac{1}{303}10\cdot20}\left ( 16\cdot10^{-6}\right ) ^{2}\cdot68.\,\allowbreak54}=\allowbreak2.\,\allowbreak985\times 10^{-3}m\thickapprox3mm$ ]
first of all we would stress again that the model we present here have the engieering character of approximations , but include direct possibilities for a development by simple taking next terms of expansions into account .
a modification of boundary conditions which would improve the transient regimes at both ends of the y - dependence is also possible .
newertheless in this simple modeling we observe some important characteristic features of real convection phenomenon as almost parallel streamlines and isotherms in the stability region ( as , for example in visualizations of interferometric study from @xcite ) .
it follows from functional parameter @xmath5 behaviour inside the domain and small contribution of cubic therm in the expresion for temperature ( @xmath146 ) .
our explicit solution form and parameter values estimation allows to conclude that : \1 .
the streamlines and isotherms of the flow are almost paralel to the vertical heating plate surface in the domain of stability , \2 .
velocity values of the fluid flow at starting edge of the plate are nonzero , \3 .
the set of boundary conditions yields in the complete set of the solution parameter including the local grashof number and hence , the characteristic linear dimension length @xmath8 in normal to plate direction @xmath0 , \4 .
the sesults allow to descibed the natural heat transfer phenomenon for given fluid in therms only the temperature difference @xmath310 and the plate heigth @xmath311 which are novel in comparison with former theories .
99 y.jaluria.:natural convection heat and mass transfer ; pergamon press , oxford , 1980 .
e .. schmidt and w.beckmann , das temperatur- und geschwindigkeitfeld vor einer wrme abgebenden senkrechten plate bei natrlicher konvektion , tech mech .
u. thermodynamik , bd.1 , nr .
10 , okt.1930 , pp.341 - 349 and bd . 1 , nr.11 , nov .
1930 , pp.391 - 406 .
h.c.li and g. p. peterson , experimental studies of natural convection heat transfer of al2o3/diwater nanoparticle suspensions ( nanofluids ) , hindawi publishing corporation , advances in mechanical engineering , volume 2010 , article i d 742739 | the model under consideration is based on approximate analytical solution of two dimensional stationary navier - stokes and fourier - kirchhoff equations .
approximations are based on the typical for natural convection assumptions : the fluid noncompressibility and bousinesq approximation .
we also assume that ortogonal to the plate component ( @xmath0 ) of velocity is neglectible small .
the solution of the boundary problem is represented as a taylor series in @xmath0 coordinate for velocity and temperature which introduces functions of vertical coordinate ( @xmath1 ) , as coefficients of the expansion .
the correspondent boundary problem formulation depends on parameters specific for the problem : grashoff number , the plate height ( @xmath2 ) and gravity constant .
the main result of the paper is the set of equations for the coefficient functions for example choice of expansion terms number .
the nonzero velocity at the starting point of a flow appears in such approach as a development of convecntional boundary layer theory formulation . |
the fast rise , slow decay of subpulses in grb is a common feature . there could be many ways to explain it ( e.g. impulsive energy infusion followed by slower cooling or light echoing ) .
it is therefore desirable to discriminate among the different models with quantitative tests and predictions whenever possible . in a previous paper ( eichler and manis 2007 , hereafter em07 )
, it was suggested that fast rise , slow decay subpulses constitute a qualitative manifestation of baryons being accelerated by radiation pressure .
more generally , the basic idea can apply to any flow in which a light , fast fluid imparts energy to a clumpy , denser component of the flow by overtaking the clumps from the rear , but for convenience in this discussion we refer to the fast light component as photons that scatter off the clumps .
it was proposed that the fast rise of a subpulse is the stage where a cloud of baryons scatters photons into a progressively narrowing beaming cone of width @xmath3 , where @xmath4 is the bulk lorentz factor of the accelerating cloud .
this narrowing of the @xmath3 cone causes brightening as long as @xmath4 remains below @xmath5 , where @xmath6 is the viewing angle offset between the observer s line of sight and the velocity vector of the scattering cloud .
once the scattering cloud accelerates to a lorentz factor exceeding @xmath5 , the viewer is no longer inside the beaming cone and apparent luminosity begins to decline .
if the cloud accelerates with roughly constant radiative force , as is reasonable to suppose over timescales that are short compared to the hydrodynamic expansion time , then the decline in luminosity is considerably slower than the rise time , because the acceleration time increases so dramatically as the velocity approaches c. it was shown in em07 that the spectral peak frequency as seen by the observer remains roughly constant during the rising phase and , well into the declining phase , softens as @xmath7 , as reported by ryde ( 2004 ) .
the spectral softening of the pulse has been carefully studied by norris and coworkers , who have noted that the asymmetry of the subpulse increases with decreasing frequency and that the width of the subpulse scales roughly as the frequency to the power -0.4 ( fenimore et al 1995 ) in the batse energy range .
this represents additional information , as the result of ryde is in principle consistent with symmetric pulses . in this letter
, we derive the light curves as a function of both time and frequency .
we show that the asymmetry of the subpulse decreases with frequency and that the hypothesis of em07 is quantitatively consistent with the formulation of fenimore et al ( 1995 ) . the basic assumption in our hypothesis - that a scattering screen can _ enhance _ the detected signal - presupposes that the unscattered radiation is beamed and directed slightly away from the observer s line of sight , so that the scattering of photons into the line of sight creates a `` flash - in - the - pan '' type brightening .
this assumption is non - trivial , but has been suggested as being an explanation for the amati relation ( 2002 ) in earlier papers ( eichler and levinson 2004 , 2006 ; levinson and eichler 2005 ) . in this series of papers , it was suggested that a significant fraction of all grb are actually brighter and harder in spectrum than they appear to be , and that they appear dimmer and softer because we , the observers , are viewing the burst from a slightly offset angle relative to the direction of the fireball .
the interpretation of the subpulses given here and in em07 is thus in accord with this picture .
the equations describing matter that is being accelerated by highly collimated radiation pressure were presented in em07 . here
we apply the solutions described in em07 to calculate the profile of a subpulse as a function of photon energy .
we assume that the differential primary photon spectrum @xmath8 has the form @xmath9exp(@xmath10 ) , where @xmath11 is the photon energy in the frame of the central engine .
this form is consistent with a comptonized thermal component .
it does not , however , exclude a power law photon spectrum produced further downstream by internal shocks . after scattering off a baryon clump that moves with velocity @xmath12 , the photon energy as seen by an observer at angle
@xmath6 is @xmath13=e_o(1-\beta)/(1-\beta cos\theta).\ ] ] together with the solution for the accelerating trajectory @xmath14 given in em07 , the source / observer frame invariance of the number of photons @xmath15 scattered within energy interval de and time interval dt , and solid angle @xmath16 , equation ( 1 ) determines the light curve n(e , t ) as a function of observed photon energy e and observer time t. in figure 1 the subpulse light curves are plotted for three different frequencies .
it is clear that the pulse width is larger and the rise - fall asymmetry is more pronounced at lower frequencies , as reported by fenimore et al .
( 1995 ) and references therein . in figure 2 the width is plotted as a function of photon energy . at high energies , which correspond to the batse measurements used by these authors ,
the width is seen to scale approximately as the photon energy to the power @xmath17 , as reported by fenimore et al .
, above @xmath18 kev . similar calculations with varying values for the low energy power law index , @xmath19 , of the primary spectrum show that this dependence is weakly dependent on @xmath19 and on viewing angle .
for a viewing offset angle of 10 degrees , the width depends on @xmath20 , with @xmath21 when @xmath22
with the sensitivity @xmath23 at @xmath24 . for viewing offset of 15 degrees , the value of @xmath25 is increased by about 0.033 so that a given range of @xmath25 is occupied by a somewhat lower ( i.e. more negative ) range of @xmath19 than for smaller viewing offsets . for an extended beam ,
some contribution from larger offsets is inevitable , but a synthesis of light curves from extended beams is deferred for future work .
it can be seen from figure 2 that the value of @xmath25 increases with @xmath26 , and the range of @xmath26 that corresponds to batse sensitivity depends on cosmological redshift , larger z implies larger intrinsic values of @xmath26 , hence steeper e dependence of the pulse width , over a given range of observed photon energies . finally , the primary source spectrum , which we argue is not a direct observable , is somewhat uncertain .
altogether , the range of @xmath27 is consistent with the ranges @xmath28 , @xmath29 , @xmath30 , and @xmath31 .
it is predicted that the dependence of width on e weakens ( i.e. @xmath25 decreases ) at lower photon energies , and this should be testable with detectors that are more sensitive at lower energies , such as the gamma ray burst monitor .
as the acceleration time is inversely proportional to the radiation flux on the scatterer , it is clear , all other things being equal , that the rise time of the pulse and spectral lag are inversely proportional to source luminosity , as observed ( e.g. gehrels et al . , 2006 ) .
of course , scatter in other variables , such as the distance from the source of illumination , optical depth of the scatterer etc .
, creates scatter in the constant of proportionality .
if the scattering is isotropic ( or backwardly biased due to high optical depth ) in the scattering frame , it follows from equation ( 1 ) that the scattered radiation , averaged over angle , is a factor of 2 ( or more ) softer than the primary emission . as explained in em07 , the other half of the energy goes into the acceleration of the scatterer . on the other hand , at most viewing angles @xmath6 , the scattered radiation is harder than scattered radiation after the scatterer has reached terminal lorentz factor @xmath32 if @xmath33 . as the scattered radiation during the acceleration phase of the scatterer is likely to be the most time dependent , it may be possible to separate out this component from the other two . the extent to which the scattering affects the spectrum depends , of course , on the fraction of primary radiation that is scattered . equating the scattered photon energy with the baryon afterglow energy , and applying the results of eichler and jontof - hutter ( 2005 ) , which estimated the afterglow efficiency
, we may tentatively estimate that about 30 percent of the primary emission is scattered , about half of that 30 percent going into baryons and the other half ending up in a scattered subpulse component .
clearly there is variation in the scattered fraction as well as uncertainty in theoretical inferences of the baryon energy from afterglow calorimetry , so this estimate should be considered rough and preliminary . the time - integrated spectrum at a _ given _ viewing angle can be different from the average , because the scattered radiation is not isotropic but , rather , beamed in an ever narrowing cone as the scatterer accelerates .
consider a primary emission spectrum that is a delta function @xmath34 . at a given @xmath35 and a given observed photon energy @xmath36 ,
a monochromatic primary spectrum @xmath37 is , after scattering , monochromatic at photon energy @xmath38 $ ] given by equation ( 1 ) , so the contribution to the emitted power at energy e comes only at @xmath39 the time integrated energy @xmath40 of the scattered radiation at observed photon energy e and viewing angle @xmath6 is @xmath41 = @xmath42 where @xmath43 is the power of the scattered radiation as observed at photon energy e in the frame of the primary source , is to make it a positive quantity .
] t is the elapsed time in the frame of the source ( and in any case the variable of integration ) , @xmath44 is has front - back symmetry in the frame of the scatterer .
a high , time varying optical depth would be more complicated . ]
@xmath45 ( em07 ) . making the simplifying assumption that the scattered radiation is isotropic in the frame of the scatterer ,
i.e. that @xmath46 , using the transformation for emitted power @xmath47 ( equation 4.97a in rybicki and lightman , 1979 ) and evaluating @xmath48 in units of @xmath11 from equation ( 1 ) , it follows that @xmath49^{1/2}/e_o[1-\mu]^{3/2}.\ ] ] in the limit that @xmath50 is sufficiently below unity that @xmath51 does not depend significantly on e , @xmath52 is proportional to @xmath53 , i.e. @xmath54 . that this is softer than the average over all viewing angles - for which @xmath55 - can be understood as the result of most of the emission at large t going into a narrower cone than the one the observer is on , so that the observer sees only the soft fringes of this dominant component .
also note that this result assumes that the scatterer s acceleration proceeds indefinitely .
if the scatterer reaches a terminal velocity @xmath56 , then the observer would not see any of the primary radiation originally at @xmath11 scattered to an energy below @xmath57 . while the scattered radiation is not the only observable component ,
the hypothesis that it comprises a significant fraction of the total fluence of many grb is broadly consistent with the tendency of the low energy spectral index to not exceed 0 .
the question of how much radiation is scattered , before and after the scatterers have reached terminal velocity , remains somewhat open at the quantitative level , but the gbm data may provide the opportunity to address these questions quantitatively as well as qualitatively .
we have presented evidence that subpulses in @xmath0-ray bursts are photons that are scattered into our line of sight by scatterers with lower lorentz factors than the frame in which the prescattered photons had zero net momentum .
because scattering can never increase the intensity of a beam of photons , the hypothesis presupposes that the observed intensity is lower in the observer s direction than in the direction of the beam , and that the scattering by the slower baryons broadens the beam enough that it engulfs the observer s line of sight .
it is this widening of the beam that allows the observer to see enhanced flux .
this fits the picture already put forth to explain the amati relation .
as most grb with known redshifts have spectral peaks and energies that are both below those of the hardest , brightest grb , it would follow , according to our interpretation , that a large fraction of , perhaps most , grb are observed from an offset viewing angle .
the question then arises as to why there are so many offset observers relative to those that are within the @xmath58 beaming cone of the primary radiation . in an earlier paper ( eichler and levinson 2004 ) , it was proposed that the complex shape of the primary beam - e.g. an annular shape - would allow a comparable number of viewers just off ( i.e. within several times @xmath58 ) of the periphery of the primary beam to the number of viewers within this periphery .
it was shown that a thick annulus , in which the separation between the inner and outer radii is comparable to half the outer radius , gives a distribution of offsets that is consistent with the observed distribution of spectral peaks .
there may be several reasons for grb jets to have an annular or otherwise complex morphology : it could be that baryons are entrained in the flow from ( or are fed neutrons by ) the confining walls ( levinson and eichler 2003 ) and that much of the liberated energy is due to dissipation associated with this entrainment . or
, it could be that dissipation from shocks associated with wall impact could preferentially liberate energy from near the confining walls ( begelman , private communication ) .
here we suggest another simple mechanism that would give the scattered @xmath0-radiation some measure of annular bias , depending on the fraction scattered : consider the region of flow where the @xmath0-rays make their last scattering off baryons within the flow .
if , as seems more likely than not , the baryons are clumpy , it is likely that the clumps are optically thick , while the interclump medium is optically thin . in this case the photons are most likely to make their last scattering off the surface of a clump .
if the clumps are moving more slowly than the primary fireball , then the photons are most likely to overtake them from the rear , and , because the clumps are optically thick , the photon is likely to emerge from its last scattering from the rear end of the clump .
it is then obscured from a viewer who lies along the velocity vector of the clump , just as sunlight scattering off the moon is obscured to a viewer on the dark side of the moon .
the viewers best positioned to see the back side illumination are those who see photons that are emitted backward in the frame of the cloud - i.e. those that are offset by more than @xmath3 from the velocity vector of the clump .
this , we suggest , could be a reason so many grb are observed from an offset angle of more than @xmath3 .
some subpulses have such fast rises that they can be interpreted as the @xmath3 shadow of an optically thick cloud narrowing from above to below the offset angle of the observer , @xmath6 , as the cloud accelerates ( em07 ) .
the very sudden rise is then attributed to the observer emerging from the shadow of the clump . to summarize
, we suggest that spectral lags from long bursts are the result of high @xmath4 radiation ( where @xmath4 is the lorentz factor of the frame in which the radiation is isotropic ) impacting slower baryons from behind , and scattering off them while accelerating them . as in ( em07 ) , the quantitative agreement with the data is excellent .
the inverse correlation between lag and luminosity ( e.g. gehrels et al 2006 ) follows from that between acceleration time and luminosity .
the assumptions needed to make the general scheme work are minimal .
fluid in the frame of the clump .
however , in the simplest internal shock model , where the average electron energy , magnetic field , and blue shift all scale with @xmath4 , the spectral peak varies as a high power of @xmath4 . ] in any case , the minimal conclusion to be drawn from the quantitative success of this model in explaining spectral lags of long grb is that baryons are still undergoing considerable acceleration by the time the fireball as a whole is becoming optically thin .
were energy systematically removed from baryons beyond the photosphere ( e.g. because they collided with other scatterers in their shadow ) , one would expect negative spectral lags .
it is predicted that the time integrated spectra of the subpulses should be slightly softer than the primary emission , and harder than the emission that is scattered at terminal lorentz factor .
the glast / fermi gbm monitor offers the potential opportunity for unraveling these three components . for short hard bursts ( shb ) ,
the subpulses are about a factor of 10 to 30 shorter than for long ones and the spectral lags are much smaller .
this difference can perhaps be attributed to the difference in timescale over which the baryons are undergoing acceleration .
for example shb are likely to be observable somewhat closer to the central engine , being unobscured by the envelope of a massive host star , and at a wider angle ( eichler , guetta , and manis 2008 ) .
if this is indeed the case , then we may be able to see baryonic clumps at an earlier stage of their acceleration , when the acceleration time is considerably shorter .
there are many unknowns in this hypothesis - e.g. the optical depth of the baryons , their point of injection and their covering factor - on which the qualitative nature of the subpulses may depend , and future work will focus on the question of whether reasonable ranges for these parameters can explain the wide diversity of grb light curves and spectra .
this work was supported by the joan and robert arnow chair of theoretical astrophysics , the us - israel binational science foundation and the israeli science foundation s center of excellence program .
99 amati , l. , et al .
, 2002 , a & a , 390 , 81 eichler , d. & jontof - hutter , d. 2005 , apj , 635 , 1182 eichler , d. & levinson , a. , 2004 , apj .
614 , l13 eichler , d. & levinson , a. , 2006 , apj , 649 , l5 eichler , d. & manis , h. , 2007 , apj , 669 , l65 eichler , d. guetta , d. & manis , h. , 2008 , apj ( submitted ) fenimore , e. e. , in t zand , j. j. m. , norris , j. p. , bonnell , j. t. & nemiroff , r. j. , 1995 , apj , 448 , l101 gehrels , n. et al .
2006 , nature 444 , 1044 levinson , a. & eichler , d. , 2003 , apj , 594 , l19 levinson , a. & eichler , d. , 2005 , apj , 629 , l13 ryde , f. , 2004 , apj , 611 , l41 | a quantitative theory of spectral lags for @xmath0-ray bursts ( grbs ) is given .
the underlying hypothesis is that grb subpulses are photons that are scattered into our line of sight by local concentrations of baryons that are accelerated by radiation pressure . for primary spectra that are power laws with exponential cutoffs , the width of the pulse and its fast rise , slow decay asymmetry is found to increase with decreasing photon energy , and the width near the exponential cutoff scales approximately as @xmath1 , with @xmath2 , as observed .
the spectral lag time is naturally inversely proportional to luminosity , all else being equal , also as observed . |
in this paper we present a discovery of significantly extended regions of star formation in some early - type galaxies ( etgs)galaxies usually thought to lie on the passive side of galaxy bimodality .
bimodality in terms of morphology and color has been known since the earliest studies of galaxies , but it was not until the massive datasets of the sloan digital sky survey ( sdss ) that fuller implications in terms of galaxy evolution became evident @xcite .
optical colors reflect the mean age of stellar populations and are therefore sensitive only to high levels of continuous sf @xcite .
if the sfr per unit stellar mass ( specific sfr ) drops below some threshold , optical colors become uniformly red and sdss photometry can not distinguish a truly passive galaxy from one that also contains a young population . these limitations
are alleviated when ultraviolet ( uv ) photometry , dominated by young stars , is available .
early results from _ galex _ showed that a surprisingly high fraction ( 15% ) of optically red sdss etgs exhibit strong uv excess @xcite .
@xcite found strong far - uv ( fuv ) excess even when selecting etgs with no h@xmath6 emission in sdss spectra .
is this uv excess due to star formation , as assumed by @xcite ?
while sf and molecular gas have been studied in nearby early - type galaxies for some time ( e.g. , @xcite ) , their significance as a possible _ phase _ in galaxy evolution or a _ mode _ of galaxy growth requires the large samples we have today . before considering such far - reaching implications one must ask whether other explanations for the uv flux exist ? after all ,
nearby ellipticals are known to exhibit a moderate uv excess ( the `` uv upturn '' , @xcite ) , that comes from old stars ( presumably hot horizontal branch ) , and not massive young stars @xcite .
also , a continuum from a weak agn could in principle produce an uv excess @xcite . with @xmath2 fwhm ,
_ galex _ imaging marginally resolves sdss galaxies at @xmath1 ( angular diameter @xmath7 ) , which is why we turned to _ high - resolution _ fuv imaging with the solar blind channel ( sbc ) of the acs .
images of our sample of massive etgs with strong uv excess and no obvious optical signs of sf reveal a surprise : they are dominated by _ extended _ star formation on scales of 1050 kpc , and with rates of up to 2 @xmath8 .
our sample is selected from the sdss dr4 main spectroscopic survey ( @xmath9 ) matched to _ galex _
medium imaging survey ir1.1 ( @xmath10 ; ab magnitudes throughout ) .
the details of sdss and _ galex _ data and the matching procedure are given in @xcite .
from the matched sdss/_galex _ catalog we select optically quiescent early - type galaxies ( qetgs ) in the following way : ( 1 ) redshift @xmath11 to yield a sample with larger angular sizes , ( 2 ) @xmath12-band light concentration ( ratio of 90 and 50% petrosian radii ) @xmath13 to select dominant spheroids ( fig .
[ fig : conc ] ) , ( 3 ) no detectable h@xmath6 emission based on dr4 version of @xcite bpt classification ( `` no h@xmath6 '' category in @xcite ) . note that no color selection has been applied to select qetgs . out of @xmath14 qetgs , one - fifth has rest - frame fuv - optical colors ( fig .
[ fig : cmd ] ) bluer than @xmath15 .
@xcite show that nearby rc3 ellipticals ( without lenticulars ) , where the uv excess is known to come from classical uv upturn ( old populations ) , are redder than this limit .
in contrast , we are interested in qetgs with _
strong _ uv excess , so we select galaxies with @xmath16 .
there are 60 such galaxies from which we exclude blends , obvious optical disturbances , late - type contaminants , and e+a post - starbursts ( based on h@xmath17 index ) , to arrive at a final _ hst _ sample of 30 .
the uv - optical ( @xmath18 ) color range of our sample can be seen from figure [ fig : cmd ] .
the sample is presented with symbols , while greyscale represents all sdss-_galex _ galaxies at @xmath19 ( the underlying population ) . by selection
, the sample is bluer than the @xmath18 red sequence , with two galaxies ( @xmath20 ) lying squarely in the @xmath18 blue sequence .. while no optical color cut has been explicitly applied , our strong uv excess sample has distinctly _ red _ optical color ( fig .
[ fig : conc ] ) , placing them firmly in the optical red sequence .
similar confinement to the red sequence is seen in @xmath21 colors ( not shown ) . @xcite and
@xcite study blue early - type galaxies in sdss selected by atypically blue @xmath21
. however , our sample galaxies are _ redder _ than their @xmath21 cuts .
with fwhm of @xmath2 , _ galex _ makes it difficult to pinpoint the origin of the fuv light at @xmath22 .
in contrast , acs / sbc on the _ hst _ places 80% of point source energy in @xmath23 @xcite .
targets were observed with one orbit through the long - pass filter f125lp ( @xmath24 ) .
processing was performed using multidrizzle with sbc - optimized parameters , and smoothed using adaptsmooth @xcite .
twenty - nine targets were successfully imaged , and each produced a detection , either of a compact central source or of extended structures ( or both ) .
to our surprise , 22 galaxies ( 76% ) revealed an extended uv morphology and an additional three had uv patches within several arcsec of the nucleus . in all cases
the extended uv emission is structured , and thus results from _ star formation _ ( unlike a diffuse component one might expect from an old population ) .
the uv extent is typically larger than the optical size , though mostly contained within a radius containing 90% _ optical _ petrosian flux .
we divide extended structures into strong ( 15 ) and weak ( 7 ) based on visual appearance and provisional flux measurements .
these are labeled in figures by circles and diamonds respectively . in all galaxies
save four a compact central source is present as well , which we divide into strong ( 14 ; plus sign ) and weak point sources ( 11 ; dots ) .
several strong extended uv morphologies are illustrated in figure [ fig : images ] .
galaxy ( a ) has a weak central source , while others exhibit a strong central source .
insets show @xmath25-band sdss images with acs fields indicated .
as required by sample selection , optical images appear like etgs , though , after the fact , hints of structure are visible in ( a ) , ( d ) and a couple of other galaxies .
the galaxy in panel ( a ) of figure [ fig : images ] shows multiple star forming features , including an inner ring and what look like flocculent spiral arms .
two other galaxies in the sample show flocculent features .
panel ( b ) shows a galaxy with a wide sf ring .
such morphology is most common in our sample ( eight other cases ) .
next is one of four galaxies showing thin concentric rings .
the last ( panel d ) is the most striking in terms of surface brightness , with two bright spiral arms emanating from what appears to be a co - rotation ring , connected by a fainter outer lindblad resonance ring @xcite .
the optical image shows a hint of red spiral arms .
two other galaxies in the sample appear to have spiral arms , albeit less prominent .
we first return to the plot showing concentration vs. optical color ( fig .
[ fig : conc ] ) .
two of the optically bluest sources do not have extended structures , emphasizing the disconnect between the uv excess and the optical colors . on average , galaxies with a strong central source tend to be more optically concentrated than those with only a weak central source , which are in turn more highly concentrated than those with no central source . figure [ fig : cmd ] shows the @xmath18 color - magnitude diagram .
sample galaxies span a wide range in absolute magnitude , with a median of @xmath26 , which is close to @xmath27 @xcite . similarly to concentration , there exists the trend between the strength of central source and absolute magnitude . not surprisingly , this trend is also present vs. stellar mass and , even more strongly , velocity dispersion .
compared to all uv - detected qetgs , the uv - strong ones have somewhat lower stellar masses ( median is @xmath28 ) , with a clear deficit at @xmath29 . in other words ,
most massive etgs do not exhibit strong uv excess .
two galaxies stand out as low luminosity ( and mass ; @xmath30 ) .
they exhibit less extended ( yet still patchy ) uv morphology without a central source . switching from @xmath18 to the less dust - sensitive @xmath31 will allow us to place the sample in a better sf history context .
figure [ fig : afuv ] plots fuv attenuation against @xmath31 color .
first focusing on the color , we see that the majority of uv excess etgs fall in the green valley ( @xmath32 ) . @xmath33
basically `` resolves '' the _ optical _ red sequence into a green valley and _ uv - optical _ red sequence .
every red galaxy had to pass through the green valley at some point , and presumably most green valley galaxies are transiting for the first time @xcite .
after a galaxy gets on a red sequence it can enter the green valley again if it experiences a star formation episode ( `` rejuvenation '' ) . from figure
[ fig : afuv ] one notices that sample galaxies in the uv red sequence lack strong extended structures , while the contrary is true for bluer galaxies .
thus , we conclude that uv excess galaxies are powered by star formation if they are in the green valley or bluer , but they are dominated by a central source ( perhaps an agn ) if despite blue @xmath18 their @xmath31 places them in the red sequence .
the dust attenuation of sample galaxies is smaller than that of other galaxies of the same color , which may hold clues to their evolutionary path .
different uv morphologies have no significant differences in terms of dust attenuation . to distinguish the effects of dust from those of sf history , we switch from @xmath31 to ( dust - corrected ) specific sfr ( @xmath34 ) .
the specific sfr is plotted against the far - uv physical size in figure [ fig : fwhm ] , with more actively star - forming galaxies on the left .
our sample now spans a narrower range in @xmath35 than it did in @xmath31 .
two galaxies that sat in the blue sequence owed such blue colors to very low dust attenuation , and actually have `` transitional '' specific sfrs as do other sample galaxies . galaxies without extended uv emission tend to have lower specific sfrs , in agreement with inferences from color .
sample galaxies span a range of sfrs from 0.01 to 2 @xmath36 , with a median of 0.4 @xmath8 .
uv sizes in figure [ fig : fwhm ] are _ galex _ fwhm measurements for both the sample and the underlying population ( to avoid biasing one measurement ) .
bursty galaxies are typically less massive and smaller compared to those with @xmath37 . beyond that , sf activity decreases as do the uv sizes
. however , sample galaxies are on average significantly larger than the underlying galaxy sample of the same specific sfr .
this study was motivated by the results of @xcite in which qetgs were selected to study the classical , old - population uv upturn , but instead resulted in much bluer samples that did not adhere to relations expected for classical uv upturn .
@xcite suggested that the blue uv - optical extent came from a low - level agn contamination , while other studies ( e.g. , @xcite ) attributed blue colors in etgs to star formation .
our current study clearly demonstrates that the majority of uv excess etgs ( even though optically quiescent ) indeed have uv morphology consistent with sf .
this is especially the case for galaxies with @xmath38 , a cut originally proposed by @xcite to distinguish sf from the old - population uv upturn .
so why is there little or no h@xmath6 when from the fuv flux one expects h@xmath6 s / n of around 40 ? as figure [ fig : images ] shows , the @xmath39 sdss fiber may entirely miss the uv - bright regions .
however , in some cases there appears to be uv emission within the fiber diameter and still no h@xmath6 .
thus , there is a possibility that h@xmath6 emission is genuinely suppressed in the central parts or throughout these galaxies .
this would happen if no sf took place over the last 10 myr ( the lifetime of massive stars producing ionizing emission ) , or if few such stars are being produced ( e.g. , @xcite ) .
planned resolved h@xmath6 imaging should settle this issue .
now that we established the nature of the uv emission , we turn to the origin of star - forming gas .
we consider these scenarios : ( 1 ) observed sf is what remains of the original , continuous sf ; ( 2 ) sf results from a gas accreted from a minor gas - rich merger or ( 3 ) accreted directly from igm , after a period of quiescence . in scenarios 2 and 3
galaxy is again in the green valley after it was already on the red sequence , while in 1 it is transiting from blue to red sequence for the first time .
while we selected the sample to have highly concentrated light profiles , high concentrations are typical for most green valley galaxies , so this can not be used to distinguish scenario 1 from the other two .
the fact that strong uv - excess etgs typically have blue sequence - like masses ( @xmath40 ) may suggest scenario 1 , because if already passive etgs are being rejuvenated , why would this happen less often to massive etgs ?
sauron sees a similar phenomenon in nearby etgs , where slow rotators do not show infrared signatures of sf @xcite .
presence / absence of sf may be another form of e - e dichotomy @xcite , in which more luminous ( @xmath41 ) ellipticals are slow rotators and have x - ray emitting gas , possibly maintained by radio mode agn @xcite , which prevents gas accretion . in our sample , etgs with sf all have @xmath42 .
under scenario 1 our uv - strong galaxies will have properties like other green - valley galaxies , most of which presumably transition from blue to red and are not rejuvenated red galaxies @xcite .
however , both low dust attenuation ( fig .
[ fig : afuv ] ) and large uv size ( fig .
[ fig : fwhm ] ) argue that this can not be the case for the majority of the sample .
low dust attenuation makes our sample more similar to @xmath31 red sequence than to green - valley galaxies . and
if red galaxies acquired gas from dwarfs , merging would not increase the dust content significantly , since dwarfs are almost dust - free ( e.g. , @xcite ) .
similarly , the uv sizes of our sample are larger than of the underlying galaxies with similar specific sfr , suggesting an external origin . together
, this implies that most ( but possibly not all ) of uv - excess etgs are rejuvenated systems .
rings , in one form or other , are present in 15 out of 22 galaxies having extended uv .
rings in disk galaxies are commonly attributed to either bar - induced resonances @xcite or head - on collisions @xcite . however ,
collisional rings , in addition to being rare , are usually off - center @xcite , which is not the case here . in a nearby sauron elliptical ngc2974 , uv rings detected by _ galex _ were explained as resonant features due to a hypothetical , optically undetected bar @xcite .
if ring structures in our sample are due to resonances , the bars that induce them would need to be more common than observed in s0 galaxies ( @xmath43 , @xcite ) . of galaxies in the sample , the galaxy in figure [ fig : images ] ( d ) most clearly shows classical resonance features , and the bar is present .
other ringed galaxies show no obvious evidence of bars in sdss images .
perhaps some other , less - investigated process introduces rotational asymmetry and forms rings @xcite .
resonances have been classically studied in disk galaxies with existing gas , while here we are probably dealing with newly - accreting gas , which may behave differently .
regardless of the origin of the rings , we still need to identify the source of the cold gas .
recently , @xcite advocated that minor gas - rich mergers ( scenario 2 ) can fully explain the extent and frequency of nuv - excess etgs .
they also construct n - body simulations @xcite of minor mergers , but these do not produce ring - like uv morphologies that we see . instead
, their young stars are highly concentrated in the nuclear region of the host
. however , their hosts are kinematically more similar to ellipticals , while as found by @xcite and indicated here , the only type of etgs that allow extensive sf may be s0/fast - rotators .
also , our selection against h@xmath6 in the central region may bias against central starbursts .
@xcite simulations also predict that there should exist _ stellar _ signatures of mergers , such as shells and tidal tails , but the current optical data are not adequate to search for them .
the sfrs of our etgs are similar to those of gas - rich dwarfs with stellar 20 times lower masses . ignoring dwarf galaxy duty cycles and merger timescales
, this would imply very minor mergers . finally , regarding scenario 3 ( igm accretion , @xcite )
, we first ask why does igm lead to sf rejuvenation and not just the continuation of the original sf that formed the bulk of stellar population .
many recent models require agn feedback both to shut down sf and to maintain their quiescence @xcite .
however , if the agn stopped being active , a galaxy could possibly re - acquire cold gas , producing cyclical phases of agn feedback and star formation .
our galaxies , showing at best only very weak agn activity , may currently be in the renewed star - forming phase .
( in contrast , in more massive etgs , @xmath5 , the accretion may be permanently disrupted . ) large uv sizes ( fig .
[ fig : fwhm ] ) and similarities with xuv - disk phenomenon @xcite point towards this scenario , but we currently can not strongly favor igm accretion over minor mergers as the main source of gas in our sample .
in closing we turn to the nature of the uv emission of the compact _ central _ source .
while the compactness of uv emission suggests an agn , acs / sbc resolution can not outright rule out a highly nucleated stellar bulge .
we see a correlation between the strength of the central source and velocity dispersion that may ( through accretion rate being a function of @xmath44 ; @xcite ) suggest an agn , but again does not exclude stellar origin . while emission lines are too weak ( by selection ) to allow secure individual bpt classification , averaging the measurements results in strong and weak central sources landing in the agn ( seyfert ) part of the bpt diagram , and those with no central source in the sf / agn composite region . finally , galaxies with no extended uv , where we can measure central uv color have @xmath45 , which is more consistent with a blue continuum source @xcite than an old stellar population .
early - type galaxies with strong far - uv excess are dominated by sf , except when they are red in the @xmath31 , in which case the fuv comes from a compact source , possibly an optically weak agn . extended sf around etgs has recently been found in some _ individual _ nearby galaxies , but these have stellar masses one to two orders of magnitude smaller @xcite .
we extended sf in galaxies up to @xmath46 .
however , strong uv excess is rare in yet more massive etgs possible consequence of agn feedback and a reflection of an ` e - e ' dichotomy . in the majority of cases
the sf gas is probably external in origin ( minor mergers or igm accretion ) , making these etgs ` rejuvenated ' .
most common uv morphology are rings , which likely result from resonances ( from yet undetected bars or other disturbances )
. a smaller fraction of uv - excess etgs may be more typical green - valley galaxies , where we see the declining original sf activity .
similar conclusions regarding the heterogeneity of evolutionary paths of transiting galaxies ( but not necessarily etgs ) were recently reached by @xcite based on hi content of nearby ( @xmath47 mpc ) galaxies .
we thank stefano zibetti for help with adaptsmooth and ray lucas for help with acs data . based on observations made with the nasa / esa hubble space telescope and supported with nasa grant hst - go-11158.03 .
aguerri , j. a. l. , mndez - abreu , j. , & corsini , e. m. 2009 , , 495 , 491 ageros , m. a. , et al .
2005 , , 130 , 1022 appleton , p. n. , & struck - marcell , c. 1996 , fundamentals of cosmic physics , 16 , 111 blanton , m. r. , et al .
2001 , , 121 , 2358 bregman , j. n. , hogg , d. e. , & roberts , m. s. 1992 , , 387 , 484 brinchmann , j. , charlot , s. , white , s. d. m. , tremonti , c. , kauffmann , g. , heckman , t. , & brinkmann , j. 2004 , , 351 , 1151 buta , r. , & combes , f. 1996 , fundamentals of cosmic physics , 17 , 95 code , a. d. 1969 , , 81 , 475 cortese , l. , & hughes , t. m. 2009 , , 400 , 1225 croton , d. j. , et al .
2006 , , 365 , 11 dale , d. a. , et al . 2007 , , 655 , 863 donas , j. , et al .
2007 , , 173 , 597 donovan , j. l. , et al .
2009 , , 137 , 5037 ford , h. c. , et al .
2003 , , 4854 , 81 greggio , l. , & renzini , a. 1990 , , 364 , 35 jeong , h. , bureau , m. , yi , s. k. , krajnovi , d. , & davies , r. l. 2007 , , 376 , 1021 kannappan , s. j. , guie , j. m. , & baker , a. j. 2009 , , 138 , 579 kaviraj , s. , et al .
2007 , , 173 , 619 kaviraj , s. , peirani , s. , khochfar , s. , silk , j. , & kay , s. 2009 , , 394 , 1713 kauffmann , g. , et al .
2003 , , 341 , 54 kauffmann , g. , et al .
2007 , , 173 , 357 kere , d. , katz , n. , weinberg , d. h. , & dav , r. 2005 , , 363 , 2 kormendy , j. , fisher , d. b. , cornell , m. e. , & bender , r. 2009 , , 182 , 216 lee , j. c. , et al .
2009 , , 706 , 599 lynds , r. , & toomre , a. 1976 , , 209 , 382 martin , d. c. , et al .
2007 , , 173 , 342 masters , k. l. , et al .
2010 , arxiv:1001.1744 peirani , s. , crockett , r. m. , geen , s. , khochfar , s. , kaviraj , s. , & silk , j. 2009 , arxiv:0912.2629 rich , r. m. , et al .
2005 , , 619 , l107 salim , s. , et al .
2007 , , 173 , 267 schawinski , k. , et al .
2009 , , 396 , 818 schwarz , m. p. 1984
, , 209 , 93 shapiro , k. l. , et al .
2010 , , 92 strateva , i. , et al .
2001 , , 122 , 1861 thilker , d. a. , et al . 2007 , , 173 , 538 thilker , d. a. , et al . 2010 , arxiv:1003.4985
yi , s. k. , et al .
2005 , , 619 , l111 wyder , t. k. , et al .
2007 , , 173 , 293 zibetti , s. 2009 , arxiv:0911.4956 | in recent years argument has been made that a high fraction of early - type galaxies in the local universe experience low levels ( @xmath0 ) of star formation ( sf ) that causes strong excess in uv flux , yet leaves the optical colors red .
many of these studies were based on _ galex _ imaging of sdss galaxies ( @xmath1 ) , and were thus limited by its @xmath2 fwhm .
poor uv resolution left other possibilities for uv excess open , such as the old populations or an agn . here
we study high - resolution far - ultraviolet _
hst_/acs images of optically quiescent early - type galaxies with strong uv excess .
the new images show that three - quarters of these moderately massive ( @xmath3 ) early - type galaxies shows clear evidence of extended sf , usually in form of wide or concentric uv rings , and in some cases , striking spiral arms .
sdss spectra probably miss these features due to small fiber size .
uv - excess early - type galaxies have on average less dust and larger uv sizes ( @xmath4 kpc ) than other green - valley galaxies , which argues for an external origin for the gas that is driving the sf .
thus , most of these galaxies appear ` rejuvenated ' ( e.g. , through minor gas - rich mergers or igm accretion ) . for a smaller subset of the sample ,
the declining sf ( from the original internal gas ) can not be ruled out .
sf is rare in very massive early - types ( @xmath5 ) , a possible consequence of agn feedback .
in addition to extended uv emission , many galaxies show a compact central source , which may be a weak , optically inconspicuous agn . |
conformal invariants have important roles in geometric function theory .
one of the basic conformal invariants is the modulus of a curve family , which is used to define quasiconformal maps @xcite . in euclidean spaces of dimension more than two ,
conformal maps are nothing but the restrictions of mbius maps ; for instance , see @xcite .
therefore , it is natural to study mbius invariants in the higher dimensional setting .
there are some metrics which are mbius invariant and some are not .
for example , the apollonian @xcite and the seittenranta @xcite metrics are mbius invariant whereas the quasihyperbolic @xcite and the distance ratio @xcite metrics are not .
the study of the mbius quasi - invariance property is hence natural for these metrics which are not mbius invariant . in other words
, it would be interesting if we obtain the lipschitz constants for those metrics which do not belong to the mbius invariant family .
gehring and osgood in @xcite proved that the distance ratio metric and the quasihyperbolic metric are not changed by more than a factor @xmath1 under mbius maps . in @xcite , sharp distortion properties of the distance ratio metric under mbius transformations of the unit ( punctured ) balls are obtained .
a sharp distortion property of the cassinian metric under mbius transformations of the unit ball is also recently established in @xcite .
most of the metrics mentioned in this paper belong to the family of relative metrics , some are mbius invariant and some are not . by a _ relative metric _ we mean a metric that is evaluated in a proper subdomain of @xmath2 relative to its boundary .
note that the topological operations ( boundary , closure , etc . ) are taken in the compact space @xmath3 .
one of our main objectives in this paper is to consider a relative metric , a special case of the family of metrics defined in ( * ? ? ?
* lemma 6.1 ) , the so - called _ cassinian metric _ ,
independently first studied by ibragimov @xcite and look at its mbius quasi - invariance properties .
note that the generalized relative metric defined in ( * ? ? ?
* lemma 6.1 ) is named as the _
@xmath4-relative metric _ and defined on a domain @xmath5 by the quantity @xmath6 where @xmath4 is continuous in @xmath7 and @xmath8 denotes the boundary of @xmath9 . if @xmath10 , then the corresponding relative metric @xmath11 defines the cassinian metric introduced in @xcite and subsequently studied in @xcite .
the choice @xmath12 similarly leads to the triangular ratio metric recently investigated in @xcite .
we refer to the next section for the explicit definition of the cassinian metric .
in one hand , we study distortion properties of the cassinian metric under mbius and bi - lipschitz maps in section [ sec3 ] . on the other hand
, we also focus on a general question suggested by vuorinen in @xcite about the convexity of balls of small radii in metric spaces .
this problem has been investigated by kln in different situations in a series of papers @xcite . in this context
, we study convexity properties of the cassinian metric balls in section [ sec4 ] .
section [ sec5 ] is devoted to the inclusion properties of the cassinian metric balls with other related metric balls .
throughout the paper we use the notation @xmath13 for the euclidean @xmath14-dimensional space ; @xmath15 for the one point compactification of @xmath2 . the euclidean distance between @xmath16
is denoted by @xmath17 . given @xmath18 and @xmath19 , the open ball centered at @xmath20 and radius @xmath21 is denoted by @xmath22 .
denote by @xmath23 , the unit ball in @xmath2 .
consequently , we set @xmath24 , the upper half - space .
let @xmath5 be an arbitrary domain .
cassinian metric _
, @xmath25 , on @xmath9 is defined by @xmath26 note that the quantity @xmath25 defines a metric on @xmath9 ; see ( * ? ? ?
* lemma 3.1 ) .
geometrically , the cassinian metric can be defined in terms of maximal cassinian ovals ( see ( * ? ? ?
* sections 3.1 - 3.2 ) and references therein ) in the domain @xmath9 in a similar fashion as the apollonian metric is defined in terms of maximal apollonian balls @xcite .
we end this section with the definitions of the hyperbolic metric , the quasihyperbolic metric and the distance ratio metric used in the subsequent sections .
the _ hyperbolic metric _ of the unit ball @xmath27 is defined by @xmath28 where @xmath29 denotes the family of rectifiable curves joining @xmath20 and @xmath30 in @xmath27 .
let @xmath5 be an arbitrary domain .
quasihyperbolic metric _
@xcite is defined by @xmath31 where @xmath29 denotes the family of rectifiable curves joining @xmath20 and @xmath30 in @xmath9 and @xmath32 , the shortest euclidean distance from @xmath33 to @xmath8 .
the quasihyperbolic metric was introduced by gehring and palka in 1976 and subsequently studied by gehring and osgood ; see @xcite , as a generalization of the hyperbolic metric of the upper half plane to arbitrary proper subdomains of @xmath34 .
let @xmath5 .
for any two points @xmath35 , the _ distance ratio metric _ , @xmath36 , is defined as @xmath37 where @xmath38 .
this form of the metric @xmath39 , which was first considered by vuorinen in @xcite , is a slight modification of the original distance ratio metric introduced by gehring and osgood in @xcite .
this metric has been widely studied in the literature ; see , for instance , @xcite .
one of our objectives in this section is to study the distortion property of the cassinian metric under mbius maps from a punctured ball onto another punctured ball .
distortion properties of the cassinian metric of the unit ball under mbius maps has been recently studied in @xcite .
[ lip - bn ] let @xmath40 and @xmath41 be a mbius map with @xmath42 .
then for @xmath43 we have @xmath44 the equalities in both sides can be attained . if @xmath45 , the proof is trivial ( see @xcite ) .
now , assume that @xmath46 .
let @xmath47 be the inversion in the sphere @xmath48 , where @xmath49 note that the sphere @xmath48 is orthogonal to @xmath50 and that @xmath51 .
in particular , @xmath47 is a mbius map with @xmath52 .
recall from @xcite that @xmath53 then @xmath54 is an orthogonal matrix ( see , for example , ( * ? ?
* theorem 3.5.1(i ) ) ) .
in particular , @xmath55 we will need the following property of @xmath47 ( see , for example , @xcite ) : @xmath56 and @xmath57 denote by @xmath58 .
now we have two choices for @xmath59 .
* case i. * @xmath60 .
it follows from ( [ mob1 ] ) and ( [ mob2 ] ) that @xmath61 * case ii . * @xmath62 .
proof of this case follows from the proof of ( * ? ? ?
* theorem 4.1 ) .
the lower estimate follows by considering the inverse mapping @xmath63 . to see the sharpness ,
consider the map @xmath47 defined by ( [ inversionmap ] ) . for @xmath64 , choose the points @xmath65 and @xmath66 in such a way that @xmath67 the image points of @xmath20 and @xmath30 under @xmath47 is given by @xmath68\setminus\{a , e_1\}.\ ] ] now , the cassinian distance between @xmath69 and @xmath70 is @xmath71 the lower bound can be seen by considering the inverse of @xmath47 and hence the conclusion follows .
gehring and osgood proved that the quasihyperbolic and the distance ratio metrics are not changed by more than a factor @xmath1 under mbius transformations ( see ( * ? ? ? * theorem 4 ) ) .
naturally , one can ask the similar question for the cassinian metric .
unfortunately , such a finite constant does not exist for the cassinian metric in arbitrary proper subdomains of @xmath34 .
indeed , the reason is clear from theorem [ lip - bn ] when @xmath72 ( see also ( * ? ? ? * theorem 4.1 ) ) .
however , if we replace mbius mappings by bi - lipschitz mappings of @xmath34 , the following fact guarantees that distortion constant exists .
* let @xmath73 be an @xmath74-bilipschitz mapping , that is @xmath75 for all @xmath76 , which maps @xmath77 onto @xmath78 .
then @xmath79 for all @xmath80 .
this section focuses on the local convexity properties of the cassinian metric ball .
we define the metric ball as follows : let @xmath81 be a metric space . then the set @xmath82 is called the _ @xmath83-metric ball _ of the domain @xmath9 .
a metric ball with respect to the cassinian metric is called a _ cassinian ( metric )
ball_. before entering into the discussion on the convexity properties , we describe the cassinian ball of a domain @xmath9 in terms of cassinian balls of @xmath84 fixing a centre in @xmath9 .
the following proposition is a consequence of the proof of ( * ? ? ?
* theorem 1.1 ) with respect to the cassinian metric .
let @xmath77 and @xmath85 .
then @xmath86 suppose that @xmath87 .
then @xmath88 for all @xmath89 .
choose @xmath90 such that @xmath91 as @xmath90 , it is clear that @xmath92 and hence @xmath93 . hence @xmath94 .
conversely , let @xmath95 and suppose that @xmath96 such that @xmath97 clearly , @xmath98 . hence @xmath99 and the proof is complete .
let @xmath100 be a domain .
we say that @xmath9 is _ convex _ if the line segment @xmath101 $ ] joining any pair of points @xmath20 and @xmath30 entirely contained in @xmath9 .
a domain @xmath9 is _ strictly convex _ if @xmath102\cap \partial d=\{x',y'\}$ ] for any pair of points @xmath103 .
we now begin with studying local convexity properties of the cassinian ball . for @xmath104 , we call these cassinian balls as _ the cassinian disks_. [ loc - con ] let @xmath105
. then * the cassinian disk @xmath106 is convex if and only if @xmath107 $ ] . *
the cassinian disk @xmath106 is strictly convex if and only if @xmath108 . *
( a ) * without loss of generality we can assume that @xmath109 .
let @xmath33 be an arbitrary point in @xmath110 .
consider the circle ( with respect to the cassinian metric ) @xmath110 for a fixed @xmath111 . from the definition of the cassinian disk
, it follows that the boundary @xmath110 is an euclidean circle with center @xmath112 and radius @xmath113 . therefore , @xmath114 is convex for @xmath115 and not convex for @xmath116 . *
( b ) * when @xmath117 the center of the above euclidean circle becomes @xmath118 , and thus @xmath119 is not strictly convex .
let @xmath120 .
then * the cassinian ball @xmath121 is convex if and only if @xmath107 $ ] . * the cassinian ball @xmath121 is strictly convex if and only if @xmath108 . with radii @xmath122 and @xmath123.,width=245,height=264 ] in the punctured space the cassinian balls are convex with small radius , but the same is not true in general .
the next result shows that this is not the case even in convex domain .
[ local conv hn ] let @xmath124 .
there exists @xmath125 such that @xmath126 is not convex .
it is sufficient to consider the case @xmath104 . for a given @xmath21
we choose @xmath127 and consider cassinian disk @xmath128 with radius @xmath21 . to show that @xmath128 is not convex we choose two points @xmath129 and @xmath130 such that @xmath131 and @xmath132 .
we choose @xmath133 .
now by the geometry of the cassinian ovals @xmath134 let @xmath135 .
again by the geometry of the cassinian ovals @xmath136 now @xmath137 and we choose @xmath138 .
we obtain @xmath139 and thus @xmath140 with radii @xmath141 and @xmath142.,width=377 ] the proof of proposition [ local conv hn ] suggest that the radius of convexity for the cassinian balls @xmath128 in @xmath143 depends on @xmath144 and @xmath21 .
we pose the following conjecture .
[ convexity conj 1 ] there exists @xmath145 such that @xmath128 is convex for all @xmath146 and @xmath147 $ ] .
computer experiments suggest that conjecture [ convexity conj 1 ] holds true for constant @xmath148 .
we pose also some more general conjectures . in a bounded convex domain @xmath149 , @xmath150 is convex for all @xmath151 and @xmath19 .
let @xmath19 .
if @xmath149 is a starlike domain with respect to @xmath151 , then @xmath150 is starlike with respect to @xmath20 .
the purpose of this section is to study inclusion properties of the cassinian metric balls in proper subdomains of @xmath0 with other related metric balls . in other words ,
we consider the problems of the following type : _ for given @xmath152 and @xmath153 , we find optimal radii @xmath154 depending only on @xmath20 and @xmath155 such that @xmath156 where @xmath83 is a metric other than the cassinian metric defined on @xmath9 .
_ [ c+euclidean ] let @xmath5 and @xmath85 .
assume that @xmath157 for @xmath153 .
then the following inclusion property holds : @xmath158 where @xmath159 and @xmath160 .
the radii @xmath21 and @xmath111 are best possible .
moreover , @xmath161 as @xmath162 .
it is clear that for @xmath163 , @xmath164 by the definition of the cassinian metric @xmath25 we have @xmath165 which implies @xmath166 hence the second inclusion holds .
now we prove the first inclusion .
let @xmath167 .
then @xmath168 and by the monotone property @xmath169 in particular , if @xmath170 with @xmath171 , then @xmath172 .
clearly , one can see that @xmath173 we finally show that radii @xmath21 and @xmath111 are best possible . for this , consider the domain @xmath174 and @xmath85 .
let us denote by @xmath175 the line through points @xmath176 and @xmath20 .
we set @xmath177 with @xmath178 .
now @xmath179 implying @xmath180 which shows that @xmath21 is best possible .
similarly , @xmath181 implying @xmath182 which shows that @xmath111 is best possible .
this completes the proof of our theorem .
* theorem 3.8)[jb - inclusion ] if @xmath77 is open , @xmath85 and @xmath153 , then @xmath184 where @xmath185 and @xmath186 .
the formulas for @xmath21 and @xmath111 are the best possible expressed in terms of @xmath155 and @xmath187 only .
we first prove the second inclusion . by ( * ? ? ?
* theorem 3.4 ) we have @xmath191 and from theorem [ c+euclidean ] , @xmath192 now for @xmath193 , using the above estimates we have , @xmath194 . for the proof of the first inclusion we use lemma [ jb - inclusion ] together with theorem [ c+euclidean ] to conclude that @xmath195 by lh@xmath196spital rule it follows that @xmath161 as @xmath162 .
[ c+euclidean+punctured ] for @xmath198 , let @xmath174 , @xmath85 and @xmath153 .
then the following holds : @xmath188 where @xmath199 and @xmath200 .
the radii @xmath21 and @xmath111 are best possible .
moreover , @xmath161 as @xmath162 .
suppose that @xmath201
. then @xmath202 . on simplification , we get @xmath203 , then @xmath204 where the first inequality follows from ( [ eqn5.7.1 ] ) and the last inequality follows from the fact that @xmath205 .
otherwise , @xmath206 where the inequality follows from ( [ eqn5.7.1 ] ) and the first inclusion follows .
now suppose that @xmath207 .
this implies , by theorem [ c+euclidean ] , @xmath208 . if @xmath209 , then @xmath210 otherwise @xmath211 where the second inequality follows from the fact that @xmath212 now , @xmath213 and hence the proof of the second inclusion follows . by lh@xmath196spital rule
it follows that @xmath214 to show that radii @xmath21 and @xmath111 are best possible , we consider the same construction as did in the proof of theorem [ c+euclidean ] .
for the same choice of @xmath129 and @xmath130 , it is easy to verify that @xmath215 which shows that @xmath21 is best possible . similarly , it can be verified that @xmath216 which shows that @xmath111 is best possible .
next , we discuss the inclusion relation ( [ inclusion ] ) when @xmath83 is the hyperbolic metric of the unit ball @xmath27 . in the unit ball @xmath27 , the @xmath219 metric and the @xmath220 metric are comparable and is given by the relation @xmath221 for @xmath222 ; see ( * ? ? ?
* theorem 7.56 ) .
the second inequality reduces to equality when @xmath223 .
it immediately follows that for @xmath224 and @xmath19 , @xmath225 this leads to the following : by theorem [ c - j inclusion ] , @xmath230 and by ( [ rhojbn ] ) , the second inclusion follows with @xmath231 . again from ( [ rhojbn ] ) and theorem [ c - j inclusion ] , we have @xmath232 on simplifying , we get @xmath233 with @xmath234 . by lh@xmath196spital rule
it is easy to see that @xmath235 this completes the proof of our theorem . in order to discuss the relation ( [ inclusion ] ) when @xmath251 , for a domain @xmath252 , we recall the useful inequality ( * ? ? ?
* lemma 2.1 ) @xmath253 it follows immediately from ( [ jkd ] ) and theorem [ c+euclidean+punctured ] that for @xmath198 , let @xmath174 , let @xmath85 and @xmath153 .
then the following inclusion relation holds : @xmath255 where @xmath256 and @xmath200 .
the radii @xmath21 and @xmath111 are sharp and @xmath161 as @xmath162 . by ( [ jkd ] ) and theorem [ c - j inclusion ] we have @xmath265 and the first inclusion follows .
again from theorem [ c - j inclusion ] and lemma [ jkinclusion_general ] we have @xmath266 by lh@xmath196spital rule it follows easily that @xmath161 as @xmath162 .
hence the proof of our theorem is complete . | in this paper we prove a sharp distortion property of the cassinian metric under mbius transformations of a punctured ball onto another punctured ball .
the paper also deals with discussion on local convexity properties of the cassinian metric balls in some specific subdomains of @xmath0 .
inclusion properties of the cassinian metric balls with other hyperbolic - type metric balls are also investigated .
in particular , several conjectures are also stated in response to sharpness of the inclusion relations .
+ * 2010 mathematics subject classification*. 30c35 , 30c20 , 30f45 , 51m10 . * keywords and phrases . * mbius transformations , cassinian metric , distortion property , inclusion property , metric balls , convexity . |
spontaneous creation of particles or charged pairs from an external gauge field or a curved spacetime is one of the most prominent phenomena in quantum field theory .
a strong electric field produces pairs of charged particles and antiparticles , known as the schwinger mechanism @xcite .
the more remarkable phenomenon is the emission of all species of particles from black holes , known as hawking radiation @xcite . under an influence of strong backgrounds
the vacuum may spontaneously breakdown due to quantum fluctuations and virtual pairs can be separated either by the energy of the fields or the causality of spacetimes .
an accelerating detector measures a thermal spectrum of the unruh temperature determined by the acceleration , known as the unruh effect.@xcite the spectrum and characteristics for these effects are summarized in table [ ta1 ] .
heisenberg and euler found the one - loop effective action for an electron in a constant electromagnetic field@xcite and schwinger introduced the proper - time integral method to express the effective action in scalar and spinor quantum electrodynamics ( qed),@xcite which is now known as the heisenberg - euler or schwinger effective action .
the most distinct feature of the heisenberg - euler or schwinger action is poles of the proper - time representation of the action in an electric field .
thus , the one - loop effective action has not only the vacuum polarization ( the real part ) but also the vacuum persistence amplitude ( twice the imaginary part ) .
the vacuum persistence amplitude is a consequence of spontaneous production of charged pairs from the dirac sea .
notice that the schwinger effect is the particle - hole theory , in which virtual particles from the dirac sea tunnel through a tilted potential barrier due to the electric field , and does not include the coulomb attraction of pairs due to the homogeneity of produced pairs .
in this paper we recapitulate the schwinger effect in curved spacetimes , such as a de sitter @xmath0 space , an anti - de sitter @xmath1 space , and an extremal or near - extremal reissner - nordstrm ( rn ) black hole .
one motivation for studying the schwinger effect in @xmath2 is to unveil the interplay between the maxwell theory as a @xmath3 gauge and the quantum gravity effect at one - loop level .
another motivation to study qed in @xmath2 is the near - horizon geometry of a near - extremal black hole@xcite and the scalar @xmath4-wave in the nariai - geometry of a rotating black hole,@xcite which are summarized in table [ ta2 ] .
further , it would be interesting to investigate whether charged black holes may have the schwinger effect different from hawking radiation .
we also provide a thermal interpretation of the schwinger effect in @xmath5 , which has recently been introduced by cai and kim@xcite and is a qed analog of the unruh effect in @xmath5.@xcite the schwinger effect from an extremal rn black hole has the same spectrum as in @xmath6 since the near - horizon geometry of the extremal black hole is @xmath7 as shown in table [ ta2 ] . the schwinger formula from the extremal black hole is similarly given a thermal interpretation.@xcite interestingly , the schwinger effect from a near - extremal black hole is factorized into the schwinger formula in @xmath6 and that in @xmath8 with the acceleration of the surface gravity due to small non - extremality of black hole .
we find the schwinger formula in @xmath9 in any dimension .
a passing remark is that the holographic schwinger effect is the particle picture of charged pairs including the coulomb attraction of the pairs.@xcite
we consider the planar coordinates for a @xmath11-dimensional ds space , @xmath12 , @xmath13 the electromagnetic field in a curved spacetime measured in a local frame @xmath14 is given by the two - form tensor @xmath15 .
we assume a constant electric field along the @xmath16-direction measured by a co - moving observer with the @xmath11-velocity @xmath17 .
then , @xmath18 in the local frame , @xmath19 dx^1 $ ] and @xmath20 in the metric ( [ ds met ] ) .
thus , the vector potential is given by @xmath21 , which has the minkowski limit @xmath22 .
first , in @xmath10 the schwinger formula ( mean number ) for charged spinless scalars is given by the dimensionless instanton action@xcite @xmath23 the schwinger formula ( [ ds sch ] ) can be interpreted in terms of the effective temperature introduced by cai and kim@xcite @xmath24 where @xmath25 is the unruh temperature for accelerating charge , @xmath26 is the gibbons - hawking temperature,@xcite and @xmath27 is the effective mass in @xmath10 , which are respectively @xmath28 it is interesting to compare the effective temperature ( [ ck ds tem ] ) with the effective temperature for an accelerating observer in @xmath10@xcite @xmath29 by solving the field equation and using the bogoliubov transformation , the pair - production rate is found@xcite @xmath30 where the upper ( lower ) sign is for scalars ( fermions ) and the dimensionless instanton actions are @xmath31 } , \quad s_{\lambda } = 2 \pi \frac{qe}{h^2}. \label{ds2 act}\end{aligned}\ ] ] the prefactor @xmath32 is the density of states and the square bracket in eq .
( [ ds2 act ] ) is present only for scalars but vanishes for fermions .
noting @xmath33 , the leading term of eq .
( [ ds2 sch ] ) is the schwinger formula ( [ ds sch ] ) from the instanton action . without the density of states , the quantity in eq .
( [ ds2 sch ] ) is the mean number of created pairs . in the in - out formalism
, the vacuum persistence amplitude ( integrated action density ) is related to the mean number of pairs as @xmath34 , \label{vac per}\end{aligned}\ ] ] which is the pressure from quantum gas .
the first logarithm in eq .
( [ vac per ] ) is the standard qed action with the mean number @xmath35 while the second one is a correction due to a charged vacuum in @xmath9 and has the character of spinor qed regardless of spins . employing the gamma - function regularization in ref . , the qed action density for scalars is@xcite @xmath36.\end{aligned}\ ] ] similarly , we find the qed action density for fermions @xmath37 here , @xmath38 denotes the principal value and the subtracted terms renormalize the vacuum energy and the charge .
the bosonic and fermionic current density is given by @xmath39 note that the current density comes from the second quantized field theory in curved spacetime , which is equivalent to @xmath40 .
in the limit of @xmath41 , i.e. , @xmath42 and @xmath43 , the current density accumulated over the hubble time @xmath44 increases as @xmath45 and exhibits the infrared hyperconductivity , which has been observed in ref . .
the schwinger effect in @xmath46 has been studied in refs . .
we now study the schwinger effect in @xmath11 dimensions , in which the field equation for charged scalars takes the form @xmath47 where @xmath48 and @xmath49 is the @xmath50-dimensional momentum transverse to the electric field .
then , the positive and negative frequency solutions for the in - vacuum ( @xmath51 ) are given by the whittaker function as @xmath52 where @xmath53 the positive and negative frequency solutions for the out - vacuum ( @xmath54 ) are given by the confluent hypergeometric function as @xmath55 the schwinger effect in a constant electric field in @xmath12 should be independent of @xmath56 and @xmath57 due the symmetry of the spacetime and the field .
in fact , the integration of the longitudinal momentum @xmath58 of the out - vacuum solution @xmath59 gives the density of states @xmath60 independent of dimensions.@xcite therefore , the pair - production rate per spacetime volume is @xmath61 where the upper ( lower ) sign is for charged scalars ( fermions ) and the dimensionless instanton actions are @xmath62 } , \quad s_{\lambda } = 2 \pi \frac{qe}{h^2 } \bigl ( \frac{\frac{qe}{h}}{\sqrt { \bigl ( \frac{qe}{h } \bigr)^2 + { \bf k}_{\perp}^2 } } \bigr ) .
\label{d - ds ins}\end{aligned}\ ] ] here , the square bracket in eq .
( [ d - ds ins ] ) vanishes for fermions . in the case of @xmath10 ,
the transverse momentum vanishes and the integration becomes unity and the schwinger formula ( [ ds2 sch ] ) is recovered .
further , in the limit of the minkowski spacetime ( @xmath22 ) , eq .
( [ d - ds ] ) recovers the schwinger formula ( @xmath63 in table [ ta1 ] ) @xmath64 finally , in the limit of the zero field @xmath65 , eq .
( [ d - ds ] ) reduces to the pure ds radiation for scalars ( fermions ) in the planar coordinates @xmath66}.\end{aligned}\ ] ]
the two - dimensional @xmath67 space has a constant curvature @xmath68 , the planar coordinates and the coulomb potential for a constant electric field @xmath69 the schwinger formula for charged scalars is given by another dimensionless instanton action@xcite @xmath70 which is an analytical continuation of the instanton action ( [ ds sch ] ) in @xmath10 . employing the effective temperature by cai and kim@xcite
, one may give a thermal interpretation for the schwinger formula ( [ ads sch ] ) as @xmath71 where the unruh temperature for accelerating charge and the effective mass in @xmath5 are @xmath72 the binding nature of a pair in the @xmath67 space increases the effective mass while the @xmath9 space intrinsically separates the pair and decreases the effective mass .
similarly , one obtains the density of states and the pair - production rate in @xmath6 @xmath73 where the upper ( lower ) sign is for charged scalars ( fermions ) and the dimensionless instanton actions are @xmath74}.\end{aligned}\ ] ] the square bracket vanishes for fermions . note that pairs are produced when the breitenlohner - freedman stability bound @xmath75 is violated .
the qed action density is for scalars @xmath76 and for fermions @xmath77 the bosonic and fermionic current density is given by @xmath78
the rn black hole with the mass @xmath79 and charge @xmath80 has the metric and the coulomb potential @xmath81 which has the outer and inner horizons @xmath82 as summarized in table [ ta2 ] , the near - horizon geometry of the rn black hole ( [ rn bh ] ) is approximately given by the rindler space , @xmath83 when @xmath84 while it is accurately approximated by the @xmath67 space , @xmath7 when @xmath85 .
interestingly , in the tortoise coordinate @xmath86 for an extremal black hole @xmath87 , the radial motion of a charged scalar in the spherical harmonics @xmath88,@xcite @xmath89 \phi_{l } = 0\end{aligned}\ ] ] has in the asymptotic region @xmath90 the effective potential for the klein - gordon equation in the coulomb potential with an angular momentum @xmath91 and a charge @xmath92 @xmath93 where @xmath94 thus , the schwinger formula from the extremal black hole is approximately given by that from a supercritical point charge in the minkowski spacetime . in the tunneling picture for hawking radiation ,
virtual pairs are created through the horizon .
thus , the near - horizon geometry of the black hole plays an important role in understanding the schwinger effect or hawking radiation . for a non - extremal black hole
, the near - horizon geometry has the rindler coordinates@xcite @xmath95 where @xmath96 and @xmath97 near the horizon @xmath98 with the surface gravity @xmath99 . a physically intuitive way to understand the pair production from the background field is to investigate the poles of the hamilton - jacobi equation for a charged particle .
the hamilton - jacobi action for each spherical harmonics @xmath100 can give the particle production as the decay rate of the vacuum . in the phase - integral method , the mean number is given by the leading term@xcite @xmath101 where @xmath102 is the action evaluated along a contour @xmath103 in the complex plane of @xmath104 .
the simple pole at the outer horizon @xmath105 recovers hawking radiation of charged particles @xmath106 where the hawking temperature and the electrostatic potential on the horizon are @xmath107
as shown in table [ ta2 ] , the near - horizon geometry of a near - extremal black hole has @xmath7 with the metric@xcite @xmath108 where coordinates are stretched by smallness parameter @xmath109
as @xmath110 the hawking radiation for the parametrization ( [ bh par ] ) is given by @xmath111 except for particles with energy close to the chemical potential , i.e , @xmath112 , hawking radiation is exponentially suppressed to zero . by taking the limit @xmath113 the near - horizon region can be stretched infinitely to @xmath7 .
this procedure differs from the amplified near - horizon geometry of a non - extremal black hole as @xmath83 .
this means that the schwinger effect from @xmath7 is very accurate for the near - extremal black hole .
the action for the hamilton - jacobi equation for the radial motion after separating the spherical harmonics is given by @xmath114 note that there are two finite poles at @xmath115 , the outer horizon , and @xmath116 , the inner horizon as well as an infinite pole at @xmath117 .
then , the contour integral gives the leading term for the schwinger effect@xcite @xmath118 where @xmath119 is the sum of residues at @xmath120 and @xmath121 is the residue at @xmath117 . from the field equation , the exact schwinger formula from the geometry ( [ near geom ] )
is given by@xcite @xmath122 where @xmath123 .
the mean number for the schwinger effect has the thermal interpretation@xcite @xmath124 where @xmath125 with @xmath126 , the unruh temperature for accelerating charge on the horizon .
note that @xmath127 and @xmath128 are the temperature for the schwinger effect in @xmath6 in sec .
[ qed ads2 ] .
the schwinger effect from the near - extremal black hole is the product of the schwinger effect in @xmath6 and a correction due to the hawking radiation from the non - extremality .
it has been further observed that the schwinger effect has the factorization@xcite @xmath129 the reason for the schwinger effect in @xmath130 is that the measure of the near - extremality is in fact a measure for the rindler space .
the schwinger effect of scalar qed in the rindler space has been studied in ref . .
we have studied the spontaneous creation of charged pairs from a constant electric field in @xmath2 and from a charged rn black hole , which is a nonperturbative quantum effect at one - loop level .
the qed action in @xmath2 exhibits the interplay of qed and the quantum gravity effect .
the schwinger formula has a thermal interpretation of the unruh temperature for accelerating charge and the gibbons - hawking temperature for @xmath10 and the corresponding curvature for @xmath6 .
we have found the schwinger formula in @xmath9 space of any dimension , which has the correct limit of the minkowski spacetime and the pure @xmath9 space .
the near - horizon geometry of charged rn black hole is @xmath7 for an extremal black hole while it is @xmath83 for a non - extremal black hole .
thus , the schwinger effect for extremal black hole has the same thermal interpretation as in @xmath6 .
the near - extremal black hole has , however , an additional schwinger effect in @xmath130 .
the author would like to thank rong - gen cai , chiang - mei chen , w - y .
pauchy hwang , hyun kyu lee , don n. page , christian schubert , and yongsung yoon for useful discussions and pisin chen for warm hospitality during the second lecospa symposium .
the participation of lecospa symposium was supported by asia pacific center for theoretical physics ( apctp ) .
this research was supported by basic science research program through the national research foundation of korea(nrf ) funded by the ministry of education ( 15b15770630 ) . | we revisit the schwinger effect in de sitter , anti - de sitter spaces and charged black holes , and explore the interplay between quantum electrodynamics and the quantum gravity effect at one - loop level .
we then advance a thermal interpretation of the schwinger effect in curved spacetimes .
finally , we show that the schwinger effect in a near - extremal black hole differs from hawking radiation of charged particles in a non - extremal black hole and is factorized into those in an anti - de sitter space and a rindler space with the surface gravity for acceleration . |
we consider canonical hamiltonian systems in the form @xmath2 where @xmath3 is a smooth real - valued function .
our interest is in researching numerical methods that provide approximations @xmath4 to the true solution along which the energy is precisely conserved , namely @xmath5 the study of energy - preserving methods form a branch of _ geometrical numerical integration _ , a research topic whose main aim is preserving qualitative features of simulated differential equations . in this context ,
symplectic methods have had considerable attention due to their good long - time behavior as compared to standard methods for odes @xcite . a related interesting approach based upon
exponential / trigonometric fitting may be found in @xcite .
unfortunately , symplecticity can not be fully combined with the energy preservation property @xcite , and this partly explains why the latter has been absent from the scene for a long time . among the first examples of energy - preserving methods we mention discrete gradient schemes @xcite which are defined by devising discrete analogs of the gradient function . the first formulae in this class had order at most two but recently discrete gradient methods of arbitrarily high order have been researched by considering the simpler case of systems with one - degree of freedom @xcite . here
, the key tool we wish to exploit is the well - known line integral associated with conservative vector fields , such us the one defined at , as well as its discrete version , the so called _ discrete line integral_. interestingly , the line integral provides a means to check the energy conservation property , namely @xmath6 & = h\displaystyle \int_0 ^ 1 \nabla^t h(y(t_0+\tau h ) ) j^t \nabla h(y(t_0+\tau h ) ) \d \tau = 0 , \end{array}\ ] ] with @xmath7 , that can be easily converted into a discrete analog by considering a quadrature formula in place of the integral .
the discretization process requires to change the curve @xmath8 in the phase space @xmath9 to a simpler curve @xmath10 ( generally but not necessarily a polynomial ) , which is meant to yield the approximation at time @xmath11 , that is @xmath12 , where @xmath13 is the order of the resulting numerical method . in a certain sense
, the problem of numerically solving while preserving the hamiltonian function is translated into a quadrature problem .
for example , consider the segment @xmath14 , with @xmath15 $ ] , joining @xmath16 to an unknown point @xmath17 of the phase space .
the line integral of @xmath18 evaluated along @xmath19 becomes @xmath20 now assume that @xmath21 is a polynomial of degree @xmath22 in the generalized coordinates @xmath23 and in the momenta @xmath13 .
the integrand in is a polynomial of degree @xmath24 in @xmath25 and can be exactly solved by any quadrature formula with abscissae @xmath26 in @xmath27 $ ] and weights @xmath28 , having degree of precision @xmath29 .
we thus obtain @xmath30 to get the energy conservation property we impose that @xmath31 be orthogonal to the above sum , and in particular we choose ( for the sake of generality we use @xmath32 in place of @xmath33 to mean that the resulting method also makes sense when applied to a general ordinary differential equation @xmath34 ) @xmath35 formula defines a runge kutta method with butcher tableau @xmath36 , where @xmath25 and @xmath37 are the vectors of the abscissae and weights , respectively .
the stages @xmath38 are called _ silent stages _ since their presence does not affect the degree of nonlinearity of the system to be solved at each step of the integration procedure : the only unknown is @xmath17 and consequently defines a mono - implicit method .
mono - implicit methods of runge kutta type have been researched in the past by several authors ( see , for example , @xcite for their use in the solution of initial value problems ) .
methods such as date back to 2007 @xcite and are called @xmath39-stage trapezoidal methods since on the one hand the choice @xmath40 , @xmath41 , @xmath42 leads to the trapezoidal method and on the other hand all other methods evidently become the trapezoidal method when applied to linear problems .
generalizations of to higher orders require the use of a polynomial @xmath19 of higher degree and are based upon the same reasoning as the one discussed above . up to now
, such extensions have taken the form of runge kutta methods @xcite .
it has been shown that choosing a proper polynomial @xmath19 of degree @xmath43 yields a runge kutta method of order @xmath44 with @xmath45 stages .
the peculiarity of such energy - preserving formulae , called hamiltonian boundary value methods ( hbvms ) , is that the associated butcher matrix has rank @xmath43 rather than @xmath39 , since @xmath46 stages may be cast as linear combinations of the remaining ones , similarly to the stages @xmath38 in . as a consequence ,
the nonlinear system to be solved at each step has dimension @xmath47 instead of @xmath48 , which is better visualized by recasting the method in block - bvm form @xcite . in the case where @xmath3 is not a polynomial ,
one can still get a _
practical _ energy conservation by choosing @xmath39 large enough so that the quadrature formula approximates the corresponding integral to within machine precision . strictly speaking , taking the limit as @xmath49 leads to limit formulae where the integrals come back into play in place of the sums .
for example , letting @xmath50 in just means that the integral in must not be discretized at all , which would yield the _ averaged vector field _
method @xmath51 , ( see @xcite for details ) . in this paper
we start an investigation that follows a different route . unlike the case with hbvms
, we want now to take advantage of the previously computed approximations to extend the class in such a way to increase the order of the resulting methods , much as the class of linear multistep method may be viewed as a generalization of ( linear ) one step methods .
the general question we want to address is whether there exist @xmath39-step mono - implicit energy - preserving methods of order greater than two .
clearly , the main motivation is to reduce the computational cost associated with the implementation of hbvms .
the purpose of the present paper is to give an affermative answer to this issue in the case @xmath40 .
more specifically , the method resulting from our analysis , summarized by formula , may be thought of as a nearly linear two - step method in that it is the sum of a fourth order linear two - step method , formula , plus a nonlinear correction of higher order .
the paper is organized as follows . in section [ def_methods ]
we introduce the general formulation of the method , by which we mean that the integrals are initially not discretized to maintain the theory at a general level . in this section
we also report a brief description of the hbvm of order four , since its properties will be later exploited to deduce the order of the new method : this will be the subject of section [ analysis_sec ] .
section [ discr_sec ] is devoted to the discretization of the integrals , which will produce the final form of the methods making them ready for implementation .
a few test problems are presented in section [ test_sec ] to confirm the theoretical results .
suppose that @xmath17 is an approximation to the true solution @xmath8 at time @xmath11 , where @xmath52 is the stepsize of integration .
more precisely , we assume that * @xmath53 with @xmath54 ; * @xmath55 , which means that @xmath17 lies on the very same manifold @xmath56 as the continuous solution @xmath8 .
the two above assumptions are fulfilled if , for example , we compute @xmath17 by means of a hbvm ( or an @xmath57-hbvm @xcite ) of order @xmath58 .
the new approximation @xmath59 is constructed as follows .
consider the quadratic polynomial @xmath60 that interpolates the set of data @xmath61 .
expanded along the newton basis @xmath62 defined on the nodes @xmath63 , @xmath64 , @xmath65 , the polynomial @xmath19 takes the form ( for convenience we order the nodes as @xmath66 ) @xmath67 as @xmath68 ranges in the interval @xmath27 $ ] , the @xmath69-length vector @xmath70 describes a curve in the phase space @xmath9 . the line integral of the conservative vector field @xmath18 along the curve @xmath71 will match the variation of the energy function @xmath3 , that is @xmath72^t \nabla h(\gamma(\tau ) ) \ , \d \tau \\[.5 cm ] & \hspace*{-1.5 cm } \displaystyle = ( y_2-y_0)^t \int_0 ^ 1 \nabla h(\gamma(\tau ) ) \ , \mathrm{d}\tau + 2(y_2 - 2y_1+y_0)^t \int_0 ^ 1 ( 2\tau-1 ) \nabla h(\gamma(\tau ) ) \ , \mathrm{d}\tau .
\end{array}\ ] ] the energy conservation condition @xmath73 yields the following equation in the unknown @xmath74 @xmath75 the method we are interested in has the form @xmath76 , where @xmath77 is implicitly defined by the following nonlinear equation in the unknown @xmath78 : @xmath79 where the residual @xmath80 is defined as @xmath81 a direct computation shows that any solution @xmath82 of also satisfies . in the next section we will show that admits a unique solution @xmath83 satisfying the order condition @xmath84
. such a result will be derived by regarding as a perturbation of the hbvm of order @xmath85 and , in turn , by comparing the two associated numerical solutions . to this end and to better explain the genesis of formula and the role of the integrals therein , a brief introduction of the hbvm formula of order four is in order .
suppose that both @xmath17 and @xmath86 are unknown ( so now @xmath17 is no longer given a priori as indicated by assumption ( @xmath87 ) ) : let us call them @xmath88 and @xmath89 respectively .
for to be satisfied , we can impose the two orthogonality conditions @xmath90 \displaystyle u_2 - 2u_1+y_0 = \eta_2 h j \int_0 ^ 1 ( 2\tau-1 ) \nabla h(\gamma(\tau ) ) \ , \mathrm{d}\tau , \end{array } \right.\ ] ] giving rise to a system of two block - equations ( the curve @xmath91 is as in with @xmath88 and @xmath89 in place of @xmath17 and @xmath86 ) . setting the free constants @xmath92 and @xmath93 equal to @xmath94 and @xmath95 ,
respectively , confers the highest possible order , namely @xmath85 , on the resulting method : @xmath96 ( see @xcite for details).$ ] of length @xmath97 , we have scaled the constants @xmath92 and @xmath93 by a factor two with respect to the values reported in @xcite . ]
furthermore , it may be shown that the internal stage @xmath88 satisfies the order condition @xmath98 .
evidently , the implementation of on a computer can not leave out of consideration the issue of solving the integrals appearing in both equations .
two different situations may emerge : * the hamiltonian function @xmath3 is a polynomial of degree @xmath22 . in such a case ,
the two integrals in are exactly computed by a quadrature formula having degree of precision @xmath99 .
* @xmath3 is not a polynomial , nor do the two integrands admit a primitive function in closed form .
again , an appropriate quadrature formula can be used to approximate the two integrals to within machine precision , so that no substantial difference is expected during the implementation process by replacing the integrals by their discrete counterparts .
case ( a ) gives rise to an infinite family of runge - kutta methods , each depending on the specific choice ( number and distribution ) of nodes the quadrature formula is based upon ( see @xcite for a general introduction on hbvms and @xcite for their relation with standard collocation methods ) . for example , choosing @xmath39 nodes according to a gauss distribution over the interval @xmath27 $ ] results in a method that precisely conserves the energy if applied to polynomial canonical hamiltonian systems with @xmath100 and that becomes the classical @xmath94-stage gauss collocation method when @xmath40 . on the other hand , choosing a lobatto distribution yields a runge - kutta method that preserves polynomial hamiltonian functions of degree @xmath101 and that becomes the lobatto iiia method of order four when @xmath40 .
the method resulting from case ( b ) are undistinguishable from the original formulae in that they are energy - preserving up to machine precision when applied to any regular canonical hamiltonian system
. stated differently , may be viewed as the limit of the family of hbvms of order four , as the number of nodes tends to infinity .
for this reason the limit formulae have been called @xmath57-hbvm of order @xmath85 ( see @xcite ) . in the present context ,
@xmath17 being a known quantity , the unknown @xmath78 in can not in general satisfy , at the same time , both orthogonality conditions in .
however , since @xmath17 may be thought of as an approximation of order four to the quantity @xmath88 in , should we only impose the first orthogonality condition , namely @xmath102 we would expect the residual @xmath80 ( the right hand side of ) to be very small .
below , it is not difficult to show that actually implies @xmath103 .
this aspect is further emphasized in the numerical test section ( see table [ tab2 ] ) . ]
this suggests that a solution to that yields an approximation of high order to @xmath104 may be obtained by allowing a small deviation from orthogonality in .
this is accomplished by setting @xmath105 , and by tuning the perturbation parameter @xmath106 in such a way that be satisfied : this evidently gives @xmath107 and we arrive at .
results on the existence and uniqueness of a solution of as well as on its order of accuracy will be derived by first analyzing the simpler nonlinear system @xmath108 obtained by neglecting the correction term @xmath109 .
for @xmath110 we set ( see ) @xmath111 and ( see ) @xmath112 in the following @xmath113 will denote the @xmath94-norm .
[ lem1 ] there exist positive constants @xmath114 and @xmath115 such that , for @xmath116 , system admits a unique solution @xmath117 in the ball @xmath118 of center @xmath16 and radius @xmath114 .
we show that constants @xmath119 exist such that the function defined in satisfies the following two conditions for @xmath120 : * @xmath121 is a contraction on @xmath118 , namely @xmath122 * @xmath123 .
the contraction mapping theorem can then be applied to obtain the assertion .
let @xmath118 a ball centered at @xmath16 with radius @xmath114 .
we can choose @xmath124 and @xmath114 small enough that the image set @xmath125,~ z\in b(y_0,\rho),~h\le h'_0\}$ ] is entirely contained in a ball @xmath126 which , in turn , is contained in the domain of @xmath127 .
is an open simply connected subset of @xmath9 containing @xmath118 while , from the assumption ( @xmath87 ) , decreasing @xmath1 causes the point @xmath17 to approach @xmath16 . ]
we set @xmath128 from and we have @xmath129 and hence @xmath130 consequently ( a ) is satisfied by choosing @xmath131 and @xmath132 . concerning ( b ) ,
we observe that @xmath133 hence @xmath134 with @xmath135 bounded with respect to @xmath1 . since @xmath136 vanishes with @xmath1 ( see ( [ lcond ] ) ) , we can always tune @xmath115 in such a way that @xmath137 .
[ lem2 ] the solution @xmath117 of satisfies @xmath138 . under the assumption ( @xmath87 ) ,
may be regarded as a perturbation of system , since @xmath17 and @xmath88 are @xmath0 and @xmath139 close to @xmath140 respectively .. ] since @xmath141 , we can estimate the accuracy of @xmath117 as an approximation of @xmath142 by evaluating its distance from @xmath89 .
let @xmath143 be the underlying quadratic curve associated with the hbvm defined by , namely @xmath144 considering that ( see ) @xmath145 from the first equation in and we get @xmath146 & \displaystyle + 8hj \int_0 ^ 1 \nabla^2 h(\tilde \gamma(\tau))\tau(\tau-1 ) \,\d
\tau \cdot ( u_1-y_1 ) + o(||u_1-y_1||^2 ) \\[.4 cm ] = & \displaystyle u_2+o(h^5 ) .
\end{array}\ ] ] if @xmath1 is small enough , @xmath89 will be inside the ball @xmath118 defined in lemma [ lem1 ] . the lipschitz condition yields ( see ) @xmath147 and hence @xmath148 .
the above result states that defines a method of order @xmath85 which is a simplified ( non corrected ) version of our conservative method defined at . in section [ test_sec ] the behavior of these two methods will be compared on a set of test problems .
we now state the analogous results for system .
[ th1 ] under the assumption ( @xmath87 ) , for @xmath1 small enough , equation admits a unique solution @xmath82 satisfying @xmath149 .
consider the solution @xmath117 of system .
we have ( see ) @xmath150 and @xmath151 hence , by virtue of , @xmath152^t \left[\int_0 ^ 1 ( 2\tau-1)\nabla h(\tilde \gamma(\tau ) ) \ , \d \tau + o(h^5)\right]=o(h^5).\ ] ] since @xmath153 is bounded with respect to @xmath1 , it follows that , in a neighborhood of @xmath117 , system may be regarded as a perturbation of system , the perturbation term being @xmath154 .
consider the ball @xmath155 : since @xmath156 , and @xmath157 , this ball is contained in @xmath158 defined in lemma [ lem1 ] and the perturbed function @xmath159 is a contraction therein , provided @xmath1 is small enough .
evaluating the right - hand side of at @xmath160 we get @xmath161 which means that property ( b ) listed in the proof of lemma [ lem1 ] , with @xmath117 in place of @xmath16 , holds true for the perturbed function @xmath162 , and the contraction mapping theorem may be again exploited to deduce the assertion .
as was stressed in section [ def_methods ] , formula is not operative unless a technique to solve the two integrals is taken into account .
the most obvious choice is to compute the integrals by means of a suitable quadrature formula which may be assumed exact in the case where the hamiltonian function is a polynomial , and to provide an approximation to within machine precision in all other cases .
hereafter we assume that @xmath163 is a polynomial in @xmath23 and @xmath13 of degree @xmath22 .
since @xmath164 has degree two , it follows that the integrand functions appearing in the definitions of @xmath165 and @xmath80 at and have degree @xmath166 and @xmath167 respectively and can be solved by any quadrature formula with abscissae @xmath26 in @xmath27 $ ] and weights @xmath28 , having degree of precision @xmath168 . in place of we now consider the equivalent form suitable for implementation @xmath169 where @xmath170 notice that from we get @xmath171 that is , @xmath172 is a linear combination , actually a weighted average , of the approximations @xmath16 , @xmath17 and @xmath86 .
therefore , since @xmath173 ( see lemma [ lem2 ] and theorem [ th1 ] ) , we may look at this term as a nonlinear correction of the generalized linear multistep method @xmath174 if @xmath163 is quadratic , we can choose @xmath175 , @xmath41 , @xmath176 , @xmath177 , @xmath178 and @xmath179 , that is we can use simpson s quadrature formula to compute the integrals in and . since , in such a case , @xmath180 , method becomes @xmath181 that is , the standard milne - simpson s method . in all other cases
@xmath172 will differ in general from @xmath182 , @xmath183 and may be regarded as an off - point entry in formula . in the sequel
we will denote the method defined at by @xmath184 and its linear part , defined at , by @xmath185 .
of course , the choice of the abscissae distribution influences the energy preserving properties of the method @xmath184 , as is indicated in table [ nodes - distribution - table ] .
.energy preserving properties of method @xmath184 for some well - known distributions of the nodes @xmath186 . [ cols="^,^,^,^",options="header " , ] we finally consider the non - polynomial hamiltonian function @xmath187 that defines the well known kepler problem , namely the motion of two masses under the action of their mutual gravitational attraction .
taking as initial condition @xmath188 yields an elliptic periodic orbit of period @xmath189 and eccentricity @xmath190 .
we have chosen @xmath191 .
though the vector field fails to be a polynomial in @xmath192 and @xmath193 , we can plan to use a sufficiently large number of quadrature nodes to discretize the integrals in so that the corresponding accuracy is within the machine precision . under this assumption , and taking aside the effect of the floating point arithmetic , the computer will make no difference between the conservative formulae and their discrete counterparts .
the left picture in figure [ kepler_fig1 ] explains the above argument .
it reports the error @xmath194 in the hamiltonian function for various choices of the number of lobatto nodes , and precisely @xmath195 .
we see that the error decreases quickly as the number of nodes is incremented and for @xmath196 it is within the epsilon machine .
the use of finite arithmetic may sometimes cause a mild numerical drift of the energy over long times , like the one shown in the upper line in the right picture of figure [ kepler_fig1 ] .
this is due to the fact that on a computer the numerical solution satisfy the conservation relation @xmath197 up to machine precision times the conditioning number of the nonlinear system that is to be solved at each step . to prevent the accumulation of roundoff errors
we may apply a simple and costless _ correction _ technique on the approximation @xmath198 which consists in a single step of a gradient descent method ( see also @xcite ) .
more precisely , the corrected solution @xmath199 is defined by @xmath200 which stems from choosing as @xmath201 the value that minimizes the linear part of the function @xmath202 .
the bottom line in the right picture of figure [ kepler_fig1 ] shows the energy conservation property of the corrected solution . produced by methods @xmath184 , with @xmath195 .
parameters : stepsize @xmath203 , integration interval @xmath204 $ ] .
right picture .
roundoff errors may cause a drift of the numerical hamiltonian function ( upper line ) which can be easily taken under control by coupling the method with a costless correction procedure like the one described at .,title="fig:",width=253,height=188 ] produced by methods @xmath184 , with @xmath195 .
parameters : stepsize @xmath203 , integration interval @xmath204 $ ] .
right picture .
roundoff errors may cause a drift of the numerical hamiltonian function ( upper line ) which can be easily taken under control by coupling the method with a costless correction procedure like the one described at .,title="fig:",width=253,height=188 ]
we have derived a family of mono - implicit methods of order four with energy - preserving properties .
each element in the family originates from a limit formula and is defined by discretizing the integral therein by means of a suitable quadrature scheme .
this process assures an exact energy conservation in the case where the hamiltonian function is a polynomial , or a conservation to within machine precision in all other cases , as is also illustrated in the numerical tests .
interestingly , each method may be conceived as a @xmath0 perturbation of a two - step linear method .
l. brugnano , f. iavernaro and d. trigiante , _ analisys of hamiltonian boundary value methods ( hbvms ) : a class of energy - preserving runge - kutta methods for the numerical solution of polynomial hamiltonian dynamical systems _ , ( 2009 ) ( submitted ) ( arxiv:0909.5659 [ arxiv:0909.5659 ] ) .
l. brugnano , f. iavernaro and d. trigiante , _ hamiltonian boundary value methods ( energy preserving discrete line integral methods ) _ , jour . of numer .
industr . and
appl . math . * 5 * , 12 ( 2010 ) , 1737 ( arxiv:0910.3621 [ arxiv:0910.3621 ] ) .
l. brugnano , f. iavernaro , d. trigiante , _ the lack of continuity and the role of infinite and infnitesimal in numerical methods for odes : the case of symplecticity _ , applied mathematics and computation ( to appear ) , doi : 10.1016/j.amc.2011.03.022[doi : 10.1016/j.amc.2011.03.022 ] , ( arxiv:1010.4538 ) . | we introduce a family of fourth order two - step methods that preserve the energy function of canonical polynomial hamiltonian systems .
each method in the family may be viewed as a correction of a linear two - step method , where the correction term is @xmath0 ( @xmath1 is the stepsize of integration ) .
the key tools the new methods are based upon are the line integral associated with a conservative vector field ( such as the one defined by a hamiltonian dynamical system ) and its discretization obtained by the aid of a quadrature formula .
energy conservation is equivalent to the requirement that the quadrature is exact , which turns out to be always the case in the event that the hamiltonian function is a polynomial and the degree of precision of the quadrature formula is high enough .
the non - polynomial case is also discussed and a number of test problems are finally presented in order to compare the behavior of the new methods to the theoretical results . ordinary differential equations , mono - implicit methods , multistep methods , canonical hamiltonian problems , hamiltonian boundary value methods , energy preserving methods , energy drift .
65l05 , 65p10 . |
in 1978 fisher and de gennes @xcite realized that there should be a so called `` thermodynamic '' casimir effect .
this means that a force emerges when thermal fluctuations are restricted by a container .
thermal fluctuations extend to large scales in the neighbourhood of critical points . in the thermodynamic limit ,
in the neighbourhood of the critical point , various quantities diverge following power laws . e.g.
the correlation length , which measures the spatial extension of fluctuations , behaves as @xmath7 where @xmath8 is the reduced temperature and @xmath9 the critical temperature . @xmath10 and @xmath11 are the amplitude of the correlation length in the high and low temperature phase , respectively . while @xmath10 and @xmath11 depend on the microscopic details of the system , the critical exponent @xmath12 and the ratio @xmath13 are universal .
this means that they assume exactly the same values for all systems within a given universality class .
a universality class is characterized by the spatial dimension of the system , the range of the interaction and the symmetry of the order parameter .
the modern theory of critical phenomena is the renormalization group ( rg ) .
for reviews see e.g. @xcite . here
we consider the xy universality class in three dimensions with short range interactions .
this universality class is of particular interest , since the @xmath6-transition of @xmath5he is supposed to share this universality class .
the most accurate experimental results for critical exponents and universal amplitude ratios for a three dimensional system have been obtained for this transition ; for a review see @xcite .
the critical behaviour is modified by a confining geometry .
if the system is finite in all directions , thermodynamic functions have to be analytic .
i.e. a singular behaviour like eq .
( [ xipower ] ) is excluded . as a remnant of such singularities there remains a peak in the neighbourhood of the transition . with increasing linear extension the hight of the peak increases and the temperature of the maximum approaches the critical temperature .
this behaviour is described by the theory of finite size scaling ( fss ) .
for reviews see @xcite . in general the physics in the neighbourhood of the transition is governed by the ratio @xmath14 , where @xmath15 is the linear extension of the container and @xmath16 the correlation length of the bulk system .
furthermore it depends on the geometry of the container and on the type of the boundary conditions that the container imposes on the order parameter . for a review on experimental studies of @xmath5he near the @xmath6-transition in confining geometries
see @xcite . here
we study thin films .
thin films are finite in one direction and infinite in the other two directions . in this case
singular behaviour is still possible .
however the associated phase transition belongs to the two - dimensional universality class .
i.e. in the case of @xmath17 symmetry , a kosterlitz - thouless ( kt ) transition @xcite is expected . in @xcite
we have confirmed the kt - nature of this transition and have studied the scaling of the transition temperature with the thickness of the film . recently @xcite we determined the finite size scaling behaviour of the specific heat of thin films .
here we investigate the thermodynamic casimir force in thin films in the three dimensional xy universality class . from a thermodynamic point of view
, the casimir force per unit area is given by @xmath18 where @xmath15 is the thickness of the film and @xmath19 is the excess free energy per area of the film , where @xmath20 is the free energy per area of the film and @xmath21 the free energy density of the thermodynamic limit of the three dimensional system ; see e.g. @xcite .
finite size scaling predicts that the casimir force behaves as @xmath22^{1/\nu})\ ] ] where @xmath23 is a universal finite size scaling function . )
we shall use @xmath24 in the following . ] in @xcite @xmath5he films of thicknesses up to 588 @xmath25 have been studied .
these experiments show clearly that the thermodynamic casimir force is indeed present . throughout it
is negative . in the low temperature phase of the three dimensional bulk system it shows a pronounced minimum .
the data are essentially consistent with the prediction eq .
( [ fsscasimir ] ) .
the minimum of @xmath23 is located at @xmath26 . it has been a challenge for theorists to compute the finite size scaling function @xmath23 .
krech and dietrich @xcite have computed it in the high temperature phase using the @xmath27-expansion up to o(@xmath27 ) .
this result is indeed consistent with the measurements on @xmath5he films .
deep in the low temperature phase , the spin wave approximation should provide an exact result .
it predicts a negative non - vanishing value for @xmath23 .
however the experiments suggest a much larger absolute value for @xmath23 in this region . until recently a reliable theoretical prediction for the minimum of @xmath23 and its neighbourhood was missing . using a renormalized mean - field approach
the authors of @xcite have computed @xmath23 for the whole temperature range .
qualitatively they reproduce the features of the experimental result .
however the position of the minimum is by almost a factor of 2 different from the experimental one .
the value at the minimum is wrongly estimated by a factor of about 5 .
only quite recently monte carlo simulations of the xy model on the simple cubic lattice @xcite provided results for @xmath23 which essentially reproduce the experiments on @xmath5he films @xcite .
these simulations were performed with lattices of a thickness up to @xmath28 @xcite and up to @xmath29 @xcite .
the authors of @xcite pointed out that for these lattice sizes corrections to scaling still play an important role .
the purpose of the present work is to get accurate control over the leading corrections to scaling , allowing us to compute @xmath23 with reliable error bars . as the first step in this direction
we simulate the improved two - component @xmath0 model instead of the xy model . for a precise definition of these models
see the section below .
this way we avoid leading corrections to scaling which are @xmath30 with @xmath31 @xcite ; similar values for the exponent are obtained with field - theoretic methods ; for a review see e.g. @xcite .
in order to mimic the vanishing order parameter that is observed at the boundaries of @xmath5he films , dirichlet boundary conditions with vanishing field are imposed . these lead to corrections @xmath32 @xcite .
these corrections can be eliminated by replacing the thickness @xmath15 by an effective one @xmath33 , where @xmath34 @xcite for the model that we have simulated .
by @xmath35 to account for surface corrections , was first discussed by capehart and fisher @xcite in the context of the surface susceptibility of ising films . ]
this paper is organized as follows : first we define the model and the observables that we have measured .
next we discuss the finite size scaling behaviour of the casimir force .
in particular , we discuss corrections to scaling caused by the dirichlet boundary conditions .
we outline the method used to compute the casimir force .
we discuss the simulations that have been performed and analyze our data .
we compare our results with experiments @xcite , previous monte carlo simulations @xcite and the @xmath27-expansion @xcite .
finally we summarize and conclude .
we study the two component @xmath0 model on the simple cubic lattice .
we label the sites of the lattice by @xmath36
. the components of @xmath37 might assume the values @xmath38 .
we simulate lattices of the size @xmath39 and @xmath40 . in 1 and
2-direction we employ periodic boundary conditions and free boundary conditions in 0-direction .
this means that the sites with @xmath41 and @xmath42 have only five nearest neighbours .
this type of boundary conditions could be interpreted as dirichlet boundary conditions with @xmath43 as value of the field at @xmath44 and @xmath45 . note that viewed this way , the thickness of the film is @xmath46 rather than @xmath15 .
this provides a natural explanation of the result @xmath34 obtained in @xcite and might be a good starting point for a field theoretic calculation of @xmath47 .
the hamiltonian of the two component @xmath0 model , for a vanishing external field , is given by @xmath48\ ] ] where the field variable @xmath49 is a vector with two real components .
@xmath50 denotes a pair of nearest neighbour sites on the lattice .
the partition function is given by @xmath51 \
, \exp(-{\cal h}).\ ] ] note that following the conventions of our previous work , e.g. @xcite , we have absorbed the inverse temperature @xmath52 into the hamiltonian . in the limit @xmath53 the field variables are fixed to unit length ; i.e. the xy model is recovered .
for @xmath54 we get the exactly solvable gaussian model . for @xmath55
the model undergoes a second order phase transition that belongs to the xy universality class .
numerically , using monte carlo simulations and high - temperature series expansions , it has been shown that there is a value @xmath56 , where leading corrections to scaling vanish .
numerical estimates of @xmath57 given in the literature are @xmath58 @xcite , @xmath59 @xcite and most recently @xmath60 @xcite .
the inverse of the critical temperature @xmath61 has been determined accurately for several values of @xmath6 using finite size scaling ( fss ) @xcite .
we shall perform our simulations at @xmath62 , since for this value of @xmath6 comprehensive monte carlo studies of the three - dimensional system in the low and the high temperature phase have been performed @xcite . at @xmath62 one gets @xmath63 @xcite .
since @xmath62 is not exactly equal to @xmath57 , there are still corrections @xmath64 , although with a small amplitude . in fact , following @xcite , it should be by at least a factor 20 smaller than for the standard xy model . in @xcite
we find for @xmath65 by fitting the data for the second moment correlation length in the high temperature phase @xmath66 \;\;,\ ] ] where @xmath67 . here
we shall use @xmath68 @xcite as value of the critical exponent of the correlation length .
recent experiments on the @xmath6-transition of @xmath5he suggest a slightly smaller value : @xmath69 @xcite .
this discrepancy is however not crucial for the present study .
note that in the high temperature phase there is little difference between @xmath70 and the exponential correlation length @xmath71 which is defined by the asymptotic decay of the two - point correlation function .
following @xcite : @xmath72 for the thermodynamic limit of the three - dimensional system .
the reduced free energy density is defined as @xmath73 i.e. compared with the free energy density @xmath74 , a factor @xmath75 is skipped .
note that in eq .
( [ hamiltonian ] ) @xmath52 does not multiply the second term . therefore , strictly speaking
, @xmath52 is not the inverse of @xmath75 . in order to study universal quantities
it is not crucial how the transition line in the @xmath52-@xmath6 plane is crossed , as long as this path is not tangent to the transition line .
therefore , following computational convenience , we vary @xmath52 at fixed @xmath6 .
correspondingly we define the ( internal ) energy density as the derivative of the reduced free energy density with respect to @xmath52 .
furthermore , to be consistent with our previous work @xcite , we multiply by @xmath76 : @xmath77 it follows @xmath78 which can be easily determined in monte carlo simulations . from eqs .
( [ fdef1],[edef1 ] ) it follows that the free energy density can be computed as @xmath79
let us discuss the scaling behaviour of the reduced excess free energy . since we study an improved model we ignore corrections @xmath30 in the following .
we take into account leading corrections due to the boundary conditions by replacing the thickness @xmath15 of the film by @xmath80 at the appropriate places .
we split the free energies in singular ( s ) and non - singular ( ns ) parts : @xmath81 where @xmath82 $ ] is a universal finite size scaling function and @xmath83^{1/\nu}$ ] .
following rg theory the non - singular part is not affected by finite size effects .
however it is not clear a priori how dirichlet boundary conditions affect the non - singular part of the free energy .
therefore we allow for @xmath84 and @xmath85 . taking the derivative with respect
to @xmath15 we get the thermodynamic casimir force per area @xcite @xmath86 where @xmath87 .
here we follow essentially the approach of @xcite . for an alternative method
see @xcite . on the lattice
the thickness @xmath15 assumes integer values .
therefore we approximate the derivative for half - integer values of @xmath15 as @xmath88 correspondingly we define @xmath89 where @xmath90 is the energy per area of a thin film and @xmath91 the energy density of the three dimensional system .
analogous to eq .
( [ integratef ] ) we compute @xmath92 where @xmath93 is chosen such that @xmath94 and hence the casimir force vanishes .
in @xcite we have studied the specific heat of thin films in the two component @xmath0 model at @xmath65 . to this end , we had determined the energy density for the three dimensional thermodynamic limit and for films of the thicknesses @xmath95 , @xmath96 and @xmath97 for a large number of @xmath52-values . in order to compute the derivative with respect to @xmath15
, we have complemented these simulations by ones for @xmath98 , @xmath99 and @xmath100 .
the monte carlo algorithm that has been used is the same as in @xcite : an update - cycle is composed of a metropolis sweep , a few overrelaxation sweeps and single @xcite and wall @xcite cluster updates .
one sweep means that a local update is performed at each site of the lattice ones . as random number generator
we have used the simd - oriented fast mersenne twister algorithm @xcite . in table
[ statistics ] we have summarized the statistics of our runs . in total these simulations took about 3 years of cpu - time on a single core of a quad - core opteron(tm ) 2378 cpu ( 2.4 ghz ) .
._[statistics ] we characterize our new simulations . in the first column we give the thickness @xmath15 of the film . in the second column
we give the linear size @xmath101 of the lattice in the other two directions . in the third and fourth column
we give the upper and lower bound of the interval in @xmath52 that has been simulated . in the fifth column we give the step size @xmath102 that we used .
e.g. @xmath103 , @xmath104 and @xmath105 means that @xmath106 , @xmath107 , @xmath108 have been simulated .
finally , in the last column we give the number of measurements ( stat ) that we have performed for each of the simulations .
_ [ cols="^,>,<,<,<,^",options="header " , ] @xmath109 .
we get the @xmath110 , @xmath111 with @xmath112/d.o.f.@xmath113 . as our final result
we take @xmath114 where the error bar covers the statistical error and the uncertainty of @xmath47 .
our result for @xmath115 can be compared with those given in the literature .
the experimental works @xcite give @xmath116 ( no final result for @xmath117 is quoted ) and @xcite @xmath118 and @xmath119 . in @xcite and @xcite
the convention @xmath120 is used . in order to convert to @xmath121^{1/\nu}$ ] we have used @xmath122 for @xmath5he at vapour pressure as discussed in section 4.2 of @xcite . in his monte carlo study of the xy model on the simple cubic lattice , hucht
@xcite finds @xmath123 and @xmath124 .
the authors of @xcite have used two different anstze for the corrections to scaling . using the first one , they arrive at @xmath125 and @xmath126 and using the second one at @xmath125 and @xmath127 . note that the anstze used by @xcite , see eqs .
( 20,21,22,23 ) of @xcite , provide only an overall @xmath37-independent factor ; therefore they do not allow for any correction to scaling of @xmath115 .
our results for @xmath117 is in good agreement with both the experiment @xcite as well as with previous monte carlo studies of the xy model @xcite . on the other hand ,
the position of the minimum @xmath115 differs by several times the quoted error bar from both the experiment @xcite as well as from monte carlo studies of the xy model @xcite .
finally , in figure [ analplot ] we take into account the analytic corrections that we have detected fitting the position @xmath128 of the minimum of the casimir force . to this end , we have replaced the argument @xmath129 by @xmath130 .
the coefficient of the analytic correction is taken from the fit where we have fixed @xmath131 and @xmath132 .
now we find an almost perfect match between the curves obtained from @xmath133 , @xmath134 and @xmath135 .
krech and dietrich @xcite have computed the finite size scaling function @xmath2 in the high temperature phase using the @xmath27-expansion up to o(@xmath27 ) . in figure [ epsilon_plot ]
we plot their result for the xy universality class ( @xmath136 ) setting @xmath137 .
for comparison we plot our results for @xmath133 , @xmath138 and @xmath139 .
we have taken into account leading boundary corrections by replacing @xmath15 by @xmath140 , where we have taken @xmath131 .
comparing with the @xmath27-expansion we can estimate the systematical error caused by setting @xmath141 in eq .
( [ integration ] ) : we read off from the @xmath27-expansion that @xmath142 for our choices of @xmath93 for @xmath133 , @xmath134 and @xmath135 . taking into account this error , we see a good agreement of our monte carlo results and the @xmath27-expansion down to @xmath143 .
the curve obtained from the @xmath27-expansion flattens as the critical point is approached . at the critical point
the slope vanishes .
in contrast , in our case , the curve steepens as the critical point is approached .
the authors of @xcite have computed the finite size scaling function @xmath2 using a renormalized mean field approach .
while this approach correctly reproduces qualitative features of the finite size scaling function @xmath2 it fails to give quantitatively accurate results . in particular
the authors of @xcite find @xmath144 and @xmath145 .
i.e. the position of the minimum is overestimated by about a factor of 2 and its value by a factor of about 5 .
the authors of @xcite incorporate fluctuation effects into their mean - field analysis .
qualitatively they get the finite size scaling function @xmath2 for the whole range of @xmath37 right .
they adjust the two parameters of their solution for the low temperature phase ( eq . ( 17 ) of @xcite ) such that the minimum of @xmath2 found in the experiments @xcite is reproduced .
we did the same exercise , adjusting to our result for the minimum of @xmath2 .
we find no good match between @xmath2 computed in @xcite and ours .
our minimum is much more peaked than that of @xcite .
finally we compare our result for the finite size scaling function @xmath2 with experiments @xcite . in @xcite films of thicknesses between @xmath146 and @xmath147
have been studied . in figure [ experimentplot ]
we have plotted the data obtained from capacitor 1 which corresponds to the thickness @xmath148 of the film at temperatures @xmath149 .
this set of data is the smoothest among the five sets given in @xcite .
the results of @xcite are , in the range of temperatures we are interested in , less precise than those of @xcite . in the tables provided by the authors @xcite the finite size scaling function @xmath2
is given as a function of @xmath150 . in order to compare with our results
we have converted this to @xmath151 , using @xmath122 . for comparison
we give our result obtained from @xmath138 , where we have taken into account the boundary correction by replacing @xmath15 by @xmath152 and the leading analytic correction as discussed above .
furthermore , we give the asymptotic value @xcite @xmath153 obtained from the spin wave approximation . as already observed by the authors of @xcite
there is a qualitative agreement among the result obtained from monte carlo simulations of lattice models and the experiment .
there is a reasonable agreement of the position of the minimum @xmath115 , as already discussed above . for @xmath154
our result is in good agreement with that of the experiment .
in contrast for @xmath155 the experimental value of @xmath2 is clearly smaller than ours .
for @xmath156 the experimental result ( capacitor 1 of @xcite ) assumes @xmath157 and is decreasing again for smaller values of @xmath37 .
this is clearly different from the prediction ( [ spinwave ] ) of the spin wave approximation . in ref .
@xcite it was argued that this discrepancy could be explained by fluctuations of the surface resulting in @xmath158 this goes indeed in the right direction , can however not fully explain the difference between the experimental result and the theoretical prediction ( [ spinwave ] ) .
we have simulated the improved two component @xmath0 model on the simple cubic lattice .
this model shares with the @xmath6-transition of @xmath5he the three dimensional xy universality class .
we consider the thin film geometry . in order to mimic the vanishing order parameter at the surface of @xmath5he films near the @xmath6-transition
, we impose dirichlet boundary conditions with a vanishing field . restricting the system to a finite geometry leads to an effective force called thermodynamic casimir force .
its behaviour in the neighbourhood of the critical point is characterized by a universal finite size scaling function .
we have computed this function and have compared our result with that obtained from experiments on films of @xmath5he @xcite , previous monte carlo simulations of the xy model on the simple cubic lattice @xcite , field theoretic methods @xcite and mean field approaches @xcite .
the thermodynamic casimir force is given as minus the derivative of the excess free energy of the film with respect to its thickness @xmath15 .
on the lattice , this is approximated by the finite difference of films of the thickness @xmath159 and @xmath160 , where @xmath159 is integer .
in general it is impossible to compute free energies from a single monte carlo simulation . to circumvent this problem ,
divide and conquer strategies are employed .
different strategies have been proposed in @xcite and @xcite . here
we essentially follow @xcite : we compute the derivative of the excess energy with respect to @xmath15 for a dense grid of temperature values in the neighbourhood of the critical point . the corresponding result for the free energy
is then obtained by numerical integration . as discussed in ref .
@xcite , corrections to scaling are numerically quite large for the thicknesses that can be studied at present .
the main purpose of the present work is to get better control over these corrections as in previous work @xcite . to this end , we have studied the @xmath0 model at @xmath65 . in this model leading corrections to scaling ( and finite size scaling ) which are @xmath30 , with @xmath31 , are suppressed at least by a factor of 20 compared with the xy model @xcite .
boundary effects lead to corrections @xmath32 .
these corrections can be cast into the form @xmath140 . in @xcite
we have determined @xmath34 for the two component @xmath0 model at @xmath65 by using a finite size scaling study at the critical point of the three dimensional system .
we have verified that this choice of @xmath47 indeed eliminates the leading boundary correction in the scaling of the temperature of the kosterlitz - thouless transition @xcite and the specific heat of thin films @xcite . also here
we confirm that corrections can be essentially eliminated by replacing @xmath161 by @xmath140 with @xmath34 .
remaining discrepancies can be fitted by analytic corrections .
essentially we confirm the results obtained by previous monte carlo simulations of the three dimensional xy model @xcite for the finite size scaling function @xmath2 .
the main discrepancy with these previous works is the position of the minimum @xmath162 compared with @xmath163 @xcite and @xmath164 @xcite .
we should note that the monte carlo simulation of lattice models is at the moment the only theoretical method that allows for a quantitatively accurate calculation of @xmath2 in the low temperature phase not too far from the critical point .
the @xmath27-expansion gives correctly the behaviour in the high temperature phase .
the spin wave approximation gives the exact result in the limit @xmath165 .
the mean field calculation of @xcite reproduces only qualitatively the features of the scaling function .
quantitatively it is not satisfactory : the position of the minimum is wrongly estimated by a factor of almost @xmath166 and its value by a factor of about 5 .
qualitatively the monte carlo studies of lattice models nicely reproduce the finite size scaling function obtain from the experimental data @xcite for films of @xmath5he .
for @xmath167 there is a very good quantitative agreement between the two .
in contrast , for @xmath168 the value obtained from the experiment is clearly smaller than that of the monte carlo studies . at large values of @xmath169
the spin wave approximation should become exact .
also in this regime , the experiments produce numbers that are too small compared with the theoretical one .
the authors of @xcite explain this discrepancy by fluctuations of the surface of the @xmath5he film .
their result indeed reduces but not completely eliminates the difference between experiment and theory
. therefore it might be interesting to perform monte carlo simulations of a lattice model that incorporates fluctuations of the surface of the film .
this work was supported by the dfg under the grant no ha 3150/2 - 1 . kosterlitz j m and thouless d j , _ ordering , metastability and phase transitions in two - dimensional systems _ 1973 j. phys .
c * 6 * 1181 ; kosterlitz j m , _ the critical properties of the two - dimensional xy model _ , 1974 j. phys .
c * 7 * 1046 campostrini m , hasenbusch m , pelissetto a , and vicari e , _ theoretical estimates of the critical exponents of the superfluid transition in he4 by lattice methods _ , 2006 phys .
b * 74 * 144506 [ cond - mat/0605083 ] saito m , _ an application of finite field : design and implementation of 128-bit instruction - based fast pseudorandom number generator _ , phd thesis , dept . of math .
, graduate school of science , hiroshima university , advisor : m. matsumoto ; the numerical program and a detailed description can be found at http://www.math.sci.hiroshima-u.ac.jp/~m-mat/mt/sfmt/index.html " | we study the thermodynamic casimir effect in thin films in the three dimensional xy universality class . to this end
, we simulate the improved two component @xmath0 model on the simple cubic lattice .
we use lattices up to the thickness @xmath1 . based on the results of our monte carlo simulations we compute the universal finite size scaling function @xmath2 that characterizes the behaviour of the thermodynamic casimir force in the neighbourhood of the critical point .
we confirm that leading corrections to the universal finite size scaling behaviour due to free boundary conditions can be expressed by an effective thickness @xmath3 , with @xmath4 .
our results are compared with experiments on films of @xmath5he near the @xmath6-transition , previous monte carlo simulations of the xy model on the simple cubic lattice and field - theoretic results .
our result for the finite size scaling function @xmath2 is essentially consistent with the experiments on films of @xmath5he and the previous monte carlo simulations .
0.5 cm 0.5 cm * the thermodynamic casimir effect in the neighbourhood of the @xmath6-transition : + 0.2 cm a monte carlo study of an improved three dimensional lattice model * martin hasenbusch 0.3 cm _ institut fr physik , humboldt - universitt zu berlin _ _ newtonstr .
15 , 12489 berlin , germany
_ _ e mail : [email protected]_ 0.4 cm * keywords : * @xmath6-transition , classical monte carlo simulation , thin films , thermodynamic casimir effect |
kagra ( previously called lcgt for large - scale cryogenic gravitational - wave telescope ) is a japanese 3-km optical interferometer currently under construction in the kamioka mine .
the aim of the detector is frequent observation of gravitational waves from faraway galaxies and to obtain unique information on the universe .
the observation will be performed with other detectors in the us ( advanced ligo @xcite ) and europe ( advanced virgo @xcite and geo - hf @xcite ) , which are now being upgraded after several - year observations in the initial configurations .
kagra and the other three detectors are called second - generation gravitational - wave detectors .
there are two unique features in the detector configuration of kagra .
one is that the entire detector is constructed in the underground .
both seismic noise and gravity gradient noise are low in the underground .
in addition , the low rms motion relaxes the requirement on the interferometer control and reduces electro - magnetic noise .
the other special feature is the cryogenic operation of the interferometer .
the sapphire test masses are cooled down to 20 k and mirror thermal noise is lower than that of room - temperature detectors .
another benefit of the cryogenic operation is that the mirror causes almost no thermal lensing effect , which is one of the biggest issues for the room - temperature detectors , for the high thermal conductivity of sapphire at 20 k. the unique features of kagra come along with unique issues .
the baseline length is limited by the size of the mountain .
the floor is tilted for the water drainage system .
the mirror is made of sapphire instead of classic fused silica .
the laser power is limited by the amount of the absorbed heat that can be extracted through the suspension fibers .
the detector configuration of kagra is designed with these issues fully considered
. some of our knowledge of the underground cryogenic interferometer will be useful for a third - generation gravitational - wave detector , einstein telescope @xcite for example , which is planned to be built underground and operated in a cryogenic temperature .
figure [ fig : ifo ] shows the schematic view of the kagra detector . the optical configuration is a michelson interferometer with the high - finesse arm cavities and the folded recycling cavities .
the signal recycling cavity reduces the storage time of the signal fields resonating in the arm cavities in order to retain a broad bandwidth ; this system is called resonant - sideband extraction ( rse ) @xcite .
the four test - mass mirrors in the arm cavities are cooled down to 20 k to reduce thermal fluctuations .
these mirrors are made of sapphire and suspended by sapphire fibers .
the mass of the sapphire mirrors ranges between @xmath1 kg , depending on the availability .
the laser power after the mode - cleaner ranges between @xmath2 w , depending on the absorption of the sapphire mirror ; the less absorption , the more power we can inject .
the power recycling gain is about 10 and the finesse of the arm cavity is 1550 .
the power reflectivity of the signal recycling mirror is 85 % and the resonant condition of the signal recycling cavity is detuned by @xmath3 deg to increase the detector sensitivity to the gravitational waves from the neutron - star binaries ( the major parameters are listed in [ appa ] ) .
our primary target source is the neutron - star binaries since the event rate can be well estimated and the waveform during the inspiral is predictable . in the end , the observable distance of the neutron - star binary inspirals with kagra ranges between @xmath4 mpc , with which we will be able to detect @xmath5 neutron - star binary signals per year @xcite@xcite@xcite . here
the normal incidence of the wave is assumed in the calculation of inspiral range .
kagra project was approved by the japanese government in 2010 .
the initial phase will be the room - temperature operation without the recycling cavities .
this is called initial kagra or ikagra .
a short - term observation of ikagra is planned in 2015 .
the test masses of ikagra are 10 kg silica mirrors provided by courtesy of ligo . in the second phase ,
the test masses are replaced by the 23 kg ( or 30 kg ) sapphire mirrors to be cooled down . the suspension system is upgraded and the recycling mirrors are introduced .
this is the baseline - design kagra or bkagra .
observations with bkagra will start in @xmath6 .
figure [ fig : spectrum ] shows the estimated noise budget of kagra @xcite .
hereafter we assume the safest parameter set , namely the 23 kg mass and the 50 w laser .
the sensitivity is limited by quantum noise at most frequencies , and also suspension thermal noise is close to limit the sensitivity at low frequencies ( @xmath7 hz ) .
this reflects quite well the characteristic of the detector .
while mirror thermal noise is low for the cryogenic operation , shot noise is higher than other advanced detectors due to the low laser power for the cryogenic operation .
the lighter mass increases quantum radiation pressure noise and suspension thermal noise .
the fiber thickness is not determined by its strength but by the amount of the heat transferred from the test mass to the upper stages that are connected to the cryo - cooler .
the thick fiber increases suspension thermal noise .
the peak of suspension thermal noise at 130 hz is the vertical - mode resonance and the peaks starting at 230 hz are the violin modes .
the peak at 16 hz is the vertical resonance of the wire suspending the recoil mass .
we discuss more in detail about each noise curve in the following sections .
seismic noise and gravity gradient noise ( ggn ) are discussed in sec .
[ sec : seismic ] , thermal noise is discussed in sec .
[ sec : thermal ] , and quantum noise is discussed in sec .
[ sec : qn ] .
on the left side of fig .
[ fig : vis ] are shown the schematic views of type - a and type - b suspensions .
the type - a is a 2-story suspension with short and steady inverted pendula on the top floor , 4 geometric anti - spring ( gas ) filters in the middle , and the triple pendulum at the bottom for the cryogenic test masses .
the type - b is for the beamsplitter and the recycling mirrors . while the type - a suspension system has more gas filters and the isolation ratio is better than type - b at low frequencies , vibrations of the cryo - cooler transferred through the heat link limit the noise performance above a few hz . in the end
the noise levels of type - a and type - b are nearly equal in the observation band .
the seismic motion of the kagra site in the kamioka mine is @xmath8 times lower than that of the tama site at @xmath9 hz @xcite . the temperature and the humidity fluctuations measured in the mine are 0.01 deg per day and 0.08 % per day , respectively @xcite . on the right side of fig .
[ fig : vis ] are shown the seismic noise spectra with the type - a and type - b suspensions together with the ggn spectra , which were simulated by courtesy of jan harms at caltech using seismic data of the kamioka region provided by a research group in nikhef @xcite .
the test masses are located at least 200 m away from the surface of the mountain .
it is an advantage of the underground observatory that ggn from the surface of the mountain is suppressed .
figure [ fig : cryostat ] shows the schematic view of the cryogenic system that contains a sapphire test mass suspended by sapphire fibers in the cryostat vacuum chamber .
the outer / inner radiation shield is connected to the 1st - stage/2nd - stage of the cryo - cooler through the pure - metal heat path and cooled down to 80 k/8 k. the two stages are connected to the pulse - tube cryo - cooler ( ptc ) through the flexible heat links and the ptc is on the vibration reduction stage @xcite .
the last 3 stages of the suspension system are inside the radiation shields .
the platform is connected to the inner shield through the 7 pure aluminum heat links of 1 mm diameter , and the intermediate mass suspended under the platform is connected to the platform through the 5 heat links of 3 mm diameter .
the heat absorbed in the test mass is transferred to the intermediate mass through the suspension fiber and then transferred to the upper stages through the heat links . the transferable heat from an object at temperature @xmath10 to an object at temperature @xmath11 is given by the following equation : @xmath12 where @xmath13 , @xmath14 , @xmath15 , and @xmath16 are diameter , length , number , and thermal conductivity of the fiber or the heat link , respectively .
the thermal conductivity depends on the fiber diameter due to the surface scattering and also depends highly on the temperature .
while the thermal conductivity can be calculated using the debye model @xcite , we use the following simple equation that is derived from measurement results and is good below @xmath17 k : @xmath18 ( w / m / k ) . with the full laser power injected , about 800 w light transmits through the input test mass ( itm ) and 400 kw light is stored in the cavity .
the absorption of the sapphire substrate ranges between @xmath19 ppm / cm in a recent measurement @xcite .
the absorption of the silica / tantala doublet coatings measured at 20 k is @xmath20 ppm @xcite . to be on a safer side
, we assume 50 ppm / cm for the substrate and 1.0 ppm for the coatings . there is additional 200 mw heat introduced through the radiation from the apertures and the view ports . in total ,
about 1.2 w heat is absorbed by the test mass with the full power .
there are point - like defects on the mirror surface and about 10 ppm of the light in the cavity is scattered around to contribute to increase the temperature of the inner radiation shield .
the inner shield is coated with the diamond - like carbon ( dlc ) to reduce time to cool down the test masses before the operation .
the absorptivity of the 1.0 @xmath21 dlc coating on the cp aluminum at 10 @xmath21 is as high as 0.41 , so that the mirror can be cooled faster with the radiation from the mirror to the inner shield @xcite . on the other hand ,
the scattered light is also absorbed by the inner shield more effectively .
most of the scattered light , about @xmath22 w , is absorbed and the heat can increase the temperature of the inner shield to @xmath23 k. all of these facts taken into account , the test mass temperature would become 23 k with the full laser power . as is shown in the left panel of fig . [
fig : tn ] , mirror thermal noise , especially substrate thermoelastic noise , increases with the mirror temperature . the inspiral range of the neutron - star binary with the full - power operation decreases from 250 mpc to 230 mpc when the mirror temperature increases from 20 k to 23 k. it is then rather better to decrease the laser power . the mirror temperature can be 20 k with about 60 % of the laser power . shot noise increases and thermal noise decreases .
the inspiral range is then 240 mpc .
an alternative solution for the increasing heat absorption would be to use thicker fibers . replacing the 1.6 mm fiber by a 2.0 mm fiber , we could retain the mirror temperature of 20 k with the full laser power injected in return to an increase of suspension thermal noise
. the inspiral range would be then slightly higher , but this strategy takes away the possibility of the sensitivity improvement in the case the absorption be lower than the estimate . the mirror thermal noise curve in fig . [ fig : spectrum ] is the sum of substrate brownian noise , coating brownian noise , and substrate thermoelastic noise .
coating thermo - optic noise is supposedly low at 20 k and is ignored here .
the substrate brownian noise and coating brownian noise levels for each mirror are given by @xcite@xcite @xmath24 respectively .
here @xmath25 , @xmath26 , @xmath27 , @xmath28 , @xmath29 , @xmath30 , @xmath31 , and @xmath32 are the boltzmann constant , mirror temperature , frequency , coating thickness , beam radius , poisson ratio , young s modulus , and mechanical loss angle , respectively , with the subscript @xmath33 indicating substrate for @xmath34 and coatings for @xmath35 . as for the coatings ,
the total thermal noise level is a sum of the noise level for silica layers and the noise level for tantala layers .
.the last fractions in eqs .
( [ eq : tn1 ] ) and ( [ eq : tn2 ] ) that consist of young s modulus and poisson ratio ( in the unit of 1/pa ) .
the sapphire substrate has an advantage that these coefficients are low . [
cols="^,^,<,<,<",options="header " , ]
10 g. harry for the ligo scientific collaboration , class .
quantum grav . , * 27 * , 084006 ( 2010 ) `` advanced virgo baseline design , '' the virgo collaboration , vir-0027a-09 ( 2009 ) ; available from https://pub3.ego-gw.it/itf/tds/ b. willke _ et al .
_ , class .
quantum grav . ,
* 23 * , s207-s214 ( 2006 ) s. hild , _ et al .
_ , class .
quantum grav . , * 27 * , 015003 ( 2010 ) j. mizuno _ et al .
_ , phys .
a , * 175 * , 273 - 276 ( 1993 ) c. kim , aip conf .
proc . , 983 , 576 - 583 ( 2007 ) d. lorimer , living rev . relativity , * 11 * , 8 ( 2008 ) j. abadie _ et al .
_ , class .
quantum grav . , * 27 * , 173001 ( 2010 ) this estimated noise spectrum is slightly different from the official sensitivity curve of kagra provided in the web page ( http://gwcenter.icrr.u-tokyo.ac.jp/en/researcher/parameter ) , which is calculated with the most optimistic parameter set .
s. sato _ et al .
_ , phys .
rev.d * 69 * , 102005 ( 2004 ) http://www.nikhef.nl/en/for-nikhef-users/departments/scientific-departments/gw t. tomaru _ et al .
_ , class . quantum grav .
* 21 * , s1005-s1008 ( 2004 ) d. morelli _ et al .
_ , phys .
b , * 66 * , 195304 ( 2002 ) `` sapphire for the kagra project , '' e.hirose , k.watanabe , and n.mio , jgw - g1100626-v1 ( 2011 ) ; available from http://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/docdb/documentdatabase k. yamamoto _ et al .
_ , phys .
rev.d * 74 * , 022002 ( 2006 ) `` development of cryogenic system for large - scale cryogenic gravitational wave telescope ( 4 ) - reduction of initial cooling time of kagra mirrors - , '' y. sakakibara , _ et al .
_ , cryogenics and superconductivity society of japan , proceedings , * 2011a * , 3d - p04 ( 2011 ) y. levin , phys .
d * 57 * , 659 ( 1998 ) g. harry _ et al .
_ , class . quantum grav .
* 19 * , 897 - 917 ( 2002 ) r. flaminio _ et al .
_ , class . quantum grav .
* 27 * , 084030 ( 2010 ) i. martin _ et al .
_ , class . quantum grav .
* 25 * , 055005 ( 2008 ) i. martin _ et al .
_ , class . quantum grav .
* 27 * , 225020 ( 2010 ) `` discussion about losses in the perpendicular and parallel directions , '' k. somiya , ligo - dcc t0900033-v1 ( 2009 ) ; available from https://dcc.ligo.org/ m. cerdonio , l. conti , a. heidmann , and m. pinard , phys .
d * 63 * , 082003 ( 2001 ) k. somiya , k. kokeyama , and r. nawrodt , phys .
d * 82 * , 127101 ( 2010 ) temperature dependence of each thermal parameter is given as follows , fitted to measurement results : @xmath36 [ w / m / k ] , @xmath37 [ j/@xmath38/k ] , @xmath39 [ 1/k ] , where @xmath40 is in kelvin . t. uchiyama _ et al .
_ , in preparation p. saulson , phys
. rev .
d * 42 * , 2437 - 2445 ( 1990 ) `` the thermal noise of the virgo+ and virgo advanced last stage suspension ( the ppp effect ) , '' f. piergiovanni , m. punturo , and p. puppo , vir-015e-09 ( 2009 ) k. somiya , y. chen , s. kawamura , and n. mio , phys .
d * 73 * , 122005 ( 2006 ) a. buonanno and y. chen , phys .
d * 64 * , 042006 ( 2001 ) y. aso _ et al .
_ , article in this issue | construction of the japanese second - generation gravitational - wave detector kagra ( previously called lcgt ) has been started . in the next @xmath0 years , we will be able to observe the space - time ripple from faraway galaxies .
kagra is equipped with the latest advanced technologies .
the entire 3-km long detector is located in the underground to be isolated from the seismic motion , the core optics are cooled down to 20 k to reduce thermal fluctuations , and quantum non - demolition techniques are used to decrease quantum noise . in this paper
, we introduce the detector configuration of kagra ; its design , strategy , and downselection of parameters . |
photosynthetic organisms capture visible light in their light - harvesting complex and transfer the excitation energy to the reaction center which stores the energy from the photon in chemical bonds .
this process occurs with nearly perfect efficiency .
although the primary process occurring in the light - harvesting complex , is the electron transfer between acceptor and donor , and the transfer to the reaction center appears as secondary process , the two processes should not be separated from one another .
both processes are nothing but two parts of the total light harvesting .
a few years ago , evidence of coherent quantum energy transfer has been found experimentally @xcite .
recent experimental results @xcite demonstrated that photosynthetic bio - complexes exhibit collective quantum coherence during primary electron transfer processes that occur on the time scale of some hundreds of femtoseconds .
furthermore , the coherence in such a system exhibits a bi - exponential decay consisting of a slow component with a lifetime of hundreds of femtoseconds and a rapid component with a lifetime of tens of femtoseconds @xcite . the long - lived components are correlated with intramolecular modes within the reaction center , as shown experimentally @xcite . these results induced different theoretical considerations which are related to the role of quantum coherence in the photosynthesis .
for example , the equivalence of quantum and classical coherence in electronic energy transfer is considered in @xcite . in @xcite , the fundamental role of noise - assisted transport
is investigated . in @xcite
, it is shown that the efficiency is increased by reducing radiative recombination due to quantum coherence .
the hamiltonian of the system in these ( and many other ) papers is assumed to be hermitian although photosynthesis is a process that occurs in an open quantum system . in the description of an open quantum system by means of a non - hermitian hamilton operator ,
the localized part of the system is embedded into an environment .
mostly , the environment is the extended continuum of scattering wavefunctions , see e.g. the review @xcite .
coherence is an important ingredient of this formalism .
meanwhile the non - hermitian formalism is applied successfully to the description of different realistic open quantum systems , see the recent review @xcite .
the paper @xcite is one of the oldest references in which the resonance structure of the cross section in the regime of overlapping resonances is considered in the non - hermitian formalism . in this paper , the resonance structure of the nuclear reaction @xmath0 with two open decay channels is traced as a function of the degree of overlapping of the individual resonances by keeping constant the coupling strength between the localized part of the system and the environment of scattering wavefunctions .
the distance between the energies of the individual resonance states is varied by hand . as a result ,
two short - lived states are formed at a critical value of the degree of overlapping .
the widths of all the other states are reduced because @xmath1 has to be constant according to the constant coupling strength between system and environment .
these states are called trapped states . in some following papers
, this phenomenon is studied as a function of the coupling strength between system and environment and is called segregation of decay widths , see the recent review @xcite . in these papers
, the short - living states are called superradiant states which exist together with long - living subradiant states .
this formalism is applied also to the problem of electron transfer in photosynthetic complexes @xcite , see also @xcite . in this formalism
, the enhancement of the electron transfer is related to the existence of the superradiant state . in other papers ,
the resonance trapping phenomenon is related to singular points which exist in the formalism of non - hermitian quantum physics , see the review @xcite and the recent paper @xcite .
these singular points are known in mathematics since many years @xcite , and are called usually exceptional points ( eps ) .
most interesting new features caused by the eps in the non - hermitian physics of open quantum systems are , firstly , the non - rigid phases of the eigenfunctions and , secondly , the possibility of an external mixing ( em ) of the states of the localized part of the system via the environment . non - rigidity of the phases of the eigenfunctions of the hamiltonian and an em of the states are possible _ only _ in an open quantum system .
they are not involved explicitly in any type of hermitian quantum physics .
furthermore , superradiant and subradiant states do not appear in this formalism .
quite the contrary , phenomena that are related in , e.g. , @xcite to their existence , are an expression for nothing but the nontrivial properties of the eigenfunctions of a non - hermitian hamilton operator , such as non - rigid phases and em of the wavefunctions . in @xcite ,
the dynamics of the system and the efficiency of electron transfer are studied in a non - hermitian formalism by taking into account noise acting between donor and acceptor , while in @xcite , the role of protein fluctuation correlations in the electron transfer is investigated and the spin - echo approach is extended to include bio - complexes for which the interaction with dynamical noise is strong .
it is the aim of the present paper to provide the general formalism of non - hermitian physics of open quantum systems @xcite by inclusion of gain which simulates the acceptor , as well as of loss which stands for the donor @xcite .
when additionally the coupling of the system to a sink is taken into account , this formalism can be applied to the description of light - harvesting of photosynthetic complexes .
we underline that this formalism describes the process of photosynthesis as a whole , i.e. as a uniform process . while the first part occurs instantly , the second part of the process may last longer .
the formalism is generic . in the future
, it has to be applied to concrete systems with realistic parameters . in sect .
[ gainloss1 ] , we sketch the formalism for the study of an open quantum system with gain and loss which is basic for the description of photosynthesis . in sect .
[ gainloss2 ] , we include additionally a sink into the formalism simulated by coupling to a second environment . in both sections we provide analytical as well as numerical results .
we discuss and summarize the results in sect . [ disc ] and draw some conclusions in sect .
[ concl ] . before providing the formalism for the description of open quantum systems , it is necessary to clarify the meaning of some terms .
we will use definitions similar to those used in nuclear physics . * in nuclear physics , channel denotes the coupling of a certain state of the nucleus @xmath2 to its decay products after emission of particle @xmath3 and leaving the residual nucleus @xmath4 in a special state .
the term channel is equivalent to embedding of a localized state of the system into an environment .
the localized state in nuclear physics is the state of the nucleus @xmath2 , while the environment is the the continuum of scattering wavefunctions the particle @xmath3 . * in difference to the definition of energy and width of a nuclear state in nuclear physics
, we use the definition @xmath5 for the complex eigenvalues of the non - hermitian hamilton operator @xmath6 .
the widths of decaying states have thus a negative sign @xcite .
* the term internal mixing of the wavefunctions denotes the direct interaction between two orthogonal eigenfunctions of a hermitian hamilton operator , @xmath7 . in our calculations , it is supposed to be included in the energies @xmath8 and widths @xmath9 of the states that define the non - hermitian hamilton matrix , see e.g. eq . ( [ gain1 ] ) .
an external mixing of two eigenstates of a non - hermitian hamilton operator occurs via the environment and is thus a second - order process .
it is defined only in an open system .
* the singularity related to the coalescence @xcite of two eigenvalues of a non - hermitian hamilton operator @xmath6 is called , in recent literature , mostly exceptional point . in older papers , the equivalent expressions branch point in the complex plane or double pole of the s - matrix
are mostly used .
we sketch the features characteristic of an open quantum system with gain and loss @xcite by considering a localized 2-level system that is embedded in a common continuum of scattering wavefunctions .
one of these two states gains particles from the environment by interacting with it , while the other one loses particles to the continuum by decay . for the description of the open quantum system , we use the non - hermitian hamilton operator @xcite @xmath10 here , @xmath11 are the two complex eigenvalues of the basic non - hermitian operator coupled to the environment @xmath12 ( called also channel @xmath12 ) @xcite .
the @xmath13 are the energies of the states @xmath14 and the @xmath15 are their widths .
one of these eigenvalues describes loss characteristic of decaying states ( @xmath16 ) while the other one describes gain from the environment ( @xmath17 ) @xcite .
the @xmath18 stand for the coupling matrix elements of the two states via the common environment @xmath12 .
they are complex where re(@xmath18 ) arises from the principal value integral and im(@xmath18 ) from the residuum @xcite . the complex eigenvalues @xmath19 of @xmath20
give the energies @xmath21 and widths @xmath22 of the states of the localized part of the system @xcite .
we will consider also the non - hermitian hamilton operator @xmath23 which describes the localized part of the open system without coupling of its states via the continuum ( @xmath24 ) .
the phases of the eigenfunctions @xmath25 of @xmath26 are rigid ( like in hermitian quantum physics ) when @xmath27 .
the eigenvalues of @xmath28 are , generally , complex and may be expressed as @xmath29 where @xmath21 and @xmath22 stand for the energy and width , respectively , of the eigenstate @xmath14 .
also here @xmath30 for decaying states and @xmath31 for gaining states @xcite .
the two states may repel each other in accordance with re@xmath32 , or they may undergo width bifurcation in accordance with im@xmath32 . when @xmath33 the two states cross each other at a point that is called usually exceptional point ( ep ) @xcite .
the ep is a singular point ( branch point ) in the complex plane where the @xmath34-matrix has a double pole @xcite .
according to its definition @xcite , the ep is meaningful in an open quantum system which is embedded in _
one _ common environment @xmath35 .
correspondingly , we denote e.g. the eigenvalues by @xmath36 .
we consider now the behavior of the eigenvalues when the parametrical detuning of the two eigenstates of @xmath37 is varied , bringing them towards coalescence @xcite .
according to ( [ int6 ] ) , the condition for coalescence reads @xmath38 we consider two cases that can be solved analytically . 1 .
when @xmath39 , and @xmath40 is real , it follows from ( [ int6i ] ) the condition @xmath41 for the coalescence of the two eigenvalues , i.e. for an ep .
it follows furthermore @xmath42 eq .
( [ int6c ] ) describes the behavior of the eigenvalues away from the ep , where the eigenvalues @xmath43 differ from the original ones through only a contribution to the energy .
the widths in contrast , remain unchanged , and this situation therefore corresponds to that of level repulsion .
( [ int6d ] ) , in contrast , is relevant at the other side of the ep . here ,
the resonance states undergo width bifurcation according to im@xmath44 .
the bifurcation starts in the neighborhood of the ep .
physically , the bifurcation implies that different time scales may appear in the system , while the states are nearby in energy .
when @xmath45 , and @xmath40 is imaginary , then the condition @xmath46 together with @xmath47 , follows for the coalescence of the two eigenvalues from ( [ int6i ] ) . here
instead of ( [ int6c ] ) and ( [ int6d ] ) we have @xmath49 thus , the ep causes width bifurcation also in this case .
however , this case is realized only when @xmath50 at the ep , i.e. when gain and loss vanish at the ep .
the eigenfunctions of a non - hermitian hamilton operator are biorthogonal ( for details see @xcite ) @xmath51 in the case of the symmetric @xmath52 hamiltonian ( [ gain1 ] ) , it is @xmath53 and the eigenfunctions should be normalized according to @xmath54 in order to smoothly describe the transition from a closed system with discrete states to a weakly open one with narrow resonance states . as a consequence of ( [ eif3 ] ) , the values of the standard expressions are changed , @xmath55 @xmath56 @xmath57 while it approaches the value @xmath58 when an ep is approached .
the hamiltonian ( [ gain2 ] ) describes the system around the ep without any mixing of its states via the environment , since @xmath59 corresponds to vanishing em of the eigenstates . in order to determine quantitatively the strength of the em , we present the eigenfunctions @xmath60 of @xmath61 in the set of eigenfunctions @xmath62 of @xmath63 , @xmath64 under the condition that the @xmath65 are normalized by @xmath66 .
the coefficients @xmath67 differ from the @xmath68 .
they contain the information on the strength of em via the environment which is determined by the value of @xmath18 . for illustration
, we consider the em of the wavefunctions @xmath69 and @xmath70 around an ep in the two cases discussed in sect .
[ eigenv1 ] .
1 . @xmath71 , and @xmath72 : according to ( [ int6b ] ) , the strength of the em via the environment is determined by the differences @xmath73 of the widths ( which both have different sign ) .
it depends thus on the fluctuations of the @xmath15 .
2 . @xmath74 , and @xmath75 : according to ( [ int6b1 ] ) the strength of the em is related to the differences @xmath76 of the energies , i.e. to the fluctuations of the @xmath13 .
this case is however realized only when gain and loss vanish at the ep ( i.e. @xmath50 at the ep ) . at the eps ,
the two corresponding eigenfunctions are not orthogonal .
instead @xmath77 according to analytical and numerical results @xcite .
we underline once more that an ep is , according to its definition , related to the common environment in which the system is embedded . in other words
, it is well defined under the condition that the system is embedded in _ only one _ continuum .
the schrdinger equation @xmath78 may be rewritten into a schrdinger equation with source term @xcite , @xmath79 in this representation , the coupling @xmath18 of the states @xmath14 and @xmath80 of the localized system via the common environment of scattering wavefunctions ( em ) is contained solely in the source term .
the source term vanishes , when @xmath71 around the ep under the condition @xmath81 according to ( [ int6b ] ) , what is fulfilled when @xmath82 .
far from eps , the coupling of the localized system to the environment influences the spectroscopic properties of the system , in general , only marginally @xcite .
the influence is however non - vanishing also in this case , see e.g. the experimental results @xcite . in the neighborhood of eps , however
, the coupling between system and environment and therewith the source term play an important role for the dynamics of the open quantum system .
the reason is , according to mathematical studies , that the source term causes nonlinear effects in the schrdinger equation ( [ eif11 ] ) around an ep .
for details see @xcite .
let us consider the resonance part of the @xmath34 matrix from which the resonance structure of the cross section can be calculated , @xmath83 a unitary representation of the resonance part of the @xmath34 matrix in the case of two resonance states coupled to a common continuum of scattering wavefunctions reads @xcite @xmath84 here , the influence of the eps onto the cross section is contained in the eigenvalues @xmath85 .
the expression ( [ sm4 ] ) allows therefore to receive reliable results also when the phase rigidity is reduced , @xmath86 .
let us assume real @xmath18 and @xmath27 .
first we consider the case corresponding to the condition ( [ int6c ] ) , i.e. for large coupling strength @xmath18 of the system to the environment of continuous scattering wavefunctions . in this case
@xmath87 and , according to ( [ int6c ] ) , @xmath88 in the other case , ( [ int6d ] ) , it is @xmath89 ; @xmath90 ; and @xmath91 in both cases , @xmath92 , i.e. @xmath93 according to ( [ cro ] ) .
this result corresponds to the well - known fact that eps can not be identified in the resonance structure of the s - matrix and therefore also not in the resonance structure of the cross section .
most important is however the result that no resonances will be excited due to @xmath27 . the result @xmath92 is violated when the conditions @xmath72 and @xmath27 are not exactly fulfilled
this may happen , e.g. , under the influence of external random ( stochastic ) processes that cause fluctuations of the @xmath15 .
in such a case , @xmath94 ; and the energy ( or information ) will be transferred with an efficiency of nearly 100 % ( because no resonances can be excited under this condition in the localized part of the system ) .
results for this case can be obtained only numerically .
let us first consider the results that are obtained by using a parametric dependence of the energies @xmath13 and widths @xmath15 of the states of the localized part of the system which is analog to that used in @xcite for a decaying system .
the results are shown in fig .
[ fig1 ] . in this figure
, the existence of an ep at the parameter value @xmath95 can clearly be seen . here ,
the two states are exchanged ( figs .
[ fig1].a , b ) , the phase rigidity @xmath96 approaches the value @xmath97 ( figs .
[ fig1].c , d ) and the em of the states via the continuum increases limitless ( fig . [
fig1].e ) .
also the contour plot of the cross section ( fig .
[ fig1].f ) shows the exchange of the two states at the ep .
that means : the state with positive width turns into a state with negative width and vice versa .
the two states have therefore nothing in common with resonance states the widths of which are always negative ( or zero at most ) . nevertheless , the results shown in fig . [ fig1 ] are formally similar to those obtained and discussed in @xcite for a decaying system . in the latter case , both states which are exchanged at the ep , are of the same type : they are resonance states with negative widths .
it is interesting to see from the numerical results ( fig .
[ fig1 ] ) , that the non - hermitian hamilton operator @xmath6 can be used , indeed , for the description of these two different types of open quantum systems as suggested in sect .
[ gainloss1 ] @xcite .
additionally we show some numerical results for the case that , respectively , the distance in energy of the two states is smaller ( fig .
[ fig2 ] ) and the em as well as the widths @xmath98 differ more from one another ( fig .
[ fig3 ] ) than in fig .
the eigenvalue and eigenfunction figures ( figs .
[ fig1].a - e , [ fig2].a - e , [ fig3].a - e ) are similar to one another and show clearly the signatures of an ep at a certain critical value of the parameter @xmath99 . the contour plots ( fig .
[ fig1].f , [ fig2].f , [ fig3].f ) differ however from one another . when the states are nearer to one another in energy , the two states with negative and positive width merge ( fig .
[ fig2].f ) . under the influence of stronger em ( stronger coupling strength @xmath18 between system and environment ) as well as of a larger difference between the two values @xmath98 , the extension of the region with non - vanishing cross section is enlarged in relation to the energy ( fig . [ fig3].f ) . in any case
, the cross section vanishes around @xmath99 .
resonance states are not excited .
more characteristic for an open quantum system with gain and loss than those in sect .
[ num1a ] are the analytical results given in sect .
[ smatr1 ] . according to these results ,
the cross section is zero when @xmath27 , @xmath71 , and @xmath72 . under the influence of an ep which causes differences between the original spectroscopic values @xmath100 and the eigenvalues @xmath101 of @xmath28 , a non - vanishing cross section is expected when at least one of the conditions @xmath102 , @xmath103 , together with @xmath72 , is not fulfilled . in fig .
[ fig4 ] we show the corresponding numerical results obtained for two neighboring states , @xmath104 , and @xmath18 almost real .
we fix the energies @xmath105 and vary parametrically the widths @xmath15 , see the dashed lines in figs .
[ fig4].a , b .
the results show an ep at @xmath106 and the hint to another ep at @xmath107 . at the ep ,
the phase rigidity approaches the value zero , @xmath108 ( fig .
[ fig4].c ) , and the em of the states is extremely large , @xmath109 ( fig . [
fig4].e ) . of special interest
is the parameter range between the two eps . here , @xmath110 at @xmath111 ( figs .
[ fig4].c , d ) . at this parameter value ,
the level repulsion is maximum ; and the two eigenfunctions of the non - hermitian hamiltonian @xmath28 are ( almost ) orthogonal . an analogous result is known from calculations for decaying systems , i.e. for systems with excitation of resonance states @xcite . in these calculations ,
the energies are varied parametrically and @xmath112 is almost imaginary ( fig . 1 in @xcite ) .
therefore , the two eigenfunctions of the non - hermitian hamilton operator are ( almost ) orthogonal ( @xmath110 ) at maximum width bifurcation ( instead of at maximum level repulsion in fig .
[ fig4 ] ) . in any case , the two eigenstates of the non - hermitian operator turn irreversibly into two states with rigid phases in spite of the non - hermiticity of the hamiltonian .
this unexpected result occurs due to the evolution of the system to the point of , respectively , maximum level repulsion and maximum width bifurcation , which is driven exclusively by the nonlinear source term of the schrdinger equation , see sect .
[ schr1 ] .
the eigenfunctions of these two states are mixed . in order to receive a better understanding of this result
, we mention here another unexpected result of non - hermitian quantum physics , namely the fact that a non - hermitian hamilton operator may have real eigenvalues @xcite .
this fact is very well known in literature for a long time , for references see the review @xcite .
the corresponding states are called usually bound states in the continuum . most interesting for a physical system is the contour plot of the cross section ( fig . [
fig4].f ) .
according to the analytical results discussed in sect .
[ smatr1 ] , the cross section vanishes far from the parameter range that is influenced by an ep .
it does however not vanish completely in the parameter range between the two eps . around the ep at @xmath113 ,
the conditions for vanishing cross section are quite well fulfilled , while this is not the case around the other ep at @xmath114 . in approaching the two eps by increasing and decreasing , respectively , the value of @xmath3 the cross section vanishes around @xmath113 , and is non - vanishing around @xmath115 .
the cross section vanishes also around the point of maximum level repulsion at which the two eigenfunctions of @xmath28 are orthogonal ( and _ not _ biorthogonal ) . additionally , we performed calculations ( not shown in the paper ) with the reduced value @xmath116 in order to determine the role of em in the cross section picture .
the obtained results are similar to those shown in figs .
[ fig4].a - f .
the parameter range influenced by the two eps is however smaller when @xmath40 is reduced : it ranges from @xmath117 to @xmath118 when @xmath119 .
accordingly , the region of the non - vanishing cross section in the contour plot shrinks in relation to @xmath3 , and also in relation to the energy . in calculations with vanishing external mixing ( @xmath59 )
, the cross section vanishes everywhere .
let us consider the @xmath120 non - hermitian matrix @xmath121 here , @xmath122 and @xmath123 are the complex eigenvalues of the basic non - hermitian operator @xmath124 relative to channel @xmath35 and @xmath125 , respectively @xcite .
the two channels ( environments ) are independent of and orthogonal to one another what is expressed by the zeros in the matrix ( [ ham22 ] ) .
one of the channels may be related to gain and loss @xcite ( acceptor and donor ) considered in the previous section [ gainloss1 ] , while the other channel may simulate a sink . in this case , the two widths @xmath126 and @xmath127 have different sign relative to the first channel .
relative to the second channel however , both @xmath128 are negative according to a usual decay process of a resonance state .
the @xmath18 and @xmath129 stand for the coupling matrix elements between the two states @xmath130 of the localized part of the open quantum system and the environment @xmath35 and @xmath131 , respectively . in the case considered above , these two environments are completely different from one another and should never be related to one another . in more detail : an em of the considered states may be caused _ only by @xmath18 or by @xmath129 _ , and never by both values at the same time @xcite .
this is guaranteed when @xmath132 what is fulfilled when there is only one state in the second channel .
when there are more states , then @xmath133 should be much smaller than @xmath134 ( here we point to the general result that the values @xmath112 are related to the widths @xmath135 of the states @xcite ) .
the values @xmath98 and @xmath136 are independent of one another and express the different time scales characteristic of the two channels .
while the @xmath98 will be usually very large , the @xmath136 are generally much smaller .
accordingly , the two - step process as a whole will show , altogether , a bi - exponential decay : first the decay occurs due to the exponential quick process ; somewhere at its tail it will however switch over into the exponential decay of the slow process .
the hamiltonian which describes vanishing coupling of the states of the localized part of the open quantum system to both environments is @xmath137 by analogy to ( [ gain2 ] ) .
it does not contain any em via an environment .
the eigenvalues @xmath139 and eigenfunctions @xmath140 of ( [ ham22 ] ) are characterized by two numbers : the number @xmath14 of the state ( @xmath141 ) of the localized part of the system and the number @xmath142 of the channel ( @xmath143 ) , called environment , in which the system is embedded .
generally , @xmath144 and @xmath145 . also the wave functions @xmath146 and @xmath147 differ from one another due to the em of the eigenstates via the environment @xmath35 and @xmath125 , respectively . from a mathematical point of view , the system has therefore four states .
an ep influences the dynamics of the open quantum system also in the two - channel case . without an ep in the considered parameter range in relation to both channels
, we have @xmath148 , @xmath149 and @xmath150 . accordingly
, one has to consider effectively only two states under the influence of an ep relative to @xmath35 ( or / and relative to @xmath125 ) , the eigenvalues and eigenfunctions will be , however , different from one another , @xmath144 , @xmath145 and @xmath151 in the corresponding parameter range . we have to consider therefore effectively four states in this case . according to @xcite an ep
is defined when the system is embedded in _
one _ common environment . under this condition , it causes nonlinear processes in a physical system , which is the crucial factor for the dynamical properties of an open quantum system @xcite .
this is valid not only for systems all states of which decay ( corresponding to some loss ) , but also for systems with loss _ and _ gain , as shown in @xcite . due to the nonlinear processes occurring near to an ep , it is difficult to receive analytical solutions for the eigenvalues and eigenfunctions of ( [ ham22 ] )
. we will provide the results of some numerical simulations , above all with @xmath104 , almost real @xmath40 and almost imaginary @xmath152 , which is the most interesting and general case for a system with gain and loss that is coupled to a sink ( see the analytical results obtained with @xmath71 and @xmath72 , eq .
( [ int6b ] ) , and the corresponding results for decaying systems in @xcite ) . using ( [ ham22 ] ) , we can write down the schrdinger equation with source term for the two - channel case in analogy to ( [ eif11 ] ) for the one - channel case . the corresponding equation reads @xmath153 the source term depends on the coupling of the system to both channels , i.e. on @xmath18 and on @xmath129 .
we will consider the general case with two channels ( two environments ) in which @xmath154 .
we repeat here that , according to their definition @xcite , eps occur only in the one - channel case , i.e. only in the submatrices related either to channel @xmath12 or to channel @xmath131 .
they are not defined in the @xmath120 matrix ( [ ham22 ] ) .
however , each ep in one of the two submatrices in ( [ ham22 ] ) influences the dynamics of the open two - channel system .
this will be shown in the following section by means of numerical results for the case that there is an ep in the first channel which simulates acceptor and donor ( gain and loss ) , while the second channel being of standard type with resonance states , may or may not have an ep .
we performed some calculations for the two - channel case by starting from the calculations for the one - channel case in , respectively , fig .
[ fig2 ] and [ fig4 ] and by adding a second channel that describes decaying states ( corresponding to loss ) .
there are , of course , very many possibilities for choosing the number of states as well as the parameters for the second channel .
one possibility is to keep the parameters constant by varying the parameter @xmath3 of the first channel .
another possibility is to relate them directly to the parameter @xmath3 , or to introduce another independent parameter @xmath155 .
the choice should correspond to the physical situation considered .
the aim of our calculations is to illustrate the influence of a second channel onto the eigenvalues and eigenfunctions of @xmath138 and onto the contour plot of the cross section .
we exemplify this by choosing parameter dependent values @xmath11 in the first channel and parameter independent values @xmath156 in the second channel . in the following
, we show a few characteristic results .
we start these calculations with two channels by choosing two merged states with gain and loss according to fig .
the second channel contains two resonance states @xmath157 with negative widths and @xmath158.a , b and c , d , e , respectively , show the eigenvalues and eigenfunctions of the first channel ( fig .
[ fig2].a - e ) as well as the eigenvalues and eigenfunctions of the second channel .
the last ones are constant as function of the parameter @xmath3 what follows from the assumption of their parameter independence .
the contour plots of the cross section ( figs . [ fig2].f and [ fig5].f )
are different from one another .
common to both of them is that the cross section does not vanish in a finite range of the energy around @xmath159 for all @xmath3 . under the influence of the two states in the second channel which are exactly in this energy and parameter range
, the cross section is somewhat reduced .
it does however not vanish .
these ( and similar ) simulations show clearly the following result .
the fluctuations of the cross section which are caused by the merging of two states with gain and loss in the first channel , excite resonance states in the second channel .
this happens although the nature of both channels is completely different .
in the first channel , internal degrees of freedom of the system are not excited , while the appearance of the resonance states in the second channel occurs via excitation of internal degrees of freedom of the system . in these calculations
we start from the results shown in fig .
[ fig4 ] for the first channel .
the second channel contains only one state .
the eigenvalue and eigenfunction pictures fig .
[ fig6].a - e contain the eigenvalues and eigenfunction trajectories of fig .
[ fig4].a - e as well as those of the second channel , e.g. the energy trajectories at @xmath160 and width trajectories at @xmath161 .
the phase rigidity approaches the value @xmath12 at @xmath162 in both cases .
the corresponding contour plot of the cross section is shown in fig .
[ fig6].f .
it is related to the contour plot of the first channel ( fig .
[ fig4].f ) ; and shows additionally the parameter independent state of the second channel in the whole parameter range .
we performed further calculations with parameters similar to those used in fig .
[ fig6].a - e and show two of the corresponding contour plots in figs .
[ fig6].g , h .
both contour plots are obtained with the comparably large width @xmath163 ; the two energies @xmath164 are however different from one another .
the difference between @xmath165 and @xmath166 can clearly seen in the corresponding contour plots figs .
[ fig6].g and h. the results of these simulations with level repulsion in the first channel and one state in the second channel show the same characteristic features as those discussed above for merging states .
the fluctuations observed in the first channel are able to excite resonance states in the second channel .
the situation with two states in the second channel is richer than that with only one state because the two states can mix via the common continuum .
we show results for one special case in fig .
[ fig7 ] . as in figs .
[ fig5 ] and [ fig6 ] , the eigenvalue figures [ fig7].a , b contain the eigenvalue trajectories of both channels .
the eigenfunction trajectories refer to the existence of an ep in the second channel : the phase rigidity of states related to the second channel is independent of the parameter @xmath3 , as expected .
it is however smaller than @xmath12 .
also the mixing @xmath167 of the two wavefunctions via the common continuum in the second channel shows small deviations from the expectations .
the relation of these results for the eigenfunctions of @xmath168 to an ep in the second channel is discussed in detail in appendix [ a1 ] .
the contour plot fig .
[ fig7].f shows the same characteristic features as fig .
[ fig6].f . both , its relation to the contour plot of the first channel ( fig .
[ fig4].f ) as well to the parameter independent state of the second channel can clearly be seen .
thus , the fluctuations observed in the first channel excite resonance states in the second channel also in this case . summarizing the numerical results shown in the three figures [ fig5 ] to [ fig7 ] we state the following . gain and loss in the first channel and excitation of resonance states in the second channel are a _ uniform
_ process .
this process can be described as a whole in the formalism for the description of open quantum systems which is used in the present paper .
in our paper , we considered gain and loss in an open quantum system @xcite .
of special interest is the interplay of these two opposed processes in the neighborhood of singular points where it causes fluctuations of the cross section .
these fluctuations are observable and can excite resonance states in the system .
the time scale of the fluctuations and that of the excitation of resonance states are very different .
the fluctuations of the cross section occur quickly , without any excitation of internal degrees of freedom of the system .
the excitation of resonance states is however much slower , and internal degrees of freedom of the system are involved .
we sketched first the formalism by means of which both the interplay between gain and loss in an open quantum system @xcite and the excitation of resonance states can be described as a uniform process . in any case , the widths @xmath135 of the states are ( generally ) different from zero . in the first case we have two states with different sign of the widths ( corresponding to gain and loss ) , while the states in the second case are standard decaying ( resonance ) states with negative sign @xcite . generally , the cross section related to the two states with gain and loss vanishes .
deviations from this rule may appear in the neighborhood of singular ( exceptional ) points , where they may cause non - vanishing fluctuations of the cross section .
these fluctuations have nothing in common with resonances .
rather , they are merely deviations from the vanishing value of the cross section and are _ not _ related to the excitation of any internal degrees of freedom of the system .
they occur therefore with an efficency of nearly 100 % at a very short time scale .
the fluctuations caused by the interplay of the states with gain and loss , are observable and may excite resonance states of the system after a comparably long time ( which corresponds to the widths @xmath169 of these states ) .
the whole process of gain and loss together with the excitation of resonance states is therefore characterized by two very different time scales : the quick process which creates the fluctuations , and the slow process which is related to the excitation of resonance states .
the decay occurs thus bi - exponential .
initially , it is determined by the quick process of the interplay between gain and loss .
somewhere at its tail however , it will switch over to the slow process with excitation of internal degrees of freedom of the system .
the states with gain and loss as well as the resonance states excited by the fluctuations , can each interact via a common environment into which the corresponding states are embedded .
the two environments ( channels ) will _ never _ mix .
this request is guaranteed in our formalism due to the very different time scales of the two processes . in any case , this so - called external mixing of the states occurs additionally to the direct so - called internal mixing of the states which is supposed in our calculations to be involved in the complex energies @xmath170 of the states . according to the numerical results of our paper ,
the fluctuations are very robust .
they appear in a relatively large finite parameter range around eps . finally , we mention a few interesting results which are characteristic of open quantum systems including those considered in the present paper
. 1 . the states of an open quantum system may interact via a common environment into which the system is embedded .
this mixing is called usually external interaction .
2 . the states of an open quantum system may have positive or negative widths .
the states with positive width @xcite gain particles ( or information ) from the environment while those with negative width loose particles ( or information ) due to their coupling to the environment . as function of a parameter
, gain may pass into loss and vice versa @xcite .
the phases of the eigenfunctions of a non - hermitian operator are , generally , not rigid , see figs .
[ fig1 ] to [ fig7 ] and @xcite .
irreversible processes determine the evolution of an open quantum system up to the occurrence of orthogonal eigenstates at maximum width or level repulsion , see figs .
[ fig4 ] , [ fig6 ] , [ fig7 ] and @xcite .
we mention further that results similar to those discussed above for systems with two states , appear also in calculations for systems with more than two states , see @xcite .
in our paper , we provided some results for a two - step process which we obtained in the framework of the non - hermitian formalism @xcite for the description of open quantum systems .
the first step is the interplay between gain and loss of particles from an environment , while the second step is the excitation of a resonance state .
the two steps are treated as two parts of the whole process .
the total process might simulate photosynthesis : the first step is the capture of light in the light - harvesting complex while the second step is the transfer of the excitation energy to the reaction center which stores the energy from the photon in chemical bonds .
that means : gain simulates the acceptor for light , and loss stands for the donor which excites a resonance state and simulates the coupling to the sink .
altogether , the energy of the light is transferred to the reaction center of the light - harvesting complex .
the obtained results are very robust , and fluctuations play an important role .
the results show some characteristic features which correspond , indeed , to those discussed in the literature for the photosynthesis .
most interesting are the following results of our calculations : 1 . the efficency of energy transfer is nearly 100 % ; 2 .
the energy transfer takes place at a very short time scale ; 3 .
the storage of the energy in the reaction center occurs at a much longer time scale ; 4 .
according to points 2 and 3 , the decay is bi - exponential . in future studies
, the theoretical results have to be confirmed by application of the formalism to the description of concrete systems in close cooperation between theory and experiment . * acknowledgment * we are indebted to j.p .
bird for valuable discussions .
at an ep the two eigenfunctions of a non - hermitian hamilton operator @xmath6 are exchanged according to ( [ sec8 ] ) . in more detail : tracing the eigenfunctions of @xmath6 as function of a certain parameter @xmath155 , the two eigenfunctions jump according to ( [ sec8 ] ) at the critical parameter value @xmath171 which defines the position of the ep .
some years ago , it has been shown @xcite that the influence of the ep is not restricted to the jump occurring at @xmath172 .
it appears rather in a finite parameter range of @xmath155 in which the wavefunctions of the two states are mixed according to @xmath173 with @xmath174 .
the two wavefunctions @xmath175 vary smoothly ( i.e. without any jump of the sign of their components ) everywhere but at @xmath171 . using the representation @xmath176 @xmath177
depends on the parameter @xmath155 . after removing a common phase factor , it follows @xmath178 and @xmath179 , respectively , in approaching @xmath171 ; and @xmath180 or @xmath181 when @xmath155 is far from the critical region around @xmath172 . in between the values @xmath182 and @xmath183 ,
the angle @xmath177 varies smoothly .
corresponding to the dependence of @xmath177 on the parameter @xmath155 , also the phase rigidity ( [ eif7 ] ) depends on this parameter in a certain finite parameter range .
the phase rigidity is determined by the ratio @xmath184 which approaches @xmath97 at the ep ( due to @xmath185 ) and @xmath12 far from the ep ( because here the wavefunctions are almost real ) .
the intermediate values of the phase rigidity @xmath96 are determined by the expression ( [ app3 ] ) calculated with the actual values @xmath186 and @xmath187 .
the results of these calculation give values for @xmath96 that are between the two limiting values @xmath97 and @xmath12 . fig .
[ fig7 ] shows numerical results for the eigenfunctions of @xmath6 which are obtained for calculations with two channels and two states in the second channel .
we see the values @xmath96 for the states of the first channel ( see figs . [ fig2 ] and [ fig4 ] ) as well as those for the two states of the second channel .
they are independent of one another .
the @xmath96 related to the second channel are constant in the whole parameter range @xmath3 shown in the figures ( which is defined for the first channel ) .
they may be smaller than 1 , see fig .
[ fig7 ] .
the results obtained for the phase rigidity @xmath188 of the two states in the second channel , may be considered as a proof of the finite parameter range in which an ep influences the properties of the system .
the value @xmath188 is nothing but an expression for the distance of the system from an ep in the second channel : the larger @xmath96 , the more distant is the ep , while @xmath108 indicates that the ep is approached .
thus , the value @xmath96 in the eigenfunction pictures of the two - channel system with two states in the second channel additionally to those of the one channel system ( figs .
[ fig2 ] and [ fig4 ] ) , allows us to determine the position of the ep in the second channel .
scully , phys .
* 104 * , 207701 ( 2010 ) + e. a. sete , a. svidzinsky , h. eleuch , r. d. nevels and m. o. scully journ .
mod . opt . * 57 * , 1311 ( 2010 ) + a.a .
svidzinsky , k.e .
dorfman , and m.o .
scully , phys .
a * 84 * , 053818 ( 2011 ) d. ferrari , g.l .
celardo , g.p .
berman , r.t .
sayre , and f. borgonovi , j. phys .
c * 118 * , 20 ( 2014 ) + g.l .
celardo , g.g .
giusteri , and f. borgonovi , phys .
b * 90 * , 075113 ( 2014 ) + g.g.giusteri , g.l .
celardo and f. borgonovi , phys .
e * 93 * , 032136 ( 2016 ) we underline that we consider open quantum systems with gain and loss .
this should not be confused with the consideration of exactly balanced gain and loss in pt - symmetric systems which are neither open nor closed , but nonisolated according to the definition in , e.g. , c.m .
bender , journal of physics : conference series * 631 * , 012002 ( 2015 ) .
in contrast to the definition that is used in , for example , nuclear physics , we define the complex energies before and after diagonalization of @xmath189 by @xmath190 and @xmath191 , respectively , with @xmath192 and @xmath193 for decaying states . this definition is useful when discussing systems with _ gain _
( positive widths ) and _ loss _ ( negative widths ) .
the coalescence of two eigenvalues of a non - hermitian operator should not be confused with the degeneration of two eigenstates of a hermitian operator .
the eigenfunctions of two degenerate states are different and orthogonal while those of two coalescing states are biorthogonal and differ only by a phase , see eq .
( [ sec8 ] ) .
numerical calculations with @xmath194 have shown that , in such a case , the states of the second channel mix also via the first channel .
the obtained results are abstruse from the point of view of physics , although they are mathematically correct . | photosynthesis is the basic process used by plants to convert light energy in reaction centers into chemical energy .
the high efficiency of this process is not yet understood today . using the formalism for the description of open quantum systems by means of a non - hermitian hamilton operator ,
we consider initially the interplay of gain ( acceptor ) and loss ( donor ) . near
singular points it causes fluctuations of the cross section which appear without any excitation of internal degrees of freedom of the system .
this process occurs therefore very quickly and with high efficiency .
we then consider the excitation of resonance states of the system by means of these fluctuations .
this second step of the whole process takes place much slower than the first one , because it involves the excitation of internal degrees of freedom of the system . the two - step process
as a whole is highly efficient and the decay is bi - exponential .
we provide , if possible , the results of analytical studies , otherwise characteristic numerical results .
the similarities of the obtained results to light harvesting in photosynthetic organisms is discussed . |
in constructing a higher dimensional quantum field theory , the regularization and the continuum limit are two important keys . particularly , the problem of the limit associates with hard difficulty in the higher dimensional case .
statistical mechanics usually insists that critical behaviors of the phase transition in the theory are equivalent to those of a mean field theory , which has only a trivial fixed point .
many pioneering works on lattice gauge theories were trying to overcome the difficulty @xcite , although the continuum limit is not taken strictly . where is the continuum limit ? by using well - known 4-dimensional theories such as qed and qcd ,
is it possible to construct the continuum limit by the related critical behavior near the critical point ?
our purpose is to construct a @xmath0-dimensional pure yang - mills theory by arranging a number of 4-dimensional yang - mills theories with appropriate couplings .
a @xmath0-dimensional lattice space is decomposed into @xmath1 layers with 4 dimensions like as fig.[fig : layer ] . originally , fu and nielsen investigated the system to show dynamical dimensional reduction by using the characteristic vacuum which confines for extra dimensional directions and deconfines for 4 dimensions called a layer phase @xcite .
we do not assume that the layer phase exists , but use only the decomposition . from now , we focus on a 5-dimensional theory to study the possibility of the construction for higher dimensional field theories explicitly but an extension to general dimensions
shall be mentioned in the final section .
our starting action for @xmath2 gauge group is written as @xmath3 + \frac{\beta_{5}}{2}\sum_{p_{5}}\left[2-\tr u_{p_{5}}\right ] , \label{lact}\ ] ] where @xmath4 is proportional to a usual 4-dimensional coupling constant inside a layer and @xmath5 is a coupling constant between neighboring layers .
@xmath6 implies a plaquette inside a layer and @xmath7 does a plaquette between neighboring layers .
to display the phase diagram explicitly , two order parameters are introduced ; ( 1 ) creutz ratio ( @xmath8 ) for 4-dimensional wilson loops and ( 2 ) 5-dimensional polyakov loop ( @xmath9 ) .
since a theory with @xmath10 is equivalent to a pure 4-dimensional yang - mills theory which is one phase , we expect a phase structure such as fig.[fig : phase1 ] .
we have calculated numerically these parameters for the system and obtained the numerical phase diagram in fig.[fig : phase2 ] . in the figure , along line i we can see both 4-dimensional and 5-dimensional deconfinement transitions in the large @xmath5 . by wider calculations along lines
ii , we recognize the phase transition is of 1st order near @xmath11 .
the results of lines iii and iv suggest that both 4-dimensional and 5-dimensional deconfinement transitions in the small @xmath5 are of 2nd or weakly 1st order , because we can not see any hysteresis loop in the coupling region .
this critical point in the @xmath12 is noted as @xmath13 .
our surprising remark is that the diagram fig.[fig : phase2 ] is not quantitatively changed for various @xmath14 and its stable property may help us to take the limit @xmath15 .
our simple consideration finds our way to existence of a 4-dimensional continuum system with finite inter - layer coupling(@xmath5 ) . near the critical point ( @xmath16 ) , the inverse ( @xmath17 ) of correlation length for polyakov loop
is written as @xmath18 where the value of @xmath19 is approximately 0.6 and @xmath20 implies the critical exponent which is 0.5 in a mean field theory .
we call the picture that many gauge fields on different layers interacts each other with finite coupling as multi - layer world .
it is noted that link variables with 5-th direction behave as bi - fundamental fields with a finite coupling .
can we construct a 5-dimensional space not an internal space ?
a straightforward way uses excitation masses @xmath21 between layers corresponding to kaluza - klein(k - k ) modes , @xmath22 \right .
\nonumber \\ & + & \langle l_5\rangle^2 \left.\left [ \begin{array}{ccccc } 1 & & & & -1\\ & 0 & & 0 & \\ & & \ddots & & \\ & 0 & & 0 & \\
-1 & & & & 1 \\
\end{array } \right ] \right\ } .
\label{kkmass}\end{aligned}\ ] ] for a lower excited mode with label @xmath23 , the simple formula of the mass is obtained as @xmath24 in order to remain finite masses of k - k modes , we must keep @xmath25 finite for large @xmath26 and @xmath27 with small @xmath28 from eq.([kkm ] ) . to go through a 5-space from our 4-dimensional multi - layer world
, we need to balance inter - layer dynamics and inside - layer one , @xmath29 from eq.([defcalr ] ) , three parameters @xmath30 tunings are necessary ( see fig.[fig : env ] ) . comparing @xmath31 with eq.([kkm ] )
, a lattice spacing @xmath32 along 5th dimension can be defined by @xmath33 .
we summarize main three steps ; @xmath34 to find a second order phase transition in the meaning of 4-dimensional statistical mechanics .
@xmath35 to take a 4-dimensional continuum limit ( multi - layer world ) .
@xmath36 to compare a 4-dimensional scale with an extra dimensional scale , i.e. confinement and kaluza - klein modes . for detailed analysis ,
see ref .
@xcite .
finally , the following problems remain ; @xmath37 further detailed study for large @xmath26 and @xmath27 .
@xmath38 to estimate contribution of bi - fundamental field for @xmath8 .
@xmath39 to recover the rotational symmetry relating to anisotropy between 4-scale and 5-scale .
@xmath40 6 or higher - dimensional
extension of these decomposition is straightforward but their phase diagram analysis is not so easy . 9 m. creutz , phys .
43 ( 1979 ) 533 .
lang , m. pilch and b.s .
skagerstam , int . j. mod
a3 ( 1988 ) 1423 .
h. kawai , m. nio and yuko okamoto , prog .
88 ( 1992 ) 341 .
s. ejiri , j. kubo and m. murata , phys .
d62 ( 2000 ) 105025 .
s. ejiri , s. fujimoto and j. kubo , phys .
d66 ( 2002 ) 036002 .
y. k. fu and h.b .
nielsen , nucl .
b236 ( 1984 ) 167 . m. murata and h. so , hep - lat/0306003 . | a higher dimensional lattice space can be decomposed into a number of four - dimensional lattices called as layers . the higher dimensional gauge theory on the lattice
can be interpreted as four - dimensional gauge theories on the multi - layer with interactions between neighboring layers .
we propose the new possibility to realize the continuum limit of a five - dimensional theory based on the property of the phase diagram . |
hypernuclei are bound systems of non - strange and strange baryons . in present facilities ,
hypernuclei are created with hadronic reactions as ( @xmath4,k ) at brookhaven and kek or electroproduction ones as ( e , ek ) at tjnaf. the decay of those objects proceeds via the weak interaction which is 10 orders of magnitude slower than the strong one and violates parity , isospin and strangeness . for the very light hypernuclei
the mesonic decay mode ( @xmath5 ) is dominant , but as @xmath6 increases so does the pauli blocking acting on the outgoing nucleon , and hypernuclei mainly decay via the one - nucleon induced non - mesonic mode , @xmath7 . since the pioneering phenomenological model of block and dalitz @xcite
, many approaches have been developed to understand the dynamics of the decay , and the results have been collected in extensive review articles @xcite .
many works are based on a meson exchange picture , either using a simple one - pion - exchange mechanism @xcite , or including also heavier meson exchanges , such as the @xmath8 @xcite or the complete pseudoscalar and vector meson octets @xcite .
the effect of correlated - two pion exchange mechanism @xcite as well as the role of @xmath9 transitions , implemented in a meson - exchange picture @xcite , have also been investigated . a four quark weak transition
effective hamiltonian , corrected by qcd , which contains both @xmath10 and @xmath11 transitions , has also been applied in the study of the weak decay of hypernuclei @xcite . in general , the weak decay rates are reasonably reproduced by the models but the ratio of neutron induced ( @xmath12 ) to proton induced ( @xmath13 ) decays , @xmath2 , turns out to be smaller than the experimental value , which is of the order of 1 or larger @xcite , although new recent theoretical progress has been achieved into the solution of this puzzle @xcite .
in addition to @xmath7 , the decay can also proceed via the two - nucleon induced process , @xmath14 , originally studied in ref .
@xcite , which amounts to about 15% of the total width @xcite and can not be neglected in the experimental analysis trying to extract the ratio @xmath2 .
the high momentum of about 400 mev / c transferred in the @xmath7 reaction makes this process quite sensitive to short range physics . as a consequence , the strong baryon - baryon interaction both in the initial and final states plays a quite important role .
the purpose of the present work is to revisit the one - meson - exchange model of ref .
@xcite , with an especial interest in quantifying the effect of final state interactions , as well as the uncertainties of different model ingredients . in doing so
, we will also point out a sign error encountered in certain transitions mediated by the exchange of strange mesons which , when corrected , gives rise to a considerably increase in the @xmath2 ratio .
we will analyze the sensitivity of our results to the way final state interactions of the emitted nucleons are implemented , using different prescriptions to obtain the @xmath15 scattered wave function .
our study concludes that , with the appropriate treatment of final state interactions and with the correct sign for the contribution of the strange mesons , the one - meson - exchange model is able to describe very satisfactorily most of the measured observables , including the elusive ratio @xmath2 .
the weak transition potential is obtained by following the model of ref . @xcite . in analogy to one - boson - exchange ( obe ) based models of the strong interaction ,
the present formalism includes not only the exchange of the long - ranged pion , but also more massive mesons which account for shorter distances .
this potential has been presented in previous papers and , therefore , it is not going to be discussed here in a great detail . however , we would like to show its expression in coordinate space , which represents a compact way of including all the mesons and transition channels in the mechanism .
the @xmath16-space potential then reads : @xmath17 where the index @xmath18 runs over the different mesons exchanged ( @xmath19 represents @xmath20,k,@xmath21,k@xmath22 ) and @xmath23 over the different spin operators : central spin independent ( c ) , central spin dependent ( ss ) , tensor ( t ) and parity violating ( pv ) .
the angular dependence is represented by the @xmath24 operator , explicitly given by : @xmath25 { \hat r } & \mbox{pv ( for vector mesons ) } \ , { \rm , } \\ \end{array } \right.\ ] ] while the isospin operator , @xmath26 , takes the form @xmath27 for isovector mesons ( @xmath4,@xmath8 ) , @xmath28 for isoscalar mesons ( @xmath29,@xmath30 ) and a combination of both operators for the isodoublet ( @xmath0,@xmath1 ) .
the different pieces of the potential , @xmath31 , are found by fourier transforming the potential in momentum space . in particular , for pseudoscalar mesons the @xmath32-space potential reads : @xmath33 where @xmath34 is the fermi weak constant times the pion mass squared , @xmath35 the coupling at the strong baryon - baryon - meson ( @xmath36 ) vertex , @xmath37 and @xmath38 the parity violating ( pv ) and parity conserving ( pc ) weak couplings respectively , @xmath39 the meson mass , and @xmath40 ( @xmath41 ) the average of the baryon masses at the strong ( weak ) vertex .
we want to note here that the convention has always been to direct the momentum towards the strong vertex and , in connection to this , care must be taken when this expression is applied to @xmath0 ( @xmath1 ) exchange , since in this case vertex 1 refers to @xmath42 and vertex 2 to @xmath43 .
therefore , the combination of the strange meson transition amplitudes with the other non - strange meson contributions requires to exchange , in the former ones , the space , spin and isospin quantum numbers between the @xmath44 and @xmath45 in the initial state and between the two emitted nucleons in the final state . for a particular transition from an initial @xmath46 channel to a final @xmath47 one , this exchange operation introduces the additional factor : @xmath48 where @xmath49 is the nucleon isospin and @xmath50 the isospin assigned to the @xmath44 isospurion , which can take the values @xmath51 or @xmath52 depending on whether one is considering a change in isospin of 1/2 or 3/2 . the way to implement @xmath9 transitions in the present mechanism can be found in ref .
the factor in eq .
( [ eq : fac1 ] ) is nothing but a sign in front of the pc and/or pv amplitudes , which is : @xmath53 it is common to write the total decay as the sum of definite spin - isospin amplitudes .
for example , for s - shell hypernuclei one has@xcite : @xmath54 if we adopt the @xmath55 rule which we always do for pseudoscalar meson exchange @xcite , the kaon and pion potentials differ in a minus sign only in the @xmath56 amplitude in ( [ transitions ] ) , the only pv amplitude with @xmath57 .
in order to account for finite size effects we include a monopole form factor at each vertex , @xmath58 , where the value of the cut - off , @xmath59 , depends on the meson ( @xmath60 ) .
the updated expression for the regularized potential was given in ref .
@xcite . in previous calculations we used the cut - offs given by the jlich b interaction @xcite .
the reason was that unlike the early @xmath61 nijmegen model@xcite , which used different cut - offs depending on the irreducible representation of the baryon - baryon channel , the jlich b was the only obe model which used different cut - offs depending on the meson .
however , the nijmegen group has recently made available new baryon - baryon interactions in the strangeness @xmath62 and @xmath63 sectors @xcite .
these potentials are based on su(3 ) extensions of the models in the @xmath64 and @xmath65 sectors , which are fitted to the experimental data .
the authors of ref .
@xcite give six different models , which fit the available @xmath15 and @xmath61 scattering data equally well but are characterized by different values of the magnetic vector @xmath66 ratio , ranging from 0.4447 ( model nsc97a ) to 0.3647 ( model nsc97f ) .
the advantage of these new models is that the form factors depend on the su(3 ) type of meson .
the momentum space potential for each meson is multiplied by the regularizing factor of gaussian type , @xmath67 , with a cut - off @xmath68 for the singlet meson , @xmath69 for the non - strange members of the meson octet and @xmath70 for the strange meson . in order to accommodate to our own formalism , which uses a monopole form factor at each vertex
, we will match the gaussian to a function of the type @xmath71 ^ 2 $ ] at @xmath72 mev / c , the most relevant momentum transfer in the @xmath7 process .
since , by definition , both functional forms give 1 at @xmath73 , our alternative expression with the modified cut - offs @xmath74 , listed in table [ tab : formfac ] , gives an excellent reproduction of the gaussian nsc97 form factors up to a momentum transfer of about 600 mev / c . as it is well known
, one of the sources of uncertainty in obe models comes from the coupling constants between baryons and mesons . in the strong sector
the different interaction models use su(3 ) in order to obtain the @xmath36 couplings that are not constrained experimentally . in the weak sector , only the decay of the @xmath44 and @xmath75 hyperons into nucleons and pions can be experimentally observed . for the other mesons
, @xmath76 represents a convenient tool to obtain the pv amplitudes , while for the pc ones , we use a pole model@xcite with only baryon pole resonances .
details of how these coupling constants are derived can be found in ref .
@xcite and , in the recent work on the weak decay of double-@xmath44 hypernuclei @xcite , one can find an updated list of the s - wave ( pv ) and p - wave ( pc ) coupling constants , the latter obtained from the pole model using different parametrizations of the strong bb interaction .
a measure of the amount of parity - violation in the weak decay , is given by the asymmetry in the angular distribution of protons coming from the decay of polarized hypernuclei . this asymmetry is given by : @xmath77 where @xmath78 is the polarization of the hypernucleus , characteristic of the production reaction , and @xmath79 the hypernuclear asymmetry parameter , characteristic of the weak decay . in the expression above
, @xmath80 is the spin of the hypernucleus , @xmath81 is the @xmath80-spin operator along the direction perpendicular to the reaction plane , and @xmath82 the hypernuclear transition amplitude . in order to compare with experiments
, one has to multiply @xmath79 by the model dependent quantity @xmath78 , which has to be determined theoretically for each hypernucleus@xcite . by using a shell - model for the initial hypernucleus and assuming spherical configuration with no mixing
, one can express this amplitude in terms of two - body transitions , @xmath7 .
the dependence of weak decay observables on the deformation of the initial ( p - shell ) hypernucleus was investigated in ref .
@xcite , by means of the nilsson model with angular momentum projection .
it was found that deformation effects change these observables by about 10% from the spherical limit , a deviation that although non - negligible it is smaller than the present experimental uncertainties .
because of the lack of stable hyperon beams , access to the @xmath83 @xmath84 interaction is limited right now to the decay of hypernuclei .
hence , extracting information of the elementary weak two - body interaction requires a careful investigation of the many - body nuclear effects present in the hypernucleus . on the one hand
, one must consider that the interacting nucleon and @xmath44 hyperon are bound in the nucleus and they should be described by bound - state single - particle wave functions , obtained from appropriate mean - field or hartree - fock potentials .
note , however , that since the mass excess of 176 mev in the initial state is converted into kinetic energy of the final particles , the nucleons emerge with a large momentum of about @xmath85 mev / c and the decay is not very sensitive to the details of the single - particle wave function . in ref .
@xcite it is shown that the decay rates obtained from various realistic @xmath44 wave functions differ by at most 15% .
on the other hand , the large momentum transfer implies that the @xmath86 decay process is very sensitive to the short - range correlations induced by the strong interaction . in the initial system , one
must then replace the mean - field two - particle @xmath84 wave function by a correlated one that accounts for the effects of the strong @xmath61 interaction at short distances , which are not considered in mean - field models . correlated wave functions , obtained from the soft and hard core nijmegen @xmath61 interactions @xcite by solving the corresponding finite nucleus scattering amplitude ( @xmath87-matrix ) @xcite , were compared in ref .
the differences between the wave functions obtained with the two models are already significant below 0.75 fm and give rise to decay rates that differ by slightly more than 15% . however ,
since the present @xmath61 models are not constrained enough by the scattering data to resolve this discrepancy , we will admit this uncertainty in our initial @xmath84 wave function .
this uncertainty also justifies the use of a spin - independent parametrization for the @xmath84 correlation function which , when multiplied with the uncorrelated one , was shown to give a decay rate in between those using the soft and hard core correlated wave functions @xcite .
any realistic calculation must also take into account the fact that the two nucleons emerging from the decay feel their mutual influence , as well as that from the residual ( a-2)-particle system .
however , as mentioned before , the most important contribution to the decay comes from the so - called _ back - to - back _ kinematical situation , in which the two nucleons emerge with the largest possible momentum of around 400 mev / c . for these fast moving nucleons the distortions with the residual nucleus
should be small , and the importance of such effects are further diminished for inclusive observables such as the decay rates . in other words , one could take into account the interaction of the emitted nucleons with the residual nucleus through an optical potential .
however , the real part should play a minor role at those high energies , and the imaginary part will remove flux from the @xmath15 channel which will reappear in other multi - nucleon channels , such that the total strength or decay rate , which is the quantity we are interested in , will not be modified . we would face a completely different situation if we were interested in calculating the energy distribution of the nucleons , in which case one should consider these final state interaction effects , which can change the energy , direction and charge of the primary nucleons emitted in the weak @xmath7 transition , producing as well low energy secondary nucleons .
the monte carlo simulation of ref .
@xcite finds indeed that the final state interactions affect the nucleon distributions mostly at energies below 50 mev . in the particular case of @xmath88c
, it is found that final state interactions produce roughly 1/3 more protons at those low energies , half of which are due to charge - exchange reactions , but barely affect the distribution at higher energies and leave the decay rate intact . since in the present work we will only be interested in the decay rates
, we will limit the treatment of final state interactions to those related to the mutual influence between the two emitted nucleons .
in contrast to the strong @xmath84 interaction , the nucleon - nucleon one is much better constrained by the huge amount of scattering data and , although differences in the wave function can also be observed with various potential models , they are only significant below 0.5 fm , having a moderate influence in the decay rates , as we will see . in the literature , one finds a variety of ways of dealing with the fsi between the emitted nucleons .
some works do not include fsi , others use a phenomenological correlation function that simply multiplies the uncorrelated one , and others use various approximations to the @xmath15 scattering equation . in the next section we will analyze the sensitivity of the non - mesonic decay observables to these various choices of fsi . the details on how the scattering @xmath15 wave function is obtained from the lippmann - schwinger ( @xmath89-matrix ) equation , as well as from other simplified approximations , are given in the appendix .
we start this section by showing the weak decay observables for and obtained with the model parameters used in ref .
@xcite , but correcting the sign in the appropriate @xmath0 and @xmath1 amplitudes , as pointed out in sect .
ii . hence , we use the coupling constants of the nsc89 strong @xmath61 interaction @xcite , together with a monopole form factor of the type @xmath90 at each vertex .
each meson has a different cut - off value , @xmath44 , and , since this particular nijmegen model distinguishes cut - offs only in terms of the transition channel , we choose , as in ref .
@xcite , the values of the jlich @xmath61 interaction @xcite .
the results are collected in table [ tab : paper1 ] and we remind here the main results given in ref .
@xcite for , namely a non - mesonic rate of @xmath91 , where @xmath92 10@xmath93 s@xmath94 is the decay rate of a free @xmath44 , and a neutron - to - proton ratio @xmath95 . as we see
, the corrected model gives a slightly larger decay rate , together with a remarkable increase in the neutron - to - proton ratio .
this is due to the fact that the corrected sign gives rise to a constructive interference between the @xmath4 and @xmath0 contributions to the @xmath96 @xmath97 pv amplitude , instead of the destructive one found in ref .
@xcite and displayed in fig .
8 of that work . on the other hand ,
the incorporation of @xmath0 exchange to the @xmath4 mechanism lowers the pc amplitudes .
this was pointed out in fig .
7 of ref .
@xcite and figs . 4 and 5 of ref .
@xcite , where the contributions of all different mesons are displayed , the @xmath4 and @xmath0 ones showing a destructive interference in the pc amplitudes .
the present constructive interference in the pv channels is especially relevant in the neutron - induced rate , @xmath98 , which lacks from the pc tensor transition for @xmath99 initial states .
conversely , the proton - induced rate , @xmath100 , is dominated by the pc tensor amplitude , hence its value is lowered when @xmath0 exchange is added to the @xmath4 mechanism , although the reduction will now be more moderate due to the constructive interference in the less dominant pv transition
. altogether , the correct incorporation of @xmath0 exchange produces , in comparison to the results given in ref .
@xcite , a more moderate decrease in the total non - mesonic decay rate , hence a slightly larger rate is now obtained , and a ratio which is almost a factor of 3 larger .
we note that these effects of @xmath0 exchange have also been pointed out in ref .
@xcite , where the @xmath4 and @xmath0 mechanisms are considered together with a description of the decay in terms of quark degrees of freedom , and in the recent work of ref .
@xcite , where , in addition of @xmath4 and @xmath0 exchange , the role of correlated two - pion exchange was also studied .
results for the asymmetry of the emitted protons from the weak decay of polarized hypernuclei are also listed in table [ tab : paper1 ] .
since this is an observable tied to the interference between pv and pc amplitudes , it will also be influenced by the change of sign in the abovementioned pv @xmath0 and @xmath1 transitions .
indeed , the magnitude of the asymmetry @xmath79 has increased almost a factor of 2 from the values given in ref .
@xcite which are @xmath101 for and @xmath102 for .
the fact that the new nijmegen potentials @xcite include different form - factors depending on the meson , makes it possible to treat all the strong interaction ingredients involved in the weak decay transition @xmath103 consistently within the same model .
we remind that the strong interaction plays a role not only in the strong vertices and form factors , but also in the pc piece of the weak vertex through the pole model and in the determination of the distorted final state @xmath15 wave function . in table
[ tab : paper2 ] we show the decay observables for and obtained with the models nsc97a and nsc97f .
we observe substantial differences between both models .
these might come from the different coupling constants used in the strong and weak vertices of the transition @xmath7 amplitude , or from the different distorted wave functions . to disentangle both effects ,
we show in table [ tab : paper3 ] the results for @xmath104he , obtained using the same strong interaction model for the vertices , which we choose to be the nsc97f one , and different distorted nn wave functions obtained with four nn interaction models , namely the nsc97a and f @xcite , the nsc93 @xcite and the bonn b potential @xcite . the resulting nn wave functions in various relevant channels are shown in fig .
[ fig : pots ] for a relative momentum @xmath105 mev / c .
some minor differences appear only in the short distances , which should affect the decay rates moderately . by inspecting the results of table [ tab : paper3 ] , one can indeed see that the non - mesonic decay rates in vary from 0.3 to 0.4@xmath106 , the neutron - to - proton ratio from 0.45 to 0.5 , and the asymmetry from @xmath107 to @xmath108 . in conclusion
, it is mainly the differences associated to the strong couplings and their influence on the weak ones the origin for the differences observed in table [ tab : paper2 ] .
going back to this table , we see that there are already some differences between the results obtained with models nsc97a and nsc97f when the @xmath0 meson contribution is added to the @xmath4 meson one , but this can not explain the differences in the final results when all mesons are included .
compare , for instance , the nonmesonic decay rate in of 0.43 obtained with the nsc97a model versus the value 0.32 obtained with the nsc97f one .
in fact , by analyzing the weak couplings listed in ref .
@xcite , we can see some appreciable differences for the @xmath109 coupling constants , which are larger in the nsc97f model , thus enhancing the destructive interference of the @xmath0 meson and producing a reduced rate in this case .
we also observe that the weak vector pc @xmath110 coupling constant obtained from the nsc97a model is one order of magnitude larger than that obtained from the nsc97f one .
this is a consequence of a subtle cancellation in the pole model expression , which involves the difference between the strong @xmath110 and @xmath111 coupling constants .
although both models use the same strong @xmath110 couplings , the differences in the @xmath111 ones induced by the freedom allowed in the magnetic @xmath66 ratio may give rise , as is the case here , to substantial differences in the values of the pc weak coupling constants . the larger value found in the case of the nsc97a model
enhances the constructive interference of the @xmath30 meson contribution and explains the final difference between the decay rate obtained with the two nijmegen models .
this discussion makes also clear that one must admit this amount of uncertainty on meson - exchange models describing the nonmesonic weak decay of hypernuclei .
we should therefore conclude that the nonmesonic decay rate for ranges between 0.55 and 0.73@xmath106 , the neutron to proton ratio between 0.29 and 0.34 , and the hypernuclear asymmetry , more stable , is 0.36 . in ,
the rate obtained lies in the range 0.320.43@xmath106 , in the range 0.340.46 and the hypernuclear asymmetry parameter is @xmath112 .
we observe that our results for the nonmesonic decay rates compare satisfactorily with the experimental data , especially after considering that the size of the nonmesonic two - nucleon induced mechanism , not included here , is around 0.3@xmath106 for medium to heavy hypernuclei @xcite .
the ratio is considerably larger than that found in our previous works and it now lies practically within the lower side of the experimental errors .
we note that a larger ratio could have been obtained , as we show below , if we had used a phenomenological treatment of fsi effects or we had ignored them althogether .
however , we want to stress here that the results shown in table [ tab : paper2 ] have been obtained consistently within the same strong interaction model and that involves also the dynamical generation of the corresponding @xmath15 wave function through a lippmann - schwinger equation . the asymmetry parameter @xmath79 has increased by more than a factor of two with respect to our previous works . as it has already been mentioned in connection to the results of table [ tab : paper1 ] , this is a direct consequence of the change of sign in the @xmath0 and @xmath1 exchange pv amplitudes for transitions that do not change the spin @xmath113 , such as @xmath56 in eq .
( [ transitions ] ) .
indeed , this amplitude interferes with @xmath114 and @xmath115 and the magnitude of the latter tensor amplitude is of especial relevance in meson - exchange models , as the one employed here .
the increased asymmetry values compare less favourably with the experimental data .
while the result for is still within errors , the large negative value obtained here for is very far away from the positive small result obtained in ref .
we note that , using a weak coupling scheme , one can relate the hypernuclear asymmetry parameter @xmath79 to that associated to the elementary reaction @xmath116 , @xmath117 ( see eq . ( 9 ) in ref .
@xcite ) . in the case of , @xmath118 and for
the relation is @xmath119 .
therefore the parameter @xmath117 of the elementary reaction that can be extracted from our results in table [ tab : paper2 ] , turns out to be very similar in both hypernuclei .
however , the present experimental results seem to contradict this finding because a small positive value is obtained for @xcite , while a large negative one was found in @xcite . in ref .
@xcite arguments are given to try to understand these differences on the basis of the fact that is a s - shell nucleus and a p - shell one , the latter one allowing transitions from initial @xmath84 relative @xmath120-wave sates .
however , it was already noticed that most of the decay for p - shell nuclei comes from relative @xmath113-wave initial states @xcite and we have checked here that removing the initial @xmath120-wave amplitudes , the decay rate of using the nsc97f model gets slightly reduced to @xmath121 and the asymmetry parameter @xmath79 only changes to @xmath122 .
hence , meson - exchange models do not explain the present experimental differences between the elementary asymmetry for and , and this seems to be a new puzzle in the study of the weak decay of hypernuclei .
this poses a challenge to the theoretical models and calls for further experimental efforts .
besides valuable new information from hypernuclear weak decay experiments , it is also possible to gain direct information from the inverse reaction @xmath123 , already studied theoretically @xcite and presently under preparation at rcnp ( osaka ) .
one of the purposes of this paper was to study the influence of fsi in the @xmath7 transition .
the results of table [ tab : paper3 ] have already shown that , using nn wave functions obtained from various realistic nn potentials , one obtains similar values for the weak decay observables .
however , one finds in the literature several possibilitites for the treatment of fsi .
they are either ignored , or included through a phenomenological function , such as @xmath124 with @xmath125 fm@xmath94 , multiplying the free nn plane wave , or through a distorted wave function built up from the @xmath0-matrix rather than from the @xmath89-matrix . in fig .
[ fig : fsiall ] we show the resulting nn wave functions in various channels for a relative momentum @xmath105 mev / c , obtained from the various approaches to fsi .
the solid lines represent the real and imaginary parts of the wave function obtained from the proper solution of the lippmann - schwinger equation ( @xmath89-matrix ) .
the imaginary part is the minoritary component in all panels .
the dotted line represents the result obtained when only the real part of the @xmath89-matrix is retained in the construction of the distorted wave function from eq .
( [ eq : psipw ] ) , which is further approximated by its principal part value .
this was the prescription followed in ref .
the dashed line corresponds to a @xmath0-matrix approximation , which would be the appropriate solution for standing waves , i.e. non - propagating solutions , as is the case in the nuclear medium , where pauli blocking prevents two interacting nucleons of the fermi sea from exciting intermediate nn states that are already occupied by other nucleons .
finally , the dot - dashed line displays the product of the phenomenological correlation function with the corresponding non - interacting bessel function .
we observe important differences beween the various approaches to the treatment of fsi , especially in the case of a phenomenological correlation function .
obviously , this will have an influence on the decay observables .
as an example , we show in table [ tab : paper4 ] the results for , obtained using the nsc97f model and taking all these different treatments for the distorted nn wave functions . the rates can differ substantially , by more than a factor of two in some cases . the neutron - to - proton ratio can also be fictitiously enhanced by using the phenomenological approach to fsi effects , or no fsi at all , while the asymmetry remains more stable .
we have revisited the ome model of ref .
@xcite for the weak nonmesonic decay of hypernuclei to the light of the new nijmegen baryon - baryon potentials @xcite .
these strong interaction models influence the weak decay mechanism , not only through the coupling constants and form factors at the strong vertex involved in the @xmath127 reaction , but also via the pc piece of the weak vertex , obtained from a pole model , as well as from the corresponding correlated wave functions for the initial @xmath84 and final @xmath15 states . we have paid an especial attention to disentangle the influence of the various model ingredients involved in the decay mechanism . in particular , we have found that the uncertainties tied to the strong coupling constants are a source of uncertainty on the nonmesonic weak decay observables .
employing different strong models of the nijmegen group , and working consistently within each one , we estimate the decay rate of to be in the range 0.320.43@xmath106 , the neutron - to - proton ratio in the range 0.340.46 and the hypernuclear asymmetry parameter @xmath79 around @xmath112 . in the case of
, the decay rate turns out to be in the range 0.550.73@xmath106 , the neutron - to - proton ratio in the range 0.290.34 and the asymmetry @xmath79 around @xmath128 .
the new results for the nonmesonic rates compare favourably with the present experimental data .
the ratio has increased with respect to our previous works and it now lies practically within the lower side of the error band .
the asymmetry for is also compatible with experiment @xcite but that for disagrees strongly from the recent experimental observation @xcite .
the latter work finds a small and positive value for the elementary asymmetry parameter @xmath117 in , while that for is large and negative .
our meson - exchange model does not explain the present experimental differences and understanding this issue is one of the current challenges , both experimental and theoretical , in the study of the weak decay of hypernuclei .
we have found a tremendous influence on the weak decay observables from the way fsi are considered , especially in the case of total and partial decay rates .
a phenomenological implementation of fsi effects , or not including them at all , gives rise to decay rates that differ by more than a factor of two , and to a neutron - to - proton ratio about 20% larger from what is obtained with the more realistic calculation that uses the proper @xmath15 scattering wave function .
the differences observed in the decay rates and the ratio are much larger than the uncertainties tied to the different strong interaction models commented above .
therefore , accurate calculations of the nonmesonic weak decay of hypernuclei demand a proper treatment of fsi effects through the solution of a @xmath89-matrix using realistic nn interactions .
the authors are indebted to k. sasaki and m. oka for many clarifying discussions about the inclusion of the @xmath0-exchange mechanism .
we also thank c. bennhold for many useful suggestions and e. oset for discussions on the effects of an optical nucleon - nucleus potential .
this work has been partially supported by the u.s .
dept . of energy under grant
de - fg03 - 00-er41132 , by the dgicyt ( spain ) under contract pb98 - 1247 , by the generalitat de catalunya project sgr2000 - 24 , and by the eec - tmr program eurodaphne under contract ct98 - 0169 .
the wave function describing the relative motion of two particles moving under the influence of a two - body potential @xmath129 is obtained from the lippmann - schwinger equation : where @xmath131 is the free relative motion state , solution of the free schrdinger equation with energy @xmath132 , @xmath133 , with @xmath134 being the relative kinetic energy operator .
the positive ( negative ) sign in front of the infinitesimal quantity @xmath135 arises from the requirement that the relative motion is free in the remote past ( future ) .
alternatively , one can write the lippmann - schwinger equation in terms of the transition matrix @xmath89 , defined such that @xmath136 ( @xmath137 ) , with the result let s take for instance @xmath140 ( representing the two nucleons in the final state ) and project eq .
( [ eq : psim ] ) into coordinate space . inserting a complete set of states in the r.h.s . of eq .
( [ eq : psim ] ) and considering also the spin quantum numbers via two - particle coupled spin states , we find : where @xmath142 stands for the adjoint of the free plane wave , @xmath143 . we perform a partial wave decomposition of the wave functions @xmath144 and , working in the coupled @xmath145 representation , and ,
similarly to eq .
( 18.119b ) of ref .
@xcite , we obtain : where @xmath147 is the adjoint of the generalized spherical harmonic given by @xmath148 the partial wave decomposition for the free wave function is simply obtained by replacing in eq .
( [ eq : pwwf ] ) @xmath149 where @xmath150 is the spherical bessel function .
a similar decomposition may be written down for the @xmath89-matrix elements , @xmath151 inserting eqs .
( [ eq : pwwf ] ) , and ( [ eq : pwt ] ) into eq .
( [ eq : wftmat ] ) , using eqs .
( [ eq : spher ] ) and ( [ eq : pwfree ] ) , and carrying out the angular integration , one obtains the following equation for the partial wave components of the correlated wave function @xmath152 where the partial wave @xmath89-matrix elements fulfil the integral equation : @xmath153 both equations ( [ eq : psipw ] ) and ( [ eq : tmatpw ] ) are solved in momentum space following a numerical matrix inversion method described in ref .
isospin is easy to incorporate at the final step of our calculation by multiplying the resulting correlated wave function by the isospinor @xmath154 , allowing only for those cases that fulfil the antisymmetry requirement @xmath155 .
an alternative but approximate way of implementing the @xmath15 final state interactions is through the @xmath0-matrix operator , which obeys : @xmath156 where @xmath157 indicates the cauchy principal value .
standing waves can then be obtained by following the same procedure of eqs .
( [ eq : wftmat ] ) through ( [ eq : tmatpw ] ) but replacing the @xmath89-matrix elements by those of the real @xmath0-matrix .
finally , a simple way to include the effects of the strong interaction between the particles is to use a phenomenological correlation function , @xmath158 , such that @xmath159 this has been the approach followed by several authors , and functions such as a gaussian or of the type @xmath160 with @xmath125fm@xmath94 have been used @xcite .
a. parreo , a. ramos , and c. bennhold , phys .
c * 56 * , 339 ( 1997 ) .
m.m . block and r.h .
dalitz , phys .
11 * , ( 1963 ) 96 ; r.h .
dalitz and g. rajasekharan , phys . lett .
* 1 * , 58 ( 1962 ) .
j. cohen , prog .
* 25 * ( pergamon , 1990 ) p. 139
gibson and e.v .
hungerford , phys .
* 257 * , 349 ( 1995 ) .
e. oset and a. ramos , prog .
nucl . phys . *
41 * , 191 ( 1998 ) .
adams , phys . rev . * 156 * , 1611 ( 1967 ) .
e. oset and l.l .
salcedo , nucl . phys . *
a443 * , 704 ( 1985 ) .
b.h.j . mckellar and b.f .
gibson , phys . rev . * c30 * , 322 ( 1984 ) . j.f .
dubach , g.b .
feldman , b.r .
holstein , l. de la torre , ann .
( n.y . ) * 249 * , 146 ( 1996 ) ; l. de la torre , ph.d . thesis , univ . of massachusetts ( 1982 ) .
k. itonaga , t. ueda , and t. motoba , nucl .
a585 * , 331c ( 1995 ) ; _ ibid .
_ , _ weak and electromagnetic interactions in nuclei _ , edited by h. ejiri , t. kishimoto and t. sato ( world scientific , singapore , 1995 ) p 546 .
a639 * , 329 ( 1998 ) .
m. shmatikov , phys . lett . * b322 * , 311 ( 1994 ) ; _ ibid .
_ , nucl . phys . *
a580 * , 538 ( 1994 ) .
d. jido , e. oset and j.e .
palomar , nucl . phys . * a * , in press , nucl - th/0101051 .
a. parreo , a. ramos , c. bennhold , and k. maltman , phys . lett . *
b435 * , 1 ( 1998 ) .
cheung , d.p .
heddle and l.s .
kisslinger , phys .
* c27 * , 335 ( 1983 ) ; d.p .
heddle and l.s .
kisslinger , phys . rev . *
c33 * , 608 ( 1986 ) .
t. inoue , s. takeuchi , and m. oka , nucl .
* a597 * , 563 ( 1996 ) ; t. inoue , m. oka , t. motoba and k. itonaga , nucl . phys . *
a633 * , 312 ( 1998 ) .
k. sasaki , t. inoue , and m. oka , nucl .
* a669 * , 331 ( 2000 ) ; _ ibid .
_ , * a678 * , 455 ( 2000 ) . k. maltman and m. shmatikov , phys . lett . *
b331 * , 1 ( 1994 ) .
szymanski et al . , phys .
c * 43 * , 849 ( 1991 ) .
h. noumi et al . , phys .
c * 52 * , 2936 ( 1995 ) .
a. montwill et al . , nucl .
a234 * , 413 ( 1974 ) .
alberico , a. de pace , m. ericson , and a. molinari , phys . lett .
* b256 * , 134 ( 1991 ) .
a. ramos , e. oset , and l.l salcedo , phys .
c * 50 * , 2314 ( 1994 ) .
a. ramos , m.j .
vicente - vacas , and e. oset , phys .
c * 55 * , 735 ( 1997 ) .
alberico , a. de pace , g. garbarino and a. ramos , phys .
c * 61 * , 044314 ( 2000 ) .
a. parreo , a. ramos , n.g .
kelkar and c. bennhold , phys .
c * 59 * , 2122 ( 1999 ) .
b. holzenkamp , k. holinde , and j. speth , nucl . phys . *
a500 * , 485 ( 1989 ) .
maessen , th .
a. rijken and j.j .
de swart , phys .
c * 40 * , 2226 ( 1989 ) .
v.g.j . stoks and th.a .
rijken , phys .
c * 59 * , 3009 ( 1999 ) ; th.a .
rijken , v.g.j . stoks and y. yamamoto , phys .
c * 59 * , 21 - 40 ( 1999 ) .
a. parreo , c. bennhold and a. ramos ( to be submitted ) .
k. itonaga , t. motoba , o. richter and m. sotona , phys .
c * 49 * , 1045 ( 1994 ) .
k. hagino and a. parreo , phys .
c * 63 * , 044318 ( 2001 ) .
nagels , t.a .
rijken , and j.j .
de swart , phys .
d * 15 * , 2547 ( 1977 ) .
d. halderson , phys .
c * 48 * , 581 ( 1993 ) .
a. parreo , a. ramos , c. bennhold , and d. halderson , in _ dynamical features of nuclei and finite fermi systems _ , ( world scientific , singapore , 1994 ) p. 318 .
j. golak , k. miyagawa , h. kamada , h. witala , w. glckle , a. parreo , a. ramos and c. bennhold , phys .
c * 55 * , 2196 ( 1997 ) .
s. ajimura et al . , phys .
* 18 * , 4052 ( 2000 ) .
h. bhang et al . , phys .
* 81 * , 4321 ( 1998 ) .
s. ajimura et al .
, phys . lett . *
b282 * , 293 ( 1992 ) .
a. ramos , e. van meijgaard , c. bennhold and b.k .
jennings , nucl .
a544 * , 703 ( 1992 ) .
stoks , r.a.m .
klomp , c.p.f .
terheggen and j.j .
de swart , phys .
c * 49 * , 2950 ( 1994 ) .
r. machleidt in _ computational nuclear physics .
2 . nuclear reactions _ , edited by k. langanke , j.a .
maruhn and s.e .
koonin ( springer - verlag , berlin , 1993 ) , pp . 1 - 29 . c. bennhold and a. ramos .
c * 45 * , 3017 ( 1992 ) .
j. haidenbauer , k. holinde , k. kilian , t. sefzick , and a. w. thomas , phys .
c * 52 * , 3496 ( 1994 ) .
h. nabetani , t. ogaito , t. sato , and t. kishimoto , phys .
rev . c. * 60 * , 017001 ( 1999 ) .
t. inoue , k. sasaki , m. oka and s. takeuchi , prog .
* 137 * , 147 ( 2000 ) .
joachain , _ quantum collision theory _
( elsevier , amsterdam , 1975 ) .
haftel and f. tabakin , nucl .
a158 * , 1 ( 1970 ) . | we present an update of the one - meson - exchange ( ome ) results for the weak decay of s- and p - shell hypernuclei @xcite , paying special attention to the role played by final state interactions between the emitted nucleons .
the present study also corrects for a mistake in the inclusion of the @xmath0 and @xmath1 exchange mechanisms , which substantially increases the ratio of neutron - induced to proton - induced transitions , @xmath2 . with the most up - to - date model ingredients , we find that the ome approach is able to describe very satisfactorily most of the measured observables , including the ratio @xmath2 . # 1#2@xmath3#2 pacs number(s ) : 21.80.+a , 24.80.+y , 13.30.eg , 13.75.ev , 13.75.cs |
recent _ chandra _ observations indicate that relativistic plasma injected into the intracluster medium ( icm ) from radio sources eventually detaches from the radio source , forming bubbles of radio plasma in the icm @xcite . in the model proposed by @xcite ,
these radio ghosts survive in the icm , and provide a seed population of cosmic rays(crs ) .
diffuse non - thermal ( nt ) emission is produced when merger induced shocks re - accelerate , via the first order fermi process , this seed cr population .
current evidence suggests that these radio ghosts contain a significant population of protons as well as electrons @xcite .
since @xcite demonstrated that diffuse nt x - ray emission could be produced by either primary electrons directly accelerated at shock fronts , or secondary electrons produced during proton - proton collisions , there are two possible sources for the observed diffuse nt emission . to determine the mechanism that produces diffuse nt emission requires accurate measurement of the spectrum and location of the nt emission .
simulations by @xcite show that diffuse nt emission occurs in a broad mass range of clusters with a luminosity proportional to the x - ray temperature , making cool clusters and groups an important diagnostic for understanding which population of electrons produces diffuse nt emission . they find that spectral index of the diffuse nt emission is dependent on the electron population producing the emission , such that the spectral index of diffuse nt emission produced from primary electrons has a steep spectral index ( @xmath71.1 ) , while for secondary it is systematically flatter ( @xmath81.1 ) .
@xcite reported detection of diffuse nt x - ray and radio emission from ic1262 , a poor cluster of galaxies .
the x - ray detection was made using the _ bepposax _ medium energy concentrator spectrometer ( mecs ) detector , and the radio using the nrao vla sky survey ( nvss ) and westerbork northern sky survey ( wenss ) .
although the mecs was able to constrain the region of the nt emission , it does not have the spatial resolution to identify the source of the nt emission . in this paper , we show that the _ chandra _ advanced ccd imaging spectrometer ( acis ) has the energy range and spatial capability to detect , localize , and characterize the spectrum of nt x - ray emission from low temperature clusters .
these results are compared to simulations of cosmic - ray acceleration at shock fronts . throughout this letter , we assume a hubble constant of h@xmath9 = 65 _ h@xmath10 _ km s@xmath4 mpc@xmath4 and q@xmath9 = @xmath11 .
quoted confidence intervals are at a 90% level , unless otherwise specified .
the ic1262 galaxy cluster is a poor cluster of galaxies located at ( j2000 ) 17@xmath12 33@xmath13 01.0@xmath14 , + 43@xmath15 45@xmath6 28@xmath16 @xcite with a redshift of 0.0343 @xcite , so that 1@xmath6 = 46 h@xmath17 kpc .
it was observed by the _ chandra _ acis s3 chip on 23 august 2001 for @xmath5 31 ksec
. the total count rate of the uncleaned data is 10.2 @xmath18 0.02 cts s@xmath4 with a peak of 37 @xmath18 3 cts s@xmath4 , which is well below the saturation limit of the acis chips operating in faint mode ( 170 cts s@xmath4 ) .
figure-[fig1 ] is a color coded intensity map that shows the full 8@xmath19 [email protected] image in the 0.3 - 8.0 kev band .
the image was obtained using processing routines outlined in the ciao 2.3 science threads .
the data was cti corrected and cleaned for flares , point sources , and anomalous high background .
exposure maps were created for 17 different energy bands to ensure proper exposure correction .
background was taken from the caldb 2.21 blank - sky datasets . to construct the temperature map ( see figure-[fig3 ] ) we divided the chip into quadrants , north , south , east , and west .
spectra were extracted from pie shaped regions with radii differences such that there were at least 20000 counts in each region . in order to ensure proper background subtraction , we eliminated periods of high background , following the _ filtering lightcurves _ thread . in order to account for background variability ,
the background was normalized to the source in the 10 - 12 kev range @xcite .
the data were grouped so that there were at least 30 counts per channel .
all models included either an _ acisabs _ component or had an _ acisabs _ corrected ancillary response file in order to account for the time dependent absorption ( at low energies ) of the acis window .
results are given in table-[tbl-1 ] .
lcccccc n@xmath20 & apec & 1.27@xmath21 & 0.14@xmath21 & & 173.0422/98 + n@xmath20 & apec+apec & 1.88@xmath22 & 0.44@xmath23 & 0.83 @xmath24 & 134.0729/96 + n@xmath25 & apec & 1.34@xmath26 & 0.14@xmath27 & & 119.7284/91 + n@xmath28 & apec & 1.57@xmath29 & 0.12@xmath30 & & 57.09452/76 + n@xmath31 & apec & 1.88@xmath32 & 0.13@xmath33 & & 65.19046/59 + n@xmath34 & apec & 2.09@xmath35 & 0.12@xmath36 & & 55.56707/48 + + w@xmath20 & apec & 1.80@xmath37 & 0.25@xmath38 & & 126.3223/108 + w@xmath20 & apec+apec & 1.98@xmath39 & 0.38@xmath40 & 0.61@xmath41 & 113.5917/106 + w@xmath25 & apec & 2.09@xmath42 & 0.33@xmath43 & & 136.9889/101 + w@xmath25 & apec+apec & 2.38@xmath44 & 0.57@xmath45 & 0.28@xmath46 & 123.7163/99 + w@xmath28 & apec & 1.93@xmath47 & 0.18@xmath48 & & 118.0369/107 + + s@xmath20 & apec & 1.62@xmath49 & 0.27@xmath50 & & 203.8204/103 + s@xmath20 & apec+apec & 2.54@xmath51 & 0.63@xmath52 & 1.05@xmath53 & 147.7749/101 + s@xmath25 & apec & 1.60@xmath54 & 0.17@xmath38 & & 134.1819/97 + + e@xmath20 & apec & 1.62@xmath55 & 0.22@xmath50 & & 114.6884/101 + e@xmath20 & apec+apec & 2.08@xmath56 & 0.49@xmath57 & 0.82@xmath58 & 87.91611/99 + e@xmath25 & apec & 2.02@xmath59 & 0.23@xmath43 & & 90.03336/89 + e@xmath25 & apec+apec & 2.30@xmath60 & 0.40@xmath61 & 0.23@xmath62 & 83.43614/87 + e@xmath28 & apec & 2.37@xmath63 & 0.39@xmath64 & & 109.8093/99 +
from the flux and hardness map , figure-[fig1 ] and figure-[fig2 ] respectively , we chose four regions that have a strong gradient in flux and hardness .
these regions are shown in figure-[fig1 ] .
we considered four models of background subtraction , to ensure any nt detection would not be the result of undersubtracted particle background .
these four background subtraction techniques were : ( 1 ) standard blank - sky fields extracted from the same region on the s3 chip as the source and normalizing in the 10 - 12 kev range as described in @xcite , ( 2 ) extracting a region far to the north of the peak flux , modeling it and then using the model times a free constant in our source + background spectrum modeling , ( 3 ) freezing the background model described in ( 2 ) with a constant proportional to the difference in size of the source and background region , and ( 4 ) subtracting the size normalized background spectrum from the source spectrum . in all cases
the source model parameters are consistent within 90% confidence , indicating that all four methods produce the same results .
we modeled our four `` possible shock '' regions with a single thermal component model , a double component thermal model , and a thermal model with powerlaw .
the addition of the powerlaw component does not significantly improve the fit in the east and west regions .
in addition , figure-[fig4 ] shows that a thermal model fits the entire spectrum in the east region . in the case of the north region , adding either a second thermal component , or a power - law component improves the fit .
the single thermal model fit to the south region has clear high end residuals ( see figure-[fig4 ] ) .
although the addition of a second second thermal component improves the fit ( @xmath65 16 for a reduction of 2 degrees of freedom(dof ) ) , the addition of a powerlaw component provides the best fit ( @xmath65 23 for a reduction of 2 dof ) and reduces the residuals ( see figure [ fig4 ] ) .
although we can not rule out a thermal interpretation of the second component , we argue a non - thermal interpretation is better statistically and physically .
the addition of a powerlaw ( 37% probability of exceeding @xmath66 = 74 for 71 dof ) produces a better fit than adding a second thermal component ( 18% probability of exceeding @xmath67 for the 71 dof ) .
the 90% lower confidence of @xmath57 kev implies a shock with a mach number of @xmath53 , which is higher than observed in clusters ( * ? ? ? * and references therein ) . in order to confirm that the nonthermal detection was not simply undersubtracted particle background , we repeated our four different techniques for handling the background that we had done for a single thermal model .
as in the case with the single thermal model , the results are consistent within 90% for all the background techniques used ( see table-[tbl-2 ] ) . also , we point out that the lack of a nt detection in the west and east regions provides more evidence that the detection in the north and south regions is not simply particle background , since that would be observed isotropically across the detector . lcccccc north + & blank - sky & 1.32@xmath68 & @xmath69 & 2.36@xmath70 & 0.79@xmath71 & 52.93678/50 + & model - bkg & 1.29@xmath72 & @xmath73 & 2.30@xmath74 & 0.78@xmath75 & 54.05200/49 + & model - bkg(frzn ) & 1.31@xmath23 & 0.39@xmath76 & 2.41@xmath77 & 0.56@xmath78 & 57.22508/50 + & -local region & 1.30@xmath23 & 0.43@xmath79 & 2.40@xmath80 & 0.58@xmath81 & 54.61599/50 + + south + & blank - sky & 1.44@xmath82 & @xmath83 & 2.21@xmath1 & 0.95@xmath84 & 74.33978/71 + & model - bkg & 1.40@xmath85 & @xmath83 & 2.14@xmath86 & 0.91@xmath87 & 77.66279/70 + & model - bkg(frzn ) & 1.44@xmath88 & 0.49@xmath89 & 2.33@xmath90 & 0.59@xmath91 & 90.72980/71 + & -local region & 1.43@xmath92 & @xmath93 & 2.27@xmath94 & 0.78@xmath95 & 76.28895/71 + the radio halo detected by @xcite appears to extend to the south along the elongated x - ray isophotes and through the south region ( see figure-[fig5 ] ) .
however the @[email protected] resolution of the nvss and wenss precludes the removal of radio point sources .
we , therefore , consider two possibilities : ( 1 ) that the extended southern radio emission is dominated by radio point sources , and ( 2 ) that it is diffuse radio emission . in the case that the radio emission is dominated by radio point sources , we argue that this would have no affect on our x - ray results .
relativistic electrons diffusing from a point source at their alfvn speed , would only reach @xmath96 h@xmath17 kpc ( @xmath97 ) ( assuming a 1 @xmath98 g magnetic field ) in their radiative lifetime ( @xmath510@xmath99 years ) .
therefore any radio point source with detectable x - ray emission would appear point - like and would have been removed from the _ chandra acis _
data . in the latter case ,
the diffuse radio emission is emitted by the same electrons producing the inverse compton emission ( as in the inverse compton scattering - syntrotron model @xcite ) .
we also note in passing that there are several identified radio point sources ( including the cd galaxy ) in ic1262 , which could provide the seed population of crs necessary in the model proposed by @xcite .
a comparison of figure-[fig1 ] and figure-[fig2 ] to flux and temperature maps created from simulations of cluster mergers done by @xcite and @xcite , indicates a recent ( @xmath50.25 - 1 gyr ago ) north - south merger of a smaller subclump ( mass ratio @xmath51:2 - 1:4 ) with ic1262 .
these simulations show the general characteristics of the temperature and morphological structure : elongation of the isophotes to the south , compression to the north , and the hottest gas south of the center .
since cd galaxies in relaxed systems will reside at the bottom of the potential well ( e.g. @xcite ) , the offset of cd galaxy from the x - ray peak and its high peculiar velocity ( @xmath100v = 453 km s@xmath4 ) provide additional evidence of a merger .
the kick to initiate sloshing is parallel to the impact parameter @xcite .
as the simulations indicate a recent collision , the subclump must have passed to the west , pulling the cd galaxy to the west .
the elongation of the x - ray isophotes toward the south that overlap the diffuse radio emission open up the possibility that the elongation is due to non - thermal emission rather than a merger .
however , the evidence for a merger comes not only from the similarity of the southern x - ray isophotes to those seen in the simulations cited above , but also from the compressed northern x - ray isophotes and the very hot region that is south of center . comparing our results with the simulations done by @xcite , we argue that the localized diffuse nt emission is from a population of primary electrons .
@xcite find that primary electron reacceleration follows the shock region closely . in the central region
( r @xmath101 0.3 mpc ) , they find the greatest difference between the spectral index produced from primary electrons ( @xmath1021.1 ) and secondary electrons ( @xmath81.1 ) .
our 90% confidence photon index @xmath0 = 2.07 - 2.36 which corresponds to a spectral index @xmath103 = 1.07 - 1.36 is consistent with the spectral index that @xcite found for primary electrons . using the best fit f@xmath104-t@xmath105 of @xcite
, we find a flux of ( 10 - 48 ) @xmath106 10@xmath107 ergs @xmath3 s@xmath4 over 0.13 to 100 kev for primary electrons in a 1.94 kev cluster ( ic1262 s temperature as determined by @xcite ) . the range fluxes depends on the r@xmath108 , the ratio of electrons to protons at relativistic energies .
using the same energy range and cluster temperature we find that secondary electrons produce a flux of 1.9 @xmath106 10@xmath107 ergs @xmath3 s@xmath4 .
for our best fit photon index @xmath0 = 2.21 and 90% confidence normalization range of 7.0 - 11.1 @xmath106 10@xmath109 photons @xmath3 s@xmath4 kev@xmath4 at 1 kev , we determine a flux of ( 6 - 10 ) @xmath106 10@xmath107 ergs @xmath3 s@xmath4 for the 0.13 - 100 kev range .
this result has a lower limit 3@xmath106 greater than the results predicted for flux from secondary electrons .
although the 90% upper confidence limit is barely consistent with the primary electron model ( suggesting r@xmath110 0.01 ) , we point out that the results of @xcite is for shocks over the entire cluster , and our detection is only for a fraction of the cluster .
that is the total non - thermal emission from cluster is probably greater than our detection , further strengthening the argument that the flux is too high to come from secondary electrons . | we report the first localization of diffuse , non - thermal , x - ray emission from a nearby galaxy cluster . using _
chandra _ data , we have isolated a diffuse non - thermal x - ray component with a photon index , @xmath0 = 2.21@xmath1 and a flux of 9.5@xmath2 photons @xmath3 s@xmath4 kev@xmath4 at 1 kev , that extends from @[email protected] to @[email protected] to the south of the x - ray flux peak .
comparison to simulations implies that the diffuse non - thermal emission is produced by primary electrons , accelerated at shocks to relativistic velocities . using these results and the flux and hardness maps produced with data from the _ chandra _ advanced ccd imaging spectrometer ,
we conclude that a smaller subclump coming from the north merged with ic1262 .
the offset of the cd galaxy from the x - ray peak and large peculiar velocity indicate that the subclump s impact parameter was to the west and on the near side of ic1262 . |
pattern formation in spatially extended systems has been an object of extensive study in past few decades . the reasons for the interest in pattern formation are not far to seek .
pattern formation happens in several natural systems and scientists are interested in understanding it .
examples could be flame propagation or occurrence of spirals in a chemical reactions or patterns on the skins of animals modeled by reaction - diffusion processes .
several systems like biological patterns @xcite , charge density waves or josephson junction arrays @xcite , lasers @xcite have been studied extensively from the viewpoint of dynamics and pattern formation .
the practical importance of understanding above systems can not be overemphasized .
partial differential equations , coupled ordinary differential equations , coupled oscillator arrays and coupled maps have all been used to model different physical and biological systems and have uncovered interesting modifications of equilibrium , bifurcations and stability properties of collective spatiotemporal states .
these systems have been studied extensively@xcite . however , while we have arrived at a good understanding of low dimensional chaos in past two decades , and understood routes to chaos in several seemingly disparate systems in an unified manner , the same is not true for spatiotemporal systems .
as one would expect , there is practical relevance to these models .
certain simplified models of spatiotemporal dynamics have been studied extensively in recent past .
we would like to make a special mention of coupled map lattices which have been studied extensively in the hope that the understanding we have developed in low - dimensional systems could be useful in understanding spatiotemporal dynamics .
majority of the studies are in coupled one - dimensional maps . at times , they have been successful in modeling certain patterns observed in spatiotemporal systems .
for example , the phenomena of spatiotemporal intermittency in rayleigh - benard convection has been modeled by coupled map lattices @xcite .
they have been used to model spiral waves in b - z reaction @xcite and even in crystal growth @xcite .
we feel that these studies could be helped considerably by _
quantitatively _ identifying different phases in the system and attempting to understand transitions between those .
a lot of attention has been devoted to transition to synchronization in these systems which is easy to identify and analyze .
however , other spatiotemporal patterns are far less analyzed and understood .
it is worth noting that in several spatiotemporal systems the field is high dimensional .
for example , for a chemical reaction , we would require information about concentrations of all reactants and products at all points to be able to simulate the system .
there are several phenomena which are possible only when dynamics is higher dimensional .
on the other hand most of the existing studies in coupled map lattices are about coupled one - dimensional maps and very few involving high dimensional maps , for example coupled chialvo s map @xcite , coupled henon map @xcite , arnold s cat map @xcite , modified baker s map for ising model @xcite , etc . here
we try to study coupled higher dimensional maps .
we try to systematically change the dimensionality of the map and study coupled maps with feedback where the feedback time could be varied .
we plot the phase diagram of coupled circle map with delay in two parameter space of coupling strength @xmath1 and delay time @xmath2 .
the abbreviation sfp , frp , fp - p / c and stc denote synchronous fixed point , frozen random pattern , fixed point - periodic / chaotic cycles and spatiotemporal chaos respectively . in the inset , we magnify the phase boundary between fp - p / c regime and stc regime .
we can have only integer values of feedback time and the lines in phase diagram are guide to the eye .
the black solid circles indicate points in phase space where spatiotemporal fixed point phase is seen .
, width=302,height=226 ] in coupled map lattice ( cml ) @xcite time evolution is conflict between competing tendencies : the diffusive coupling between units tends to make the system homogeneous in space , while the chaotic map produces temporal inhomogeneity , due to sensitive dependence on initial condition .
such systems display a rich phase diagram and origin and stability of various phases in these systems are of theoretical and practical interest .
the transitions between various phases in these systems could be viewed as dynamic phase transitions .
though these are clearly non - equilibrium systems , different techniques used in the study of equilibrium phase transitions have been applied to explore the variety of phenomenologically distinct behaviors .
such non - equilibrium systems range from growths of surfaces@xcite to traffic jams in vehicular traffic and granular flows@xcite . for such analysis
, there is a need to define an order parameter that will characterize the transition .
however , not many dynamic phase transitions have been studied from this viewpoint @xcite .
most of the studies are devoted to transition in models with so - called absorbing states . in the context of synchronization
, there have been extensive studies to ascertain whether or not this transition is in the universality class of directed percolation @xcite .
transition to synchronization , locking in of the dynamics of several systems to a collective oscillation , is an important but minuscule part of overall spatially extended systems .
several non - synchronization transitions have been observed such as spiral state to spatiotemporal chaos @xcite , traveling wave to spatiotemporal chaos @xcite .
the spatiotemporal dynamics is far richer and other transitions deserve attention .
as suggested by pyragas , feedback in the dynamical systems can play a central role in controlling chaos by stabilizing upo s embedded in chaotic attractor@xcite .
the scheme has been efficiently implemented in wide range of experimental applications .
feedback leading to synchronization is a well established fact supported by several researchers@xcite .
though feedback has been an important aspect of studies in control and synchronization , not much has been said about its effect on different dynamical phases .
it is then important to find the effects of feedback in terms of alterations in system dynamics that may lead to origin of different dynamic phases . in this work , we investigate coupled maps with feedback . as mentioned before , coupled higher dimensional maps have not been studied adequately . using feedback
, we can systematically vary dimensionality of local map using feedback and study its effect on overall dynamics .
we note that the feedback as defined in our model does not change the location of fixed point of system without feedback .
we have used it extensively as a reference point .
this allows us to study spatially extended higher dimensional systems with known properties .
we investigate a model of coupled circle maps with feedback .
the strength of feedback and the delay time are the parameters in the system
. we will present a general phase diagram of the system .
in particular , we observe a novel transition from _ localized chaos to spatiotemporal chaos_. by localized chaos , we mean that the laminar and turbulent phases coexist in the same space domain for weaker couplings .
the reason to call it localized chaos is that one can see there are well defined boundaries of turbulent sites which do not mix with laminar ones .
a similar pattern has been named spatial intermittency ( si ) in previous studies and some of its properties such as eigenvalue spectrum of jacobian are found to be related to its dynamical properties@xcite .
interestingly , we find that persistence acts in effective way to characterize and quantify certain dynamical phases and transition between them in this system .
thus , persistence can be taken as excellent order parameter for certain phase transitions .
the plan of paper is as follows .
we introduce the model in the section ii . in section iii
, we make a detailed analysis of one of the phases , namely localized chaos and calculate quantities such as turbulent length distribution . in section iv , we present our investigations in transition between localized chaos regime and spatiotemporal chaos and we present evidence that persistence acts as an excellent order parameter to characterize the transition . in section
v , we demonstrate that some characteristic features in eigenvalue distributions of jacobian can distinguish between different dynamical phases .
we define model of coupled circle maps with feedback as follows .
we assign a continuous variable @xmath3 at each site @xmath4 at time @xmath0 where @xmath5 .
@xmath6 is the number of lattice points .
the evolution of @xmath3 is defined by , @xmath7 + \delta f(x_i(t-\tau ) ) \label{eq1}\end{aligned}\ ] ] the parameter @xmath1 is the coupling strength and the local dynamical update function @xmath8 is the sine - circle map , @xmath9 where , @xmath10 is the strength of nonlinearity and @xmath11 is the winding number .
the parameter @xmath12 is strength of feedback and @xmath2 is delay time .
if we increase @xmath2 , we can increase the dimension of map . in our simulations ,
we confine values of parameters to @xmath13 , @xmath14 and @xmath15 .
the coupling topology considered here is diffusive i.e nearest - neighbor connections . for coupled circle map
, we confine the dynamics in the interval @xmath16 $ ] using the following rule .
if int@xmath17 = m$ ] , @xmath18 if @xmath19 and @xmath20 if @xmath21 .
we study the system with random initial conditions .
the system is updated synchronously with periodic boundary condition .
the feedback does not change the fixed point solution of local map @xmath8 without feedback .
the fixed point solution for the local map @xmath8 is given by , @xmath22 .
the parameter values considered here yield the value of fixed point as @xmath23 .
this point acts as an excellent reference point in visualizing the dynamical behavior of system and also in symbolic dynamics of the system .
we can define the site dynamics to be laminar if @xmath24 for small enough neighborhood @xmath25 and to be turbulent otherwise .
we examine the phenomenology of delay coupled circle map with feedback under variation of coupling strengths and delay time .
various spatiotemporal patterns are clearly evident from phase diagram fig .
[ fig1 ] and the bifurcation diagram shown in fig .
note that on y axis of fig .
[ fig1 ] , the time is discrete .
the delay time @xmath2 can only be some integer .
the phase boundary of various regions can be seen .
bifurcation diagram of spatial pattern at an instant of time ( after transience ) , with respect to coupling strength @xmath1 for delay time @xmath26.,width=226,height=226 ] increasing value of @xmath2 is equivalent to increasing the dimensionality of local dynamics .
there are clearly four distinct behaviors shown by the system as seen from phase diagram fig .
the first region in phase diagram labeled as sfp ( synchronous fixed point ) where , for smaller coupling strengths @xmath27 the entire system goes to local fixed point @xmath28 .
this is an absorbing state and system settles in this state very quickly .
we have checked that this state exists at least for @xmath29 . for higher coupling strengths
, the system goes to spatiotemporal chaos as shown in phase diagram marked as stc .
these different behaviors can also be visualized using the bifurcation diagram fig .
[ fig2 ] obtained for different coupling strengths for @xmath26 .
this bifurcation diagram shows the state of all sites at a given instant of time . for very high coupling strengths
, the system shows some stability islands for some values of @xmath2 though they constitute a small part in phase diagram , we find them worth a mention . in the intermediate coupling strengths ,
the system shows a rather interesting behavior .
we can see that some of the sites approach local fixed point and some of the sites take on different values which do not change in time .
this behavior is shown in phase diagram marked as frp which stand for frozen random pattern @xcite .
this region is characterized by a spatial sequence of different domains whose size and periodicity may greatly vary . upon further increase of coupling strengths ,
one enters in a region where we can see coexistence of different dynamical regimes .
where some of the sites indeed go to fixed point , some of the sites show periodic / chaotic behavior .
this is an interesting state where one can see that different dynamical phases do not mix with each other .
the regions marked as frp and fp - p / c in the phase diagram collectively represent the dynamics which we refer to as _ localized chaos_. chaos is localized in certain dynamical regimes of turbulence and does not interfere with evolution of laminar states . in localized chaos regime , chaotic domains and periodic domains are fixed in space .
this is evident from spatial distribution pattern .
the perturbations do not propagate in space .
essentially , the dynamics is non - spreading and non - infective in nature , i.e. the spatially intermittent solutions have zero velocity components in spatial direction , and modes traveling along the lattice are absent .
this can also be considered as a case of spatial intermittency @xcite .
the important result is that spatial intermittency is seen over a large parameter regime in presence of feedback . for each feedback time
, we have a well defined region where we get the transition .
these spatiotemporal patterns are clearly evident from space - time density plots below .
each figure represent a particular region in phase diagram . [ cols= " < , < , < " , ] ( color online ) overlapping histograms of the calculated eigenvalues . in red ( gray ) ,
histogram of eigenvalues obtained by theory of pertubative analysis . in blue(black ) , histogram of numerically obtained eigenvalues of matrix @xmath30 is given .
the values for turbulent sites given by @xmath31 can also be seen away from the band @xmath32 $ ] .
the agreement between theory and numerics can be clearly seen for degenerate as well as non - degenerate eigenvalues.,width=226,height=226 ] the localized eigenvectors showing peaks at @xmath33 and @xmath34 for lattice size @xmath34 .
, width=226,height=226 ] .
similar calculations can be carried out for our jacobian @xmath35 .
the localized eigenvectors explain why localized patterns are stable .
the eigenvectors having largest eigenvalue are localized .
thus any perturbation from laminar state is likely to be growing in the direction of localized eigenvectors which does not disturb values at nearby sites .
the above analysis of eigenvalue distribution for localized chaos is restricted to demonstrate the analogy to spatial intermittency .
the eigenvalue spectrum clearly differentiates infective to non - infective phase .
the above analysis explains why the eigenvalue spectrum is different for spatiotemporal chaos and spatial intermittency .
in this paper , we have simulated coupled circle maps with feedback .
this allows us to study coupled systems where local dynamics is high dimensional .
we give an overall phase diagram and discuss various dynamical phases .
we also give an all - site bifurcation diagram and remark that there are phases which can not be found from bifurcation diagram and we need to study a detailed simulated space - time plot . in particular , we make a detailed study of a dynamical phase transition between the phase of localized chaos and spatiotemporally chaotic state .
we find various characteristics such as distribution of domains of turbulent sites .
phase transition from different dynamical phases is demonstrated .
we make an attempt to analyze this transition in a manner in which equilibrium transitions are analyzed . in order that we are able to do so ,
it is necessary to define an order parameter .
we find that local persistence acts as an excellent order parameter to characterize the phase transition .
in addition to this , conventional scaling laws pertaining to finite size simulations and offcritical scaling add credibility to use of persistence as quantifier of the transition .
interestingly , same definition of persistence used for without feedback case @xcite works for our case with feedback .
we also calculate one step jacobian of the system and find gaps in eigenvalue distribution of system which is dynamical signature of non - infectious nature of dynamics in ` localized chaos ' regime .
the important effect of feedback is that the parameter window for which spatial intermittency is seen , gets enhanced .
it should be noted that persistence acts as a good order parameter for this transition for all values of @xmath2 including @xmath36 , whereas it is not clear that dynamic characterizer of gaps in eigenvalues spectrum might work for higher value of @xmath2 .
we get this transition for several values of @xmath2 .
we have shown that persistence acts as an excellent order parameter for this transition for all values of @xmath2 .
we also try to study gaps in the eigenvalue spectrum which work well for @xmath36 and @xmath37 .
we go a step ahead in explaining the origin of such gaps in eigenvalues spectra of one step jacobian by perturbation theory analysis .
analytical work also gives an idea of when above analysis may not work .
there is a possibility to experimentally investigate dynamical phases in spatiotemporal systems .
certain electrical oscillators are well simulated with circle maps @xcite and it is relatively easy to implement feedback mechanism in electric devices . if one can define a persistence in an analogous manner in differential equations , it is possible to find persistence in spatiotemporal systems experimentally . in context of localized chaos , we would also like to point to an interesting work by sethia et .
@xcite where they observe ` chimera ' states , first observed by kuramoto @xcite .
these states have spatially disconnected regions of synchronization separated by regions of decoherence .
thus there is a strong analogy between ` chimera ' states and ` localized chaos ' .
several works study ` chimera ' under various modifications for identical oscillators.@xcite and non - identical oscillators @xcite . coupled evolution of systems with delay induces effectively nonlocal contributions in the evolution of individual system .
feedback propagates in space and time and affects the dynamics at some distant site .
it could be of interest to find some general definition of persistence applicable to these systems and investigate transition to ` chimera ' state as a dynamic phase transition .
further work in this direction is pursued . | we investigate coupled circle maps in presence of feedback and explore various dynamical phases observed in this system of coupled high dimensional maps .
we observe an interesting transition from localized chaos to spatiotemporal chaos .
we study this transition as a dynamic phase transition .
we observe that persistence acts as an excellent quantifier to describe this transition .
taking the location of the fixed point of circle map ( which does not change with feedback ) as a reference point , we compute number of sites which have been greater than ( less than ) the fixed point till time @xmath0 .
though local dynamics is high - dimensional in this case , this definition of persistence which tracks a single variable is an excellent quantifier for this transition . in most cases
, we also obtain a well defined persistence exponent at the critical point and observe conventional scaling as seen in second order phase transitions .
this indicates that persistence could work as good order parameter for transitions from fully or partially arrested phase .
we also give an explanation of gaps in eigenvalue spectrum of the jacobian of localized state .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
coupled map lattices have been investigated extensively in past few decades .
the dynamics can show very novel features when local dynamics is higher dimensional . using feedback to make the local dynamics higher dimensional
makes it possible to systematically increase dimensionality of local map without changing several other features such as location of fixed point .
investigation of phase diagram of such a system shows possibility of existence of partially arrested state in which certain sites are near the fixed point while others are not .
we study the nonequilibrium phase transition from such a state to a state of spatiotemporal chaos and propose that persistence works as an excellent order parameter for such transitions . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ |
r - parity is a discrete symmetry defined by assigning to every field the number @xmath1 ( @xmath2 - baryon ( lepton ) number , @xmath3 - spin of the particle ) @xcite .
if it is conserved then baryon and lepton number violating transitions are forbidden . in that case , the theory guarantees both proton stability and lepton universality .
however , in supersymmetric extensions of the standard model , gauge invariance and renormalizability , the two main principles of any gauge theory , do not assure r - parity conservation . at present , wide phenomenological investigations of r - parity violating processes have been undertaken ( for reviews see e.g. @xcite ) . here
we will explore the possibility of discovering the lepton number violating process of single chargino production at future lepton - lepton colliders ( see fig.1(i ) for an electron - electron collision ) .
to our knowledge this process has not yet been discussed , though lepton number violating charginos pair production in electron - electron collisions ( fig.1(ii ) ) has been considered @xcite .
let us start with electron - electron collisions .
the analysis of the muon option is analogous and will be shortly discussed whenever needed . as can be seen from fig.1(i )
, the cross section for single chargino production is proportional to @xmath4 where @xmath5 and @xmath6 are couplings involved in the following lagrangians ( @xmath7 , @xmath8 are family indices ) : @xmath9 these lagrangians are written in physical basis .
the matrix @xmath10 in eq .
( [ lag_g ] ) comes from the sneutrino mass matrix diagonalization .
if r - parity is violated , we have to take into account the mixing between the sneutrinos @xmath11 , @xmath12 , @xmath13 and the neutral higgs bosons @xmath14 , @xmath15 .
we shall , however , assume that this mixing is negligible and does not affect the results , at least at the stage of chargino production . in what follows
we shall also assume that the exchange of the lightest ( electron ) sneutrino dominates ( which is equivalent to some hierarchy assumption in the sneutrino sector ) and neglect the contribution of the heavier @xmath16 s .
we therefore set ( @xmath17 stands for electron ) @xmath18 in eq .
( [ lag_g ] ) . for more complicated cases where the interplay between sneutrino masses in propagators and appropriate elements of the @xmath19 matrix matters we refer to @xcite .
the second mixing matrix , namely @xmath20 in eq .
( [ lag_g ] ) is connected with the chargino sector and describes the weights of the wino component of the chargino fields @xcite .
since this is the only component of the charginos that couples to the electron and the sneutrino ( the charged higgsino coupling is neglected in the limit of zero electron mass ) we set for simplicity @xmath21 . this is further justified by the analysis @xcite ( in the parameter region @xmath22 gev , @xmath23 gev for both small and large @xmath24 , with @xmath25 being the higgsino and gaugino @xmath26 mass parameters , respectively , and @xmath27 a ratio of two vacuum expectation values involved in mssm ) . in general
the results should be multiplied by @xmath28 .
furthermore , with r - parity violation , additional couplings between leptons , gauginos and higgsinos @xmath29 exist , but are known to be smaller than the gauge ones @xcite .
2 we gather the cross sections for single chargino production at future electron - electron colliders with c.m . energies @xmath30 gev and @xmath31 tev as functions of the chargino mass for different sneutrino masses electron beam polarization . in reality
we can expect that @xmath32 can be achieved
. then the cross sections must be multiplied by a factor @xmath33 . ] . for the r - parity violating coupling , we have used the most conservative available upper limit @xmath34 @xcite , independently of the @xmath11 mass ( in the case of muon - muon collisions the @xmath35 couplings
would be involved ) . for sneutrino masses larger than 100 gev
this limit becomes weaker @xcite .
as can be deduced from fig.2 , with a planned annual luminosity of some 50 fb@xmath36 yr@xmath36 @xcite and with a discovery limit at a level of 10 events per year ( @xmath37 fb ) , the process is detectable for a wide range of sparticle masses . with the r - parity violating production process ( i ) we are already definitely out of the sm physics .
it is therefore interesting to investigate the possible detector signals . with r - parity non - conservation ,
the collider phenomenology is quite different from the mssm case and depends especially on the nature of the lsp ( lightest supersymmetric particle ) . in the mssm
, the stable lsp must be charge and color neutral for cosmological reasons @xcite . with r - parity violation
there are no hints about the unstable lsp
. it can be among others a sneutrino , gluino or even a chargino @xcite . here
we give an example of nonstandard phenomenology but restrict ourselves to a scenario in which charginos decay uniquely ( via sneutrino exchange ) to charged leptons .
final leptonic signals with lepton number violation and without missing energy could be detected , an interesting situation from the point of view of nonstandard physics , as there is no sm background ( see further discussion ) .
these two conditions ( charged leptons without missing energy in the final state ) require the chargino to be the second lightest supersymmetric particle ( nlsp ) with sneutrino the lsp .
this situation is schematically summarized in fig.3 .
if the chargino were the lsp its lifetime should be long enough so that it would be seen in the detector .
in other cases ( i.e. when the chargino is neither nlsp nor lsp ) the chargino would also have cascade decays to final jet states @xcite .
then , the situation would be more complicated but at least we can expect that for kinematical reasons a decay to the r - parity lepton violating lsp sneutrino would still be important and the final signal with four charged leptons could be observed ( work in progress ) .
let us discuss the scenario with nlsp chargino .
first , we should notice that , to get substantial chargino production ( e.g. @xmath38 in fig.2 ) , we are interested in the situation where at least one , let us say @xmath39 coupling is large
. then the decay of chargino to three charged leptons must be observed in the detector , as it can undergo uniquely through the same large @xmath39 coupling ( the only possible decay channel , see fig.3 ) . in fig.4 , we show the final results for the angular distribution of the final positron ( fig.3 ) for two different energies ( @xmath40 gev ) .
we have taken @xmath41 gev and @xmath42 gev .
results have been obtained using the vegas procedure .
four particles in the final state give us an 8 dimensional integration .
we have also applied the narrow width approximation where @xmath43 ( see appendix for details ) .
the solid line describes results based on eq .
( a.21 ) , when interferences between production and decay of charginos with @xmath44 are taken into account .
the dashed line describes results with factorization assumed @xcite , which means the following replacement in eq .
( a.21 ) is done @xmath45 we can see that spin correlations do not change the results substantially ( @xmath46 % for considered c.m energies and the chargino mass ) .
it is important that the positron angular distributions are not so strongly peaked in the beam directions , even for @xmath31 tev collider energy . with assumed cuts
@xmath47 enough events will be detected to investigate the process . the only sm process , which gives a four charged lepton signal without missing energy is @xcite @xmath48 . with a possible z boson decay to the lepton - antilepton pair
, it does not coincide with the process under investigation ( @xmath49 ) .
that means that we do not have to bother about the sm background contamination .
however , this cross section is large enough ( @xmath50 pb for @xmath51 tev energies ) to cover some other scenarios .
as an example let us assume that not only @xmath39 but also @xmath52 is not negligible and change the second coupling in the chargino decay channel ( fig.3 ) from @xmath39 to @xmath52 .
that means that we have now @xmath53 in the final state , and this scenario will be dominated by the sm process given above with the z decay to the muon antimuon pair . in this way we can find that @xmath54 and @xmath55 charged lepton signals are testable ( meaning sensitivity to the products @xmath56 , @xmath57 and @xmath58 , respectively ) . finally , our results can be easily applied to the muon - muon collider where another set of r - parity lepton violating couplings can be tested , namely : @xmath59 ( @xmath60 in the final state ) , @xmath61 ( @xmath62 ) and @xmath63 ( @xmath64 ) .
present experimental limits on r - parity violating couplings do not exclude large and detectable lepton number violating signals in lepton - lepton collisions .
we discuss such a possibility in conjunction with single chargino production and its subsequent leptonic decay .
if at least one @xmath5 value is large enough - in our discussion mainly @xmath39 - single chargino production in electron - electron collisions will be observable ( fig.2 ) .
if the chargino is nlsp and sneutrino the lsp , a unique lepton number violating signal of four charged leptons without missing energy could be observed .
we would like to thank h. fraas , c. blchinger , s. hesselbach ( wrzburg university ) and g. moortgat pick ( desy , hamburg ) for helpful discussions and valuable remarks .
this work was supported by the polish committee for scientific research under grant no .
j.g . would like to thank the alexander von humboldt - stiftung for fellowship .
the helicity amplitude for single chargino production @xmath65 is given by : @xmath66 , \\
t(u)&=&m_{\tilde \chi}^2-\sqrt{s}(\tilde e \mp 2 \tilde p \cos{\tilde \theta})\end{aligned}\ ] ] where @xmath67 denotes the momentum and helicity of the chargino , @xmath68 are the corresponding quantities for the incoming electrons , @xmath69 and @xmath5 denote momentum and helicity of the muon ( tau ) and @xmath70 label the c.m .
azimuthal and polar angles of a chargino with respect to the direction of the initial electron @xmath71 .
@xmath77 \\ m ( - , - \rightarrow - , + ) & = & -g \sqrt{2s } \sqrt { ( \tilde{e}+\tilde{p } ) ( \tilde{e}-\tilde{p } ) } \cos{\frac{\tilde{\theta}}{2}}\sin{\frac{\tilde{\theta}}{2 } } \lambda_{112 } \left [ \frac{1}{t - m_{\tilde{\nu}_e}^2}- \frac{1}{u - m_{\tilde{\nu}_e}^2 } \right ] .
\nonumber \\ & & \end{aligned}\ ] ] in fig .
3 we study the r - parity violating chargino decays via sneutrino exchange in the t - channel : @xmath80 .
analogously to the production the amplitude is : @xmath81 , \nonumber \\\end{aligned}\ ] ] where ( @xmath82 denote the momenta and helicities of the electron and muon , @xmath83 are analogous quantities for the final positron . using eqs .
( [ van1],[van2 ] ) we get @xmath84 \nonumber \\ & \times & \left [ -e^{i/2(\phi_2-\phi_+ ) } \sin{\frac{\theta_2}{2}}\cos{\frac{\theta_+}{2 } } + e^{-i/2(\phi_2-\phi_+ ) } \cos{\frac{\theta_2}{2}}\sin{\frac{\theta_+}{2 } } \right ] \label{norm1}\\ t(\tilde \lambda = -1/2 ) & = & \omega_t(- ) \left [ e^{i/2(\phi_1-\tilde{\phi } ) } \sin{\frac{\theta_1}{2}}\sin{\frac{\tilde{\theta}}{2}}+ e^{-i/2(\phi_1-\tilde{\phi } ) } \cos{\frac{\theta_1}{2}}\cos{\frac{\tilde{\theta}}{2 } } \right ] \nonumber \\ & \times & \left [ -e^{i/2(\phi_2-\phi_+ ) } \sin{\frac{\theta_2}{2}}\cos{\frac{\theta_+}{2 } } + e^{-i/2(\phi_2-\phi_+ ) } \cos{\frac{\theta_2}{2}}\sin{\frac{\theta_+}{2 } } \right ] \end{aligned}\ ] ] where @xmath85 denote azimuthal and polar angles of the final positron ( electron ( 1 ) , muon ( 2 ) ) which are defined with respect to the direction of the initial electron beam @xmath86 and @xmath87 is given by @xmath88 the decay width can be written as @xmath89 where @xmath90 formulae eq .
( a.13 ) describe the 3body decay of a chargino with energy @xmath91 and angles @xmath92 .
the angles of the chargino and of the particles produced by chargino decay are defined with respect to the direction of the initial electron beam @xmath93 .
angles of the decaying particles are also defined with respect to the initial cm system of colliding electrons .
we have left quantities connected with @xmath94 as independent parameters to be integrated over . from the 12 quantities describing the chargino 3body decay ,
four are eliminated by momentum conservation .
these are chosen to be the angles @xmath95 and the energies @xmath96 , namely : @xmath97 the angle @xmath98 is fixed by two relations : @xmath99 for completeness , it is trivial to compute the sneutrino decay width . for @xmath101 ( the scenario discussed in the text , see fig.4 ) only one decay channel to an @xmath102 pair is open ( for simplicity we assume that only one @xmath39 dominates ) : @xmath103 finally for the combined process of production and decay we obtain in the narrow width approximation : @xmath104 j. gluza and t. whrmann , phys .
b408 * ( 1997 ) 229 .
m. hirsch , h.v .
klapdor - kleingrothaus , s. kolb and s.g .
kovalenko , phys . rev .
* d57 * ( 1998 ) 2020 ; m. hirsch , h.v .
klapdor - kleingrothaus and s.g .
kovalenko , phys . rev .
* d57 * ( 1998 ) 1947 .
s. kuhlman _ et al . _ [ nlc zdr design group and nlc physics working group collaboration ] , hep - ex/9605011 .
j. ellis , j. s. hagelin , d. v. nanopoulos , k. olive and m. srednicki , nucl .
* b238 * ( 1984 ) 453 .. h. dreiner , s. lola and p. morawitz , phys . lett .
* b389 * ( 1996 ) 62 .
s. kawasaki , t. shirafuji and s.y .
tsai , prog .
* 49 * ( 1973 ) 1656 ; d.a .
dicus , e.c.g .
sudarshan and x. tata , phys . lett .
* b79 * ( 1985 ) 79 ; for latest applications , see e.g. : g. moortgat - pick and h. fraas , phys .
* d59 * ( 1999 ) 015016 ; g. moortgat - pick and h. fraas , acta phys . polon .
* b28 * ( 1997 ) 2395 ; g. moortgat - pick , h. fraas , a. bartl and w. majerotto , acta phys . polon .
* b29 * ( 1998 ) 1497 ; g. moortgat - pick , h. fraas , a. bartl and w. majerotto , eur .
j. * c7 * ( 1999 ) 113 ; g. moortgat - pick , h. fraas , a. bartl and w. majerotto , eur
j. * c9 * ( 1999 ) 549 . v. barger , k. cheung , b.a .
kniehl and r.j .
phillips , phys . rev .
* d46 * ( 1992 ) 3725 ; v. barger , j.f .
beacom , k. cheung and t. han , phys . rev .
* d50 * ( 1994 ) 6704 . k. koodziej and m. zraek , phys .
d43 * ( 1991 ) 3619 . | we examine single chargino production in conjunction with r - parity lepton number violation in future lepton - lepton collisions .
present bounds on r - parity violating couplings allow for a production cross section of the order of @xmath0 for a wide range of sneutrino and chargino masses .
scenarios of chargino decay which lead to purely leptonic signals in the final state and without missing energy are also discussed .
= 5000 = 5000 = 1000 @twoside -17pt eqnsel = mathrm#1#1ifundefinedmathrm#1#1 1.2 cm and |
the lep collaborations now all have high - statistics data samples which enable them to make accurate measurements of a wide range of @xmath10 qcd observables [ 1 - 9 ] - jet fractions , thrust distributions , the hadronic width of the @xmath4 , energy - energy correlations , etc .. for most of these quantities non - perturbative effects such as hadronization corrections are expected to be reasonably small and next - to - leading order ( nlo ) calculations are available in renormalization group ( rg ) improved qcd perturbation theory @xcite . for the hadronic width of the @xmath4 alone we have a next - nlo ( nnlo ) calculation available
@xcite . by comparing such calculations with the data one hopes to be able to extract @xmath11 , the fundamental su(3 ) standard model parameter , or equivalently @xmath12 , the @xmath13 scheme coupling constant .
unfortunately this intention is hampered by our rather limited current knowledge of what qcd actually predicts for any of these quantities due to the dependence of perturbative calculations at nlo on the unphysical renormalization scale @xmath14 , and more generally on the renormalization scheme ( rs ) in higher - orders . in a complete all - orders calculation of any quantity
the @xmath14-dependence would cancel and so it is an arbitary parameter , but our reliance on a truncated expansion in @xmath12 in which such a cancellation no longer quite occurs renders predictions ambiguous .
since it was first clearly formulated in ref .
@xcite this ` scheme dependence problem ' has been the subject of an extensive literature which has yet to produce consensus .
the response of theorists to the dependence of nlo predictions on @xmath14 has mainly been to advocate - in the face of the central fact that it is actually an arbitrary unphysical quantity - that some particular choice of the scale is the ` correct ' or ` best ' one , thereby removing the ambiguity .
any such choice must appeal to some criterion outside of the familiar , well - defined business of calculating feynman diagrams and no such appeal has so far been able to command universal support .
confronted with such confusion amongst their colleagues , experimentalists have tended to adopt pragmatic compromises that ` mix and match ' different proposed solutions .
specific reservations about the way this has been done at lep will be expressed in section 2 , but it is clear that any approach to the problem that increases consistency amongst observables by expanding error bars to include a ` theoretical uncertainty ' ( essentially obtained by averaging across camps of theorists ) does so by sacrificing unambiguous clarity over what is being measured , and limits the extent to which qcd could be investigated in detail . in some sense ,
all previously proposed approaches have been an attempt to second - guess the uncalculated higher - order terms in the perturbation series , in that they say this choice of scale and a nlo calculation will give a prediction _ as if _ we knew the full series - and it is likely to be a reasonable , perhaps good , prediction at that . on the other hand ,
our position is essentially that the higher - order terms are simply unknown and that we therefore should restrict attention to the more modest aim of deciding what it is actually possible to learn about the uncalculated terms from the information at our disposal .
a widespread belief already exists that the severity of the scale dependence for any particular observable is somehow related to the size of its corrections in higher - orders , although this overlooks the problem that the size of the unknown terms is itself a scheme dependent notion - there is after all always a scheme in which there are no corrections whatsoever . put another way , the total uncalculated corrections to a @xmath14-dependent nlo prediction must also depend on the scale so that their sum does not .
superficially , we have therefore posed ourselves an unphysical ( and hence uninteresting ) question .
however this @xmath14-dependence would only be a fundamental problem if it were unknown , unpredictable and uncontrollable , when in fact ( as we shall show ) the higher - orders can be split into a predictable contribution incorporating all the awkward @xmath14-dependence and a remaining piece containing all the genuinely unknown aspects .
this is really just a consequence of the fact that the @xmath14-dependence in a nlo calculation has a universal form , and so a minimal version of our claim would be that we are suggesting a simple , convenient means of comparing theory and data in which the scale dependence is trivial .
the formalism to be proposed has the advantages of being derived non - perturbatively and as far as possible entirely in terms of rs - invariant quantities .
in particular , @xmath15 is avoided in favour of @xmath16 , the dimensional transmutation parameter of qcd , as the free parameter in the theory since the former is just a theoretical construct without any direct experimental significance and is indeed strictly a different quantity in nnlo from what it was in nlo .
it remains true that @xmath17 is rs - dependent , but with the difference that the values in different schemes are exactly related by only a nlo calculation .
this @xmath17 is to be as directly related as possible to @xmath18 , any generic lep observable dependent on a single dimensionful scale @xmath19 ( i.e. @xmath20 , the @xmath21 centre of mass energy ) . because this @xmath18 will play a role analogous to @xmath15
, we shall refer to this approach as the ` effective charge formalism ' ; it was originally introduced by grunberg @xcite and has also been discussed by dhar and gupta @xcite .
unfortunately , there has been a widespread view that grunberg s approach constitutes no more than a particular choice of rs . indeed the insistence on using the rs - invariant @xmath18 rather than the rs - dependent @xmath15 in the role of coupling means that the formalism has much in common with the fastest apparent convergence ( fac ) criterion for choosing the scheme , by which the rs is chosen such that the higher corrections vanish .
this perturbative property is a nice one , often simplifying calculations , but in this paper it is entirely secondary to the clarifying benefits of non - perturbatively identifying each observable as a coupling ( from which it follows trivially ) .
we can not overemphasize that the effective charge ( ec ) formalism is both derived non - perturbatively and is more general than the adoption of a particular scheme , even if pieces of that formalism can be interpreted in terms of the fac scheme . because it should provide a way of measuring @xmath17 without any scale ambiguity whatsoever , thereby saving the formalism from merely being a useful framework for displaying data , as important an interpretation of the formalism derives from the way in which @xmath18 runs as a renormalisation group improved coupling with @xmath19 .
this evolution is described by a function @xmath22 which is simultaneously both the basis for a convenient rs - invariant measure of the importance of the higher - orders and the @xmath23-function corresponding to the fac scheme .
thus the effective charge @xmath23-function @xmath22 can , at least in principle , be experimentally measured , so that by detailed investigations of the running of @xmath18 with @xmath19 one can determine @xmath24 even in the absence of nnlo perturbative calculations .
the formalism applied in this way is non - perturbative and transcends the use of any particular scheme , with the only barrier to an accurate determination of @xmath24 being the purely experimental uncertainty .
of course at present we do not have high precision information on the running of these lep observables , and the running can only be investigated using data obtained at lower energies with different detectors and machines . in an ideal world lep
would take data at energies below @xmath25 and study the running of the observables in detail .
as we shall argue , this would enable reasonably precise determinations of @xmath3 , and more importantly would also indicate whether nnlo corrections by themselves would be sufficient , helping to establish whether the extremely arduous calculations involved are worth embarking on @xcite . even in the absence of information on nnlo corrections and the detailed running of the observables with energy ,
however , there is still information to be obtained with the effective charge formalism , given only nlo calculations .
one can exhibit the relative size of the uncalculated higher - order corrections for different quantities . in the effective charge language
these are related to how the energy dependence of the quantity at the experimental energy differs from its asymptotic dependence as @xmath26 .
if this difference can not be neglected , then nnlo calculations or measurements of the running are necessary before @xmath6 can be determined .
this can be decided in the effective charge formalism given only a nlo calculation and data at one energy .
our proposal is thus not a solution to the scheme dependence problem in the conventional sense . for a start ,
sufficiently much can be learnt about qcd independently of the absolute value of @xmath24 for the problem itself to be something of a distraction ; refining the experimental data on the relative size of higher - order corrections for different observables and providing explanations for the aberrant ones is a serious challange in its own right .
furthermore , the energy - dependence data already allows us to make tentative absolute statements in a way complementary to perturbation theory .
the scale dependence problem is not so much solved as completely outflanked .
a preliminary discussion of the effective charge formalism applied in this way has appeared in references [ 17 ] and [ 18 ] .
the plan of this paper is as follows . in section 2
we shall review the rs dependence problem and outline the traditional approach to @xmath12 extraction . in section 3
we shall introduce in detail the effective charge formalism sketched above and stress its advantages over the conventional approach .
we shall attempt to determine @xmath3 using the rather limited information on nnlo corrections and energy dependence of observables currently available .
we shall also discuss an attempt to resum higher - order terms in the effective charge beta - function using the knowledge of leading and next - to - leading logarithms in the jet resolution variable @xmath27 for jet rates in the @xmath28 or durham algorithm @xcite , where the perturbation series has an exponentiated structure @xcite .
such a next - to - leading logarithm ( nll ) resummation of the effective charge @xmath23-function is without ambiguity , and in the effective charge language the @xmath27 dependence can be extracted from the experimental measurements and compared with that of the resummed @xmath23-function , indicating whether or not the nll approximation is adequate .
recent attempts to compare resummed perturbation series with the lep data require use of an ad hoc matching procedure @xcite to include the exact nlo results , the additional problem of scale dependence still remaining .
in section 4 we give our conclusions .
consider a generic dimensionless lep observable @xmath18 , where @xmath19 denotes the single dimensionful scale on which it depends , typically the @xmath10 centre of mass energy @xmath29 . in rg
improved perturbation theory we can write without loss of generality , @xmath30 where @xmath31 denotes the rg improved coupling @xmath32 .
notice that by dividing any observable depending only on a single dimensionful scale by its possibly dimensionful tree - level perturbative coefficient ( which will be rs invariant ) , and raising to a suitable power , we can always obtain a series of the form representing a dimensionless @xmath18 .
the coupling @xmath31 and series coefficients @xmath33 depend on the rs employed . using the compact notation and conventions introduced by stevenson @xcite @xmath31
is specified by a @xmath23-function equation @xmath34 where @xmath35 , with @xmath14 the renormalisation scale and @xmath6 the dimensional transmutation mass parameter of qcd , here differing from the traditional definition by a factor @xmath36 .
@xmath37 and @xmath38 are rs invariants dependent only on the number of quark flavours @xmath39 and the number of colours @xmath40 ; for @xmath41 we have @xmath42 and @xmath43 . in all our lep determinations
we take @xmath44 and assume massless quarks .
for massive quarks the scheme dependence discussion will also go through in any mass - independent rs , i.e. one in which the coefficients @xmath45 , @xmath46 @xmath47 do not depend on the fermion masses @xcite .
the higher coefficients @xmath48 , @xmath49 , @xmath47 are rs - dependent . indeed stevenson @xcite , extending the work of stueckelberg and petermann @xcite , has shown that one may consistently use the parameters @xmath50 to label the renormalization scheme . in the conventional approach when retaining terms up to and including @xmath51 in equation one truncates the @xmath23-function of equation retaining terms up to and including @xmath52 . on integrating up the truncated equation one can define @xmath53 and correspondingly one finds for consistency @xmath54,@xmath55,@xmath56 , @xmath47,@xmath57 . in this way
the @xmath58-order truncated approximant is also labelled by the scheme variables , @xmath59 . of course when summed to all orders this dependence must cancel and a formally rs - independent sum be obtained .
the formal consistency of perturbation theory further ensures that @xmath60 so that differences between results in two different rs s are formally effects one order higher in perturbation theory .
of course @xmath61 depends on @xmath62 and @xmath63 and may well not be a small coefficient . before discussing the general situation
further let us consider the simplest @xmath64 , nlo case .
this is the accuracy to which all but one of the lep observables have been calculated . to nlo
we have @xmath65 where @xmath66 is obtained by integrating up the nlo truncation of equation @xmath67 with the boundary condition @xmath68 one obtains @xmath69 where we define the function @xmath70 for later use , so that @xmath71 . how does @xmath72
depend on @xmath73 explicitly ? to see this , consider two different rs s , @xmath74 and @xmath75 .
the connection between the corresponding renormalised couplings @xmath31 and @xmath76 will then be such that @xmath77 and we can define @xmath78 inserting into the series for @xmath79 and @xmath80 in the two rs s and equating coefficients one finds @xmath81 by integrating up @xmath82 and @xmath83 as in and taking the difference we find @xmath84 where the @xmath85 and @xmath86 terms reflect contributions beyond nlo in the @xmath23-functions . using equation and equating coefficients of corresponding powers of @xmath31
gives @xmath87 eliminating @xmath88 between equations and we finally arrive at @xmath89 this implies that @xmath90 it is useful to identify the rs - invariant combination @xcite @xmath91 since @xmath18 is a function of a single dimensionful scale @xmath19 it follows that @xmath92 with @xmath93 and @xmath94 rs - invariants .
@xmath94 is dependent only on the particular qcd observable @xmath95 .
this strongly suggests that these quantities should have a _ physical _ significance as opposed to rs - dependent quantities such as @xmath54 and @xmath96 which depend on unphysical parameters .
as we shall show in section 3 this is indeed the case , @xmath93 and @xmath94 are connected with the asymptotic ( @xmath26 ) dependence of @xmath18 on @xmath19 with @xmath97 and so @xmath93 , or equivalently @xmath94 could in principle be directly measured given unlimited experimental energies .
this observable - dependent ` physical ' quantity @xmath94 is directly connected to @xmath98 the universal dimensional transmutation parameter of qcd which we are attempting to determine . to see this , recall that @xmath99 is the variable which specifies the rs in nlo .
notice it is _ not _ sufficient to specify that one has chosen a given renormalisation scale @xmath14 in order to specify the rs even at nlo .
the renormalisation scheme is specified by @xmath73 which involves @xmath98 as well . since @xmath93 is an invariant we see that for two different rs s , @xmath74 and @xmath100 @xmath101 where @xmath102 denotes the nlo correction evaluated in the renormalisation scheme @xmath74 and @xmath103 with @xmath104 . from equation with @xmath105
we find directly that @xmath106 and so given a nlo calculation in some rs the invariant @xmath94 is exactly related to the dimensional transmutation parameter @xmath98 .
notice that @xmath107 is a _ q - independent _ quantity .
rearranging equation yields @xmath108\ ] ] the exponent is a universal @xmath14-independent constant which exactly relates the @xmath6 s in the two schemes , given nlo calculations in these two rs s .
the implication of this important result @xcite is that it does not matter which @xmath98 we choose to try to extract from the data .
if the rs is mass - independent then @xmath98 will be universal and can be exactly related to @xmath109 in any other scheme by a universal factor given exactly if we have a nlo calculation for any observable in both rs s .
thus there is no non - trivial residual scheme dependence implied in our convenient choice of @xmath3 as the fundamental parameter we wish to extract .
as an illustration consider the minimal subtraction ( @xmath110 ) renormalisation procedure and the @xmath13 or modified minimal subtraction procedure where the @xmath111 terms present in dimensional regularization are subtracted off @xmath112 is euler s constant ) . for any observable , and independent of @xmath14 @xcite , @xmath113 so using equation we have @xmath114 this relation is independent of @xmath39 or @xmath40 . for the @xmath110 and @xmath13 procedures the higher order rs - dependent @xmath23-function coefficients are identical , @xmath115 ( @xmath116 ) .
the only difference is in the subtraction procedure and hence in the definition of the renormalisation scale @xmath14 .
a choice of scale @xmath14 using the @xmath110 procedure corresponds , therefore , to the _ same _ renormalization scheme as use of the @xmath13 procedure with scale @xmath117 .
we are finally in a position to exhibit the explicit dependence of @xmath118 on @xmath73 .
in fact it is easier to equivalently consider the dependence on @xmath66 . using equation to write @xmath54 in terms of @xmath119 and @xmath73 , and using for @xmath66 in terms of @xmath73 we find @xmath120 when plotted versus @xmath96
, @xmath121 has the generic approximately inverted - parabolic shape shown in figure 1 , provided that @xmath122 . from this condition
will automatically hold provided that @xmath123 .
before we discuss this nlo scheme dependence in more detail let us complete the picture by discussing the explicit scheme dependence of the higher coefficients @xmath55 , @xmath56 , @xmath47 as well .
integrating up the @xmath23-function truncated at @xmath58 order with an infinite constant of integration ( related to the definition of @xmath6 , see section 3 ) we obtain , analogous to equation , @xmath124 , \nonumber \\ b^{(n)}(x)&\equiv&(1+cx+c_2x^{2}+c_{3}x^{3}+\ldots+c_{n}x^{n } ) .
\\end{aligned}\ ] ] this transcendental equation can then be solved explicitly for @xmath53 . to obtain the explicit rs dependence of the @xmath33 it is convenient to identify a special rs , the effective charge ( ec ) scheme @xcite , in which @xmath125 .
then @xmath126 in this scheme , and the coupling constant is the physical observable .
this choice of rs will correspond to particular values of the scheme parameters , @xmath127 .
the fact that @xmath128 follows directly from equation , @xmath129 then @xmath130 . from equation
we have that for two rs s , barred and unbarred , @xmath131
if the barred rs is chosen to be the ec scheme then @xmath132 with @xmath133 . then gives @xmath134 where @xmath135 is the inverted perturbation series . by expanding out the right - hand side as a power series in @xmath95 and equating coefficients
we obtain @xcite @xmath136 these objects are @xmath19-independent rs invariants constructed from the @xmath33 and @xmath137 in any rs . for instance given nlo and nnlo calculations of @xmath138 and since @xmath139 is known @xcite , one can construct @xmath140 as we shall see in the next section just like the nlo invariant @xmath119 , the nnlo and higher @xmath141 invariants have a physical significance , whereas quantities such as @xmath142 are intermediate rs - dependent quantities which should be eventually eliminated in favour of rs invariants .
having obtained @xmath143 from nnlo , and higher order calculations in any convenient rs , we can exhibit the explicit @xmath144 dependence of the perturbative coefficients by rearranging equation .
we have that @xmath145 the result for @xmath146 is a polynomial in @xmath147 of degree @xmath148 with coefficients involving @xmath149 and @xmath150 , such that @xmath151 .
let us now return to figure 1 and discuss how one might attempt to use it to extract @xmath3 or @xmath1 in the @xmath13 scheme .
notice first that the curve @xmath118 of equation is a _
universal _ function of @xmath73 and @xmath119 ( for given fixed @xmath39 ) , where the value of the invariant @xmath119 depends on the particular observable @xmath95 .
@xmath119 is of course directly related to @xmath3 given a nlo calculation of @xmath152 , with @xmath153 @xmath118 has a maximum at @xmath154 ( @xmath155 ) where @xmath156 .
the full - width at half - maximum in @xmath96 is approximately @xmath157 .
@xmath121 vanishes at @xmath158 ( @xmath159 ) and at @xmath160 ( @xmath161 ) . for comparison figure 1
is drawn with @xmath162 and @xmath163 .
let us suppose that the horizontal dashed line in figure 1 represents the measured experimental data . if @xmath119 is sufficiently small that the maximum @xmath164 lies above the data then the curve will cut the data at two scales @xmath165 ( as illustrated ) .
conversely , if @xmath119 is made larger the curve will be below the data line .
thus an infinite set of @xmath166 pairs fit the data perfectly .
if we wish to _ measure _ @xmath119 ( and hence @xmath3 ) we must _ specify _ @xmath73 .
this is the nlo scheme dependence problem . at nnlo one
would have a surface @xmath167 and to extract @xmath119 one would need to specify @xmath73 and @xmath48 , and so on with one extra unphysical parameter for each order in perturbation theory . at least in nlo for a given value of @xmath119 ( @xmath3 )
there is a _ maximum
_ possible @xmath121 .
in nnlo and higher one can show @xcite that for a given value of @xmath119 ( @xmath3 ) there exists a choice of @xmath168 , @xmath47 , @xmath169 such that @xmath170 has _ any _ desired positive value .
thus , for _ any _ @xmath3 we can choose a sequence of schemes such that @xmath171 , the experimentally measured value .
various `` solutions '' of the scheme dependence problem , i.e. motivations for particular choices of @xmath172 have been proposed .
\(1 ) _ principle of minimal sensitivity ( pms ) _
@xcite the idea here is that since the exact ( all - orders ) result is independent of @xmath73,@xmath48,@xmath49,@xmath173 one should choose the @xmath58-order approximation @xmath174 to mimic this property and be as insensitive as possible to the chosen value of the unphysical parameters .
that is one arranges that @xmath175 and the pms scheme is specified by @xmath176 . at nlo
one has to solve @xmath177 and this corresponds to solving the transcendental equation @xmath178 + \frac{1}{2}\frac{c}{1+ca } = \rho_{0}.\ ] ] the solution @xmath179 then gives @xmath180 the stationary pms point of @xmath118 is the maximum at @xmath181 ( this is a somewhat better approximation than the cruder @xmath182 given previously for the position of the maximum ) .
@xmath119 is adjusted so that @xmath183 .
\(2 ) _ effective charge ( ec ) scheme _
@xcite this corresponds to a choice of scheme such that @xmath184 , hence @xmath185 is an effective charge . as we have seen the scheme parameters are then @xmath186 , where the @xmath187 are the rs - invariants in equation .
once these have been computed up to @xmath188 from higher - order calculations in any technically - convenient rs ( e.g. @xmath13 ) they can be inserted in the integrated @xmath23-function equation ( 2.24 ) which can then be solved with @xmath189 for given @xmath119 .
@xmath119 can then be obtained by requiring @xmath190 .
this particular approach is sometimes also referred to as the fastest apparent convergence criterion . in this language
it appears rather artificial , but as we shall emphasise in section 3 its advantage is that the all - orders coupling constant _ and _ @xmath23-function are experimentally observable physical quantities , allowing a _ non - perturbative _ approach to supplement @xmath3 measurement , and a physical definition of the uncertainty .
it should be emphasised that ec and pms predictions in nlo are very similar since , as we have seen , @xmath128 , @xmath191 .
it is also true that in nnlo ec and pms remain close to each other @xcite . for the pms approach , however , the coupling constant and @xmath23-function are unphysical quantities , and it is not clear even that their all - orders versions are defined , given the complex nature of the coupled equations which must be solved .
\(3 ) _ physical scale _
@xcite according to this viewpoint the renormalization scale should be chosen to be close to the physical scale in the problem , @xmath105 .
if predictions in the vicinity of @xmath105 are strongly @xmath14-dependent then this is supposedly an indication that the perturbation series is intrinsically badly behaved .
this viewpoint is usually justified by noting that perturbative coefficients in higher orders will be polynomials in @xmath192 , @xmath193 to avoid unnecessarily large logarithms one should therefore set @xmath194 . whilst this is true as far as it goes
, it tacitly assumes that the nlo rs dependence is completely given by the dependence on the renormalisation scale .
in fact as we have seen @xmath146 where @xmath195
so the coefficients @xmath196 in will depend on @xmath98 .
the meaning of @xmath14 of course depends for instance on whether it is @xmath197 or @xmath198 .
in contrast we have from that @xmath199 so to avoid unnecessarily large terms we should clearly choose @xmath200 , the effective charge scheme ! to make contact with notice that @xmath201 = b \ln \left [ \frac{\mu}{q } e^{r_{1}^{rs}(q)/b } \right ] .\ ] ] so we can write as @xmath202^{l}\ ] ] in and the coefficients @xmath203 do not depend on the nlo rs choice , only on @xmath204 and the rs invariants @xmath141 , whereas in the @xmath196 will depend on @xmath205 and so have a hidden dependence on the nlo rs choice which is customarily ignored in the usual ` physical scale ' argument . applying the argument to instead
one would infer that one should set @xmath206 , which of course corresponds to the effective charge scheme , @xmath207 .
for particular cases the extra factor may be large .
recall that @xmath94 is an rs - invariant characterising how the observable runs with @xmath19 asymptotically .
different observables will have different @xmath94 s .
if it happens that @xmath208 ( @xmath209 ) then predictions will be stable and large logarithms avoided for @xmath210 , but no special physical significance should be accorded to this fact .
it is always possible to modify the subtraction procedure so that @xmath211 , since @xmath6 s in different rs s are related according to .
we conclude that a modified ` avoidance of unnecessarily large logarithms ' physical scale argument which correctly labels the nlo rs dependence actually leads to the effective charge scheme , @xmath207 . for most of the lep observables we have a differential distribution in a kinematical variable rather than a single experimental data value as in figure 1 . for instance a thrust distribution in the thrust variable @xmath212 , or a multijet rate in the jet resolution cut @xmath27 .
denoting such a generic kinematical variable by @xmath213 , we should really add an extra perpendicular axis to figure 1 and consider the @xmath213 dependence as well , @xmath214 .
the nlo coefficient will also depend on @xmath213 and so for a given choice of @xmath3 we will have @xmath215 from .
if @xmath3 is sufficiently large then the experimental data @xmath216 will intersect @xmath214 for at least one @xmath73 value for any @xmath213 and so the data can be fitted perfectly with a suitable @xmath213-dependent rs parameter choice @xmath217 .
in general there will be two intersection points as in figure 1 and so two functions @xmath218 will fit the data perfectly . since this can be done for any suitably large @xmath3 there is no unique best fit .
the opal analysis @xcite uses a mixture of approaches ( 3 ) and ( 4 ) .
the data for each observable over a range of the corresponding kinematical variable is compared with the nlo prediction with the @xmath13 scale chosen as @xmath219 , and a best fit for @xmath3 performed .
the resulting @xmath220 values are shown in figure 2(a ) , reproduced from reference @xcite .
there is evidently a considerable scatter in the @xmath12 values obtained .
one then performs a simultaneous best fit for @xmath14 and @xmath3 for each observable over a range of the kinematical variables .
this provides a second @xmath220 value .
the quoted central value of @xmath220 is then taken to lie mid - way between these extremes ( perhaps modified by hadronization ) and the difference between them is taken as indicative of the size of the uncalculated higher - order corrections . by enlarging the ` scheme dependence error ' in this way one obtains greater apparent consistency between different observables as shown in figure 2(b ) .
in our view one learns nothing useful from an analysis of this kind .
consider figure 2(a ) and the considerable scatter of @xmath12 values obtained from different observables fitted with @xmath219 .
choosing @xmath219 means that from the higher - order coefficients are given by @xmath221 where the @xmath203 depend on the invariants @xmath222 and on the nnlo and higher rs parameters @xmath150 .
we have tabulated in table 1 the @xmath223 values for some of the observables at particular values of the kinematical parameters using the nlo matrix elements of reference 10 .
( for more details see section 3b ) .
one sees that they are large and rather scattered , e.g for the jet rates @xmath224 which results in rather small ec ( or pms ) scales , @xmath225 .
as emphasized earlier no physical significance attaches to these small scales , since the scale itself is just an artifact of the unphysical subtraction procedure employed .
thus from one may expect large and rather scattered higher order corrections for the different observables with @xmath219 , which translates itself into the expectation of scattered values of @xmath220 , which indeed are observed .
if we instead choose @xmath226 ( the effective charge scale ) for each observable then the higher order coefficients are given by @xmath227 where @xmath228 does not depend on the nlo rs choice ( @xmath73 ) , only on the nnlo and higher rs parameters and the rs - invariants @xmath141 .
any scatter now observed in the extracted @xmath3 or @xmath220 for different observables will be attributable to these nnlo and higher rs invariants being large , and not to the scatter due to the nlo scheme dependence logarithms involving @xmath223 , which are avoided by the choice @xmath229 . as we shall discuss in section 3
there is still scatter when this is done but it is somewhat reduced . with the choice @xmath219 predictable scatter due to _ avoidable _
already known nlo scheme dependence logarithms is superimposed on top of the interesting scatter due to the size of the nnlo and higher invariants which is telling us about nnlo and higher order uncalculated corrections .
the choice @xmath229 removes the predictable scatter and provides genuine information on the interesting nnlo and higher - order uncalculated corrections .
the @xmath13 scales obtained by fitting to the data for an observable over some range of the associated kinematical parameters tend to be considerably smaller than @xmath230 , as for the ec or pms scale . typically for values of the kinematical variable @xmath213 well away from the 2-jet region where large logarithms will be important
, @xmath215 and @xmath231 do not vary strongly with @xmath213 .
this means that one can not obtain best fits over such a range of @xmath213 .
it is only by including the 2-jet region that one can obtain stable best fits , but then fitting for a @xmath213-independent scale over a range where @xmath215 and @xmath231 ( and presumably uncalculated higher order corrections ) are strongly @xmath213-dependent represents a rather severe compromise , and it is not obvious what one learns . in conclusion , we believe that the standard nlo analyses used to extract @xmath12 from lep data do not serve to disentangle the genuinely new and interesting nnlo and higher order uncalculated corrections from predictable higher order corrections connected with the choice of rs at nlo .
the resulting quoted values of @xmath220 ( or @xmath3 ) with attendant scheme dependence uncertainties therefore reflect the ad hoc way in which they were extracted , rather than the actual values of these parameters .
in contrast , by choosing the effective charge scale for each observable one exposes the relative importance of uncalculated nnlo and higher corrections . an effective charge formalism which can be supplemented with non - perturbative information
is described in the next section .
* a. the @xmath19 dependence of @xmath18 * for the dimensionless qcd observable @xmath18 we can define @xmath232 @xmath233 and hence @xmath234 are , at least in principle , experimentally observable quantities , although collider experiments are usually designed to make high - statistics observations at a fixed energy @xmath19 rather than examining the detailed running of observables with energy . to make contact with the standard perturbative approach we note that is the @xmath23-function equation in the effective charge scheme where the coupling @xmath235 and the @xmath23-function is @xmath22 as defined in equation .
so we have @xmath236 the effective charge @xmath23-function @xmath22 may be regarded as a physical observable .
as measured from data it will include a resummed version of the ( perhaps asymptotic ) formal perturbative series exhibited in together with the non - perturbative terms involving @xmath237 which are invisible in perturbation theory .
( here s,@xmath238,@xmath239,@xmath240, are observable - dependent constants which we shall not specify further @xcite . ) hence by comparing with data one can test how well the first few terms of the perturbative series serve to represent the observed running of @xmath18 . that is , from measurements of @xmath241 and @xmath242
, one can test the extent to which @xmath243 describes the data . a marked discrepancy
would either indicate the importance of the nnlo invariant @xmath244 and higher terms , or the relative importance of the non - perturbative contributions , or both . given a collider program dedicated to the investigation of the @xmath19-dependence of observables
the above tests would indeed be powerful and useful . the focus of present studies , however , is the accurate measurement of @xmath95 at fixed @xmath19 , i.e. @xmath245 , and the comparison of these results with nlo matrix element calculations in order to extract @xmath3 . to obtain @xmath18 we clearly need to integrate equation
the boundary condition will be the assumption of asymptotic freedom , that is @xmath246 as @xmath247 , which corresponds to the requirement that @xmath248 for @xmath249 , with @xmath250 some suitably low energy .
( equivalently , @xmath251 for @xmath249 , assuming @xmath252 or @xmath253 for @xmath254 qcd . )
any zero of @xmath234 , @xmath255 , say , would correspond to @xmath256 as @xmath257 and ultraviolet fixed point behaviour . integrating equation
we obtain @xmath258 where @xmath94 is a finite constant which depends on the way the infinite constant is chosen .
the infinite constant can be chosen to be @xmath259 where @xmath260 is _ any _ function which has the same @xmath261 behaviour as @xmath262 .
we know from equation that @xmath262 has the universal @xmath263 behaviour @xmath264 so we choose @xmath265 .
inserting this choice for @xmath260 and rearranging equation we find @xmath266.\ ] ] the first integral on the right - hand side is just @xmath267 where @xmath70 is the function defined in equation ( 2.6 ) , @xmath268.\ ] ] we define for later convenience @xmath269.\ ] ] notice that despite appearances the integrand of @xmath270 is regular at @xmath271 .
then we have @xmath272 as @xmath257 , @xmath246 so that @xmath273 and thus asymptotically @xmath274 with the constant of integration given by @xmath275 notice that in arriving at ( 3.9 ) we did not need to refer to perturbation theory , except to assume the asymptotic @xmath263 @xmath276 behaviour @xmath277 . despite this
, the constant @xmath94 obtained with the particular choice @xmath265 is precisely the @xmath94 introduced in equation .
that is @xmath278 we can see this immediately since if @xmath31 denotes the coupling in the rs with @xmath279 using the @xmath13 subtraction procedure in nlo , and with higher order @xmath23-function coefficients zero @xcite we have @xmath280 giving a well - defined all - orders coupling .
further @xmath281 @xmath282 as @xmath257 , so asymptotically as @xmath283 we have @xmath284 where the ellipsis denotes terms which vanish as @xmath247 . inserting into and using we find as @xmath257 .
so finally we have @xmath285 we could , of course , have written down the result of equation perturbatively in the ec scheme at once by simply using the integrated beta - function equation of equation ( 2.24 ) with @xmath286 , @xmath287 , and @xmath185 .
however , our purpose here has been to stress that equation holds _ beyond perturbation theory _ for the measured observable @xmath18 and @xmath288 constructed from the measured running of @xmath18 , @xmath289 .
that is , we can write a _ non - perturbative _ closed expression _ exactly _ relating the universal constant @xmath3 to observable quantites .
since @xmath290 involves an integral in @xmath18 between @xmath291 and 0 , to actually measure it would require knowlege of @xmath18 ( equivalently @xmath233 ) on the full range @xmath292 .
we of course only know @xmath18 on some finite energy range , so if we want to obtain @xmath3 from equation and measurements of @xmath18 there will always be an uncertainty related to our lack of knowledge about the behaviour of @xmath18 beyond the highest energy reached , or correspondingly related to our lack of knowledge about @xmath22 for @xmath293 , with @xmath250 the highest energy reached .
as we shall discuss in sub - section 3c we can use measurements of the running of @xmath18 to determine @xmath22 in the vicinity of @xmath294 , and perturbative qcd calculations to determine @xmath22 in the vicinity of @xmath295 .
putting these two pieces of information together we can make unambiguous statements about the validity of perturbation theory and the nature of the function @xmath22 , which can assist our attempts to determine @xmath3 .
notice the fundamental significance of the nlo perturbative coefficient @xmath152 in all of this .
knowledge of the full behaviour of @xmath18 at large @xmath19 would allow one to extract @xmath94 from equation , but this by itself is _ observable - dependent _ and so one would not test qcd .
a test is only possible if one also knows @xmath152 ( from feynman diagram calculations ) and can then obtain the _ universal _ @xmath3 from equation .
the universal @xmath3 of course depends on the number of active quark flavours , @xmath296 ; the rs invariants @xmath37 and @xmath38 occurring in equation also involve @xmath39 .
since @xmath3 in equation involves the @xmath283 behaviour of @xmath18 , one might wonder what value for @xmath39 should be taken . with three generations of quarks , @xmath297 presumably represents the asymptotic number of active flavours and so , introducing the notation @xmath298 , @xmath299 , the function @xmath260 in equation should be chosen as @xmath300 .
so finally with the explicit @xmath39 dependence exhibited becomes @xmath301 \right\ } \\ & = & b(6)\ln \frac{q}{\tilde{\lambda}^{(6)}_{\overline{ms } } } - r_{1}^{\overline{ms}}(q , n_{f}=6 ) - \int_{0}^{r(q ) } dx \left [ \frac{b(6)}{\xi(x ) } + \frac{1}{x^{2}(1+c(6)x ) } \right].\nonumber\end{aligned}\ ] ] obviously the strict asymptotic @xmath257 behaviour of @xmath18 and @xmath302 is an idealized concept since we could never actually measure it .
for instance , it presumably makes no sense to consider qcd in isolation beyond the gut energy scale . rather than using for lep observables where we have @xmath163 active flavours , it makes more sense to consider the equation for @xmath18 in a world with @xmath163 active flavours , replacing @xmath303 , @xmath304 , and @xmath305 in by @xmath306 , @xmath307 , and @xmath308 .
the decoupling theorem @xcite means that we can consider an @xmath163 version of qcd for energies below top threshold . as we shall discuss
we shall be interested in how the observed @xmath19-evolution of @xmath18 at lep energies compares with its asymptotic @xmath257 evolution in a _
hypothetical _ world with @xmath163 flavours .
this will decide how accurately we can determine the universal constant @xmath308 . to have any chance of determining @xmath305 we would require measurements around the top threshold and at higher energies , which we do not currently possess . a further subtlety connected with the effective charge formalism concerns the fact that one can always define effective charges other than @xmath309 .
more generally one could define @xmath310 where @xmath311 is any analytic function of @xmath31 .
chla has suggested that this arbitrariness is just the scheme dependence problem in disguise @xcite . in our view
this is incorrect ( see also grunberg s response @xcite ) .
in fact the arbitrariness in defining the effective charge is completely equivalent to redefining the observable to be @xmath312 , maintaining the standard definition of effective charge @xmath313 .
one can of course always consider @xmath314 as the measured observable rather than @xmath95 , and then nlo truncation of the perturbation series for @xmath314 using some rs will give different results from those obtained with @xmath95 .
such a redefinition of the observable would only be useful if some information on the uncalculated higher order corrections , i.e. knowledge of the function @xmath22 , were available to inform the choice .
let us suppose that we have measurements for a number of lep observables at a fixed energy ( @xmath245 ) , and knowledge of their nlo qcd perturbative corrections @xmath152 .
what can we learn ?
recall from equation that @xmath315 where @xmath37 , @xmath38 and @xmath152 are evaluated with @xmath163 . as @xmath257 , @xmath316 and @xmath317 at sufficiently large @xmath19 , @xmath318 and may be used to obtain @xmath308 .
we have that @xmath319.\ ] ] given only a nlo perturbative calculation we have no information on @xmath320 . from the measured observable @xmath321 and the nlo perturbative coefficient
we can then extract @xmath322.\ ] ] to relate this to the discussions of section 2 , the quantity on the left of equation is just the value of @xmath3 which would be extracted from the data at nlo choosing @xmath323 , the effective charge scale . if @xmath324 , then this nlo estimate will be accurate .
indeed the fractional error in @xmath3 will be @xmath325 the size of @xmath326 will depend on the qcd observable .
if the @xmath308 values obtained from equation exhibit significant scatter , then this indicates that @xmath320 is not negligible for at least some of the observables .
recall that @xmath320 is defined non - perturbatively in terms of the running of @xmath18 by equation ( 3.8 ) where @xmath327 . if @xmath19 is sufficiently large that @xmath233 has its asymptotic ( @xmath257 ) @xmath19-dependence , that is if @xmath328 , then @xmath329 .
thus significant scatter in the @xmath308 extracted at nlo with the effective charge scale unambiguously indicates that the current experimental @xmath19 is not yet large enough for all observables to be evolving with @xmath19 asymptotically . if we want to reliably determine @xmath308 additional information on the ` sub - asymptotic ' effects is required .
the presence of these effects is a physical fact which can be directly observed from the @xmath19-dependence of data , and can not be remedied by changing the unphysical renormalization scheme .
we stress once again how different this is from the scatter obtained with @xmath219 which will be partly due to different values of @xmath223 and hence which can be modified by changing the unphysical subtraction procedure . to make this distinction more precise , let us define @xmath330 to be the @xmath308 value extracted at nlo using an rs with @xmath331 .
the nlo coefficient @xmath332 may be taken to label the rs at nlo . using @xmath31 to denote the coupling constant in this rs we have @xmath333 and @xmath334 then using equations ( 3.20),(3.23 ) , and ( 3.24 ) it is straightforward to show that @xmath335\tilde{\lambda}^{(5)}_{\overline{ms } } \nonumber \\ & = & \exp(\delta f(r_{1},r)/b ) \tilde{\lambda}(nlo,0 ) , \\end{aligned}\ ] ] where @xmath336 @xmath337 @xmath338 is just the @xmath308 extracted choosing @xmath229 ( @xmath339 ) , given by equation ( 3.21 ) . from equation ( 3.25 )
we see that the nlo scatter in @xmath308 is factorized into two components .
the first is an _ already known _
@xmath332-dependent contribution involving @xmath340 which may be obtained exactly from ( 3.26 ) with @xmath190 , and does _ not _ involve the unknown nnlo and higher rs invariants .
the second component , @xmath341 , does not depend on @xmath332 but does involve the unknown higher invariants , and represents the irreducible uncertainty in determining @xmath3 at nlo .
the uninformative @xmath332-dependent part of the scatter can therefore be completely removed by choosing @xmath229 , which sets @xmath342 .
the scatter is then completely given by @xmath270 .
rather than focusing on the scatter in @xmath3 , we prefer for presentational reasons to concentrate on the relative size of @xmath270 .
we can define for each observable @xmath343 given a nlo perturbative calculation of @xmath152 and experimental measurements of @xmath321 we can then measure @xmath344 from the data up to @xmath345 which should be a universal constant .
changing @xmath24 merely translates the @xmath270 obtained up or down by the _ same amount _ for each observable , so we can choose an arbitrary reference value for @xmath308 .
for two different observables @xmath346 and @xmath347 , the difference @xmath348 can then be measured absolutely and so the _ relative _ size of @xmath270 may be investigated .
if @xmath349 for each observable , corresponding to small sub - asymptotic effects , then the @xmath350 points should lie on a horizontal straight line with some small scatter .
substantial deviations from a horizontal straight line therefore unambiguously indicate the presence of sizeable sub - asymptotic effects .
the scatter in @xmath350 is of course entirely equivalent to the scatter in the @xmath308 extracted at nlo with @xmath351 , these correspond to the reference @xmath308 values for which @xmath352 .
we shall return to these @xmath308 values later . if one has a nnlo perturbative calculation or measurement of @xmath321 at more than one energy then , as we shall discuss in the next subsection , one can attempt to estimate @xmath270 directly .
one can then see if for all observables a single value of @xmath308 brings @xmath350 into agreement with the directly estimated @xmath270 .
we now turn to extracting @xmath350 from lep data .
the observables we shall use are the ones appearing in table 1 .
@xmath353 and @xmath354 denote the jet fractions defined by @xmath355 and @xmath356 , where @xmath357 and @xmath358 are the cross - sections for @xmath359 resolved jets respectively , with @xmath27 the jet resolution cut and @xmath360 the total hadronic cross - section .
the e0 and d labels denote different recombination algorithms used to cluster hadrons together .
details of these algorithms and results for the nlo perturbative coefficients @xmath152 can be found in reference @xcite for the e0 algorithm , and for the d ( durham or @xmath28 ) algorithm in reference @xcite .
@xmath212 denotes the thrust variable , and @xmath361 the energy - energy correlation ( eec ) angle .
we used recent opal data on the differential distributions in these quantities @xcite .
the asymmetry in the energy - energy correlation ( aeec ) obtained by subtracting the eec measured at @xmath361 and ( @xmath362 ) was also considered .
details of the definitions of these quantities and their nlo perturbative coefficients are contained in reference @xcite . from the measured hadronic width of the @xmath4 ( @xmath363 ) and the @xmath4 leptonic decay width ( @xmath364 ) one
can obtain @xmath365 , where @xmath366 has a perturbation series of the form ( 2.1 ) with @xmath367 .
the electroweak contribution ( @xmath368 ) is from reference @xcite .
this coefficient is for massless quarks , if one includes heavy quark masses @xcite , one finds @xmath369 instead in the @xmath13 scheme with @xmath219 . within the errors
the values of @xmath350 obtained are are consistent , so we shall use massless qcd . for the experimental value we take the 1993 lep average @xmath370 @xcite .
as noted earlier the reference value of @xmath371 assumed for @xmath350 is irrelevant since we are concentrating here on the scatter of @xmath372 values .
we shall choose @xmath373 mev which corresponds to the central value obtained from a non - perturbative lattice analysis of the 1p-1s splitting in the charmonium system @xcite .
whilst at present limited by the use of a quenched approximation , such calculations could in future provide reliable estimates of @xmath17 . in figure 3
we show the @xmath350 values obtained for the lep observables discussed above . we have taken for each observable the particular values of the kinematical variables given in table 1
of course for each observable we actually have a @xmath350 distribution in the associated kinematical variable . in figures 4 - 7
we give the @xmath350 distributions for the jet rates @xmath374 , @xmath375 , @xmath376 , @xmath377 as functions of the jet resolution cut @xmath27 .
we see that for @xmath378 , @xmath270 is constant within the errors . for smaller @xmath27
there is a much stronger dependence . in this small @xmath27 region
we may expect sizeable corrections involving @xmath379 .
an attempt to resum leading and next - to - leading logarithms to all - orders for @xmath270 for @xmath376 will be discussed in sub - section 3d . in figures 4 - 7
we used published opal data @xcite without correcting for hadronization effects .
these are expected to be totally negligible for @xmath380 , but will be important in the small @xmath27 region . for thrust and eec the @xmath350 distributions are flat away from @xmath381 and @xmath382 where large corrections involving @xmath383 and @xmath384 respectively will be important .
returning to figure 3 we see that there is significant scatter , in particular between the jet rates with e0 and d recombination algorithms . to proceed further
then we need either nnlo calculations or information on the @xmath19-dependence of these observables .
the only nnlo calculation completed so far is that for @xmath5 @xcite . for @xmath385 and eec there
are published data from the jade collaboration at the petra @xmath10 machine with @xmath29 in the range 22 - 44 gev @xcite , using a comparable analysis . in the next sub - section
we shall attempt to use this data together with the lep data for these quantities to obtain absolute estimates of @xmath270 from the @xmath19-dependence .
we conclude by tabulating in table 2 the @xmath308 values obtained by fitting to lep data at nlo for these same observables at the same values of the kinematical parameters as in table 1 using @xmath386 and @xmath351 .
the @xmath351 values correspond to the @xmath308 values for which @xmath387 , and so the scatter is directly related to points in the @xmath270 plot of figure 3 .
the @xmath219 values show much greater scatter , since as discussed in section 2 the scatter in @xmath223 is superimposed on top of the physically - relevant scatter in @xmath270 . to further emphasise this , suppose that we lived in an ` asymptotic world ' where @xmath245 was sufficiently large that @xmath388 for all observables .
thus for all observables @xmath389 supposing that @xmath390 mev , then may be used to generate data ` @xmath391 ' in such an ` asymptotic world ' . by construction
the @xmath308 obtained by fitting at nlo with @xmath229 will be 200 mev for all observables , but that obtained with @xmath219 will still exhibit significant , but completely predictable , scatter .
these ` asymptotic world ' results are also tabulated in table 2 .
if we have available a nnlo perturbative calculation then the rs invariant @xmath244 defined in equation ( 2.28 ) can be obtained . for @xmath5 , @xmath392 ( @xmath163 ) using the nnlo calculation of reference @xcite .
for all the other observables nnlo corrections have yet to be calculated and so @xmath244 is unknown .
if we insert the perturbative expansion of @xmath393 ( equation ( 2.26 ) ) into ( 3.8 ) we obtain @xmath394 using the 1993 lep average data we have @xmath395 with @xmath396 @xcite .
thus with @xmath392 we have @xmath397 . adjusting the reference value of @xmath308 so that @xmath350 for @xmath5 corresponds with this @xmath398 results in @xmath399 mev .
this is of course just the @xmath3 obtained at nnlo using the ec scheme with @xmath207 and @xmath400 . to obtain estimates of @xmath270 for
the other observables one can try to use their @xmath19-dependence .
suppose that we have measurements of @xmath95 at two energies @xmath401 and @xmath402 ( @xmath403 ) , then we can construct @xmath404 and @xmath405 of equation ( 3.27 ) from the data ( with the same reference value of @xmath308 ) .
then @xmath406 where @xmath407\ ] ] is the integrand of @xmath270 and @xmath408 .
thus the integrand of @xmath270 may be measured from the @xmath19-dependence of the data , @xmath409 obviously by measuring @xmath233 at @xmath410 with sufficient accuracy one can in principle determine @xmath411 and so by suitably detailed measurement over the energy range @xmath412 $ ] one can determine @xmath413 on the interval @xmath414 $ ] .
the situation for measurements at @xmath415 and @xmath416 is shown in figure 8 for @xmath413 versus @xmath417 . the uncertainty in @xmath418 is represented by the horizontal error bar between @xmath419 and @xmath420 and the measurement errors in @xmath419 and @xmath420 themselves contribute a vertical error bar .
we have from equation ( 3.29 ) that @xmath421 so @xmath422 .
thus from a nnlo perturbative calculation we can obtain the integrand at the origin .
notice that , like @xmath94 , the rs invariant @xmath244 is connected with the asymptotic @xmath19-dependence of @xmath18
. a next - nnlo calculation would provide @xmath423 and tell us the slope of @xmath413 at the origin , @xmath424 . if nnlo perturbation theory is adequate to determine @xmath270 , then @xmath425 , corresponding to @xmath426 constant , on the range @xmath427 $ ] .
thus , if nnlo perturbation theory is adequate , then we expect @xmath428 and so from the measurements of @xmath419 and @xmath420 we can estimate @xmath429 from ( 3.32 ) , @xmath430 if nnlo calculations eventually become available for the observables one can then check explicitly whether @xmath431 .
a marked discrepancy would indicate the importance of next - nnlo and higher perturbative effects and/or large ` non - perturbative ' @xmath237 effects . given @xmath244 one
can then estimate @xmath423 from the slope of the straight line joining @xmath432 and @xmath433 and obtain an improved estimate of @xmath434 from the area under the trapezium ( see figure 8) . in this way perturbative calculations and experimental @xmath19-dependence measurements serve as complementary pieces of information .
the nnlo and higher perturbative calculations effectively provide details of the @xmath257 running of @xmath18 which could never be obtained experimentally . together
they can help to constrain the behaviour of the function @xmath435 , and hence to refine the estimates of @xmath308 .
the running of the observables with energy has been used before as a test of qcd .
for instance @xmath436 has been studied as a function of @xmath19 over the petra - lep energy range @xcite .
the apparent running has been compared with the nlo qcd expectations for various choices of renormalization scale @xmath14 , and @xmath6 .
it has even been suggested @xcite that the lep measurements may indicate slightly less running than is expected from qcd , and that light gluinos , which would modify the qcd @xmath23-function , can not be excluded .
such a claim seems ludicrously premature given our lack of knowledge of nnlo and higher qcd effects .
if we take data for @xmath436 at @xmath437 gev from jade @xcite , and at @xmath438 gev from opal @xcite , then we obtain from ( 3.34 ) , @xmath439 , which corresponds to @xmath440 .
so if nnlo perturbation theory is adequate for this observable we estimate @xmath441 , for the as yet uncalculated nnlo rs invariant . for the eec with @xmath442 measured by jade at @xmath437 gev @xcite , and by opal at @xmath438 gev
@xcite we similarly find @xmath443 , which corresponds to @xmath444 , and an estimate of @xmath445 . since @xmath350 is very insensitive to @xmath27 and @xmath361 , comparable results are obtained for other choices of these variables . in figure 9
we plot @xmath270 for @xmath5 , @xmath436 and @xmath446 with @xmath245 .
the diamonds correspond to absolute predictions .
for @xmath5 we take @xmath447 , and @xmath436 and for @xmath448 the @xmath449 values noted above obtained from @xmath19-dependence .
the crosses correspond to the @xmath372 for these observables with the reference @xmath308 chosen as 287 mev so that @xmath350 for @xmath5 agrees with @xmath398 .
there is then agreement , within the considerable errors , between @xmath350 and @xmath450 for @xmath354 and @xmath451 .
demanding consistency between @xmath350 and @xmath452 , @xmath398 then requires @xmath453 mev .
although only performed for a limited number of observables , we regard this consistency between @xmath350 and the absolute @xmath270 obtained from nnlo calculations and @xmath19-dependence measurements as very encouraging .
ideally one would like to see additional lep measurements for all observables off the @xmath454 peak at a lower value of @xmath19 to avoid the uncertainties due to combining measurements from different machines and detectors .
one could then perform an exhaustive analysis of the kind described here and obtain a reliable determination of @xmath308 .
scatter in the @xmath308 values obtained by adjusting @xmath350 to agree with @xmath450 ( obtained from @xmath19-dependence using ( 3.34 ) ) would then unambiguously indicate the importance of next - nnlo corrections or non - perturbative @xmath237 effects for some of the observables .
it is interesting that the above @xmath19-dependence estimates of @xmath244 , and the exact nnlo calculation of @xmath244 for @xmath5 , suggest @xmath244 large ( @xmath455 ) and _
negative_. this has implications for the infra - red behaviour of @xmath18 since it is consistent with fixed - point behaviour , that is @xmath456 implying @xmath457 as @xmath458 @xcite .
there has been much recent interest , both theoretical @xcite and experimental @xcite , in the possibility of resumming leading and next - to - leading logarithms ( nll ) in kinematical variables to all - orders for lep observables . by identifying the leading and next - to - leading logarithms in the perturbative coefficients @xmath459 , with @xmath460 ( where @xmath213 is @xmath461 ) , and by demonstrating generalised exponentiation one can resum these terms to all - orders . to include the exact nlo perturbative coefficient , however , some ad hoc matching prescription is then required @xcite , and anyway the problem of renormalization scale dependence still remains .
the advantage of the ec formalism here is that from the leading and next - to - leading logarithms in @xmath462 one can obtain the leading and next - to - leading logarithms in the rs invariants @xmath463 .
one can then _ unambiguously _ construct @xmath464 @xmath465 is constructed from rs - invariants and so does not involve the renormalization scale , or the nlo perturbative coefficient
. there is , therefore , no analogue of the matching prescription ambiguity and one can obtain @xmath466.\ ] ] this may be directly compared with @xmath350 . if the observed @xmath213-dependence of @xmath350 in the small-@xmath213 region disagrees with that predicted from @xmath467 then one has unambiguous evidence that the nll approximation is inadequate , that is that the neglected sub - leading logarithms and constants are important
. we shall perform this comparison here only for the 2-jet rate @xmath376 , where the exponentiation of logarithms is rather straightforward , deferring discussions of other observables .
we have @xmath468 where @xmath95 has a perturbation series of the form ( 2.1 ) .
to nll accuracy @xcite , @xmath469\ ] ] where @xmath470 .
the @xmath37 dependent term in the exponent may be absorbed into @xmath31 by a change of scale @xmath471 .
since the rs invariants @xmath472 are independent of the renormalization scale to construct them , we can equally use @xmath473 \equiv \exp(r_{0}^{nll}a).\ ] ] one has from ( 3.37 ) that @xmath474 and from ( 2.27 ) with @xmath475 , @xmath476 from the form of @xmath472 of ( 2.28 ) one sees that @xmath204 will not contribute at nll level .
we retain @xmath38 in @xmath82 so that @xmath477 . from ( 3.39 ) and ( 3.40 ) one has @xmath478 then using ( 3.41 ) one finds @xmath479 where @xmath480 if one then inserts @xmath465 of ( 3.43 ) into ( 3.36 ) and change variables from @xmath417 to @xmath481 performing the integration one finds @xmath482 in figure 10(a ) the diamond points show the absolute prediction for @xmath467 versus @xmath483 obtained using ( 3.46 ) together with the opal data of reference @xcite for @xmath376 uncorrected for hadronization effects .
the other points show the corresponding @xmath350 ( see figure 5 ) with the reference value of @xmath308 adjusted so that @xmath350 agrees with @xmath467 at the peak in @xmath467 .
notice that for @xmath484 the nll prediction has a completely different low-@xmath27 behaviour from the uncorrected data . in figure 10(b )
the same quantites using hadronization - corrected opal data for @xmath376 @xcite are displayed .
evidently , the low @xmath27 behaviour is now at least in qualitative agreement .
we conclude that in order to test the adequacy of the nll approximation we are having to compare with data in a region where hadronization corrections are sizeable . unless one is extremely confident in one s ability to estimate these corrections it is hard to draw firm conclusions .
evidently if one has a nlo perturbative qcd calculation available there will be some uncertainty in the nlo perturbative prediction for a qcd observable due to the missing uncalculated higher - order terms in the perturbation series .
this has the effect that when one compares the nlo calculation with experimental data one may anticipate that the value of @xmath3 extracted will not be universal but will exhibit some scatter for different observables due to the differing sizes of the uncalculated contributions .
the problem is compounded by the fact that the @xmath3 extracted depends also on the choice of rs at nlo , which may be labelled by @xmath332 the nlo coefficient ( or equivalently the renormalisation scale @xmath14 ) , an unphysical parameter .
we have shown in equation ( 3.25 ) that the relation between the extracted @xmath330 and the actual @xmath3 which one is trying to measure is factorized into two contributions .
one , @xmath340 , is @xmath332-dependent and known exactly from equation ( 3.26 ) , and the other , @xmath270 , is unknown but @xmath332-independent .
@xmath344 depends only on the observable and nnlo and higher rs - invariants @xmath141 .
the predictable @xmath332-dependence of @xmath330 therefore has nothing to tell us about the importance of uncalculated corrections , the irreducible uncertainty resides in the unknown @xmath270 .
some of the @xmath332-dependent logarithms in the nnlo and higher corrections can be summed up into @xmath340 and the remainder can be absorbed into rs - invariant combinations contained in @xmath270 .
the unknown @xmath270 can be isolated by choosing the particular rs where @xmath339 , @xmath229 the effective charge scale , which sets @xmath342 . from the scatter in @xmath338 for different observables one
can then infer the scatter in @xmath270 , in particular relative differences @xmath485 for observables a , b can be absolutely measured .
figure 3 indicates that these relative differences can not be neglected for a range of lep observables . for at least some of the observables @xmath270 must be sizeable , and so @xmath270 must be estimated before we can determine @xmath3 with any reliability .
in contrast , in the standard lep determinations of @xmath1 one tries to estimate the importance of uncalculated higher - order corrections by using two different ad hoc scale choices and interpreting the spread in extracted @xmath1 as indicating a ` theoretical error ' . by artificially enlarging uncertainties in this way one obtains a spurious consistency between different observables , the real scale - independent uncertainty due to @xmath270 being buried beneath the supposedly informative scale dependence of @xmath486 .
the global @xmath1 determinations obtained with these sort of analyses should therefore be treated with some scepticism . having determined that @xmath270 is not negligible one must try to estimate it to make further progress . within the effective charge formalism
one can write @xmath3 in terms of the measured observable @xmath18 , @xmath487 and @xmath270 constructed non - perturbatively from the measured running @xmath488 .
@xmath270 may also be obtained perturbatively by expanding the effective charge @xmath23-function , so that given a nnlo calculation one can estimate @xmath489 . by combining nnlo calculations and @xmath19-dependence measurements one can refine ones knowledge of @xmath344 and hence of @xmath3 .
the perturbative rs - invariants @xmath490 are connected with @xmath257 evolution of the observable @xmath18 , and hence provide information about the function @xmath22 in the vicinity of @xmath295 .
the present situation is that a nnlo calculation is available only for the hadronic width of the @xmath4 . by estimating @xmath491 for this observable , and combining petra data for jet rates and energy - energy correlations , with lep data on these observables , to obtain an estimate of @xmath270 from @xmath19-dependence , we found consistency for @xmath8 mev .
we also showed that for observables where leading and next - to - leading logarithms in kinematical variables can be resummed one can unambiguously estimate @xmath467 by resumming leading and next - to - leading logarithms in the rs invariants @xmath472 . in the conventional approach
one needs to use an ad hoc matching procedure to include the exact nlo perturbative coefficient , and the problem of scale dependence still remains . performing this analysis for @xmath376 we found qualitative agreement between the @xmath27-dependence at low @xmath27 of @xmath467 and @xmath350 using
hadronization corrected data , but a completely opposite low @xmath27 dependence using uncorrected data .
we conclude that reliable measurements of @xmath3 at @xmath10 machines will require at least nnlo perturbative calculations and/or measurements of observables at more than one energy . acquiring
such information will entail considerable experimental and theoretical effort .
the effective charge formalism allows one to efficiently harness this hard won information to refine one s knowledge of @xmath3 and the interplay between perturbative and non - perturbative effects .
the insistance on formulating everything in terms of physical quantities allows one to quantify uncertainties in a way which is impossible in approaches which choose the unphysical rs - dependence parameters according to some plausiblity argument .
an important remaining problem is to extend the approach to processes with initial state hadrons where there is an additional factorization ambiguity connected with the separation of structure functions from the hard cross - sections .
we would like to thank siggi bethke , nigel glover , andrei kataev , and paul stevenson for stimulating discussions and useful comments which have helped us to clarify our ideas .
paolo nason is thanked for supplying a computer code for generating nlo corrections to the @xmath10 matrix elements .
m.t.r . gratefully acknowledges receipt of a s.e.r.c u.k .
studentship . 99 opal collaboration , p.d.acton et al .
, cern preprint ppe/93 - 38 ( 1993 ) .
opal collaboration , m.z.akrawy et al .
, z.phys . *
c55*,1 ( 1992 ) .
opal collaboration , m.z.akrawy et al .
, z.phys . *
c49*,375 ( 1991 ) .
delphi collaboration , p.abreu et al .
, z.phys . *
c54*,55 ( 1992 ) .
aleph collaboration , d.decamp et al .
, phys.lett . * b225*,623 ( 1991 ) .
l3 collaboration , p.adeva et al .
, phys.lett . * b248*,464 ( 1990 ) .
s.bethke , invited talk at the workshop on jet studies at lep , durham 1990 , j.phys .
* g17*,1455 ( 1991 ) .
s.bethke , proceedings of the xxvi international conference on high energy physics , dallas ( 1992 ) , edited by james r.sanford , p.81 .
lep collaborations joint report , cern - ppe/93 - 157 ( 1993 ) .
r.k.ellis , d.a.ross and a.e.terrano , nucl.phys . * b178*,421 ( 1981 ) .
z.kunszt and p.nason , in `` z physics at lep1 '' ( eds .
g.altarelli , r.kleiss and c.verzegnassi ) , cern 89 - 08 ( 1989 ) .
s.g.gorishny , a.l.kataev and s.a.larin , phys.lett . * b259*,144 ( 1991 ) ; l.r.surguladze and m.a.samuel , phys.rev.lett . * 66*,560 ( 1991 ) .
p.m.stevenson , phys.rev . * d23*,2916 ( 1981 ) .
g.grunberg , phys.lett * b95*,70 ( 1980 ) ; g.grunberg , phys.rev . * d29*,2315 ( 1984 ) .
a.dhar and v.gupta , phys . rev .
* d29*,2822 ( 1984 ) .
r.k.ellis , summary talk in proceedings of the workshop on `` new techniques for calculating higher order qcd corrections '' december 1992 .
eth zurich preprint eth - th/93 - 01 ( 1993 ) .
d.t.barclay and c.j.maxwell , durham preprint dtp-92/26 ( 1992)(unpublished ) .
c.j.maxwell , proceedings of the xxvi international conference on high energy physics , dallas ( 1992 ) , edited by james r.sanford , p.905 .
s.bethke , z.kunszt , d.soper and w.j.stirling , nucl.phys . * b370*,310 ( 1992 ) .
s.catani , y.l.dokshitzer , m.olsson , g.turnock and b.r.webber , phys.lett . *
b269 * , 432 ( 1991 ) .
s.catani , g.turnock and b.r.webber cern preprint cern - th.6640/92 ( 1992 ) .
r.coquereaux , ann.phys.(n.y.),*125*,401 ( 1980 ) .
stueckelberg and a.petermann , helv.phys.acta @xmath492,449 ( 1953 ) .
w.celmaster and r.j.gonsalves , phys.rev , * d20*,1420 ( 1979 ) .
w.celmaster and r.j.gonsalves , phys.rev , * d21*,3112 ( 1980 ) .
c.j.maxwell , phys.rev . * d29*,2884 ( 1984 ) ; p.m.stevenson , phys.rev . * d33*,3130 ( 1986 ) .
o.v.tarasov , a.a.vladimirov , and a.yu.zharkov , phys.lett . * b93*,429 ( 1980 ) .
s.a.larin and j.a.m.vermaseren , phys.lett .
* b303*,334 ( 1993 ) .
j.a.nicholls , ph.d .
thesis durham university ( 1990 ) ( unpublished ) .
c.j.maxwell and j.a.nicholls , phys.lett . * b213*,217 ( 1988 ) .
j.chla , a.kataev , and s.larin , phys.lett . * b267*,269 ( 1991 ) .
the ` physical scale ' argument is widespread in the literature .
for a recent strong defence of the choice @xmath210 see : stefano catani , proceedings of the @xmath493 workshop of the infn eloisatron project , qcd at 200 tev , erice , june 1991 .
preprint cern - th .
6281/91 ( 1991 ) .
j.chla , phys.rev . * d40*,676 ( 1989 ) .
g.grunberg , phys.rev . * d40*,680 ( 1989 ) .
m.beneke , max planck institute preprint , mpi - ph-93 - 6 ( 1993 ) .
g.t hooft in ` deeper pathways in high - energy physics ' ( orbis scientiae , coral gables , 1977 ) eds .
b.kursunoglu , a.perlmutter and w.scott ( plenum , new york , 1977 ) ; in ` the why s of subnuclear physics ' ( erice 1977 ) ed . a.zichici ( plenum , new york,1977 ) . t.appelquist and j.carazzone , phys.rev . *
d11 * , 2856 ( 1975 ) .
opal collaboration , p.d.acton et al.,phys.lett . * b276*,547 ( 1992 ) .
d.bardin et al . ,
nucl.phys . *
b351 * , 1 ( 1991 ) .
k.g.chetyrkin and j.h.khn , phys.lett .
* b248 * ( 1990 ) 359 .
a.x.el-khadra , g.m.hockney , a.s.kronfeld , p.b.mackenzie , phys.rev.lett . * 69*,729 ( 1992 ) .
jade collaboration , s.bethke et al .
, phys.lett . * b213*,235 ( 1988 ) .
jade collaboration , w.bartel et al .
, z.phys * c25*,231 ( 1984 ) .
m.jezabek and j.khn , phys.lett .
* 301*,122 ( 1993 ) .
a.c.mattingly and p.m.stevenson , phys.rev.lett . * 69*,1320 ( 1992 ) ; proceedings of the dpf92 ( fermilab , nov.1992 ) ; rice university preprint de - fg05 - 92er40717 - 7 ( 1993 ) .
we thank s.bethke for supplying us with hadronization - corrected data used in reference @xcite .
figure 1 . : : the generic approximately inverted - parabolic behaviour of the nlo approximant @xmath494 versus @xmath96 .
the figure is plotted for @xmath162 and @xmath163 .
the dashed line represents measured data ( see text ) .
figure 2(a ) . : : @xmath1 obtained by fitting nlo perturbative calculations with @xmath219 to opal data . figure 2(b ) .
: : as ( a ) but with enlarged ` theoretical errors ' using @xmath219 and fits to data , as described in the text
. reproduced from reference [ 8 ] .
: : @xmath350 of equation ( 3.27 ) for various lep observables ( see text for details ) .
a reference value of @xmath495 mev is chosen .
: : the @xmath483 dependence of @xmath350 for @xmath496 .
opal data of reference [ 2 ] is used uncorrected for hadronization . a reference value of @xmath495 mev is chosen .
: : as figure 4 but for @xmath376 .
: : as figure 4 but for @xmath385 .
: : as figure 4 but for @xmath377 .
: : learning about the integrand of @xmath270 , @xmath413 of equation ( 3.31 ) .
nnlo perturbative calculations can provide @xmath422 , and measurements at energies @xmath415 and @xmath416 can provide a point away from the origin .
: : absolute estimates of @xmath270 for lep observables at @xmath497 gev are compared with the observed @xmath498 with @xmath499 mev . for eec(@xmath500 ) and @xmath436 the absolute estimates are @xmath450 of equation ( 3.34 ) obtained from @xmath19-dependence using jade data . for @xmath5 @xmath398 from the nnlo calculation of reference
[ 12 ] is used .
figure 10(a ) .
: : for the observable @xmath376 the nll approximation @xmath467 of equation ( 3.46 ) versus @xmath27 is compared with @xmath350 with @xmath308 adjusted .
opal data of refervence [ 2 ] uncorrected for hadronization effects is used .
figure 10(b ) .
: : as ( a ) but with hadronization - corrected data .
the first two columns show the central values of @xmath308 ( in mev ) extracted by comparing nlo perturbative qcd predictions for the observables of table 1 , taking @xmath219 and @xmath229 , with opal data [ 2 ] .
the third and fourth columns give the nlo @xmath308 values extracted with @xmath219 and @xmath229 , assuming @xmath501 for each observable , and an actual value of @xmath502 mev . | we show that standard next - to - leading order ( nlo ) perturbative qcd analyses used to extract @xmath0 from lep data do not serve to disentangle the completely unknown renormalization scheme ( rs ) invariant next - nlo ( nnlo ) and higher - order uncalculated corrections from those dependent on the renormalization scale in a predictable manner .
the resulting quoted values of @xmath1 @xmath2 with attendant scheme dependence uncertainties therefore reflect the ad hoc way in which they were extracted , rather than the actual values of these parameters . choosing the scale so that the nlo coefficient vanishes ( effective charge scheme ) exposes the relative size of these unknown rs invariant higher order terms .
they are seen to be sizeable for some of the lep observables , indicating that they must be estimated if @xmath3 is to be determined reliably .
this can be accomplished either from nnlo calculations , at present only available for the hadronic width of the @xmath4 ( @xmath5 ) , or non - perturbatively by writing the rs - invariant uncertainty in @xmath6 in terms of the running @xmath7 of the observable r(q ) with energy using an effective charge formalism .
the nnlo calculations for @xmath5 and lep data supplemented by lower energy petra data lead to @xmath8 mev .
we also discuss how the effective charge approach can be used to remove scale dependence from next - to - leading logarithm resummations of some lep observables .
= 11 addtoresetequationsection # 1 ( [ # 1 ] )
@=11 versim#1#20.2ex @=12 * measuring @xmath9 at lep * d.t .
barclay _ department of physics , university of illinois _ _ at chicago ,
illinois 60680 , u.s.a _
c.j .
maxwell and m.t .
reader _
centre for particle theory , university of durham _ _
durham dh1 3le , england _ |
the qcd `` hard pomeron '' is understood as the solution of perturbative qcd expansion at high energy ( @xmath2 ) after resumming the leading @xmath3 terms .
it is known to obey the bfkl equation@xcite .
it has recently attracted a lot of interest in relation with the experimental results obtained at hera for deep - inelastic scattering reactions at very low value of @xmath4 where @xmath5 is the virtuality of the photon probe @xmath6 and @xmath2 is , in this case , the c.o.m energy of the @xmath6-proton system .
interestingly enough , the proton structure functions increase with @xmath2 at fixed @xmath7@xcite in a way qualitatively compatible with the prediction of the bfkl equation .
however , the phenomenological discussion is still under way , since scattering of a `` hard '' probe on a proton is not a fully perturbative qcd process and moreover , alternative explanations based on renormalization group evolution equations do exist@xcite . on the other hand ,
the phenomenological success@xcite of parametrizations based on the bfkl evolution in the framework of the qcd dipole model@xcite is quite encouraging for a further study of its properties . beyond these phenomenological motivations , there
exist interesting related theoretical problems which we want to address in the present paper . in its 2-dimensional version ,
the bfkl equation expresses@xcite the leading - order resummation result for the elastic ( off - mass - shell ) gluon - gluon scattering amplitude in the @xmath8 transverse plane @xmath9 which depends on the energy @xmath10 and on the 2-momenta of the incoming and outgoing gluons with @xmath11 is the transferred 2-momentum .
@xmath12 is the total rapidity space available for the gluon - gluon reaction . alternatively ,
one introduces@xcite the coordinate variables via 2-dimensional fourier transforms and the gluon green function @xmath13 where 2-momentum conservation in momentum space leads to global translationnal invariance in coordinate space
. the amplitude @xmath14 is solution of the bfkl equation expressed in the 2-dimensional transverse coordinate space and explicit solutions can be obtained@xcite using conformal invariance properties of the bfkl kernel .
we will heavily use these symmetry properties in the sequel . in our paper
, we will address the problem of finding the amplitudes @xmath15 solution of processes involving @xmath16 external gluon legs where @xmath17 @xmath18 ... ,@xmath19 are their arbitrary coordinates in the plane transverse to the initial gluon - gluon direction . note that @xmath20 @xmath21 is the original bfkl amplitude .
recently , it has been shown@xcite that @xmath22 is equal , up to kinematical factors , to the number density @xmath23 of dipoles existing in the wave - function of an initial quark - antiquark pair ( onium ) after an evolution `` time''@xmath12 ( such can be interpreted @xmath12 in the bfkl equation written as a diffusion process@xcite ) . in the qcd - dipole picture@xcite , gluons are equivalent to a @xmath24 pairs ( in the @xmath25 limit ) which recombine into a collection of independent and colorless dipoles .
the elastic amplitude is thus obtained from the elementary dipole - dipole amplitude weighted by the dipole number densities of each initial state obtained after evolution time @xmath26 using conformal invariance properties of the bfkl kernel , it turns out@xcite that : @xmath27 where @xmath28(resp .
@xmath29 are the components of the dipole density ( resp .
the gluon green s function ) expanded upon the conformally invariant basis , namely : @xmath30 and @xmath31 is the value of the bfkl kernel in the ( diagonal ) conformal basis
. the corresponding eigenvectors are explicitly known@xcite to be : @xmath32 with @xmath33 are the holomorphic ( resp .
antiholomorphic ) components in the 2-d transverse plane considered as @xmath34 and @xmath35 , @xmath36 ( @xmath37 , @xmath38 ) , are the quantum numbers defining the appropriate unitary representations@xcite of the conformal group @xmath39 .
indeed the bfkl solution ( i.e. , also , the qcd dipole solution ) is given by @xmath40 which can be explicitly calculated in terms of hypergeometric functions@xcite . in order to generalize these investigations to an arbitrary number of gluons
, we shall use the qcd dipole formalism allowing to express the probability of finding @xmath41 dipoles in an initial one , i.e. the p - uple dipole density after an evolution time
@xmath42 @xmath43 @xmath44 is the solution of an integral equation which has been proposed in the paper of ref.@xcite , and approximate solutions have been worked out and applied to problems like the double and triple - qcd pomeron coupling@xcite , the qcd dipole production@xcite , hard diffraction@xcite and , more generally , to the unitarization problem@xcite . in particular ,
monte - carlo simulations of the unitarization series based on a numerical resolution of the @xmath44 integral equations have been performed@xcite . however a general expression for the solution of these equations and a physical interpretation of its properties are yet lacking .
it is the purpose of our work to provide such a solution , which is intimately related , as we shall see , to dual string amplitudes emerging from the qcd dipole picture .
our main result is the explicit expression of the p - uple dipole density distributions in the transverse coordinate plane as dual virasoro - shapiro amplitudes@xcite ( for conformal spins all equal to @xmath45 these are the dominant contributions at high @xmath26 we also give the expressions for arbitrary conformal spins ( i.e. for all the conformal components ) in terms of a well - defined generalization of shapiro - virasoro amplitudes .
the paper is organized as follows : in section * 2 * , we derive the qcd dipole equation for @xmath46 ( for zero conformal spins ) .
the solution is found in a compact form in terms of integrals over explicit conformal eigenvectors . in section * 3 * we reformulate the obtained 3-dipole vertex in terms of the koba - nielsen projective - invariant parametrization of the shapiro - virasoro amplitude @xmath47 in the following section * 4 * we show that the solution can be iterated and obtain the triple dipole density distribution in terms of @xmath48 this iteration gives the p - uple dipole distribution in terms of @xmath1 integrands . in section * 5 * we discuss the extension for arbitrary conformal spin and the possibility of finding a target - space realization of the underlying closed string picture emerging from the qcd dipole interactions .
the integral equation satisfied by the dipole pair density @xmath50 see fig.1 , is written as follows@xcite @xmath51 where the integration domain @xmath52 avoids the singularity at the initial dipole end - points @xmath53 the physical meaning of this equation is transparent : the probability of finding a pair of dipoles at given transverse coordinates at time @xmath12 is made of two terms ; the last term in ( 7 ) is the survival probability after @xmath12 of two dipoles , whereas the former corresponds to the probability of creating two new dipoles @xmath54 and @xmath55 which then are surviving till @xmath12 , each with probability @xmath23 . in both cases ,
the survival probability is given by the @xmath56-dependent sudakov - like form factor @xmath57 in fact , it is possible@xcite to reexpress this equation in a way which is explicitely independent of @xmath58 when @xmath59 multiplying both terms of eqn.(7 ) by @xmath60 and differentiating with respect to @xmath12 , one gets : @xmath61 where @xmath62 denotes a parametrization@xcite of the bfkl kernel in coordinate space , namely @xmath63 with the property@xcite @xmath64 in equation ( 8) , the @xmath65 limit can be taken without harm , thanks to the non - singular behaviour of the kernel @xmath62 after regularization .
the kernel @xmath62 is conformally invariant@xcite and its eigenvalues and eigenfunctions are given by equations ( 3 ) and ( 4 ) , respectively . in order to solve eqn.(8 ) , we shall expand the function @xmath49 on the basis ( 4 ) by writing : @xmath66 where @xmath67 are conjugate variables and @xmath68 is an auxiliary transverse coordinate variable which labels the set of eigenfunctions @xmath69 the solution of eqn.(8 ) is obtained using the known@xcite orthogonality relations of the @xmath70 eigenfunctions @xmath71 @xmath72 with @xmath73 given in ref.(5 ) and @xmath74 inserting the decomposition ( 10 ) into eqn.(3 ) , we integrate both sides of the equation by @xmath75 one finds using relations ( 11 ) @xmath76 where the normalisation factor takes into account the overcompleteness of the eigenbasis , since @xmath77 and @xmath70 correspond to the same component @xmath78 . introducing finally the components @xmath79 see ( 2 ) , for the single density distributions with their corresponding eigenvalues @xmath80 and eigenvectors @xmath81 one writes @xmath82 @xmath83 @xmath84 @xmath85 where @xmath86 and @xmath87 are auxiliary coordinates which play for the dipoles @xmath54 and @xmath55 the same r^ ole played by @xmath88 for the dipole @xmath89 , see fig.1 .
the expression ( 15 ) gives the formal solution of @xmath49 in terms of the eigenfunctions of the bfkl kernel which are given in ( 4 ) .
let us comment one after the other the 3 different building blocks appearing in ( 15 ) i)the summation @xmath90 is related to the dipoles @xmath91 @xmath92 and will be used for inserting @xmath49 in cross - section calculations .
it will also restrict the final integration over the quantum numbers @xmath93 through the poles appearing at @xmath94 in the denominators of ( 15 ) .
ii)the factors @xmath95 are the only ones depending on the coordinates @xmath96 and @xmath97 .
interestingly enough , the subsequent interaction terms involving these dipoles , for instance when computing multi - pomeron interactions in the qcd dipole picture@xcite , will be greatly simplified by using factorization and the orthogonality relations ( 11 ) .
these features are already at the root of the derivation@xcite of the equivalence ( see formula ( 1 ) ) between the elastic bfkl and qcd - dipole amplitudes .
iii)the last factor in the solution ( 15 ) , is the triple - dipole vertex @xmath98 in the qcd dipole picture .
it depends on the conformal quantum numbers and auxiliary coordinates of the 3 dipoles which are coupled together in @xmath99 it is the purpose of the next sections to give an interpretation of these vertices in terms of dual amplitudes .
let us first consider the vertex @xmath100 for conformal spins @xmath101 this vertex is physically relevant since it corresponds to the dominant contribution at high @xmath26 in fact , the method used for its evaluation will be valid for any conformal spin . inserting the definitions ( 4 ) in the expression of @xmath102 obtained from the solution ( 15 ) , one gets : @xmath103 our observation is that @xmath104 can be expressed as follows : @xmath105 where @xmath106 is a shapiro virasoro@xcite amplitude with @xmath107 external legs and @xmath108 is a known conformally - invariant tensor of the coordinates , i.e. it is completely fixed@xcite by the symmetry . in order to derive the expression ( 17 ) ,
let us recall the koba - nielsen formulation@xcite of @xmath109 , namely .
@xmath110 where the integration measure is the conformally - invariant one and the powers @xmath111 correspond to scalar products of external momenta on the target space in the closed string realizations of the shapiro - virasoro amplitudes .
note that the string tension can be chosen to be equal to 1 by scale invariance of the solution . in our case ,
the powers @xmath111 are not specified in terms of external momenta but stringent constraints are required on @xmath111 in order to satisfy the requirements of duality and conformal symmetry@xcite namely : @xmath112 it is tedious but straightforward to verify that the constraints ( 19 ) are verified for the set of @xmath111 corresponding to formula ( 16 ) given in table i. from that , one easily obtains formula ( 17 ) with @xmath113 formula ( 20 ) does not come as a surprise , since it is the universal form@xcite of a 3-point correlation function for field theories obeying global conformal invariance .
indeed it corresponds to a correlation function @xmath114 where @xmath115 @xmath116 @xmath117 are so - called `` quasi - primary '' fields@xcite of conformal dimensions @xmath118 respectively .
it is interesting to note that such fields already appear in the field - theoretical interpretation of the bfkl equation and its relation to conformal invariance properties@xcite , where correlation functions @xmath119 are introduced , with @xmath120 being a scalar ( i.e. of conformal dimension 0 ) field representing external ( reggeized ) gluons . in this sense
our result ( 17 ) can be considered as a qcd dipole model realization of the triple hard - pomeron vertex .
it is a stimulating ( but non - trivial ) challenge to confront the solution we obtain ( in particular eqns.(15 - 20 ) ) with the known results@xcite on the @xmath121 gluon vertex in the bfkl approach .
this would allow to go deeper in the equivalence of bfkl to the qcd dipole approach .
the result obtained in eqn.(20 ) for @xmath122 is a consequence of the global conformal invariance of the bfkl equation .
indeed , the @xmath106 function is conformally invariant by construction ( provided the constraints ( 19 ) are satisfied ) and the overall vertex @xmath104 of formula ( 17 ) should reflect the general structure@xcite @xmath123 where @xmath124 @xmath125 has to be coordinate - independent while its dependence on the conformal dimensions is not specified by the global @xmath126 symmetry .
this result could be expected for any conformally invariant kernel . by contrast , the specific bfkl kernel @xmath127 of formula ( 9 ) allows the determination of @xmath128 and its relation to the shapiro - virasoro amplitude @xmath106 .
let us generalize the vertex calculation to the @xmath129-dipole amplitude by considering first the exemple of @xmath130 see fig.2 . in much the same way as for @xmath131 one writes : @xmath132 @xmath133 where we have used the fact that the probability of finding three dipoles at @xmath12 can be expressed by two equivalent iterations of the bfkl kernel , namely : @xmath134 hence , the intermediate step ( the @xmath135 segment or , equivalently , the @xmath136 segment in the case of fig.2 ) is not relevant in the final integration kernel . from the symmetry of ( 24 ) , it is easy to realize that this iteration procedure will give a symmetric kernel for any number @xmath41 of produced dipoles .
the integral over the four eigenfunctions appearing in the last term of expression ( 23 ) gives the @xmath137 dipole vertex and can be cast in the following form ( for zero conformal weights ) : @xmath138}\times \ c_{8}\left ( \frac{\rho _ { \alpha \beta _ { } } \rho _ { \gamma \delta _ { } } } { \rho _ { \alpha \gamma } \rho _ { \beta \gamma } } , \frac{\rho _ { \alpha \beta } \rho _ { \gamma \delta } } { \rho _ { \alpha \delta } \rho _ { \beta \gamma } } \right ) , \tag{25}\ ] ] where the first factor is determined@xcite by conformal invariance constraints on 4-point correlation functions for quasi - primary fields of conformal dimension @xmath139
it is a matter of tedious but straightforward verification that the vertex term in equation ( 23 ) can be cast into the generic form ( 25 ) with : @xmath140 indeed the functional dependence ( 25 ) is obtained from the constrained relations @xmath141 see ( 19 ) .
the corresponding values of @xmath111 are displayed in table ii .
as well - known@xcite global conformal invariance of 4-point correlation functions does not determine the residual function @xmath142 as an arbitrary function of the @xmath126-invariant harmonic ratios of the 8 coordinates involved in the problem , see fig.2 .
however the explicit integrand ( 24 ) induced by the bfkl kernels leads to a well - defined connection with the ( koba - nielsen form of ) the shapiro - virasoro @xmath143 function .
indeed , one may write : @xmath144
^{-1}\prod_{i < j}\rho _ { ij}^{-2p_{ij } } , \tag{26}\ ] ] where the factor @xmath145 ^{-1}$ ] conventionnally means that 4 integration variables out of @xmath146 have to be omitted . formula ( 26 ) can be identified by comparison with the koba - nielsen formulation of @xmath147 namely : @xmath148
^{-1}\prod\limits_{i < j}\rho _ { ij}^{-2p_{ij}}\ \equiv \int d\rho
_ { \delta } \left| \frac{\rho _ { \beta \delta } } { \rho _ { \gamma \delta } \rho _ { \delta \alpha } } \right| ^{2 } \ c_{8 } , \tag{27}\ ] ] where the three coordinates in the @xmath126 volume factor have been choosen in order to match with the variables of @xmath149 it is clear from formulae ( 26,27 ) , that @xmath150 is the @xmath151-invariant integrand of the shapiro
vivasoro amplitude @xmath152 by simple iteration in the number of produced dipoles , it is not difficult to find the generalization of formulae ( 23)-(27 ) to the @xmath153 dipole distribution .
the key observation to determine the powers @xmath111 appearing in ( 26 ) which are known combinations of conformal dimensions @xmath154 obeying the constraints ( 19 ) is to notice that the coordinates in transverse plane are nearest neighbours along the polygonal perimeter @xmath155 ( see fig.2 for the case @xmath156 ) .
the auxiliary points @xmath157 are also connected via nearest neighbours to the previously mentioned perimeter , except for their conformal - dimension factors @xmath158 which are fixed by the constraints .
indeed , we note that the 4 legs connected to a given coordinate @xmath159 in the polygonal perimeter are such that the corresponding sum @xmath160 is always zero . the powers attached to the auxiliary points
are fully determined by @xmath126 invariance ( for the set of constraints ( 19 ) ) .
the shapiro - virasoro amplitudes which are obtained for @xmath153 dipole distributions lead naturally to the question of a closed - string interpretation of the high - energy limit of perturbative qcd .
indeed , these amplitudes appear in the context of a closed string moving in a minkowskian @xmath161 target space@xcite .
more generally , such amplitudes appear as a consequence of vertex operator constructions in conformal field theories@xcite and are related to the existence of an ( anomalous ) infinite - dimensional algebra associated with local conformal invariance , namely the virasoro algebra .
moreover in the case of a critical target - space dimension @xmath162 the fock space on which the quantum string theory is realized is spanned by positive normed states ( no ghosts ) with full reparametrization invariance .
this connection has both a practical and conceptual interest for qcd calculations .
first , the many and much explored mathematical properties of dual amplitudes may lead to a simplification of qcd dipole computations for given processes , e.g. `` hard '' diffraction , multi - pomeron contributions , etc .. second , there is a possibility of building an effective theory of qcd in the high - enrgy limit , which could be based on a string theory ( instead of a field theory ) .
this would allow the computation of string loop contributions and thus induce an effective theory of interacting qcd pomerons .
however , the variables which appear as conformal exponents @xmath111 of the integrands are not directly expressed as scalar products of momenta in a minkowskian @xmath161 target space .
they are complex numbers , see , e.g. ( 26 ) obeying constraints which are not directly expressed as on - mass shell and momentum conservation constraints as for the closed string@xcite .
even if such a target - space interpretation is possible , an analytic continuation in the imaginary direction ( implied by the quantum numbers of the conformal eigenvectors ( 4 ) ) is to be performed .
it is thus useful to pass in review the properties of shapiro - virasoro amplitudes in this context and to see which are those to be completed for a full closed string theory to be valid . a first consequence of the solutions ( 17 ) , ( 25 ) , and their iterations , is that duality properties exist in the dipole formulation of qcd vertices .
indeed , by construction , shapiro - virasoro amplitudes are meromorphic but forbid the existence of multiple pole singularities coming from dual channels .
for instance , in fig.2 , the @xmath163 and @xmath164 channels can not both together bring singularities to the amplitude . as usual in dual theories
, there exists intricate relations between different ways of describing the amplitudes depending on the series of pole contrbutions which are choosen for their expansion .
an interesting example of such a duality property has been provided by the equivalence of the `` t - channel '' bfkl elastic 4-gluon amplitude with the `` s - channel '' qcd dipole description of the same amplitude@xcite .
further application of this fruitful concept are expected from our results . as a practical consequence of our identification of the multiple - dipole vertices with integrands of standard shapiro - virasoro amplitudes in the case of zero conformal spins
, one may use some tools which are developed in the string theoretical formalism to generalize our investigations to the case of general ( integer or half - integer ) conformal spins as follows ; one can consider in general amplitudes of the form : @xmath165 ^{-1}\prod\limits_{i < j}^n \ \rho _ { ij}^{-p_{ij } + \frac { n_{ij}}2 } \bar { \rho } _ { ij}^{-p_{ij } + \frac { { \tilde n}_{ij}}2 } , \tag{28}\ ] ] where @xmath166 are integers .
interestingly enough , in the framework of string theory , this corresponds to consider external excited states of the bosonic string@xcite .
moreover , the same techniques allow to connect closed string to open string tree amplitudes which may allow to extend to the multiple - vertex calculations the conformal - block structure initially identified in the bfkl 4-point amplitudes@xcite . in the seminal paper of ref.@xcite
, it has been noticed that it was not straightforward to extend the ( already beautiful ) global conformal symmetry @xmath39 to the infinitely dimensional conformal group in 2 dimensions . in other words , only the 6 generators @xmath167 @xmath168 @xmath169 ( 3 holomorphic and 3 non - holomorphic ) of the virasoro algebra were expected to generate the symmetry algebra of the bfkl kernel .
the results we obtain indicate that the algebra can probably be extended to the infinite series of locally conformal generators , i.e. the whole virasoro algebra , at least in the qcd dipole representation . as usual , the symmetry is expected to be anomalous due to the possibility of a central charge@xcite ( conformal anomaly ) at the quantum level of consistency .
this issue will depend on the interpretation of a suitable @xmath170independent target - space representation of the exponents @xmath171 see , for @xmath172 , table i and ii ) .
for instance , if an embedding @xmath173 in a minkowskian @xmath161 space is allowed , this will determine the central charge to be related to @xmath174 and the critical dimension to be @xmath175 by compensation of the ghost contribution due to reparametrization symmetry@xcite .
this interesting issue deserves certainly more study .
we want to thank andrzej bialas and henri navelet for a fruitful collaboration on the qcd theory of dipoles which initiated the present work helping the author to find the way out in two noticeable occasions .
christophe royon and samuel wallon are acknowledged for stimulating discussions . 99 l.n.lipatov , _ sov . j. nucl .
* 23 * ( 1976 ) 642 ; v.s.fadin , e.a.kuraev and l.n.lipatov , _ phys .
_ * b60 * ( 1975 ) 50 ; e.a.kuraev , l.n.lipatov and v.s.fadin , _ sov.phys.jetp _ * 44 * ( 1976 ) 45 , * 45 * ( 1977 ) 199 ; i.i.balitsky and l.n.lipatov , _ sov.j.nucl.phys .
_ * 28 * ( 1978 ) 822 a recent general review of the experimental and phenomenological situation with opening views on the future can be found in : _ future physics at hera _ ,
proceedings of the workshop 1995/96 ( g.ingelman , a.de roeck , r.klanner eds .
, http://www.desy.de/ heraws96 ) .
j.bartels , _ zeit .
_ * c60 * ( 1993 ) 471 ; _ phys .
_ * b298 * ( 1993 ) 204 ; j.bartels , m.w " usthoff _ zeit . fur . phys .
_ * c66 * ( 1995 ) 157 ; j.bartels , l.n.lipatov , m.w " usthoff _ nucl .
phys . _ * b464 * ( 1996 ) 298 .
a.m. polyakov , _ zh . eksp .
_ lett.*12 * ( 1970 ) 538 , * 66 * ( 1974 ) 23 ; a.a .
migdal , _ phys .
_ * b44 * ( 1972 ) 112 ; a.a .
belavin , a.m. polyakov and a.b .
zamolodchikov , _ nucl .
_ * b241 * ( 1984 ) 333 ; | using the qcd dipole picture of bfkl dynamics and the conformal invariance properties of the bfkl kernel in transverse coordinate space , we show that the @xmath0 dipole densities can be expressed in terms of dual shapiro - virasoro amplitudes @xmath1 and their generalization including non - zero conformal spins .
we discuss the possibility of an effective closed string theory of interacting qcd dipoles . service de physique thorique cea - saclay f-91191 gif - sur - yvette france |
the study of the emission characteristics from rotation powered pulsars has been a subject of long standing interest .
our knowledge of the fundamental properties ( e.g. , mass , spin , and magnetic field ) of the underlying neutron star stems from detailed spectral and timing investigations . to facilitate an understanding of the mechanism by which the loss of rotational energy is converted into high energy radiation ,
many observational and theoretical studies have sought to determine the relationship between the x - ray luminosity , @xmath0 , and the rate of rotational energy loss or spin down power , @xmath1 .
indeed , a correlation of the form @xmath9 was found in einstein data by seward & wang ( 1988 ) .
subsequent studies using a larger sample of pulsars led to a relation of the form , @xmath10 based on rosat data ( see becker & trmper 1997 ) and @xmath11 based on asca data ( see saito 1998 ) .
recently , a reanalysis of 39 pulsars based on data obtained from several x - ray satellites by possenti et al .
( 2002 ) led to an intermediate relation @xmath12 , similar to that of seward & wang ( 1988 ) .
while the deduced existence of a correlation between @xmath0 and @xmath1 suggests that the observed x - rays are produced by a process which taps the rotational energy of the neutron star , a detailed description of the mechanism remains elusive .
this is , in part , a result of the fact that the data from different satellites are obtained in different energy ranges .
the results can be affected , for example , by interstellar absorption especially for those pulsars studied in the soft x - ray regime of rosat ( 0.1- 2.4 kev ) .
in addition , the total x - ray luminosity is composed of contributions from both the pulsed and non - pulsed components , and these components are likely to reflect physical conditions in diverse spatial environments .
the x - rays radiated by rotation powered pulsars could include contributions from 5 distinct components with different origins : + 1 . non - thermal non - pulsed diffuse radiation can be emitted from a pulsar wind nebula .
the radiation will be discussed in 2 , and we suggest that this emission component is the major contribution to the non - pulsed x - ray radiation in the asca observations of pulsars .
non - thermal non - pulsed radiation from the pulsar magnetosphere could be important as suggested by becker et al .
( 2004 ) if the angular resolution of x - ray detector is sufficiently high to separate this contribution from that of the nebula .
although such a component may be present in the asca data that we use , due to its low angular resolution , it is not likely to dominate the contribution from the pulsar wind nebula .
for example , tennant et al . (
2001 ) have found a non - thermal non - pulsed component from the crab pulsar by using the high - resolution camera of the chandra x - ray observatory .
they estimate that the x - ray luminosity of this component is @xmath13 , whereas , the non - thermal non - pulsed x - ray luminosity from the crab nebula detected by asca ( saito et al .
1997a ) is @xmath14 .
the specific origin of the non - thermal non - pulsed component , not associated with the nebula , is not clear .
although it contaminates the emission of the nebula , it is small .
provided that this contribution is small for all the pulsars we study in this investigation , this component should not affect our analysis .
the non - thermal pulsed radiation component is generally believed to be produced from the pulsar magnetosphere .
this follows from the fact that the motion of charged particles within the light cylinder is strongly affected by the magnetic field , and the radiation of charged particles in the open field lines should be pulsed .
this component could be produced in the vicinity of the polar cap as a result of inverse compton scattering of higher order generation pairs of particles on soft photons emitted by the neutron star ( zhang & harding 2000 ) or in the outer magnetosphere as a result of synchrotron radiation of downward cascades from the outer gap electron / positron particles ( cheng & zhang 1999 ) .
the thermal non - pulsed radiation from the pulsar surface can contribute to the soft x - ray bands , typically , characterized by @xmath15 kev ( cheng & zhang 1999 ) . in this paper , we have used the asca data on pulsars so this component will not contribute to the 2 - 10 kev asca energy band .
the thermal pulsed radiation from the pulsar surface likely originates in a polar cap on the neutron star surface and can contribute to the hard x - ray band ( @xmath16 kev ) . in analyzing the pulsed emission from pulsars , we follow the model of cheng & zhang ( 1999 ) where the thermal pulsed x - ray emission is included in their model prediction . thus , the x - ray emission observed from pulsars in the asca sample primarily consists of contributions from components 1 , 3 , and 5 .
we have assumed that component 2 is not important and component 4 does not significantly contribute to the 2 - 10 kev energy band .
on the other hand , becker & trmper(1997 ) have used rosat data and obtained a linear relation between the x - ray luminosity consisting of components 2 and 3 and spin - down power . in the present paper ,
we concentrate on the x - ray properties in the energy band of asca ( 2 - 10 kev ) .
the x - rays , then , consists of a thermal component from the polar cap , a non - thermal pulsed component from the magnetosphere and a non - thermal , non - pulsed emission from the nebula .
that is , we have explicitly assumed that the pulsar magnetosphere and the pulsar wind nebula are primarily responsible for the pulsed and non - thermal non - pulsed component respectively .
since the pulsed and non - pulsed emission have different origins , the relationship between the total x - ray luminosity and spin down power is not expected , in general , to be represented by a single power law .
in fact , the non - thermal pulsed and non - pulsed emission components are distinguished by different spectral signatures and conversion efficiencies . to determine their relative importance and their relationship to the spin down power
, we have reexamined the pulsar data obtained in the 2 - 10 kev energy band of the asca satellite , focusing on separating the non - pulsed and pulsed emission data . in the next section ,
we provide a simplified description of the emission from a pulsar wind nebula , based upon earlier work by chevalier ( 2000 ) , where the relativistic pulsar wind interacts with the interstellar medium . in 3 , the x - ray properties of the pulsed and non - pulsed components are collected and the relation between their luminosities with spin down power and photon spectral index are determined . with these results in hand ,
the pulsar wind nebula model is used to explain the occurrence of the tail - like emission morphology in geminga and psr b 1757 - 24 in 4 .
finally , we summarize and conclude in the last section .
a theoretical description of the interaction between a pulsar and its nebula , as applied to the crab nebula , was first outlined in a seminal paper by rees & gunn ( 1974 ) . in this model ,
the central pulsar generates a highly relativistic particle dominated wind which passes through the medium in the supernova remnant , forming a shock front .
the electrons and positrons in the shock are envisioned to be accelerated to a power law energy distribution and to radiate synchrotron photons in the downstream region . since the magnetic field in the nebula derives from the pulsar , a magnetohydrodynamical ( mhd ) model , rather than a hydrodynamical model , of the pulsar wind nebula
was required .
such a model , in the steady state and spherically symmetric approximation , was developed by kennel & coroniti ( 1984 ) for the crab nebula .
a hot relativistic positronic plasma flow was assumed to be terminated by a strong mhd shock , decelerating the flow and producing a non - thermal distribution of electrons .
the magnetization parameter , @xmath17 , introduced to describe the efficiency for conversion of energy contained in the pulsar wind into synchrotron radiation , was represented by the ratio of the electromagnetic energy flux to the particle kinetic energy flux defined as : = b^24n_w mc^2 . here
@xmath18 is the particle number density , @xmath19 is the magnetic field , @xmath20 is the lorentz factor of relativistic particles in the wind , @xmath21 is the particle mass , and @xmath22 is the speed of light . as pointed out by kennel & coroniti ( 1984 ) , the magnetization parameter must be small ( for the crab nebula , @xmath23 ) in order that sufficient compression occurs in the shock for transformation of the bulk flow energy to random particle motion and to the subsequent production of the observed synchrotron radiation .
since the magnetic energy density must dominate the particle energy density just inside the light cylinder , the wind nebula must have evolved from a high @xmath17 to a low @xmath17 state to produce the observed properties .
the manner in which this transformation takes place is central for detailed pulsar wind models and is not understood in detail .
nevertheless , the low @xmath17 model can account for the basic high energy properties of the crab , including the position of the optical wisps , the spectrum of the nebula , and the size of the nebula as a function of wavelength ( van den bergh & pritchet 1989 ) .
in the present paper , we concentrate on the x - ray properties of the pulsar wind nebula , using a simple one zone model similar to that developed by chevalier ( 2000 ) for a description of the wind interaction with the surrounding medium in a variety of environments . although the model does not provide a description of the spatial details associated with equatorial tori and polar jets observed in _
chandra _ studies of young pulsar wind nebulae , it can explain the global x - ray properties of the nebula to a first approximation .
in general , the energy in the shock waves is stored in the magnetic field as well as in the proton ( ion ) and electron particle components .
we assume that the fractional energy density of the magnetic field , @xmath24 , is @xmath25 ( kennel & coroniti 1984 ) .
provided that the spin down power is eventually carried away by pulsar wind particles and assuming equipartition of energy between the electrons and protons , the fractional energy densities in the proton and electron components should satisfy @xmath26 . for a given @xmath24 in the shock , the magnetic field at the shock termination radius ,
@xmath27 is estimated as @xmath28 , where @xmath29 is the spin down power of the pulsar . here ,
@xmath30 and @xmath31 are the spin period and its derivative , and @xmath32 is the moment of inertia of the neutron star ( @xmath33 ) . at the shock front , the electrons attain a power law distribution corresponding to @xmath34 for @xmath35 , where @xmath36 is the lorentz factor and @xmath37 . an estimate for @xmath38
can be obtained by equating the synchrotron cooling timescale to the electron acceleration timescale .
the former timescale is given by @xmath39 , and the latter timescale is given by @xmath40 , leading to @xmath41 , where @xmath42 is the electron mass and @xmath43 is the thompson cross section .
the synchrotron power of an electron with lorentz factor @xmath36 is ( blumenthal & gould 1970 ) p()=43_t c^2b^2 8 .
following chevalier ( 2000 ) , the total rate of particles ejected by pulsars at the lorentz factor @xmath36 is ( chevalier 2000 ) ( ) = ( p-1)_m^p-1(_wmc^2)^-1 ^-p . the resulting total number of radiating particles becomes @xmath44 , where @xmath45 is a characteristic timescale . for a young pulsar with a supernova remnant similar to the crab , @xmath45 is identified with the age of the nebula while for old pulsars or millisecond pulsars , @xmath45 is comparable to the flow timescale within the nebula .
thus , at the emitting frequency of an electron @xmath46 , the luminosity of the particle is about @xmath47 which should be less than @xmath1 .
the detailed properties of the x - ray nebula are a function of @xmath48 , @xmath24 , @xmath27 , @xmath1 , and the two critical frequencies @xmath49 and @xmath50 where @xmath49 is the frequency radiated by the electrons with the lorentz factor of @xmath51 , and
_ c=18em_ec/ _ t^2 t^2 b^3 is the electron synchrotron cooling frequency ( see chevalier 2000 and references therein ) .
the x - ray luminosity and spectral index depend on the inequality between @xmath52 and @xmath50 . for typical values of @xmath53 and @xmath54 g
, @xmath55 hz is always less than @xmath52 . in this case
, the x - ray luminosity from the total emission of particles per unit frequency can be calculated according to the synchrotron spectral profiles which have been presented in figure 1 of the paper by sari , piran & narayan ( 1998 ) . for @xmath56
, the luminosity per unit frequency can be obtained following equations ( 7 ) and ( 8) of sari , piran & narayan ( 1998 ) l_(_m/ _ c)^-1/2 ( /_m)^-p/2l_,max , where @xmath57 , and @xmath58 hz .
this expression corresponds to the fast cooling regime , and is equivalent to equation ( 9 ) in chevalier ( 2000 ) reflecting our use of the same form of particle number injection rate by a pulsar ( see equation 3 in this section ) . on the other hand ,
if @xmath59 ( denoted as slow cooling ) , the synchrotron x - ray luminosity is estimated as ( sari , piran & narayan 1998 ) l_(/_m)^ -(p-1)/2l_,max .
it can be seen that the x - ray luminosity in the fast cooling regime is larger than that in the slow cooling regime by a factor @xmath60 if @xmath52 is in the slow cooling regime . since the range of @xmath5 due to shock acceleration lies between 2 and 3 ( see achterberg et al .
2001 ; lemoine & pelletier 2003 and references therein ) , the photon index in the slow cooling regime is expected to lie in the range from 1.5 to 2 whereas the photon index in the fast cooling regime is expected to be in the range of 2 to 2.5 . within the theoretical framework ,
the energy of the electrons is converted to x - ray radiation solely at the shock termination radius . for pulsar motion which is subsonic ,
the determination of this radius is obtained via the balance between the wind ram pressure and total magnetic and particle pressure within the nebula ( rees & gunn 1974 ) . in this case , the wind bubble will be nearly centered about the position of the pulsar .
rees & gunn ( 1974 ) estimated @xmath61 cm , which is consistent with the size of the inner x - ray ring of the crab nebula ( @xmath62 pc , see kennel & coroniti 1984 ; weisskopf et al .
the termination shock picture is not necessarily restricted to slowly moving pulsars since even for pulsars moving at several hundred km s@xmath63 the motion can be subsonic in , for example , regions where the gas has been heated to temperatures of @xmath64 k by a supernova shock .
the radius of the termination shock can be estimated as : r_s ( b^2c)^1/2~610 ^ 14_34 ^ 1/2b_mg^-1 cm , where @xmath65 is the pulsar spin - down power in units of @xmath66 ergs s@xmath63 and @xmath67 is the magnetic field strength in the nebula in milligauss . with the observed values of the crab pulsar and its nebula ( @xmath68 and @xmath69 ) , consistency of the shock radius
is easily achieved . for supersonic motion ,
the nebula will form a bow shock morphology . in this case ,
the termination shock radius is given by the balance of the ram pressure between the wind particles and the medium at the head of the bow shock : = 12_ism v_p^2 , where @xmath70 is the density of the interstellar medium and @xmath71 is the pulsar s proper motion velocity .
the termination shock radius is given as : r_s ( 2_ism v_p^2c)^1/2~310 ^ 16_34 ^ 1/2n_1 ^ -1/2v_p,100 ^ -1 cm , where @xmath72 , and @xmath73 are the number density in the interstellar medium in units of 1 particle @xmath74 and the pulsar space velocity in units of 100 km s@xmath63 . as an example , caraveo et al .
( 2003 ) recently discovered a bow shock structure coincident with the geminga pulsar based on observations obtained with _ xmm - newton_. an estimate of the termination radius for geminga follows from its spin down power , @xmath75 , distance of @xmath76 pc , and proper motion velocity of @xmath77 km s@xmath63 ( bignami & caraveo 1993 ) , leading to a radius of @xmath78 cm which is consistent with the observational constraint on geminga s compact x - ray nebula . the general features of the emission region surrounding rotation powered pulsars in the galaxy will be bracketed by these two wind nebula morphologies . since the range in magnetic field strengths in the shock can vary from @xmath79 g to as high as @xmath80 g in the vela nebula and @xmath81 g in the crab nebula for the very young pulsars within supernova remnants , a range in the conversion efficiency of the spin down energy into nebula luminosity
is expected .
the relation between the nebula x - ray luminosity and the pulsar s spin down power will be deferred until 3 .
in the above we have discussed the pulsar wind nebula models in different contexts within a one - zone approximation .
the luminosities and spectral properties of the nebulae can be determined by eqs .
( 5 ) and ( 6 ) in the fast and slow cooling regime respectively . in the fast cooling case ,
the efficiency of converting spin down power to nebula luminosity is higher than that in the slow cooling case . assuming that @xmath82 , @xmath83 and @xmath84 ( see bednarz & ostrowski 1998 ) , the ratio of the x - ray luminosity , @xmath85 , to the spin down power at @xmath58 hz can be estimated as @xmath86 for the fast cooling case and @xmath87 for the slow cooling case if @xmath88 s , and @xmath89 g. the spectral properties in these two regimes differ since the photon index in x - rays is @xmath90 , and @xmath91 for fast and slow cooling respectively .
since @xmath5 varies from 2 to 3 , @xmath92 ( slow cooling ) and @xmath93 ( fast cooling )
. these are typical of the photon indices of observed x - ray nebulae ( see table 1 of chevalier 2000 ) .
as described in 2.1 , the fast and slow cooling regimes are distinguished by the x - ray and cooling frequencies . since the cooling frequency is dependent on the magnetic field and the flow time scale of the nebula , it is possible that different regions of the nebula may correspond to different cooling regimes .
for example , the inner region of the nebula may correspond to the slow cooling regime , whereas the fast cooling may be more appropriate in the outer region of the nebula . in this case , the power law index of the x - ray spectrum will steepen by @xmath94 between the inner and outer region .
this may be consistent with the spectral results from the high spatial resolution observations of the nebulae surrounding psr b1823 - 13 obtained with the _ xmm - newton _ satellite ( gaensler et al .
2003a ) . because the spatial resolution of current x - ray detectors is limited the emission from compact nebulae of size @xmath95 may be contaminated by the non - thermal x - ray emission from the pulsar magnetosphere .
however , the pulsar magnetospheric radiation is assumed to be pulsed , whereas the nebula contribution is non - pulsed .
therefore , the separation of the non - thermal x - ray emission into its pulsed and non - pulsed components is essential for determining their relative contributions .
since the pulsed and non - pulsed emission components originate from different regions , their x - ray luminosities are not expected to exhibit a similar relationship to the spin down power .
for example , wang & zhao ( 2004 ) recently found that the correlation of the observed pulsed x - ray luminosity with spin down power is not well matched to the linear form found by becker & trmper ( 1997 ) for the total luminosity in the rosat energy band ( 0.1 - 2.4 kev , see figure 8 of wang & zhao 2004 ) .
we note , however , that studies in this lower energy band are incomplete since pulsars can emit a significant fraction of their luminosity at higher energies . to examine the x - ray luminosity- spin down power relation at higher energies ( 2 - 10 kev ) , kawai et al .
( 1998 ) selected a sample of bright nebulae and found a relation in the form @xmath96 .
since the individual components were not separately analyzed , the result was insufficient to establish the relation between the pulsed or non - pulsed emission with spin down power . over 50 rotation powered pulsars
have been detected in the x - ray band ( becker & aschenbach 2002 ) , but only a fraction of them have been resolved with pulsed non - thermal components . here , we have selected 23 x - ray pulsars with both pulsed and non - pulsed emission measurements obtained from the asca mission ( listed in table 1 ) .
the pulsed and non - pulsed x - ray luminosities of these selected pulsars are directly taken from the cited references .
this pulsar sample includes 19 normal pulsars and 4 millisecond pulsars whose observational data are taken from the recent references ( see table 1 ) . since the asca satellite does not have high spatial resolution , the x - ray luminosity of the pulsars in the asca field is composed of emission from the pulsar s magnetosphere and compact pulsar wind nebula . for reference ,
the total pulsed plus non - pulsed x - ray luminosity in the asca energy range ( 2 - 10 kev ) is plotted versus spin down power in figure 1 .
a correlation is found which is consistent with the form @xmath11 found by saito ( 1998 ) , but the best fit form of this correlation is found to be @xmath98 . here , the error in the power law exponent represents @xmath99 corresponding to the scatter in the observed data points , which may reflect variations in @xmath48 , @xmath24 , @xmath20 , and uncertainties in distance from pulsar to pulsar .
our best fit power law relation is consistent with the conclusion of possenti et al .
( 2002 ) , who used a sample of 39 pulsars observed mainly by rosat and data from asca , rxte , bepposax , chandra and xmm - newton .
although possenti et al .
( 2002 ) considered their result ( @xmath100 ) statistically unacceptable , our result is not subject to the uncertainties associated with the normalization of different satellite data to obtain the x - ray luminosity between 2 - 10 kev ( cf .
table 1 of possenti et al .
2002 ) , for which the extrapolation relied on the uncertain photon index in the rosat energy band .
the x - ray luminosity associated with the pulsed emission component is illustrated versus spin down power in figure 2 .
a correlation separate from the total luminosity is found which is inconsistent with either the form @xmath101 or @xmath102 . the best fitting function to the data
is found to be @xmath103 , which significantly deviates from the @xmath104 power law relation proposed by saito ( 1998 ) .
such a relation is consistent with the relation @xmath105 derived from the theoretical x - ray magnetospheric emission model of cheng & zhang ( 1999 ) , however , this latter result is not without uncertainties since the inclination angle of the magnetic field with respect to the rotation axis and the viewing angle are not well determined .
the observed conversion efficiencies are found to range from @xmath106 , which is not in conflict with model predictions ( cheng , gil & zhang 1998 ; cheng & zhang 1999 ) .
a correlation is also found to exist between the non - pulsed x - ray luminosity and the spin down power as shown in figure 3 .
the data points are consistent with the previous asca relation of the type @xmath107 , but the correlation is also consistent with @xmath108 .
upon comparison to the results from fig .
1 , this power law relation is a consequence of the fact that the pulsar emission in the asca sample is dominated by the non - pulsed radiation component . the conversion efficiency for the non - pulsed component overlaps with that for the pulsed component ,
but extends to efficiencies as high as 0.1 .
generally , the x - ray luminosity of pulsar wind nebulae ( chevalier 2000 ) is a nonlinear function of the spin down power .
as can be seen from equation ( 9 ) of chevalier ( 2000 ) , the nebula s x - ray luminosity follows from l_x _ e^p-1 _
b^(p-2)/4_w^p-2r_s^-(p-2)/2 ^(p+2)/4 . here , @xmath48 and @xmath24
are assumed to be constant , but @xmath109 from eqs .
( 7 ) and ( 9 ) .
although the above equation corresponds to the fast cooling regime , the dependence of @xmath110 on @xmath1 for the slow cooling regime is unchanged because the ratio of the x - ray luminosities in these two regimes depends only on the cooling frequency @xmath50 , which is independent of @xmath1 .
we note that @xmath20 also depends on the spin down power so the explicit dependence of @xmath0 on @xmath1 remains to be theoretically determined .
to estimate this dependence we make use of the results of ruderman ( 1981 ) and arons ( 1983 ) , who argued that large fluxes of protons ( ions ) could also be extracted from the neutron star and accelerated in the parallel electric field in the magnetosphere .
the initial poynting flux can be converted into particle thermal and kinetic energy well within the termination radius . since
both electrons and protons are basically accelerated by the low frequency electromagnetic wave generated by the pulsar , they will be accelerated to the same relativistic speed as they are bound by the strong electrostatic force .
hence , the lorentz factor of the electrons and protons are the same , leading to the result that the protons may have carried away most of the spin down power ( coroniti 1990 ) .
thus , we can obtain a form for the spin down power from ~n _ wm_pc^2 , where @xmath111 is the outflow current from the surface .
this current should be of the order of the goldreich - julian current ( goldreich & julian 1969 ) given as @xmath112 . since @xmath113 , we find @xmath114 , leading to @xmath115 .
therefore , we obtain the relation @xmath116 , where @xmath5 generally varies between 2 and 3 .
the relation deduced from the non - pulsed x - ray luminosity and spin down power of observed pulsars may result from a relatively high electron energy index in the nebula .
the spectral observations of the x - ray emission provide additional evidence in support of the pulsar wind nebula interpretation .
in particular , the photon indices of both the pulsed and non - pulsed emission components , where determined , for x - ray pulsars are listed in table 2 . in this data compilation ,
the observational results are taken from the asca data and also from the recent observations by _ bepposax , chandra _ and _ xmm - newton_. in figure 4 , the photon index of the pulsed component is plotted versus the ratio of the isotropic pulsed x - ray luminosity to pulsar spin down power , @xmath117 .
the results reveal that the photon index of the pulsed emission component generally varies from 1.1 - 1.9 ( except for psr 0218 + 4232 ) .
such a range is consistent with the pulsed component originating in the pulsar s magnetosphere ( wang et al .
1998 ; cheng , gil , & zhang 1998 ) .
the values of @xmath118 vary over a wide range from @xmath119 to @xmath120 with no apparent correlation of the photon index with respect to efficiency .
we note that the actual pulsed x - ray luminosities depend on the beaming of the pulse and , therefore , the efficiency of spin down power to pulsed x - ray luminosity is uncertain .
the real efficiency may only scatter around an apparent efficiency , which is based on an assumed constant solid angle . unless the beaming conspires with the photon index to produce a correlation , the rough distribution of the data presented in figure 4 would not be expected to change . in summary , the efficiency and the photon index do not appear to show any correlation , which is consistent with the model predictions of cheng & zhang ( 1999 ) .
on the other hand , cheng & zhang ( 1999 ) suggest that the photon index sensitively depends on the local properties of the stellar magnetic field .
we also illustrate the relation between the efficiency of pulsar spin down power to non - pulsed x - ray luminosity , @xmath121 , and the photon index of the non - pulsed component in figure 5 .
it can be seen that the photon index of the non - pulsed emission component shows two distinct regimes .
in particular , there is a grouping of data points with @xmath122 for low efficiencies ( @xmath123 ) and a separate grouping of data points with @xmath124 at higher efficiencies .
this tendency may reflect a larger photon index in the fast cooling regime in comparison to the slow cooling regime as discussed in 2.3 .
specifically , we showed that the fast cooling regime and slow cooling regime is separated by @xmath125 or @xmath126 where @xmath50 is the cooling frequency and @xmath52 is the observed x - ray frequency respectively . in the former regime ,
the photon indices are in the range from 2 - 2.5 and in the latter regime , the photon indices range from 1.5 to 2 . typically , the slow cooling region has a lower efficiency than the fast cooling region by a factor @xmath127 .
the actual efficiency of an individual pulsar depends on its corresponding parameters ( e.g. @xmath128 ) .
evidence in favor of such a relation is suggested by the difference between the crab , which is in the fast cooling regime with an efficiency of @xmath129 and geminga , which is in the slow cooling regime with an efficiency of @xmath130 . in table 3 ,
the termination radii derived from the images of some identified pulsar wind nebulae are listed .
although the determination of the radius is imprecise , we have adopted two procedures for its estimation . for the nebulae which have ring structures ( e.g. , crab nebula ) ,
the termination radius is chosen as the size of the inner ring .
on the other hand , for the extended sources with a bright point - like central sources , the termination radius is taken to be the size corresponding to the scale in which about 90% of the observed counts are included .
for comparison , the theoretical termination radii for some nebulae are estimated according to eqs .
( 7 ) ( for the crab nebula ) and ( 9 ) ( other nebulae ) as discussed in 2.2 and also listed in table 3 . with the knowledge of the termination radius of the nebulae
, the x - ray luminosity can be estimated assuming typical values for the parameters ( @xmath20 , @xmath5 , @xmath48 , @xmath24 ) . the observed and estimated x - ray luminosities of these compact nebulae are given in table 3 .
for these estimates , we have taken @xmath27 as the observed values , and assumed @xmath131 , and @xmath132 for all pulsars .
we find that the model luminosities and the observed luminosities are in rough agreement with each other to within a factor of 4 , which taking into account the possible variations of these parameters amongst the pulsars , suggests that the simple one - zone model provides a reasonable approximation for the x - ray luminosity level .
the x - ray emitting region with a characteristic frequency ( @xmath133 ) may exhibit a tail - like spatial structure provided that the pulsar velocity exceeds that of the termination shock front and the nebula magnetic field is sufficiently low . in this case
, the distance traversed by the pulsar within the synchrotron cooling timescale can be taken as a lower limit of the elongation length .
specifically , the synchrotron cooling time in the x - ray band is @xmath134 s where @xmath135 is the magnetic field in the emission region in milligauss .
thus , the typical cooling time is @xmath136 s for @xmath137 , and the length of the x - ray elongated feature is about @xmath138 cm for a pulsar moving at a velocity of 100 km s@xmath63 with respect to the interstellar medium .
the x - ray images of some pulsar wind nebulae , indeed , reveal extensive x - ray tails . in table 3 , several sources with x - ray tail features are included .
seven cases are thought to be the consequence of a bow shock formed by the high velocity pulsar ( e.g. , geminga , psr b1823 - 13 , b1757 - 24 , b1957 - 20 , j1747 - 2958 , j1124 - 5916 and b1853 + 01 ) .
the two pulsars j1930 + 1852 and b0453 - 685 exhibit elongated structures and may also be bow shock structures .
on the other hand , the crab and vela nebulae exhibit an outflow structure ( x - ray jet ) .
geminga is one of the most studied pulsars because it is one of the very nearby pulsars and is the first pulsar detected in optical , x - ray and gamma - ray , but not in radio ( bignami & caraveo 1996 ) wavelengths .
based on asca observations , saito ( 1997a ) found that the non - pulsed x - ray luminosity was @xmath139 s@xmath63 and the pulsed x - ray luminosity was @xmath140 s@xmath63 in the 2 - 10 kev band .
based on more recent xmm - newton observations , caraveo et al .
( 2004 ) has reported the detection of a power law component ( @xmath141 s@xmath63 in 2 - 8 kev band ) , which consists of the pulsed as well as the non - pulsed component . the pulsed component in 2 - 6
kev is @xmath142 , which is consistent with asca observation .
however , it is still not clear if the non - pulsed component originates from the pulsar or the nebula .
the recent _ chandra _ observations of the geminga pulsar show a x - ray bow shock structure(caraveo et al .
2003 ) . the image of the field including the geminga pulsar reveals a compact point - like x - ray source of size 20 " or ( @xmath143 cm ) with two elongated tails extending to 2 ( caraveo et al .
since the x - ray tails have a non - thermal x - ray spectrum ( @xmath144 ) , these features may be produced by synchrotron emission in the bow shock between the pulsar wind and the interstellar medium .
the spin down power of the geminga pulsar is relatively low compared to other pulsars , with @xmath145 . assuming a proper motion velocity of @xmath146120 kms@xmath63 ( bignami & caraveo 1993 ) and a number density of 1 particle @xmath74 in the interstellar medium , the termination shock radius of the pulsar wind nebula is @xmath147 cm .
this scale is consistent with the observational constraint on the compact x - ray nebula for an assumed distance of @xmath76 pc .
the point - like source has a x - ray luminosity of @xmath148 in the energy range from 0.3 to 5 kev , and the total luminosity from the two tails is @xmath149 ergs s@xmath63 ( caraveo et al .
the pulsed and non - pulsed emission of the point - like source was not resolved in the _ xmm - newton _ observations , but the results from 3 ( see table 1 ) , reveal that the non - pulsed luminosity of geminga is @xmath150 . interpreting this luminosity as arising from the compact x - ray nebula , we find that @xmath151 which is low compared to other pulsar wind nebula .
we suggest that the nebula is in the slow cooling regime where the photon spectral index , @xmath91 .
the observed photon index of the tail is 1.6 , corresponding to @xmath152 , and its observed length ( @xmath153 cm ) suggests that the magnetic field of the surrounding medium is weak ( @xmath154 g , see also caraveo et al .
thus , the inferred fractional magnetic energy density is @xmath155 . assuming a fractional electron energy density of @xmath156 yields a nebula luminosity of @xmath157 in the 2 - 10 kev band , which modestly overestimates the contributions from the compact nebula and x - ray tails by a factor of @xmath158 .
another example where the model can be compared to the observational data is the x - ray tail associated with psr b1757 - 24 observed by kaspi et al .
( 2001 ) with the _ chandra _ satellite .
psr b1757 - 24 is a 124 ms radio pulsar discovered near the supernova remnant ( snr ) g5.4 - 1.2 ( manchester et al .
1985 ) with a spin down power @xmath159 ( manchester et al . 1991 ) . assuming an association between the pulsar and g5.4 - 1.2 , and that the pulsar s characteristic age is a good estimate of its true age , a proper motion of 75 mas yr@xmath63 is inferred .
this corresponds to a transverse space velocity of @xmath160 km s@xmath63 for a distance of 5 kpc ( frail , kassim , & weiler 1994 ) .
however , recent interferometric observations have failed to detect the implied proper motion ( gaensler & frail 2000 ) , which suggests that the pulsar is older than its characteristic age or that the assumed pulsar birth place is incorrect . here , we take the proper motion velocity of psr b1757 - 24 to be @xmath161 km s@xmath63 ( gaensler & frail 2000 ) .
the observed 2 - 10 kev luminosity is about @xmath162 corresponding to a conversion efficiency of @xmath163 , with the x - ray tail extending to @xmath164 ( 0.5 pc for the distance of 5 kpc ) . the observed photon index , @xmath165 , yielding @xmath152 in the slow cooling regime .
assuming an interstellar medium number density , @xmath166 @xmath74 , the termination shock radius is @xmath167 cm , with the known velocity @xmath161 km s@xmath63 . taking the typical pulsar wind nebula parameters , @xmath168 ( corresponding to @xmath169 g at @xmath27 ) , and @xmath170
, we find the x - ray luminosity from 2 - 10 kev is @xmath171 using the one - zone model .
the predicted x - ray tail length is given by @xmath172 cm .
both these estimates for the x - ray luminosity and tail length are in approximate accord with the observed values .
the non - thermal non - pulsed x - ray emission of rotation powered pulsars has been investigated in the context of emission from a pulsar wind nebula .
we have confirmed that this emission can significantly contribute to the total x - ray emission , thereby , increasing the x - ray emission above that produced in the pulsar magnetosphere .
a reexamination of asca data for pulsars separated into the non - pulsed and pulsed emission components indicates that their non - thermal components , individually , are correlated with spin down power . in particular , the dependence is steeper for the non - pulsed component with @xmath173 as compared to the relation , @xmath174 for the pulsed component .
our results are similar to that discovered by becker & trmper(1997 ) , who found that the non - thermal ( pulsed and non - pulsed ) component satisfied the linear relation , @xmath175 , based on rosat data in 0.1 - 2.4 kev range . in this paper
, we have used the asca data in the 2 - 10 kev range , which includes the contribution of the additional thermal pulsed component with @xmath176 kev from the polar cap ( cheng & zhang 1999 ) . in this case , the dependence of the x - ray luminosity on spin down power is steeper . because of the poor angular resolution of asca
, the possible contribution of non - thermal non - pulsed emission from the magnetosphere is mixed in with that of the nebula . within the framework of a one zone model for the pulsar wind developed by chevalier ( 2000 ) , the power law relation between the non - pulsed luminosity and spin down power provides an important diagnostic for the electron energy distribution in the shock resulting from the interaction of the relativistic wind with the interstellar medium .
specifically , @xmath177 where @xmath5 is the power law index of the electron energy distribution , suggesting that the index characterizing these pulsars is relatively high , @xmath178 . the observational results also show that the conversion efficiency of spin down power to x - ray luminosity is a function of spin down power varying from @xmath119 to 0.1 .
the efficiencies in the lower ( upper ) end of the range can be understood in terms of emission in the slow ( fast ) cooling regime where the x - ray frequency is less ( greater ) than the electron synchrotron cooling frequency .
the spectra from pulsar wind nebulae are distinctly non - thermal , described by a photon power law index , @xmath179 or @xmath180 for the fast or slow cooling regime respectively .
this implies a correlation between the photon index and efficiency of conversion of spin down power to non - pulsed x - ray luminosity with steeper photon power law indices associated with higher conversion efficiencies .
although the photon power law indices inferred from the pulsars studied have large uncertainties , there are hints from the observed data that such a correlation exists .
we note that the low values of @xmath6 and @xmath181 inferred for the extended emission observed from geminga and b 1757 - 24 are consistent with such an interpretation .
future observed determinations of the photon power law indices could further verify this correlation , thereby , providing an additional diagnostic tool for probing the pulsar and its environment .
finally , the pulsar wind nebula interpretation may also apply to the spatially extended x - ray emission observed in binary systems containing a pulsar .
for example , the extended x - ray emission associated with the millisecond pulsar b1957 + 20 ( stappers et al .
2003 ) may be interpreted within such a framework .
the application of such models to recycled pulsars in the galaxy will be the subject of a future investigation .
we are grateful to the anonymous referee for the fruitful suggestions .
this work is partially supported by the nsf through grant ast-0200876 , by a rgc grant of the hong kong government , and by the national natural science foundation of china under grant 10273011 .
achterberg , a. , gallant , y.a . ,
kirk , j.g . & guthmann , a.w .
2001 , , 328 , 393 arons , j. 1983 , in electron - positron pairs in astrophysics , ed .
m. l. burns , a. k. harding , & r. ramaty ( new york : aip ) , 113 becker , w. , & trmper , j. 1997 , , 326 , 682 becker , w. et al . , 2004 , astro - ph/0405180 bednarz , j. & ostrowski , m. 1998 , phys . rev .
lett . , 80 , 3911 bignami , g. f. , & caraveo , p.a .
1993 , , 361 , 704 bignami , g. f. , & caraveo , p.a .
1996 , ann .
astron . astrophys .
34 , 331 blumenthal , g. & gould , k. 1970 , rev . mod .
phys . , 42 , 237 camilo , f. et al .
2002 , , 579 , l25 caraveo , p.a .
2003 , science , 301 , 1345 cheng , k. s. , gil , j. & zhang , l. 1998 , , 493 , l35 cheng , k. s. , & zhang , l. 1999 , , 515 , 337 chevalier , r. a. 2000 , , 539 , l45 coroniti , f. v. 1990 , , 349 , 538 finley , j.p .
1998 , , 493 , 884 frail , d. a. , kassim , n. e. , & weiler , k. w. 1994 , aj , 107 , 1120 gaensler , b. m. , & frail , d. a. 2000 , nature , 406 , 158 gaensler , b. m. et al . 2003a , , 588 , 441 gaensler , b. m. et al .
2003b , , 594 , l111 gaensler , b. m. et al .
2004 , , submitted ( astro - ph/0312362 ) goldreich , p. , & julian , w.h .
1969 , , 157 , 859 gotthelf , e.v . &
kaspi , v.m .
1998 , , 497 , l29 gotthelf , e.v .
2000 , , 542 , l37 greiveldinger , c. et al .
1996 , , 465 , l35 halpern , j. p. , & wang , f.y .- h .
1997 , , 477 , 905 halpern , j. p. et al .
1998 , , 552 , l125 helfand , d. j. , gotthelf , e. v. , & halpern , j.p .
2001 , , 556 , 380 hessels , j.w.t .
2004 , , submitted ( astro - ph/0403632 ) hirayama , m. et al .
1996 , , 48 , 833 hirayama , m. et al .
2002 , , 333 , 603 hughes , j.p .
2001 , , 559 , l153 kaaret , p. et al .
2001 , , 546 , 1159 kaspi , v.m .
2001 , , 562 , l163 kawai , n. , tamura , k. , & shibata , s. 1998 , iau symp .
188 , the hot universe , ed .
k. koyama , s. kitamoto , & m. itoh ( dordrecht : kluwer ) , 265 kennel , c. f. , & coroniti , f. v. 1984 , , 283 , 694 kulkarni , s. r. , & hester , j. j. 1988 , , 335 , 801 lemoine , m. & pelletier , g. 2003 , , 589 , l73 lu , f.j .
2002 , , 548 , l49 manchester , r. n. , damico , n. , & tuohy , i. r. 1985 , mnras , 212 , 975 manchester , r. n. et al .
1991 , mnras , 253 , 7 marsden , d. et al .
1997 , , 491 , l39 marshall , f. e. et al .
1998 , , 499 , l179 murray , s.s .
2002 , , 568 , 226 pavlov , g. g. et al , 2001 , , 552 , l129 petre , r. , kuntz , k.d . &
shelton , r.l .
2002 , , 579 , 404 pivovaroff , m. j. , kaspi , v. m. , & gotthelf , e.v .
2000 , , 528 , 436 possenti , a. et al . 2002 , ,
387 , 993 rees , m. j. , & gunn , j. e. 1974 , , 167 , 1 reynolds , s.p .
1994 , , 271 , l1 ruderman , m. 1981 , in iau sym .
95 , pulsars , ed .
w. sieber & r. wielebinski ( dordrecht : reidel ) , 87 saito , y. 1998 , ph.d .
thesis , univ . of tokyo saito , y. et al .
1997a , in neutron stars and pulsars : thirty years after discovery , ed . n. shibazaki et al .
( tokyo : university academy press ) , 295 saito , y. et al . 1997b , , 477 , l37 sari , r. , piran , t. & narayan , r. 1998 , apj , 497 , l17 seward , f. d. , & wang , z. 1988 , , 332 , 199 shibata , s. et al .
1997 , , 483 , 843 stappers , b. w. et al .
2003 , science , 299 , 1372 takahashi , m. et al .
2001 , , 554 , 316 tennant .
2001 , , 554 , l173 torii , k. et al .
1997 , , 489 , l145 torii , k. et al .
1999 , , 523 , l69 torii , k. et al .
2001 , , 551 , l151 toscano , m. et al .
1999 , , 307 , 925 van den bergh , s. & pritchet , c.j .
1989 , , 338 , l69 wang , f. y .- h . , & halpern , j.p .
1997 , , 482 , l159 wang , f. y .- h .
1998 , , 498 , 373 wang , q. d. , & gotthelf , e. v. 1998 , , 494 , 623 wang , w. , & zhao , y. 2004 , , 601 , 1038 weisskopf m. c. et al .
2000 , , 536 , l81 willingale , r. et al .
2001 , , 365 , l212 zhang , b. , & harding , a.k . 2000 , , 532
, 1150 l c c c c c c c l psr & @xmath30 ( s ) & @xmath182 & @xmath183 & @xmath1 & @xmath184 & @xmath185 & @xmath186 & reference + j0631 + 1036 & 0.288 & @xmath187 & 1.0 & @xmath188 & @xmath189 & @xmath190 & @xmath191 & 1 + j1811 - 1926 & 0.065 & @xmath192 & 5.0 & 7.0@xmath193 & 1.6@xmath194 & 1.9@xmath195 & 1.4@xmath196 & 2,3 + b0531 + 21 & 0.033 & 4.2@xmath197 & 2.0 & 4.5@xmath198 & 1.1@xmath199 & 6.8@xmath200 & 1.0@xmath201 & 4 + b0833 - 45 & 0.089 & 1.25@xmath197 & 0.3 & 6.9@xmath193 & 1.9@xmath195 & 1.5@xmath202 & 1.8@xmath203 & 4,5,6 + b0633 + 17 & 0.237 & 1.1@xmath204 & 0.16 & 3.2@xmath196 & 8.2@xmath205 & 4.6@xmath205 & 3.6@xmath206 & 7,8 + b1706 - 44 & 0.1025 & 9.3@xmath204 & 1.82 & 3.4@xmath193 & 6.8@xmath202 & 4.1@xmath202 & 2.7@xmath207 & 4 + b1509 - 58 & 0.150 & 1.54@xmath208 & 4.3 & 1.8@xmath199 & 4.0@xmath194 & 1.3@xmath194 & 2.7@xmath194 & 4 + b1951 + 32 & 0.0395 & 5.8@xmath209 & 2.5 & 3.7@xmath193 & 6.2@xmath195 & 6.2@xmath202 & 5.6@xmath203 & 4 + b1046 - 58 & 0.124 & 9.6@xmath204 & 2.98 & 2.0@xmath193 & 9.5@xmath202 & 5.5@xmath202 & 4.5@xmath202 & 4,9 + b1929 + 10 & 0.227 & 1.16@xmath209 & 0.17 & 3.9@xmath203 & 1.54@xmath210 & 5.6@xmath205 & 9.8@xmath205 & 10 + b0656 + 14 & 0.385 & 5.5@xmath204 & 0.76 & 3.8@xmath196 & 1.7@xmath211 & 1.0@xmath211 & 7.0@xmath212 & 4 + b0540 - 69 & 0.05 & 4.8@xmath197 & 49.4 & 1.5@xmath198 & 8.3@xmath213 & 1.3@xmath213 & 7.0@xmath193 & 11 + b0950 + 08 & 0.253 & 2.3@xmath214 & 0.12 & 5.6@xmath207 & 4.6@xmath205 & 1.6@xmath205 & 3.0@xmath206 & 10 + b1610 - 50 & 0.232 & 4.93@xmath197 & 7.26 & 1.6@xmath193 & 9.6@xmath195 & 3.0@xmath195 & 6.6@xmath203 & 9 + b1055 - 52 & 0.197 & 5.83@xmath209 & 1.53 & 3.0@xmath196 & 2.7@xmath210 & 2.0@xmath210 & 7.0@xmath206 & 12 + b1853 + 01 & 0.267 & 5.4@xmath197 & 2.02 & 8.0@xmath215 & 1.4@xmath195 & 1.4@xmath202 & 1.3@xmath203 & 4 + j2229 + 6114 & 0.0516 & 7.8@xmath204 & 3.0 & 2.2@xmath201 & 1.7@xmath195 & 4.0@xmath202 & 1.3@xmath203 & 13 + b0537 - 69 & 0.016 & 5.13@xmath204 & 47 & 4.8@xmath198 & 2@xmath213 & 1.7@xmath200 & 1.8@xmath193 & 14,15 + j1846 - 0258 & 0.32 & @xmath216 & 19 & 1@xmath199 & 1@xmath213 & 4@xmath194 & 1@xmath213 & 16 + b1937 + 21 & 0.00156 & 1.05@xmath217 & 3.6 & 1.1@xmath193 & 5.7@xmath202 & 2.5@xmath202 & 3.2@xmath207 & 17 + j2124 - 3358 & 0.005 & 1.08@xmath217 & 0.25 & 3.5@xmath203 & 4.8@xmath205 & 1.6@xmath205 & 3.2@xmath206 & 18 + b1821 - 24 & 0.003 & 1.6@xmath218 & 5.1 & 2.2@xmath193 & 6.5@xmath195 & 9.4@xmath202 & 5.5@xmath195 & 19 + j0437 - 47 & 0.0058 & 2.0@xmath217 & 0.18 & 4.2@xmath203 & 1.3@xmath210 & 4.0@xmath205 & 9.0@xmath206 & 4 + l c c c c c l psr & @xmath1 & @xmath219 & @xmath220 & @xmath221 & @xmath222 & reference + b1823 - 13 & 2.8@xmath213 & @xmath223 & @xmath224 & @xmath225 & @xmath226 & 1 + j0537 - 6910 & 4.8@xmath227 & @xmath228 & @xmath229 & @xmath230 & @xmath231 & 2 + b0540 - 69 & 1.5@xmath227 & 0.01 & 0.05 & @xmath232 & @xmath233 & 3 + b1509 - 58 & 1.8@xmath199 & @xmath234 & @xmath235 & @xmath236 & @xmath237 & 4 + j0218 + 4232 & 2.5@xmath200 & 4.8@xmath238 & 2@xmath239 & 0.61@xmath240 & @xmath241 & 5 + j0631 + 1036 & 5.4@xmath194 & @xmath242 & @xmath243 & 2.3@xmath244 & @xmath245 & 6 + j1811 - 1926 & @xmath246 & 3@xmath238 & @xmath247 & 1.65@xmath248 & @xmath249 & 7 + b0531 + 21 & 4.5@xmath227 & 1.6@xmath250 & 0.03 & 1.6@xmath251 & 2.1@xmath252 & 8 + b0833 - 45 & @xmath253 & 2@xmath254 & 3@xmath239 & 1.5@xmath255 & 1.67@xmath256 & 9 + b0633 + 17 & 3.2@xmath194 & 1.4@xmath254 & 1.1@xmath257 & 1.0@xmath244 & 1.6@xmath251 & 10,11 + b1055 - 52 & 3@xmath194 & 7@xmath254 & 2@xmath257 & 1.5@xmath255 & 1.9@xmath251 & 12 + b1937 + 21 & 1.1@xmath213 & 2@xmath238 & @xmath258 & 1.71@xmath259 & 0.8@xmath260 & 13 + b1821 - 24 & 2.2@xmath213 & 5@xmath238 & 3@xmath261 & 1.2@xmath244 & @xmath262 & 14 + j2229 + 6114 & 2.2@xmath199 & 2@xmath254 & @xmath80 & 1.51@xmath263 & 1.6@xmath251 & 15 + j1105 - 6107 & 2.5@xmath213 & 6@xmath238 & 1.6@xmath261 & 1.8@xmath264 & 2.2@xmath264 & 16 + b1706 - 44 & 3.4@xmath213 & 1.4@xmath238 & @xmath80 & 1.9@xmath260 & 1.7@xmath244 & 17 + b1757 - 24 & 2.6@xmath213 & @xmath163 & @xmath80 & 1.6@xmath265 & 1.0@xmath260 & 18 + j0205 + 6449 & 2.6@xmath199 & @xmath119 & @xmath266 & 1.1@xmath264 & 1.9@xmath251 & 19 + b1929 + 10 & 3.9@xmath195 & 3@xmath238 & @xmath267 & 1.79@xmath268 & & 20 + b0656 + 14 & 3.8@xmath194 & 3@xmath238 & 2@xmath239 & 1.5@xmath269 & & 21 + j2021 + 3651 & @xmath270 & 3@xmath238 & @xmath243 & 1.5@xmath255 & 1.7@xmath255 & 22 + j1747 - 2958 & 2.5@xmath213 & @xmath163 & 0.02 & 1.8@xmath269 & 2.1@xmath269 & 23 + b1853 + 01 & @xmath271 & @xmath242 & 1.5@xmath261 & @xmath272 & @xmath273 & 24 + l c c c c c c c c l psr & @xmath1 & @xmath274 & @xmath275 & @xmath276 & @xmath277 & @xmath18 & @xmath71 & @xmath278 & reference + & & cm & cm & & & @xmath74 & km s@xmath63 & cm & + b1823 - 13 & 2.8@xmath213 & 6@xmath279 & & 3@xmath195 & 2@xmath195 & 1 & & 7@xmath280 & 1 + j0537 - 6910 & 4.8@xmath227 & 4@xmath279 & & 1.8@xmath213 & 3@xmath213 & & & & 2 + b0540 - 69 & 1.5@xmath227 & 3@xmath279 & & 7@xmath213 & 4@xmath213 & & & & 3 + j1811 - 1926 & @xmath246 & 2@xmath279 & & 1.4@xmath194 & @xmath66 & & & & 4 + b0531 + 21 & 4.5@xmath227 & 4@xmath279 & 3@xmath281 & @xmath282 & 4@xmath213 & 10 & 123 & @xmath283 & 5 + b0833 - 45 & @xmath253 & @xmath284 & 2@xmath281 & 1.8@xmath195 & @xmath285 & 1 & 65 & 8@xmath286 & 6 + b0633 + 17 & 3.2@xmath194 & 5@xmath286 & 4@xmath287 & 4@xmath205 & @xmath288 & 1 & 120 & 4@xmath279 & 7 + j2229 + 6114 & 2.2@xmath199 & 4@xmath279 & & 1.3@xmath195 & 3@xmath195 & & & & 8 + j1105 - 6107 & 2.5@xmath213 & 4@xmath286 & & 4@xmath195 & 3@xmath195 & & & & 9 + b1706 - 44 & 3.4@xmath213 & 3@xmath279 & & 3@xmath202 & 4@xmath202 & & & & 10 + b1757 - 24 & 2.6@xmath213 & 4@xmath286 & @xmath289 & 3@xmath202 & 5@xmath202 & 1 & @xmath290 & @xmath283 & 11 + j0205 + 6449 & 2.6@xmath199 & @xmath284 & & 3@xmath194 & 3@xmath194 & & & & 12 + b1957 + 20 & @xmath291 & 5@xmath286 & 4@xmath286 & 1.6@xmath211 & @xmath292 & 1 & 220 & 4@xmath279 & 13 + j2021 + 3651 & @xmath270 & 8@xmath279 & & 3@xmath203 & 2@xmath195 & & & & 14 + j1747 - 2958 & 2.5@xmath213 & 3@xmath286 & 7@xmath287 & 5@xmath194 & @xmath66 & 0.3 & 600 & 2@xmath280 & 15,16 + j1124 - 5916 & @xmath282 & @xmath284 & 2@xmath279 & 4@xmath196 & @xmath66 & 0.5 & 450 & 6@xmath279 & 17 + b1853 + 01 & @xmath271 & 10@xmath293 & 3@xmath286 & 6@xmath202 & 4@xmath202 & 5 & 375 & @xmath283 & 18 + j1930 + 1852 & 2@xmath213 & @xmath284 & & @xmath285 & 7@xmath202 & 1 & & @xmath283 & 19 + b0453 - 685 & @xmath282 & 6@xmath279 & & 6@xmath194 & 2@xmath194 & 0.4 & & @xmath294 & 20 + j0538 + 2817 & 4@xmath194 & 8@xmath286 & 3@xmath287 & 6@xmath211 & 5@xmath211 & 0.5 & 385 & & 21 + | the general properties of the non - thermal non - pulsed x - ray emission of rotation powered pulsars are investigated in the context of a pulsar wind nebula model .
an examination of the observed x - ray emission from a sample of 23 pulsars in the energy range between 2 - 10 kev reveals that the relation of x - ray luminosity , @xmath0 , to the pulsar spin down power , @xmath1 , is steeper for the non - pulsed component than for the pulsed component .
specifically , @xmath2 for the non - pulsed component , whereas @xmath3 for the pulsed component .
the former relation is consistent with emission from a pulsar wind nebula model in which @xmath4 where @xmath5 is the power law index of the electron energy distribution .
the relation for the pulsed component , on the other hand , is consistent with a magnetospheric emission model .
in addition , the photon spectral index , @xmath6 , was found to be correlated to the conversion efficiency of spin down power to non - pulsed x - ray emission with greater efficiencies for @xmath7 than for @xmath8 . such a relation
is naturally understood within the framework of a pulsar wind nebula model with the former relation corresponding to the emission of x - rays in the fast cooling regime and the latter relation corresponding to emission in the slow cooling regime .
the x - ray properties of pulsar wind nebulae are sensitive to the physical conditions ( e.g. , the density and magnetic field ) of the interstellar medium and can lead to important differences between the x - ray emission characteristics ( luminosity , photon spectral index and emission morphology ) of pulsars in various environments .
such wind nebulae can contribute to the non - thermal symmetric emission morphology ( point - like ) and elongated emission morphology ( tail - like ) from sources similar to geminga and psr b1757 - 24 .
# 1 # 1#2#1 # 2 |
coupled oscillator models exhibit complex dynamics that has been observed in a wide range of different fields including physical @xcite and biological @xcite models .
synchronization @xcite , clustering @xcite , chaos @xcite and spontaneous switching between different cluster states @xcite have all been observed in such systems .
other studies have examined coupling between two or more systems that may individually be chaotic , and a wide variety of types of synchronization have been found and analysed ; see for example @xcite .
we examine phase oscillator models that are appropriate if the coupling between oscillators is weak compared to the attraction onto the limit cycle ( e.g. @xcite ) .
although the coupling structure and strength are important for the dynamical behaviour of the system , the exact coupling function ( which represents the nonlinearities in the oscillators and the coupling ) has a subtle effect on the collective behaviour of the system .
research into the dynamics of coupled nonlinear oscillators has long explored the question `` what is the dynamics of a given system ? '' .
a less frequently asked , but also very interesting , question is `` how can we design a coupled system to have specific dynamics ? '' .
this latter question was considered by @xcite who designed cluster states with a prescribed clustering by giving explicit conditions on the coupling function and its first derivative to have a stable cluster state with a specific clustering .
they demonstrate specific coupling functions that give stable cluster states for any partition of the oscillator into groups , regardless the number of oscillators and the size of each cluster . in this paper
we go beyond @xcite in four ways .
firstly , we examine three - cluster states and show that not only stable cluster states , but also cluster states with specific transverse stability properties can be designed by suitable choice of coupling function . secondly , we give some results on how the transverse stability can be varied independently of the tangential stability and hence exhibit possible bifurcation scenarios from transversely stable clustering .
thirdly , we show examples of how nontrivial cluster states with three inequivalent clusters can be joined into a heteroclinic network .
finally , we generalize some of the bifurcation results to more general multi - cluster states with an arbitrary numbers of clusters .
we use a fourier representation of the coupling function associated with a system of @xmath0 oscillators to design general three - cluster states , as in @xcite . the rest of the paper is organized as follows
; for the remainder of this section we recall some of the notation and previous results on existence and stability of periodic cluster states .
we define a notion of inequivalence of clusters within a cluster state and consider some sufficient conditions for clusters to be inequivalent .
section [ sec:3cluster ] recalls and extends some basic results on the appearance of tangentially stable but transversely unstable three - cluster states .
we present in theorem [ thm : transstab ] a characterization of transverse stability , and in corollary [ cor : couplingfunction ] a result on transverse bifurcation of three - cluster states .
section [ sec : heteroclinic ] presents what we claim is the smallest possible cluster state with three inequivalent nontrivial clusters ( requiring at least @xmath2 oscillators ) and gives some examples how these may be connected into robust attracting heteroclinic networks @xcite .
finally , section [ sec : conclusion ] discusses some consequences of this work , including a generalization of corollary [ cor : couplingfunction ] . in this paper
we consider @xmath0 phase oscillators that are all - to - all coupled and governed by the following generalization @xcite of kuramoto s model system of coupled phase oscillators : @xmath3 where @xmath4 is the phase of the @xmath5 oscillator , @xmath6 and @xmath7 is a @xmath8-periodic nonlinear _ coupling function _ that we assume is smooth and represented by a truncated fourier series as in @xcite : @xmath9 where @xmath10 ( @xmath11 ) and @xmath12 ( @xmath13 ) are the real coefficients and @xmath14 is the number of fourier modes .
note that the coupling function @xmath1 derived from weakly coupled nonlinear phase oscillators will typically have several non - zero modes in its fourier series , even if the oscillators are close to hopf bifurcation @xcite .
conditions on the coupling function and its first derivative that ensure the existence and stability of desired cluster states in the system ( [ eq : coupledoscillatorsystem ] ) are derived in @xcite .
note that the system is invariant under `` spatial '' symmetries @xmath15 acting by permutation of the components and `` temporal '' symmetries @xmath16 given by @xmath17 for any @xmath18 @xcite .
we now look at periodic cluster states in a bit more depth .
consider a partition @xmath19 into @xmath20 clusters , where @xmath21 ; each of @xmath22 form a cluster of size @xmath23 for @xmath24 and @xmath25 .
we say a cluster is a _ multi - cluster _ if @xmath26 .
the @xmath27th cluster is said to be _ nontrivial _ if @xmath28 . a periodic orbit @xmath29 of ( [ eq : coupledoscillatorsystem ] ) defines an associated _ clustering _ @xmath30 $ ] where @xmath31 and @xmath32 being the same cluster @xmath22 if and only if @xmath33 for some @xmath34 .
it is possible that periodic cluster states can have additional symmetries associated with non - trivial phase shift symmetry @xmath16 ; more precisely , it is known that they can be classified according to decompositions of the form @xmath35 where @xmath36 are all positive integers .
[ thm : asiso ] the subsets of @xmath37 that are invariant for ( [ eq : coupledoscillatorsystem ] ) because of symmetries of @xmath38 corresponding to isotropy subgroups in the conjugacy class of @xmath39 where @xmath40 , @xmath41 and @xmath42 denotes the semidirect product .
the points with this isotropy have @xmath43 clusters that are split into @xmath44 groups of @xmath45 clusters of the size @xmath46 .
the clusters within these groups are cyclically permuted by a phase shift of @xmath47 .
the number of isotropy subgroups in this conjugacy class is @xmath48 $ ] .
we say two clusters of @xmath30 $ ] are _ equivalent _ if there is a symmetry in @xmath38 that maps one cluster to the other .
otherwise they are said to be _
a clustering is said to be _ phase non - degenerate _ if the phase difference between two different clusters is only attained by those clusters ; more precisely we say the cluster phases @xmath49 are _ non - degenerate _ if @xmath50 for all @xmath51 with @xmath52j , @xmath53 , unless @xmath54 and @xmath55 . as a metric on the circle .
] as noted in @xcite , a sufficient condition for this is that the phase differences are rationally independent of each other and of @xmath8 .
the following more general statement follows from the definitions and theorem [ thm : asiso ] ; note that if the phase differences are rationally independent this implies ( c ) .
suppose @xmath56 is a cluster state .
any of the following is a sufficient condition for all clusters to be inequivalent : * @xmath0 is prime and at least one cluster is nontrivial . *
no two clusters have the same size * the clustering is phase non - degenerate * proof : * in case ( a ) , theorem [ thm : asiso ] gives that the only possible factorizations @xmath57 have @xmath58 or @xmath59 .
the latter case is ruled out if one of the @xmath60 . in case
( b ) , this also implies that @xmath58 in theorem [ thm : asiso ] .
finally in case ( c ) , if @xmath61 then there will be clusters which can be chosen such that @xmath62 and @xmath63 ( @xmath64 if @xmath65 , otherwise we can choose @xmath66 .
this implies that the clustering will be phase degenerate and conversely , phase non - degeneracy implies that @xmath58 and the clusters are inequivalent . given a partition @xmath67 let us now define the subspace @xmath68 in which clustering occurs as follows : @xmath69=@xmath70 : if there exists @xmath27 such that @xmath71 then @xmath72 because ( [ eq : coupledoscillatorsystem ] ) is equivariant under the action of @xmath15 by permutation of the oscillators and by the circle group @xmath16 ( [ eq : tsymm ] ) , these cluster states @xmath69 are simply fixed point subspaces for groups conjugate to @xmath73 , i.e. with @xmath58 . the system ( [ eq : coupledoscillatorsystem ] ) can be simplified on @xmath69 as follows ; as if @xmath31 is in the @xmath27th cluster then we say @xmath74 and the system reduces to @xmath75 for @xmath76 .
the dynamics of a periodic multi - cluster state can then be expressed as @xmath77 for @xmath24 where @xmath78 represents the relative phases of clustering and @xmath79 is the frequency of the periodic orbit .
this means that : @xmath80 where @xmath24 .
computing the difference between the first equation and the remaining ones gives : @xmath81 .
defining @xmath82 as in @xcite means that we can rewrite ( [ eq : frequency ] ) as @xmath83 equations ( [ eq : existenceequations ] ) are conditions that that the coupling function should satisfy for existence of such a periodic cluster state .
we also refer to @xcite for a discussion of stability where it is shown that a cluster state is linearly stable if and only if it is linearly stable in both of the following senses : * * tangential stability : * ( also called inter - cluster stability @xcite ) to determine the stability of the state to the change in its phases that respects the clustering we consider the linearized stability for perturbations @xmath84 , and writing @xmath85 , gives the following : @xmath86 where , as stated in @xcite , @xmath87 is the matrix : @xmath88\ ] ] where @xmath89 is the kronecker delta .
the matrix @xmath87 has @xmath20 eigenvalues ( including one trivial value ) and for tangential stability we require that all the other tangent eigenvalues have negative real parts .
this means that a @xmath90 cluster is tangentially stable if : @xmath91 * * transverse stability : * ( also called intra - cluster stability @xcite ) the stability of a periodic state to changes of phases that change the clustering and is obtained by linearizing ( [ eq : coupledoscillatorsystem ] ) about @xmath92 , which gives the following : @xmath93 where ( as stated in @xcite ) @xmath94 is given by @xmath95.\ ] ] this has @xmath96 real eigenvalues that are negative for a transverse stable cluster state ; put otherwise , a cluster state is transversely stable if : @xmath97 where @xmath98 is the number of nontrivial clusters ( i.e. those with more than one oscillator ) .
a coupling function associated with a stable cluster state will satisfy ( [ eq : existenceequations ] ) , ( [ eq : tangentstability ] ) , and ( [ eq : transversestability ] ) and we observe that the number of degrees of freedom on choosing the @xmath99 will always be surplus to requirements @xcite if enough fourier modes are chosen .
note that the above holds for cluster states regardless of whether the clusters are inequivalent or not .
if two clusters are equivalent then the transverse eigenvalues for those clusters will be equal ; if the clusters are inequivalent then generically the transverse eigenvalues for those clusters will be unequal .
we now derive conditions for tangential and transverse stability of three - cluster states for the system ( [ eq : coupledoscillatorsystem ] ) of globally coupled phase oscillators .
consider a periodic state where all clusters are nontrivial ( i.e. the clusters have sizes @xmath100 for @xmath101 , where @xmath102 ) .
this implies that @xmath103 ; we explore the special case @xmath2 in more detail in section [ subsec : n=6 ] .
recall from @xcite that the condition for existence of a three - cluster state is : @xmath104 the tangential stability is determined by ( [ eq : tangentstability ] ) , namely @xmath105 where @xmath106 and @xmath107 transverse stability is determined by the eigenvalues : @xmath108 where the multiplicities of @xmath109 are @xmath110 respectively .
our first new result is the following sufficient condition for tangential stability of three - clusters : suppose there is a periodic three - cluster state such that @xmath111 for all @xmath112 .
then the cluster is tangentially stable with complex contracting eigenvalues .
* proof : * if @xmath111 then all terms in ( [ eq : mueqn ] ) are negative and so @xmath113 .
moreover , note that ( [ eq : oldnu ] ) can be written in the form @xmath114 .
\end{split } \label{eq : newnu}\ ] ] hence if @xmath111 then all terms in ( [ eq : newnu ] ) are positive and so @xmath115
. hence the eigenvalues ( [ eq : tangenteigenvalues2004 ] ) are complex with negative real parts and so the cluster is tangentially stable .
the next result demonstrates that the tangential and transverse stability can be set independently of each other .
we define @xmath116 without loss of generality ( renumbering the clusters if necessary ) let us assume that @xmath117 and we demonstrate the following : [ thm : transstab ] suppose that @xmath1 is such that there is a periodic three - cluster state with non - trivial clusters such that the clusters are tangentially stable and assume that @xmath118 .
then we can classify the transverse stability as follows : * if @xmath119 then @xmath120 ( all clusters unstable ) . *
if @xmath121 then @xmath122 and @xmath123 ( one stable cluster ) .
* if @xmath124 then @xmath125 and @xmath126 ( two stable clusters ) .
* if @xmath127 then @xmath128 ( all clusters stable ) .
* proof : * these conclusions follows from noting that the conditions on @xmath129 ensure that ( [ eq : transeigs3 ] ) have zero , one , two or three positive transverse eigenvalues as stated .
note that this is independent of the number of oscillators , although if one or more of the clusters are trivial , the transverse exponent for that cluster is not defined .
if @xmath130 for any @xmath31 then the state will be at a bifurcation point and the stability is not determined at linear order .
theorem [ thm : transstab ] can be used to prove the following corollary about bifurcation of three - cluster states involving changes in transverse stability .
[ cor : couplingfunction ] suppose that @xmath1 is such that there is a periodic three - cluster state with non - trivial clusters , such that the clusters are tangentially stable .
then there is a parametrized family of coupling functions @xmath131 and parameter values @xmath132 such that for all values of the parameter @xmath133 the cluster state remains with the same phases and tangentially stability and : 1 . if @xmath134 then @xmath120 ( all clusters unstable )
. 2 . if @xmath135 then @xmath122 and @xmath123 ( one stable cluster ) .
3 . if @xmath136 then @xmath125 and @xmath126 ( two stable clusters ) .
4 . if @xmath137 then @xmath128 ( all clusters stable ) .
* proof : * for this specific three cluster state with relative phases @xmath138 there will be a @xmath139 such that @xmath140 for all @xmath141 .
now consider a smooth compactly supported periodic function @xmath142 such that @xmath143 for all @xmath144 with @xmath145 , @xmath146 and @xmath147 .
one can verify that @xmath148 , @xmath149 for all @xmath112 and @xmath133 , @xmath150 and @xmath151 for all @xmath152 .
hence the existence condition ( [ eq : existenceequations ] ) and the tangential stability conditions do not depend on @xmath153 , while the cases of theorem [ thm : transstab ] translate into the cases depending on @xmath153 . in this section
we consider properties of cluster states with nontrivial and inequivalent clusters for @xmath2 , before applying the results from the previous section to give sufficient conditions for a nontrivial stable three cluster state for ( [ eq : coupledoscillatorsystem ] ) . by restricting to phase non - degenerate ( and hence inequivalent ) clusters
we do not consider the cases @xmath61 in theorem [ thm : asiso ] .
it does not appear to be easy to characterise the solutions of the system of equation for the fourier coefficients analytically .
for @xmath2 recall that ( [ eq : coupledoscillatorsystem ] ) can be written as @xmath154 the symmetries of interchange of the @xmath155 oscillators , @xmath156 , gives nine possible isotropy subgroups corresponding to possible cluster states that are listed in the table [ tab : fixedpointsubspacesofs_6 ] ; each periodic cluster state will reside in precisely one of these invariant subspaces .
.conjugacy classes of isotropy subgroups and representative fixed - points subspaces for the action of @xmath156 on the phase space for @xmath2 globally coupled oscillators corresponding to inequivalent clusters .
note that @xmath157 represents the trivial group while the number of conjugate groups and the number of point in the group orbit under @xmath156 are given in the last two columns .
[ cols="^,^,^ , > , > " , ] left : the inset shows for case 1 the coupling function @xmath1 and a timeseries representing the phase difference of the @xmath27th oscillator relative to the 6th oscillator ; @xmath158 as a function of @xmath34 .
observe that the six oscillators synchronize into three clusters for most of the time , but there are short times when the clusters break along a connection .
right : although the cycle in case 2 has an additional unstable direction , the trajectory still appears to approach a heteroclinic cycle between three symmetrically related states . in both cases , i.i.d .
white noise of amplitude @xmath159 is added to each component of the ode.,title="fig:",width=302 ] left : the inset shows for case 1 the coupling function @xmath1 and a timeseries representing the phase difference of the @xmath27th oscillator relative to the 6th oscillator ; @xmath158 as a function of @xmath34 .
observe that the six oscillators synchronize into three clusters for most of the time , but there are short times when the clusters break along a connection .
right : although the cycle in case 2 has an additional unstable direction , the trajectory still appears to approach a heteroclinic cycle between three symmetrically related states . in both cases , i.i.d .
white noise of amplitude @xmath159 is added to each component of the ode.,title="fig:",width=302 ] the parameters listed as cases 0 , 1 and 2 in table [ tab:3clustereg ] all yield three - cluster states of type @xmath160 with varying numbers of positive transverse eigenvalues .
while case 0 gives a stable cluster the dynamics for cases 1 and 2 are more subtle .
as illustrated in figure [ fig : examplecase12 ] , a randomly chosen initial condition evolves towards a heteroclinic cycle that connects three symmetrically related periodic cluster states within the same invariant subspace .
note that any phase non - degenerate @xmath160-clustered periodic orbit will have a representative that exists within the subspace @xmath161 where @xmath162 note moreover that this @xmath163 invariant subspace will contain six distinct clustered periodic orbits given by cyclically permuting the phases of the clusters , due to the clusters being inequivalent .
more specifically , suppose there is a point on a phase non - degenerate periodic @xmath160-cluster @xmath164 for some @xmath165 .
without loss of generality one can choose @xmath166(in fact , one can assume that @xmath167 ) and phase non - degeneracy means that @xmath168 . as a consequence , @xmath169 and @xmath170
are also points on periodic @xmath160-clusters , where @xmath171 and although these are also within @xmath161 , the phase non - degeneracy means that they are distinct points ; there are three more that are in the same subspace which we write as @xmath172 the relative location of these six equilibria can be seen in figure [ fig:222clusterdiag ] ( left ) calculated using xppaut @xcite . for the coupling function in case 1 , table [ tab:3clustereg ] reveals that each of the @xmath173 has a single positive transverse eigenvalue corresponding to instability of one of the clusters but is otherwise stable .
the unstable manifold will therefore be contained within a fixed point subspace of symmetry @xmath174 where one of the clusters is broken , but the numerical results in figure [ fig : examplecase12 ] indicate that this unstable manifold is within the stable manifold of one of another @xmath173 . if we write * @xmath175 * @xmath176 * @xmath177 then one can verify that there will be a sequence of connections ( heteroclinic orbit ) ( a ) from @xmath178 to @xmath169 that is transverse within @xmath179 , ( b ) from @xmath169 to @xmath170 that is transverse within @xmath180 and ( c ) from @xmath170 to @xmath178 that is transverse within @xmath181 as show in figure [ fig:222clusterdiag](right ) @xmath182 similarly there is a symmetrically related heteroclinic cycle @xmath183 that connects the remaining equilibria within @xmath161 . turning to case 2 in figure [ fig : examplecase12 ]
we note that there are two transversely unstable directions from each of the @xmath173 in figure [ fig:222clusterdiag ] and `` accordingly '' a continuum of directions by which the trajectory can leave a neighbourhood of the @xmath173 .
these need no longer be within any of the invariant subspaces @xmath184 . as can be seen for this case , the resulting dynamics `` nonetheless '' seems to return repeatedly to a cycle between the @xmath173 suggesting that the cycle is a milnor attractor .
the results in section [ sec:3cluster ] ( concerning clustering behavior and bifurcations ) apply to systems with any number of coupled oscillators , regardless of the size of each cluster
. however , these results are essentially local in phase space .
more precisely , a given coupling function may admit a variety of different cluster states of varying stability , and there may be constraints on the possible cluster states and/or their stabilities
. it would be interesting to understand the nature of such constraints , but we leave this for future work . by considering properties of three - cluster states with equal sized but inequivalent clusters , we find in section [ sec:3cluster ] a new type of robust heteroclinic attractor for @xmath2 oscillators .
our results on the existence of robust connections for these heteroclinic attractors still rely on numerical observation of robust connections - it is a challenge to characterise coupling functions that give rise to such cycles in a more analytical ( or geometric ) manner .
this does not seem to be an easy task , even if one restricts to cycles with one - dimensional unstable manifolds , i.e. between states that have clusterings consisting only of pairings .
[ thm : couplingfunctiongen ] suppose that @xmath185 is such that there is a periodic @xmath20-cluster state @xmath186 with non - trivial clusters of size @xmath187 , such that the clusters are tangentially stable .
then there is a parametrized family of coupling functions @xmath131 and with real parameter @xmath152 and parameter values @xmath188 such that for all values of the parameter @xmath153 a nearby cluster state exists , is still tangentially stable and moreover : * proof : * the proof is similar to that of corollary [ cor : couplingfunction ] ; we make use of the fact that the transverse exponent of the @xmath27th cluster can be written in the form @xmath191.\ ] ] the nontrivial assumption of the clusters mean one can choose a compactly supported perturbation @xmath142 with @xmath146 and @xmath147 so that @xmath192 , @xmath193 and @xmath194 are independent of @xmath153 while @xmath195 . | in this paper we examine robust clustering behaviour with multiple nontrivial clusters for identically and globally coupled phase oscillators .
these systems are such that the dynamics is completely determined by the number of oscillators @xmath0 and a single scalar function @xmath1 ( the coupling function ) .
previous work has shown that ( a ) any clustering can stably appear via choice of a suitable coupling function and ( b ) open sets of coupling functions can generate heteroclinic network attractors between cluster states of saddle type , though there seem to be no examples where saddles with more than two nontrivial clusters are involved . in this work
we clarify the relationship between the coupling function and the dynamics .
we focus on cases where the clusters are inequivalent in the sense of not being related by a temporal symmetry , and demonstrate that there are coupling functions that give robust heteroclinic networks between periodic states involving three or more nontrivial clusters .
we consider an example for @xmath2 oscillators where the clustering is into three inequivalent clusters .
we also discuss some aspects of the bifurcation structure for periodic multi - cluster states and show that the transverse stability of inequivalent clusters can , to a large extent , be varied independently of the tangential stability . |
instantons are believed to be an essential component of the long - distance physics of the qcd vacuum , and the lattice provides a unique opportunity to gain insight into their role . in this study
, we will for the first time quantitatively examine the extent to which instanton degrees of freedom , contained within standard monte - carlo generated gauge - field configurations , can maintain the characteristic features of the mass and renormalisation functions of the non - perturbative quark propagator . in order to isolate the effects of instanton degrees of freedom a uv filter is required to remove topologically non - trivial fluctuations .
a variety of filters have been used , including cooling@xcite , ape smearing@xcite , hyp smearing @xcite and stout link smearing @xcite , among others .
these algorithms can suffer from destruction of the instanton content of the vacuum , and so in this work we use over - improved stout - link smearing @xcite , a form of smearing tuned to preserve instantons .
section [ sec : smearing ] briefly describes these smearing methods .
we then seek to quantify the effects of smearing on the lattice gauge fields in section [ sec : smearingeffect ] .
we produce configurations dominated by instanton - like objects , and compare to the phenomenological instanton liquid model @xcite , which models the vacuum as composed of a constant number of instantons and anti - instantons of constant radius .
we will then briefly introduce the fat link irrelevant clover(flic ) overlap action @xcite in section [ sec : overlap ] , an improved fermion action with a lattice deformed version of chiral symmetry , which removes the problem of additive mass renormalisation of the quark propagator .
results will be compared for smeared and unsmeared configurations in section [ sec : results ] , and conclusions summarised in section [ sec : conclusion ] .
it has long been known @xcite that smearing the lattice reveals objects which approximate classical instantons .
however , it has been known for nearly as long that excessive smearing can destroy the same instanton - like objects or distort their structure .
we can understand this behaviour by explicitly considering the single instanton solution , @xcite @xmath0 where @xmath1 is the instanton radius , @xmath2 the pauli matrices and @xmath3 the center of the instanton . the wilson gauge action can be expanded in powers of @xmath4 as@xcite @xmath5\mathrm { , } \nonumber \\ \quad \end{aligned}\ ] ] where @xmath6 $ ] for arbitrary @xmath7 . inserting the instanton solution of eq .
( [ instsol ] ) into eq .
( [ wilsonactexp ] ) , we acquire @xmath8.\ ] ] the source of the problem is clear ; although at first order this is equal to the continuum instanton action , the leading order error term is strictly negative and inversely proportional to @xmath1 .
the use of a plaquette - based smearing algorithm will reduce the size of instantons in decreasing the action and ultimately enable lattice artifacts to spoil and remove the instanton .
this situation is not ameliorated by using improved actions such as the symanzik action @xcite this issue can be mitigated using over - improved stout - link smearing , a form of smearing designed to preserve instantons .
this was first implemented in ref .
@xcite , although here we follow the work ref .
@xcite , which uses a slightly modified combination of links .
one introduces a new parameter , @xmath9 , defining @xmath10,\ ] ] with @xmath11 and @xmath12 the @xmath13 and @xmath14 plaquettes .
we note that a value of @xmath15 gives the wilson plaquette action and @xmath16 gives the symanzik improved action @xcite . again substituting the instanton solution into this
, one acquires @xcite @xmath17.\ ] ] a negative value of @xmath9 will preserve instanton - like objects by making the first order error term positive .
this has however , simultaneously introduced a new problem ; we have removed the possibility of shrinking objects and replaced it with that of enlarging , and so smearing can distort the topological structure of the lattice if used excessively with large negative values of @xmath9 . following ref .
@xcite , we adopt the small value @xmath18 .
explicitly , a sweep of over - improved stout - link smearing is implemented by replacing all links on the lattice with a smeared link , defined by @xmath19 where @xmath20 with @xmath21 then we define @xmath22,\end{aligned}\ ] ] where @xmath1 is a parameter controlling the level of smearing and @xmath23 are staples ; the 3 links in the @xmath24 plane forming the @xmath11 plaquette with @xmath25 removed , and @xmath26 are analogously defined for @xmath27 rectangles , illustrated in fig .
[ pic : stapledef ] . and
@xmath26 ] following the work in ref .
( @xcite ) , we choose @xmath29 and @xmath18 , as these provide values of @xmath30 close to 1 , and preserve instanton - like objects on the lattice with size above the dislocation threshold of @xmath31 .
we note however , that small objects can still be destroyed by the smearing process .
pair annihilation can also remove them from the lattice .
we wish to find a smearing level such that the configurations are dominated by topological objects as similar to continuum instantons as possible . at the same time
, one needs to be wary of distorting their topological structure by enlarging or pair annihilating these objects .
we will quantify the effects of smearing in order to choose an optimal balance between these two effects . the work in ref .
@xcite suggests that after just 20 sweeps of over - improved stout link smearing , topological objects found on the lattice closely approximate instantons . here
we adopt a similar approach to that taken in ref .
@xcite , searching the lattice for sites which are local maxima of the action in their surrounding hypercube @xcite .
these are then taken as the approximate centre of an ( anti-)instanton , around which we fit the classical instanton action density , @xmath32 where @xmath33 , @xmath1 and @xmath3 are fit parameters , noting that @xmath3 is not restricted to lattice sites .
the parameter @xmath33 is introduced as lattice topological objects are expected to have a higher action than classical instantons .
we wish to determine the fit by using the shape of the action density around a local maximum , rather than the height .
we can then compare data obtained for the radii , @xmath1 , of instanton candidates from this to the relationship between the radius and topological charge at the centre of an ( anti-)instanton , @xmath34 where @xmath35 for an ( anti-)instanton .
@xmath36 at the fitted values of @xmath3 are found using linear interpolation from neighbouring hypercubes to find an extremum inside the hypercube containing @xmath3 .
this data will provide the basis for our investigation of the effects of smearing .
simulations are performed on @xmath37 dynamical flic @xmath38 configurations , with a lattice spacing of @xmath39 fm , corresponding to a spatial extent of @xmath40 fm .
periodic boundary conditions are used .
up to 300 sweeps of smearing are investigated . to calculate the action and topological charge densities used in our fits we use the @xmath41 five - loop improved action and charge densities defined in ref .
@xcite it should be noted that at low levels of smearing we expect to fit a large number of false positives ; local maxima of the action corresponding to noise . the degree
to which fitted results concur with eq .
( [ topcharge ] ) will thus be a key first test .
this correspondence is graphed for a single configuration in fig .
[ fig : rhovsq ] for various smearing levels .
the number of instanton candidates starts out large and distributed fairly evenly around sizes of 2 - 8 lattice units , with little correlation to the predicted charge lines of eq .
( [ topcharge ] ) .
this quickly changes as the number of smearing sweeps increases , eventually leading to a very close fit .
the number of instanton candidates also drops off rapidly at first , then steadily decreases . by the 50 sweep mark
we can be confident that almost all local maxima found closely approximate an instanton near the centre .
notably , instanton candidates with low radii have systematically lower topological charge at the centre than predicted by eq .
[ topcharge ] .
this is due to their proximity to the dislocation threshold of @xmath42 , under which objects are shrunk .
we define the squared distance from the theoretical relationship of eq .
( [ topcharge ] ) , @xmath43 , as the minimum distance of each point from the theoretical relationship , i.e. for an instanton candidate with radius @xmath44 and topological charge at the centre @xmath45 , @xmath46\mathrm{.}\ ] ] the average value of this is plotted for data from 10 configurations in fig . [ fig : dist ] , as a function of the number of sweeps of overimproved smearing , @xmath47 .
this confirms our earlier observations ; before 50 sweeps , @xmath43 decreases very rapidly as noise is removed from the lattice .
after 50 sweeps , the decrease is characteristically slower .
we have seen in eq .
( [ eq : slsmac ] ) that smearing carries the risk of distorting the vacuum structure by enlarging topological objects .
this concern is confirmed in fig .
[ fig : radius ] , illustrating the average radius of instanton candidates as a function of smearing . after an initial rapid drop , attributable to false positives being rapidly removed by the smearing algorithm
, there is a small but steady increase in the average radius of the instanton candidates found . in fig .
[ fig : number ] , we plot the average density of instanton candidates , @xmath48 , where @xmath49 is the number of instanton candidates found on a configuration and @xmath50 the lattice volume .
this shows a rapid decrease until around 70 sweeps in the early stages of smearing as false positives are rapidly eliminated .
eventually , instanton / anti - instanton annihilation becomes the dominant factor , slowing as the population becomes sparse .
again , the regime of 50 sweeps characterises this transition . also of note
is the density of @xmath51 at this point . in the instanton liquid model
, phenomenological constraints set the instanton radius to be around @xmath52 and density to @xmath53 , leading to a packing fraction , the proportion of the vacuum composed of instantons , of @xmath54 .
here , @xmath55 is the 4-volume of an instanton of radius @xmath1 . on the lattice
, we have found a density of between @xmath56 and @xmath57 , reaching an equivalent density to the instanton liquid model at approximately 160 sweeps of smearing . in the second half of this study , we use configurations with less than 100 sweeps of smearing , leading to a density much higher than used in the instanton liquid model . , of topological charge density at the centre of ( anti-)instantons is explored as a function of smearing sweeps , @xmath58 .
plotted are the fraction of lattice sites within @xmath59 ( upper ) and @xmath60 ( lower ) of the fitted radius of an instanton candidates centre with topological charge of the same sign.,title="fig : " ] , of topological charge density at the centre of ( anti-)instantons is explored as a function of smearing sweeps , @xmath58 .
plotted are the fraction of lattice sites within @xmath59 ( upper ) and @xmath60 ( lower ) of the fitted radius of an instanton candidates centre with topological charge of the same sign.,title="fig : " ] in order to investigate this further , we have plotted the topological charge coherence , the proportion of lattice sites with topological charge of the same sign as at the centre , within @xmath59 , @xmath61 , and @xmath60 , @xmath62 , of the fitted instanton radius , in fig .
[ fig : cohere ] . at @xmath59 of the instanton radius ,
almost all sites are charge coherent after a relatively small number of smearing sweeps , whereas at @xmath60 of instanton radius , only @xmath63 of the sites are charge coherent , even after a large amount of smearing .
this suggests that our fitted values of @xmath1 may be over - estimating the true radius of instantons .
we thus calculate the packing fraction using the topological charge found at the centre and eq .
( [ topcharge ] ) .
this is plotted in fig .
[ fig : packing ] .
we note that a packing fraction exceeding 1 is not an issue , as instanton candidates may overlap , particularly at low levels of smearing .
we find a packing fraction of between @xmath64 and @xmath65 . once again , this is notably higher than the instanton liquid model value of @xmath66 , suggesting that we have found a far `` fuller '' vacuum on the lattice .
aggregating these measures , it is clear that between 50 and 100 sweeps of smearing gives an optimal balance , whereby the gauge field is dominated by instantons without excessive distortion .
we will thus calculate quark propagators at 0 , 30 , 50 , 80 and 100 sweeps of smearing .
the overlap fermion operator @xcite is defined in the massless case by @xmath67 with @xmath68 the matrix sign function , and the overlap kernel , @xmath69 , any reasonable hermitian dirac operator with mass parameter @xmath70 , governing the resolution of toplogically non - trivial field structures @xcite .
the overlap is an explicit solution of the ginsparg - wilson relation @xcite , @xmath71 and will thus have a lattice - deformed version of chiral symmetry .
this is sufficient to prevent additive quark mass renormalisation @xcite , greatly simplifying propagator analysis @xcite .
the massive overlap operator is then given by @xmath72 where the overlap mass parameter @xmath73 is defined to represent a bare quark mass @xmath74 we use the flic action @xcite as the overlap kernel , as studies have shown it to have superior spectral properties , accelerating calculation of the overlap operator , and redcued lattice discretisation errors @xcite .
we project low modes of the kernel and calculate their contribution to the propagator explicitly , greatly reducing the condition number of the matrix square root .
notably , we have found results consistent with the work of neuberger @xcite , who predicts that in the presence of a sufficiently smooth background field , the eigenvalues of the wilson kernel are maximally displaced from @xmath75 at @xmath76 . in fig .
[ eigenflow ] , we have plotted the 50 lowest lying eigenvalues of the flic kernel on a configuration with 50 sweeps of smearing and we observe that the region where the spectrum becomes dense(i.e . ignoring the isolated low - lying topological nodes )
is indeed maximally separated from 0 near @xmath76 .
we choose to perform all calculations at @xmath76 .
we consider nine values of @xmath73 , given in table [ mutable ] .
.the nine values of @xmath73 considered , with the corresponding bare mass in physical units [ cols="^,^",options="header " , ] in order to define an overlap quark propagator , we naturally wish to preserve the most important properties of the continuum propagator . as a simple consequence of chiral symmetry , in the massless case the continuum quark propagator obeys @xmath77
we also note that eq .
( [ gwrel ] ) implies @xmath78 suggesting an appropriate form for the overlap propagator , @xmath79 it is also important to maintain the correct continuum limit , and so given @xmath80 it is then natural to define the massless bare overlap propagator as @xcite @xmath81 we can then construct the bare massive overlap propagator simply by adding a bare mass to the inverse of eq .
( [ eq : baremassoverlap ] ) , @xmath82 then recalling the definition of the overlap mass parameter in eq .
( [ eq : mudef ] ) , we have @xmath83 due to the lack of additive mass renormalisation , the quark propagator on the lattice will have as its general form @xmath84 with @xmath85 the non - perturbative mass function and @xmath86 containing all renormalisation information .
this can be defined as having the inverse @xmath87 where @xmath88 is defined@xcite by the tree level propagator , calculated with @xmath89 , @xmath90 comparing eq .
( [ propingenform ] ) to eq .
( [ propgenform ] ) , we see @xmath91 multiplying eq .
( [ propingenform ] ) by @xmath88 and taking the trace provides @xmath92 to determine @xmath93 and @xmath94 we follow ref .
@xcite and define @xmath95 and @xmath96 by @xmath97 such that @xmath98,\end{aligned}\ ] ] where @xmath99 and @xmath100 are the extent of spin and colour indices . comparing to eq .
( [ propingenform ] ) , we find @xmath101 defining @xmath102 we thus have @xmath103
we fix to landau gauge by maximizing the @xmath104 improved gauge fixing functional@xcite @xmath105 using a fourier transform accelerated algorithm@xcite .
to avoid gribov copy issues , we first gauge fix configurations with 100 sweeps of smearing and then use these as a preconditioner for the same configurations with lower levels of smearing@xcite . the matrix sign function is calculated using the zolotarev rational polynomial approximation @xcite .
we average data over spatial symmetries and choose @xmath106 as the renormalization point for @xmath107 , with @xmath108 .
we apply a cylinder cut @xcite to data with a radius of @xmath109 , and estimate errors using a second - order single - elimination jack - knife method .
@xmath107 is plotted in fig .
[ z1271 ] for the lightest quark mass , @xmath110 , for various levels of smearing .
this reveals the well - known shape of the renormalization function for the unsmeared case ; dipping in the infrared region and rising in the ultraviolet limit .
after just 30 sweeps of smearing however , different behaviour is seen .
the renormalisation function is more tree like , dipping half the amount and approximating @xmath111 for @xmath112 . the renormalisation function rises earlier to this plateau value , with similar behaviour seen at all levels of smearing .
this appears to confirm our earlier observations ; after 30 sweeps most short - range behaviour is removed , and the gauge field structure of the lattice is much simpler .
the physics responsible for the drop in @xmath107 is removed early as all curves agree . to further show this ,
@xmath107 is plotted as a function of @xmath113 at given values of @xmath114 in fig .
[ z1271indq ] . we have chosen to plot against @xmath113 as this is proportional to the smearing radius .
this shows a large change from 0 to 30 sweeps of smearing , affirming that the physics responsible for the drop in @xmath107 is removed early in the smearing process , after which smearing has relatively little effect .
higher masses are illustrated for all @xmath114 in fig .
[ fig : zhigh ] . remarkably little sensitivity to the quark mass is observed .
we note that the lowest two @xmath114 values are purely timelike , and thus have comparatively large error bars due to the lack of symmetries for averaging . [ h ! ! ] we now plot the mass function for a quark mass of @xmath110 in fig .
[ m1271thirty ] .
the unsmeared data reveals the expected shape ; a large effective quark mass in the infrared region created through dynamical mass generation , tapering to an approximate plateau in the ultraviolet region , where logarithmic corrections produce a running quark mass higher than the input bare mass , illustrated by the dot - dash line .
looking at the lowest level of smearing considered , 30 sweeps , we note that the smeared results consistently have smaller error bars , particularly in the ultra - violet region , indicating short distance physics is a significant source of noise .
although the instanton content of each ensemble is different , its impact on the mass function remains similar .
we also note that , as expected , in the ultraviolet region the running quark mass is now barely higher than the input bare mass , due to the spoiling of short distance perturbative physics at this range . at the lowest momenta
, we find perfect agreement with the unsmeared case , but as we increase momentum there is some suppression of dynamically generated mass . we attribute this to the destruction of some topological objects by the smearing algorithm , both those smaller than the dislocation threshold and those eliminated by pair annihilation .
smearing also has the effect of enlarging objects , again removing dynamically generated mass from this region .
indeed one anticipates a significant smearing dependence in the regime of @xmath115 .
we now plot all levels of smearing , in fig . [
m1271hundred ] . even at high levels of smearing , we have maintained the qualitative shape of the mass function , although increasing smearing results in a larger loss of dynamically generated mass . as the level of smearing increases , the mass function becomes flat for lower values of momenta .
this is in accord with the increase of the smearing radius , the distance within which physics is suppressed .
the most dramatic shift is between the unsmeared case and 30 sweeps .
this suggests that , to at least some extent , smearing has changed the mass renormalization , and so in order to acquire results with the same physical mass we must shift our value of @xmath73 for the smeared results . in fig .
[ mshifted ] we have plotted smeared results at a value of @xmath116 , chosen in order to match unsmeared results in the ultraviolet limit .
this reveals that not all loss of dynamical mass generation is due to a changing mass renormalization .
although we have retained the majority of long - range physics , it is clear that smearing , particularly after a large number of sweeps , has removed some important aspects of vacuum structure , creating a gap between unsmeared and smeared results .
@xmath85 is plotted as a function of @xmath113 at fixed values of @xmath114 in fig .
[ m1694indq ] , for a quark mass of @xmath117 .
this shows an approximately linear dependence on @xmath113 and thus the smearing radius .
this suggests that the increase in smearing radius is responsible for loss of dynamical mass generation by removing topological objects from the lattice . at higher momenta corresponding to smaller distances
, smearing has removed all relevant physics from the lattice early , and so an increase in the smearing radius has a smaller effect .
higher masses are shown in fig .
[ fig : mhigh ] for all @xmath114 , using the same bare mass for both smeared and unsmeared configurations .
these show broadly similar results at all masses , though the gap between levels of smearing becomes clearer for larger masses , which should be more sensitive to the disruption of short distance physics .
this reinforces the necessity of maintaining a high density of topological objects in the qcd vacuum and reinforcing the danger of destroying topological objects with excess smearing .
the largest gap , however , remains between no smearing and 30 sweeps .
we have used over - improved stout - link smearing , to reveal an underlying gauge field structure resembling an instanton liquid .
after around 50 sweeps we can be confident the lattice is dominated by instanton - like objects .
our calculations of the non - perturbative mass function on smeared configurations reveals that it retains its qualitative shape at even high levels of smearing .
there is some loss of dynamical mass generation , increasing with smearing , which can be attributed to a thinning of the vacuum through the destruction of instanton / anti - instanton pairs by the smearing algorithm .
regardless , we have shown that a gauge configuration consisting solely of instanton - like objects can accurately reproduce the majority of the long - range behaviour of the quark propagator , and thus conclude that instantons are the primary mechanism responsible for the dynamical generation of mass .
we wish to thank craig roberts for discussions inspiring this quantitative evaluation of the role of qcd instantons in the quark propagator .
we also thank peter moran for his input .
this research was undertaken with the assistance of resources at the nci national facility in canberra , australia , and the ivec facilities at murdoch university ( ivec@murdoch ) and the university of western australia ( ivec@uwa ) .
these resources were provided through the national computational merit allocation scheme , supported by the australian government .
this research is supported by the australian research council .
m. albanese _ et al . _
[ ape collaboration ] , phys .
b * 192 * , 163 ( 1987 ) .
a. hasenfratz and f. knechtli , phys .
d * 64 * , 034504 ( 2001 ) [ hep - lat/0103029 ] . c. morningstar and m. j. peardon , phys .
d * 69 * , 054501 ( 2004 ) [ hep - lat/0311018 ] .
m. garcia perez , a. gonzalez - arroyo , j. r. snippe and p. van baal , nucl .
b * 413 * , 535 ( 1994 ) [ hep - lat/9309009 ] .
p. j. moran and d. b. leinweber , phys .
d * 77 * , 094501 ( 2008 ) [ arxiv:0801.1165 [ hep - lat ] ] .
e. v. shuryak , nucl .
b * 203 * , 93 ( 1982 ) .
e. v. shuryak , nucl .
b * 203 * , 116 ( 1982 ) .
e. v. shuryak , nucl .
b * 203 * , 140 ( 1982 ) .
w. kamleh , d. h. adams , d. b. leinweber and a. g. williams , phys .
d * 66 * , 014501 ( 2002 ) [ hep - lat/0112041 ] .
a. a. belavin , a. m. polyakov , a. s. schwartz and y. .s .
tyupkin , phys .
b * 59 * , 85 ( 1975 ) .
d. -j .
kusterer , j. hedditch , w. kamleh , d. b. leinweber and a. g. williams , nucl .
b * 628 * , 253 ( 2002 ) [ hep - lat/0111029 ] .
g. t hooft , phys .
lett . * 37 * , 8 ( 1976 ) .
g. t hooft , phys .
d * 14 * , 3432 ( 1976 ) [ erratum - ibid .
d * 18 * , 2199 ( 1978 ) ] .
k. symanzik , nucl .
b * 226 * , 187 ( 1983 ) .
p. j. moran and d. b. leinweber , phys .
d * 78 * , 054506 ( 2008 ) [ arxiv:0801.2016 [ hep - lat ] ] .
s. o. bilson - thompson , f. d. r. bonnet , d. b. leinweber and a. g. williams , nucl .
* 109a * , 116 ( 2002 ) [ hep - lat/0112034 ] .
r. narayanan and h. neuberger , phys .
b * 302 * , 62 ( 1993 ) [ hep - lat/9212019 ] .
r. narayanan and h. neuberger , nucl .
b * 412 * , 574 ( 1994 ) [ hep - lat/9307006 ] .
r. narayanan and h. neuberger , phys .
lett . * 71 * , 3251 ( 1993 ) [ hep - lat/9308011 ] .
r. narayanan and h. neuberger , nucl .
b * 443 * , 305 ( 1995 ) [ hep - th/9411108 ] .
p. j. moran , d. b. leinweber and j. zhang , phys .
b * 695 * , 337 ( 2011 ) [ arxiv:1007.0854 [ hep - lat ] ] .
p. h. ginsparg and k. g. wilson , phys .
d * 25 * , 2649 ( 1982 ) .
hasenfratz , nucl .
b * 525 * , 401 ( 1998 ) [ hep - lat/9802007 ] .
j. skullerud , d. b. leinweber and a. g. williams , phys .
d * 64 * , 074508 ( 2001 ) [ hep - lat/0102013 ]
. f. d. r. bonnet _ et al . _ [ cssm lattice collaboration ] , phys .
d * 65 * , 114503 ( 2002 ) [ hep - lat/0202003 ] .
j. i. skullerud and a. g. williams , phys .
d * 63 * , 054508 ( 2001 ) [ hep - lat/0007028 ] .
j. m. zanotti _ et al . _ [ cssm lattice collaboration ] , phys .
d * 65 * , 074507 ( 2002 ) [ hep - lat/0110216 ] .
w. kamleh , p. o. bowman , d. b. leinweber , a. g. williams and j. zhang , phys .
d * 71 * , 094507 ( 2005 ) [ hep - lat/0412022 ] .
h. neuberger , phys .
d * 61 * , 085015 ( 2000 ) [ hep - lat/9911004 ] .
r. g. edwards , u. m. heller and r. narayanan , phys . rev .
d * 59 * , 094510 ( 1999 ) [ hep - lat/9811030 ] .
c. t. h. davies et al .
d * 37 * , 1581 ( 1988 ) d. s. roberts , w. kamleh and d. b. leinweber , pos lattice * 2012 * , 261 ( 2012 ) .
d. s. roberts , p. o. bowman , w. kamleh and d. b. leinweber , phys .
d * 83 * , 094504 ( 2011 ) [ arxiv:1011.1975 [ hep - lat ] ] .
f. d. r. bonnet , p. o. bowman , d. b. leinweber , a. g. williams and d. g. richards , austral .
j. phys .
* 52 * , 939 ( 1999 ) [ hep - lat/9905006 ] .
j. e. hetrick and p. .de forcrand , nucl .
suppl . * 63 * , 838 ( 1998 ) [ hep - lat/9710003 ] .
chiu , t .- h .
hsieh , c .- h . huang and t .- r .
huang , phys .
d * 66 * , 114502 ( 2002 ) [ hep - lat/0206007 ] | we quantitatively examine the extent to which instanton degress of freedom , contained within standard monte - carlo generated gauge - field configurations , can maintain the characteristic features of the mass and renormalisation functions of the non - perturbative quark propagator .
we use over - improved stout - link smearing to isolate instanton effects on the lattice . using a variety of measures ,
we illustrate how gauge fields consisting almost solely of instanton - like objects are produced after only 50 sweeps of smearing .
we find a full vacuum , with a packing fraction more than three times larger than phenomenological models predict .
we calculate the overlap quark propagator on these smeared configurations , and find that even at high levels of smearing the majority of the characteristic features of the propagator are reproduced .
we thus conclude that instantons contained within standard monte - carlo generated gauge - field configurations are the degrees of freedom responsible for the dynamical generation of mass observed in lattice qcd . |
in [ k ] khovanov introduced a homology theory for links in @xmath0 that was a categorification of the jones polynomial . in [ aps ] asaeda , przytycki and sikora extended this theory to links embedded in i - bundles .
their homology theory incorporated some of the topology of the i - bundle into their invariant .
turner and turaev showed in [ t ] that the homology from [ aps ] could be recreated using embedded surfaces as elements of the chain groups instead of decorated diagrams . in this paper
we accomplish that in the case of i - bundles over orientable surfaces .
section 2 contains definitions and explains the skein relations on surfaces that are used .
section 3 defines the grading on the chain groups and which surfaces generate the chain groups .
the boundary operator is defined in section 4 and it is also shown that it is well - defined with respect to the relations . in section 5
it is proved that the boundary operator squared is equal to zero , and thus the boundary operator together with the chain groups form a chain complex . finally , in section 6 it is shown that the homology produced from the chain complex coincides with the homology from [ aps ]
let @xmath1 be a surface properly embedded in a 3-manifold @xmath2 . a boundary circle of @xmath1 is said to be * inessential *
if it bounds a disk in @xmath2 , otherwise it is said to be * essential*. if @xmath1 is an oriented surface and @xmath3 is an oriented boundary component of @xmath1 then the orientation of @xmath1 is * compatible * with the orientation of @xmath3 if the boundary orientation of @xmath3 from @xmath1 agrees with the orientation of @xmath3 .
two oriented boundary curves of an orientable connected surface are * compatible * if both curves are compatible with the same orientation on the surface .
if @xmath1 is a connected unoriented orientable surface and @xmath3 is an oriented boundary component of @xmath1 then there is exactly one orientation for the other boundary curves to be oriented compatibly with @xmath3 . [ cols="^,^,^,^,^ " , ] note that under @xmath4 the associated state is nt affected , thus for example if t @xmath5 tt by changing a smoothing before applying @xmath4 , then after applying @xmath4 the boundary circles behave the same way , and an inessential boundary circle turns into two inessential boundary circles by placing a bridge .
the following 21 items show what @xmath6 is in each of the cases when the boundary circles are affected as in the previous table . 1
. note @xmath7 = . after a bridge
is placed there are two trivial boundary curves in the top .
this has euler characteristic equal to 0 , and thus it is a compressible annulus .
compress the annulus to get two disks , each with a dot .
2 . @xmath7 = .
when a bridge is placed there are two non - trivial boundary components in the top .
this is an incompressible annulus with a dot , so it is trivial in the quotient .
@xmath8 = . after a bridge
is placed there are two trivial boundary curves in the top .
this is a compressible annulus .
compress the annulus to get disk with dot , disk + disk , disk with dot .
4 . @xmath8 = . after a bridge
is placed there are two non - trivial boundary curves in the top .
this is an incompressible annulus , so have unoriented annulus = average of oriented annuli .
@xmath9 = . after a bridge
is placed there is a non - trivial boundary curve in the top and a trivial boundary curve in the top .
compress the neck that is near the trivial boundary curve to get an annulus , oriented same way as the original annulus and a disk with a dot .
@xmath9 = . after a bridge
is placed there are two non - trivial boundary curves on the top .
one can only compress and separate boundary curves if we have at least 4 non - trivial and we only have three , so we have a surface that is trivial in the quotient by lemma [ pairofpants ] 7
. refer to 5 8 .
refer to 6 9 .
@xmath10 = . after a bridge
is placed there is one trivial boundary component .
now we have two dots on the same component , so it is trivial in the quotient . 10 .
@xmath11 = . after a bridge
is placed there is one trivial component .
these two disks combined to make a disk with a dot . 11 .
@xmath12 = . after a bridge
is placed there is one trivial boundary component .
this leaves us with a disk .
@xmath13 = . placing
a bridge would result in a trivial boundary component in the top .
thus the original boundary components must have been parallel .
therefore the bridge falls into the category of ( eo ) since they are oriented the same way .
thus the result is trivial in the quotient
@xmath13 = . placing a bridge results in one non - trivial boundary curve on the top .
thus we have an incompressible pair of pants which is trivial in the quotient .
@xmath14 = . after placing a bridge
there is one trivial boundary component .
thus the original non - trivial curves were homotopic .
compress upon the disk that is present near the trivial curve on top .
this results in a disk on top with a dot and an annulus on the bottom + disk on top with an annulus with a dot on the bottom which is equivalent to just having a disk with a dot in the quotient .
@xmath14 = . after a bridge
is placed there is one non - trivial boundary component . as in 13
, we have an incompressible pair of pants which is trivial in the quotient .
refer to 12 17 .
refer to 13 18 .
@xmath15 = . after a bridge
is placed there is one non - trivial boundary curve on the top .
note bridging to a disk does nt change the annulus , except it adds a dot , which makes the foam trivial in the quotient .
refer to 18 20 .
@xmath16 = . after a bridge
is placed there is a non - trivial boundary component on top . absorbing a disk does nt change annulus , so we get the same annulus with the same orientation .
refer to 20 by examining the list and the table , we can see that @xmath17 in each case .
thus note @xmath18 thus @xmath4 is a chain map , as desired . given a link diagram @xmath19 , @xmath20 @xmath21 by @xmath22 .
let @xmath23 , @xmath24 , @xmath25 be the indices coming from the [ aps ] theory and let @xmath26 be an enhanced kauffman state from [ aps ] . then note clearly @xmath27 since the smoothings stay the same under @xmath4 .
we also have , @xmath24 ( @xmath28 ) @xmath29 + 2 ( # positive trivial circles @xmath30 # negative trivial circles ) @xmath31 similarly , @xmath24 ( @xmath30 ) @xmath32 + 2 ( # positive trivial circles @xmath30 # negative trivial circles ) @xmath33 also ,
@xmath24 ( @xmath280 ) @xmath34 + 2 ( # positive trivial circles @xmath30 # negative trivial circles ) @xmath35an annulus with bottom boundary curve oriented in the positive direction ) finally , @xmath24 ( @xmath300 ) @xmath36 + 2 ( # positive trivial circles @xmath30 # negative trivial circles ) @xmath37an annulus with bottom boundary curve oriented in the negative direction ) then note all non - trivial circles that are present in a smoothing of the diagram appear in the bottom of the foam , so the @xmath25-grading is also preserved under @xmath4
. the proof will proceed by induction on the number of crossings in the diagram .
assume @xmath19 has zero crossings . therefore the boundary maps are all the zero map .
thus the chain groups are also the homology groups .
note @xmath4 is an isomorphism on the chain groups since it takes generators to generators , so it is also an isomorphism on homology in this case .
let @xmath19 be a diagram in @xmath38 , with @xmath39 crossings and inductively assume @xmath22 is an isomorphism for all diagrams with less than @xmath39 crossings .
note we have a relation between the short exact sequences coming from the two theories .
the diagram commutes since @xmath40 and @xmath41 are defined identically and the same is true for @xmath42 and @xmath43 .
0 & & |c_i+1,j+1,s(d _ ) & ^| & |c_i , j , s(d_p ) & ^| & |c_i-1,j-1,s(d_0 ) & & 0 + & & _ & & _ & & _ & & + 0 & & c_i+1,j+1,s(d _ ) & ^ & c_i , j , s(d_p ) & ^ & c_i-1,j-1,s(d_0 ) & & 0 + this induces the long exact sequence : & |h_i+1,j-1,s(d_0 ) & ^ & |h_i+1,j+1,s(d _ ) & ^| _ * & |h_i , j , s(d_p ) & ^| _ * & |h_i-1,j-1,s(d_0 ) & ^ & |h_i-1,j+1,s(d _ ) & & + & & & _ _ * & & _ _ * & & _ _ * & & _ _ * + & h_i+1,j-1,s(d_0 ) & ^ & h_i+1,j+1,s(d _ ) & ^ _ * & h_i , j , s(d_p ) & ^ _ * & h_i-1,j-1,s(d_0 ) & ^ & h_i-1,j+1,s(d _ ) & &
+ all @xmath22 , except the middle one , are isomorphisms by the inductive assumption .
also , the diagram commutes since @xmath4 is a chain map .
note by the five lemma the middle @xmath4 is an isomorphism .
thus by induction given a link diagram @xmath19 , @xmath20 @xmath21 by @xmath22 .
since asaeda , przytycki and sikora proved invariance for the @xmath44 homology and @xmath45 by the previous theorem we obtain , @xmath46 is an invariant under reidemeister moves 2 and 3 and a reidemeister 1 move shifts the indices in a predictable way .
* asaeda , marta m. , jozef h. przytycki and adam s. sikora , * categorification of the kauffman bracket skein module of i - bundles over surfaces * , algebr .
4 ( 2004 ) 1177 - 1210 * bar - natan , dror , * khovanov s homology for tangles and cobordisms * , geom .
9(2005 ) 1443 - 1499 *
bar - natan , dror , * on khovanov s categorification of the jones polynomial * , algebr .
. 2 ( 2002 ) 337 - 370 mr1917056 * khovanov , m. , * a categorification of the jones polynomial * , duke math . j. 101 ( 2000 ) 359 - 426 mr1740682 * manturov , v. , * additional gradings in khovanov homology * , arxiv:0710.3741v1 [ math.gt ] * turaev , vladimir and paul turner , * unoriented topological quantum field theory and link homology * , arxiv : math/0506229v2 * viro , oleg , * remarks on definition of khovanov homology * , arxiv : math/0202199v1 [ math.gt ] | this paper introduces a homology theory for links in i - bundles over an orientable surface .
the theory is unique in that the elements of the chain groups are surfaces instead of diagrams .
it is then shown this theory yields the same results as the homology theory constructed by asaeda , przytycki and sikora . |
graph reconstruction is an old and extensive research topic .
it dates back to the _ reconstruction conjecture _ raised by kelly and ulam in 1941 ( see @xcite ) , which asserts that every graph on at least three vertices is uniquely determined by its collection of vertex deleted subgraphs . as a natural extension of the reconstruction conjecture ,
numerous papers considered either reconstruction of structures other then graphs ( a research topic proposed by ulam in 1960 ) , or reconstructions of graphs from other information . in the first direction , reconstructed objects include colored graphs , hypergraphs , matroids , relations , and other classes . in the second direction ,
the `` information '' may be @xmath8-vertex deleted subgraphs , edge - deleted subgraphs , elementary contractions , spanning trees , etc . in addition , various papers considered reconstruction of _ parameters of the graph _ instead of its full structure .
such parameters include the order , the degree sequence , planarity , the types of spanning trees , and many others ( see the surveys @xcite for references ) . in this paper , we study the problem of reconstructing the geometric structure of a set of points in the plane from its geometric tree graph . _
tree graphs _ were defined in 1966 by cummins @xcite in the context of listing all spanning trees of a given connected graph effectively .
the _ tree graph _ @xmath9 of a graph @xmath10 has the spanning trees of @xmath10 as its vertices , and two spanning trees are adjacent if one can be obtained from the other by deleting an edge and adding another edge .
these graphs were studied in a number of papers and were shown to be hamiltonian and to have the maximal possible connectivity ( see , e.g. , @xcite ) . in 1996 ,
avis and fukuda @xcite defined the _ geometric tree graph _ , as the counterpart of tree graphs in the geometric graph setting .
let @xmath0 be a finite point set in general position in the plane .
the _ geometric tree graph _
@xmath3 is defined as follows .
the vertices of @xmath3 are the simple ( i.e. , non - crossing ) spanning trees ( ssts ) of @xmath1 .
two such vertices are adjacent in @xmath3 if they differ in exactly two edges , i.e. , if one can be obtained from the other by deleting an edge and adding another edge .
geometric tree graphs were shown to be connected @xcite , and upper and lower bounds on their diameter were established @xcite .
we study a reconstruction problem for geometric graphs : is the geometric tree graph @xmath3 sufficient for `` reconstructing '' the structure of @xmath1 ? in a sense , this question is a geometric counterpart of the work of sedlek @xcite , who studied the question whether a graph can be reconstructed from its spanning trees .
as we deal with a geometric setting , we seek to reconstruct the _ geometric structure _ of the graph .
let @xmath0 be a finite set of points in general position in the plane .
the _ geometric structure _ of the complete graph @xmath1 as a geometric graph includes , for any pair @xmath2,[c , d]$ ] of vertex - disjoint edges , the information whether they cross or not .
our main result is the following : [ thm : main ] for any finite set @xmath0 of points in general position in the plane , the geometric structure of @xmath1 can be reconstructed from the geometric tree graph @xmath3 . while the proof of the theorem is elementary , it is rather complex , and consists of several stages : 1 . * maximal cliques in @xmath3 .
* we study thoroughly the structure of maximal cliques in @xmath3 .
we divide these cliques into two types , called `` union max - cliques '' and `` intersection max - cliques '' , and show that given a maximal clique in @xmath3 , one can determine its type .
this study spans section [ sec : max - cliques ] .
* stars and brushes in @xmath3 . *
we show how to identify the vertices of @xmath3 that correspond to spanning stars and spanning brushes ( i.e. , spanning trees of diameter 3 with a single internal edge ) , by examining the max - cliques to which they belong .
the stars are determined only up to an automorphism of @xmath1 ( obviously , one can not do better ) , and once they are fixed , the brushes are determined uniquely .
this part of the proof is presented in section [ sec : stars - and - brushes ] .
* the geometric structure of @xmath1 .
* we show how the geometric structure of @xmath1 can be derived from information on the brushes in @xmath3 .
this part is presented in section [ sec : geom - structure ] . in the last part of the paper , section [ sec : general ]
, we consider abstract ( i.e. , non - geometric ) graphs , and show that a variant of the argument developed in sections [ sec : max - cliques ] and [ sec : stars - and - brushes ] can be used to prove the following result : [ thm : main - general ] for any @xmath11 , the automorphism group of the tree graph of @xmath12 is isomorphic to @xmath13 .
our treatment of the geometric reconstruction problem ( i.e. , @xmath1 from @xmath3 ) falls short of this .
it leaves open the ( quite implausible ) possibility that the geometric tree graph @xmath3 has an automorphism @xmath14 , other than the identity , that fixes each star and each brush .
this leaves open , for further research , the following question .
is this true that for any finite set @xmath0 of points in general position in the plane , we have @xmath15 , where @xmath3 is treated as an abstract graph , whereas @xmath1 is treated as a geometric graph ?
in this section we study the structure of maximal ( with respect to inclusion ) cliques in the geometric tree graph @xmath3 .
we divide the maximal cliques into two types , called u - cliques and i - cliques , and our ultimate goal is to determine , given a maximal clique in @xmath3 , what is its type .
we start in section [ sec : sub : notations ] with a few definitions and notations , to be used throughout the paper . in sections
[ sec : sub : max - cliques - basics ] and [ sec : sub : geom - meaning ] we describe a classification of the maximal cliques into two types , presented originally in @xcite , and discuss basic properties of both types .
in order to distinguish between general combinatorial considerations and geometric arguments specific to ssts , we start in section [ sec : sub : max - cliques - basics ] with a general combinatorial framework , and leave the geometric arguments to section [ sec : sub : geom - meaning ] . in sections [ sec : sub : char - degenerate ] and [ sec : sub : identify - degenerate ] we study _ degenerate _ maximal cliques , i.e. , maximal cliques of size 2 . in section [ sec : sub : char - degenerate ] we give a geometric characterization of the situation when a maximal clique is degenerate , and in section [ sec : sub : identify - degenerate ] we show how to identify whether a given degenerate maximal clique is a u - clique or an i - clique . finally , in section [
sec : sub : identify - non - degenerate ] we show how to determine whether a given _ non - degenerate _
maximal clique is a u - clique or an i - clique .
the following notations and conventions are used throughout the paper . the straight line that passes through points @xmath16
is denoted by @xmath17 .
the vertex and edge sets of a graph @xmath10 are denoted by @xmath18 and @xmath19 , respectively .
since the set @xmath0 is fixed , we shall always identify a spanning subgraph of @xmath1 ( and , in particular , a spanning subtree ) with its set of edges .
in our study we shall extensively use _ maximal cliques _ of @xmath3 .
these are defined as follows : a _ max - clique _ in a graph @xmath10 is a maximal ( with respect to inclusion ) clique included in @xmath10 . since any max - clique is a complete graph on its vertex set , we shall identify a max - clique with its set of vertices .
we shall use the following observation on the structure of @xmath3 , proved by avis and fukuda @xcite .
[ claim : g(p)-connected ] for any set @xmath0 of points in general position in the plane , @xmath3 is connected and its diameter is @xmath20 . a generic combinatorial way to treat a tree graph is to consider a base set @xmath21 and a graph @xmath10 whose vertices are @xmath22-subsets of @xmath21 ( not necessarily all the @xmath22-subsets ) , such that two vertices @xmath23 are adjacent if and only if @xmath24 .
( in the case of the ( geometric ) tree graph of a point set @xmath0 , we have @xmath25 , @xmath26 , and the vertices of @xmath10 are the sets of edges of ( simple ) spanning trees of @xmath1 . )
let @xmath27 be two adjacent vertices of @xmath10 .
denote @xmath28 a third vertex @xmath29 is a common neighbor of @xmath30 and @xmath31 if and only if : 1 .
@xmath32 and @xmath33 . in this case
, @xmath34 is obtained from @xmath35 by adding a single element .
2 . @xmath36 and @xmath37 . in this case
, @xmath34 is obtained from @xmath38 by removing a single element .
( it is easy to see that in any other case , either @xmath39 or @xmath40 . ) if @xmath41 are both common neighbors of @xmath30 and @xmath31 , such that @xmath34 satisfies ( 1 ) and @xmath42 satisfies ( 2 ) , then clearly , @xmath43 .
hence , if two common neighbors of @xmath30 and @xmath31 are themselves neighbors , then either both satisfy ( 1 ) or both satisfy ( 2 ) . on the other hand ,
it is clear that any two common neighbors satisfying ( 1 ) are themselves neighbors , and the same holds for ( 2 ) .
thus , any pair @xmath44 of adjacent vertices of @xmath10 is included in at most two max - cliques : 1 .
@xmath45 , and 2 .
@xmath46 .
for any two elements @xmath47 , the union @xmath48 is constant ( and equal to @xmath38 ) .
likewise , for any two elements @xmath49 , the intersection @xmath50 is constant ( and equal to @xmath35 ) .
this is the motivation behind the following definition .
a max - clique of the first type will be called a union max - clique , or a _ u - clique _ , and a max - clique of the second type will be called an intersection max - clique or an _ i - clique_. [ remark : three - determine ] it is clear that given two vertices @xmath44 of a max - clique @xmath34 , we can not determine whether @xmath34 is a u - clique or an i - clique .
however , if we are given a third vertex @xmath51 , we can determine the type , according to which one of the equalities @xmath52 , @xmath53 holds .
moreover , once we determine that @xmath34 is , say , an i - clique , we know that @xmath54 as this is the unique i - clique that includes @xmath30 and @xmath31 . hence , three vertices of a max - clique determine it uniquely .
if @xmath55 , we say that @xmath56 is a _ degenerate u - clique_. similarly , if @xmath57 , we say that @xmath56 is a _ degenerate i - clique_. for a pair of adjacent vertices @xmath44 , there are four possible situations : 1 .
@xmath58 and @xmath59 . in this case , there are exactly two max - cliques that contain @xmath30 and @xmath31 .
2 . @xmath58 and @xmath60 . in this case
, there is a unique max - clique that contains @xmath30 and @xmath31 , namely @xmath61 .
( the i - clique that contains @xmath30 and @xmath31 is degenerate . ) 3 . the same as ( 2 ) , with the roles of @xmath61 and @xmath62 interchanged .
4 . @xmath63 . in this case , the set @xmath56 itself is a max - clique ( that is , both a u - clique and an i - clique ) . as we shall see in the sequel , this situation can not occur in our geometric setting .
now we turn to the geometric graph @xmath3 and show additional properties of max - cliques that follow from its geometric structure . in order to make our notation suggestive
, we denote now the two adjacent vertices of @xmath3 by @xmath64 , and the edges in their symmetric difference by @xmath65 and @xmath66 .
in @xmath3 , u - cliques and i - cliques have a geometric meaning
. 1 . * @xmath67 . *
consider the graph @xmath68 .
obviously , it is a connected graph with a unique cycle .
this cycle contains @xmath69 and @xmath70 , and is simple if and only if @xmath69 and @xmath70 do not cross .
( note that @xmath70 can not cross another edge of @xmath71 , as both these edges belong to the sst @xmath72 . ) as shown above , @xmath61 consists of all vertices of @xmath3 that are obtained from @xmath73 by removing a single edge .
since the vertices of @xmath3 are the edge sets of ssts , we can say that removing an edge ( other than @xmath74 ) from @xmath73 results in an element of @xmath67 if and only if the unique cycle of @xmath73 is simple , and the removed edge belongs to that cycle .
note that if @xmath74 cross , then removal of any edge other than @xmath74 from @xmath73 results in a non - simple graph , and thus , the only elements of @xmath67 are @xmath71 and @xmath72 .
* @xmath75 . *
consider the graph @xmath76 .
obviously , it is a simple forest with two connected components . as shown above , @xmath62 consists of all vertices of @xmath3 that are obtained from @xmath77 by adding a single edge .
since the vertices of @xmath3 are the edge sets of ssts , we can say that adding an edge to @xmath77 results in an element of @xmath62 if and only if that edge makes the forest @xmath77 into a simple spanning tree of @xmath1 .
the geometric interpretation allows us to characterize the cases when u - cliques and i - cliques are degenerate .
[ claim : deg - u - clique ] let @xmath64 be ssts such that @xmath78 and @xmath79 .
the u - clique @xmath67 is degenerate if and only if @xmath69 and @xmath70 cross . by the geometric interpretation ,
if @xmath69 and @xmath70 cross then the only vertices of @xmath67 are @xmath71 and @xmath72 , and thus , it is degenerate .
if @xmath69 and @xmath70 do not cross , then @xmath73 is a _ simple _ connected graph on @xmath80 vertices with @xmath80 edges .
( note that @xmath73 is simple since as @xmath64 are ssts , the only edges in @xmath73 that may cross each other are @xmath74 . )
thus , @xmath73 has a cycle of order at least 3 .
removal of any edge from this cycle gives rise to a vertex in @xmath67 .
therefore , in this case @xmath67 is non - degenerate . [
prop : deg - i - clique ] let @xmath0 be a finite set of points in the plane , no three on a line , @xmath81 .
let @xmath64 be ssts of @xmath1 such that @xmath78 and @xmath79 .
the i - clique @xmath75 is degenerate only in the following case : the convex polygon @xmath82 has three consecutive vertices @xmath83 ( i.e. , @xmath84 $ ] and @xmath85 $ ] are edges of @xmath82 ) , such that : 1 .
the triangle @xmath86 contains no other points of @xmath0 .
2 . @xmath87 $ ] and @xmath88 $ ] .
@xmath89 \in t_1
\cap t_2 $ ] . assume that ( 1)(3 ) hold , and let @xmath90 .
@xmath91 is connected , and thus , has an edge @xmath92 that emanates from @xmath93 .
note that as @xmath83 are consecutive vertices of @xmath82 and the triangle @xmath86 contains no other points of @xmath0 , any edge of @xmath1 that emanates from @xmath93 ( other than @xmath69 and @xmath70 ) must cross @xmath89 $ ] .
thus , either @xmath94 , @xmath95 , or @xmath92 crosses @xmath89 $ ] .
the latter is impossible , since @xmath89 \in t_1
\cap t_2 $ ] , and as shown above , @xmath96 is included in any element of @xmath75 . if @xmath94 , then @xmath97 , which implies @xmath98 ( as both have the same number of edges ) .
similarly , if @xmath95 then @xmath99
. therefore , the only elements of @xmath75 are @xmath71 and @xmath72 , i.e. , @xmath75 is degenerate . in the other direction ,
assume that @xmath75 is degenerate .
as mentioned above , the graph @xmath100 is a simple forest with two connected components .
color the vertices of one component white and the vertices of the other component black , and call an edge _ colorful _ if its endpoints are of different colors .
since @xmath77 is planar , it can be extended to a triangulation @xmath101 of @xmath82 with vertex set @xmath0 . as @xmath101 is connected , it contains a colorful edge @xmath92 .
a triangle in @xmath101 to which @xmath92 belongs clearly contains another colorful edge @xmath102 .
addition of either @xmath92 or @xmath102 to @xmath77 results in a simple tree , and thus , gives rise to a vertex of @xmath75 .
( note that @xmath92 and @xmath102 can not cross edges of @xmath77 since they belong to a triangulation that extends @xmath77 . )
since @xmath75 is degenerate , this implies that all other edges of @xmath101 are not colorful .
we claim that this can happen only in the case described in the statement of the proposition .
consider the edges @xmath103 .
if @xmath92 is not a boundary edge of @xmath82 then it belongs to another triangle in @xmath101 .
the other triangle must contain an additional colorful edge , contradicting the assumption that only @xmath92 and @xmath102 are colorful .
the same holds for @xmath102 , and thus , both @xmath92 and @xmath102 are boundary edges of @xmath82 .
denote their common vertex by @xmath93 and their other endpoints by @xmath16 , respectively .
it is clear that condition ( 1 ) above holds for @xmath83 , since @xmath101 is a triangulation of @xmath0 and @xmath104 is one of its triangles . to see that condition ( 2 ) holds , note that @xmath105 and @xmath106 are the only vertices of @xmath75 , and thus , are equal to @xmath71 and @xmath72 .
as @xmath107 for @xmath108 , the edges @xmath103 must coincide with @xmath69 and @xmath70 .
finally , since @xmath81 , @xmath89 $ ] is a diagonal of @xmath82 , and thus , in the triangulation @xmath101 it belongs to another triangle @xmath109 ( see figure [ fig8 ] ) .
this triangle is monochromatic ( as otherwise , there are at least four colorful edges in @xmath101 ) , hence @xmath110 are of different colors .
we apply a flip to the triangulation @xmath101 , replacing the triangles @xmath111 by @xmath112 .
the resulting triangulation includes an additional colorful edge @xmath113 $ ] that does not cross any edge of @xmath77 , except possibly for @xmath89 $ ] .
since by assumption , there are only two colorful edges that do not cross edges of @xmath77 , we must have @xmath89 \in \tilde{t}$ ] , which means that condition ( 3 ) holds .
[ cor : at - least - one - non - degenerate ] each edge of @xmath3 is contained in at least one non - degenerate max - clique .
let @xmath114 \in g(p)$]$ ] for edges of @xmath3 , like is commonly used for geometric graphs , although @xmath3 is treated as an abstract graph . ] and denote @xmath78 and @xmath79 .
if @xmath74 do not cross then @xmath67 is non - degenerate by claim [ claim : deg - u - clique ] .
if @xmath74 cross then @xmath75 is non - degenerate , since by the proof of proposition [ prop : deg - i - clique ] , if @xmath75 is degenerate , then the only edges that can be added to @xmath96 to form a simple tree share a vertex , which is not the case for @xmath74 that cross in an interior point .
[ lemma : i - clique - not - leaf ] let @xmath115 be a sst of @xmath3 , and let @xmath116 be an i - clique that contains @xmath115 .
denote the common intersection of pairs of elements of @xmath116 by @xmath77 , and let @xmath117 . if @xmath92 is not a leaf edge of @xmath115 , then @xmath118 .
we begin the proof with the argument used in the proof of proposition [ prop : deg - i - clique ] .
namely , we consider the graph @xmath77 , which is a simple forest with two components .
we color its components black and white , and extend it to a triangulation @xmath101 of @xmath0 .
the triangulation contains a colorful edge @xmath2 $ ] , and consequently , another colorful edge @xmath119 $ ] in the same triangle .
assume w.l.o.g . that @xmath120 is black , @xmath121 and @xmath122 are white
now we would like to use the assumption that @xmath92 is not a leaf edge of @xmath115 .
this assumption implies that each connected component of @xmath77 has at least two vertices .
consequently , any vertex @xmath123 is an endpoint of at least one monochromatic edge of @xmath101 ( as otherwise , @xmath93 would be isolated in @xmath77 ) . if both @xmath2 $ ] and @xmath119 $ ] are boundary edges of @xmath82 , then ( since @xmath124 is a triangle in @xmath101 ) these are the only edges of @xmath101 that emanate from @xmath120 .
this is impossible , as they are both colorful .
hence , we can assume w.l.o.g .
that @xmath2 $ ] is a diagonal of @xmath82 , and thus , belongs to two triangles , @xmath125 and @xmath126 .
( note that either @xmath127 or @xmath128 is equal to the vertex @xmath122 mentioned above . )
we consider several cases , according to the colors of @xmath127 and @xmath128 : 1 . * case 1 : @xmath127 and @xmath128 are white . * ( this case is illustrated in figure [ fig2p2 ] . )
consider the neighbors of @xmath120 in @xmath101 . since @xmath101 is a triangulation
, all these vertices lie on a path in @xmath101 . as stated before
, since @xmath120 is black , at least one of its neighbors must be black .
on the other hand , some of its neighbors ( including @xmath129 ) are white .
hence , at least one of the edges in the path connecting the neighbors of @xmath120 is colorful ( see figure [ fig2p2 ] ) .
in addition , the edges @xmath2,[a , c],[a , d]$ ] are colorful .
thus , @xmath101 contains at least four colorful edges , and each of them gives rise to a vertex of @xmath116 . therefore ,
* case 2 : @xmath127 and @xmath128 are black .
* since the black vertices @xmath130 are all neighbors of the white vertex @xmath121 , the argument of case 1 applies , with the roles of @xmath120 with @xmath121 , black and white , interchanged .
* case 3 : @xmath127 and @xmath128 have different colors .
* assume w.l.o.g .
that @xmath127 is white and @xmath128 is black . in this case , the edges @xmath2,[a , c],$ ] and @xmath131 $ ] are colorful .
we further divide this case into three subcases , according to whether @xmath132 $ ] and @xmath131 $ ] are diagonals of @xmath82 or not , and show that in each case , @xmath101 contains at least one additional colorful edge . 1 .
* case 3a : @xmath131 $ ] is a diagonal of @xmath82 . * in this case , @xmath131 $ ] belongs to an additional triangle of @xmath101 ( i.e. , other than @xmath126 ) .
if this triangle is @xmath133 , then the edge @xmath134 $ ] is colorful , implying @xmath118 ( see left part of figure [ fig3p2 ] ) .
if this triangle is @xmath135 for some @xmath136 , then , as @xmath121 and @xmath128 have different colors , one of the edges @xmath137 $ ] and @xmath138 $ ] is colorful , again implying @xmath118 ( see right part of figure [ fig3p2 ] ) .
* case 3b : @xmath132 $ ] is a diagonal of @xmath82 . *
the argument of case 3a applies , with the roles of @xmath120 and @xmath121 , @xmath127 and @xmath128 , black and white , interchanged .
* case 3c : both @xmath132 $ ] and @xmath131 $ ] are boundary edges of @xmath82 . * in this case , @xmath139 is a convex quadrilateral , and thus , its diagonal @xmath134 $ ] lies inside it ( see figure [ fig5p2 ] ) . as both @xmath125 and @xmath126 are triangles in @xmath101 , they do not contain points of @xmath0 , hence @xmath134 $ ] does not cross any edge of @xmath77 .
( note that @xmath2 $ ] is not an edge of @xmath77 , since it is colorful . ) since @xmath134 $ ] is colorful , the graph @xmath140\}$ ] is an sst , hence belongs to @xmath116 .
therefore , @xmath118 , which completes the proof of the lemma . using the lemma , we can identify whether a given degenerate clique is a u - clique or an i - clique .
[ prop : identify - degenerate ] let @xmath141 $ ] be an edge in @xmath3 that constitutes a degenerate clique .
( this actually means that there is only one max - clique that includes @xmath142 . )
if @xmath141 $ ] is included in a max - clique of 3 vertices , then the degenerate clique @xmath141 $ ] is an i - clique .
if @xmath141 $ ] is included in a max - clique of at least 4 vertices , then the degenerate clique @xmath141 $ ] is a u - clique . by corollary [ cor : at - least - one - non - degenerate
] , each edge is included in at least one non - degenerate max - clique .
hence , @xmath141 $ ] is included in a max - clique of size at least 3 .
if the degenerate clique @xmath141 $ ] is an i - clique , then , by proposition [ prop : deg - i - clique ] , @xmath143 is a forest with two connected components .
one of them is an isolated vertex @xmath93 that lies on the boundary of @xmath82 , and the two neighbors of @xmath93 on the boundary of @xmath82 , @xmath16 , are adjacent in @xmath143 . in such a case ,
the unique cycle in the graph @xmath144 is the triangle @xmath145 . by the geometric characterization of section [ sec : sub : geom - meaning ]
, this implies that the size of the u - clique that includes @xmath141 $ ] is 3 .
if the degenerate clique @xmath141 $ ] is a u - clique , then , by claim [ claim : deg - u - clique ] , the unique cycle of @xmath144 includes the edges @xmath146 and @xmath147 , and these edges cross .
this implies that @xmath148 is not a leaf edge of @xmath149 , and thus , by lemma [ lemma : i - clique - not - leaf ] , the size of the i - clique that includes @xmath141 $ ] is at least 4 .
our next goal is to identify whether a given non - degenerate max - clique is a u - clique or an i - clique .
this identification is somewhat more complex , and requires some preparations . for any sst @xmath150 ,
define a graph @xmath151 as follows : @xmath152 is the set of max - cliques that contain @xmath115 ( including degenerate cliques ) .
two max - cliques are adjacent in @xmath151 if and only if their intersection is a single edge ( that obviously has @xmath115 as one of its endpoints ) .
note that by remark [ remark : three - determine ] , two non - identical max - cliques may intersect in at most two vertices ( since three mutually adjacent vertices determine a max - clique uniquely ) . as all vertices of @xmath151 include @xmath115 , it follows that any two non - adjacent vertices of @xmath151 intersect in @xmath115 only .
[ thm : d_t ] assume that @xmath153 .
for any @xmath150 , the graph @xmath151 is connected . before we present the proof of the theorem
, we show how it can be used ( along with several of the previous lemmas ) to identify the types of max - cliques in @xmath3 . by the discussion in section [ sec : sub : max - cliques - basics ] , each edge of @xmath3 belongs to exactly one u - clique and exactly one i - clique ( one of them possibly degenerate ) .
this implies that @xmath151 is 2-colorable .
indeed , we can color all its vertices that are u - cliques white and all its vertices that are i - cliques black .
if two vertices are adjacent , both include the same edge @xmath141 \in g(p)$ ] , and thus , one is a u - clique and the other is an i - clique , so they have different colors . using theorem [ thm : d_t ]
, we can conclude that @xmath151 is connected and 2-colorable , which implies that the 2-coloring is unique , in the sense that fixing the color of any vertex determines the colors of all other vertices .
this will allow us to determine the types of all max - cliques , by the following four - step process : 1 .
pick an sst @xmath115 that belongs to a degenerate clique @xmath142 .
( it is easy to show that such a @xmath115 always exists .
we show this in lemma [ lemma : exist - degenerate ] below . ) using proposition [ prop : identify - degenerate ] , identify whether @xmath142 is a u - clique or an i - clique .
2 . consider the vertices of @xmath151 and determine for each of them whether it is a u - clique or an i - clique .
( this is possible by the explanation above . since we know the `` color '' of the vertex @xmath142 of @xmath151 , we can determine the colors of all other vertices . )
3 . consider a neighbor @xmath154 of @xmath115 .
note that the edge @xmath155 \in g(p)$ ] belongs to a max - clique @xmath34 that is a vertex of both @xmath151 and @xmath156 .
determine whether @xmath34 is a u - clique or an i - clique .
( this is possible , as by the previous step we can determine the type for all max - cliques that are vertices of @xmath151 . ) using this information about @xmath157 , determine for each vertex of @xmath156 whether it is a u - clique or an i - clique .
4 . repeat step 3 with a `` new '' vertex of @xmath3 every time until the types of all max - cliques are determined .
( since @xmath3 is connected by claim [ claim : g(p)-connected ] , we indeed reach all vertices of @xmath3 in this way . ) therefore , in order to determine the types of all max - cliques we have to prove a simple lemma on the existence of degenerate max - cliques , and to prove theorem [ thm : d_t ] .
we provide these two items now .
[ lemma : exist - degenerate ] for any set @xmath0 , @xmath153 , of points in general position in the plane , there exists an sst that belongs to a degenerate u - clique in @xmath3 . by the erds - szekeres theorem
@xcite , there exist four points in @xmath0 in convex position .
denote these points @xmath158 , such that @xmath159 $ ] is a convex quadrilateral ( in this order ) .
consider the tree @xmath115 that includes the edges @xmath2,[b , d],[d , c]$ ] , all edges that connect @xmath120 to each @xmath160 that lies above the straight line @xmath161 and all edges that connect @xmath127 to each @xmath160 that lies below @xmath161 , as shown in figure [ fig6p4 ] .
clearly , @xmath115 is an sst of @xmath1 .
addition of the edge @xmath132 $ ] to @xmath115 creates a self - crossing cycle , and thus , by the discussion in section [ sec : sub : geom - meaning ] , @xmath162 belongs to a degenerate u - clique .
suppose @xmath163 .
choose edges @xmath164 \in c$ ] , @xmath165 \in c'$ ] . in order to prove that @xmath34 is connected to @xmath42 by a path in @xmath151
, it suffices to find a sequence @xmath166 of sst s , such that @xmath167 , @xmath168 , and for each @xmath169 , there is a max - clique @xmath170 of @xmath3 that includes both @xmath171 $ ] and @xmath172 $ ] .
this will be done in four steps
. 1 . * step 1 .
* extend the sst @xmath115 to a triangulation @xmath101 of @xmath82 . recall that @xmath101 is a 2-connected graph .
* step 2 . *
if @xmath173 , leave @xmath149 as it is . if not , @xmath101 contains the graph @xmath174 .
since @xmath101 is 2-connected , there is another edge @xmath175 in @xmath101 that connects the two components of @xmath77 .
define @xmath176 .
@xmath177 is an sst of @xmath1 .
note that @xmath177 is included in the i - clique @xmath178 ( since @xmath179 ) .
+ do the same for @xmath180 : leave it , if @xmath181 , or replace it by some @xmath182 , such that @xmath183 .
+ step 2 allows us to restrict our attention to the case where both @xmath149 and @xmath180 are included in @xmath101 ( and all intermediate sst s will be included in @xmath101 , as well ) .
* step 3 . *
assume @xmath184 , @xmath185 .
if @xmath186 , then @xmath164 $ ] and @xmath165 $ ] belong to the same i - clique .
suppose @xmath187 . since @xmath115 is a simple tree , there is a unique simple path in @xmath115 whose edges are ( in this order ) @xmath188 , with @xmath189 and @xmath190 . for @xmath191
, choose an edge @xmath192 of @xmath101 other than @xmath193 that connects the two components of @xmath194 , and define @xmath195 .
in addition , let @xmath196 and @xmath197 .
now we only have to connect @xmath198 with @xmath199 for @xmath200 .
( note that the desired sequence @xmath166 will be the concatenation of the sequences connecting @xmath201 to @xmath202 , @xmath202 to @xmath203 etc .
, and @xmath204 will be the sum of their lengths . )
* step 4 . *
suppose @xmath101 is a triangulation of @xmath82 with vertex set @xmath0 .
let @xmath205 be sst s of @xmath1 that are included in @xmath101 .
assume that both @xmath149 and @xmath180 are adjacent to @xmath115 in @xmath3 , @xmath206 , @xmath207 , @xmath187 , and @xmath103 share a vertex .
suppose @xmath208 , e'=[x',y]$ ] , and let @xmath209 .
denote the edges of @xmath115 that emanate from @xmath210 by @xmath211 .
removal of all edges that emanate from @xmath210 divides @xmath115 into @xmath212 connected components : @xmath116 that includes @xmath213 , @xmath214 that includes @xmath215 , @xmath216 ( @xmath217 ) that includes the second endpoint of @xmath218 , and @xmath219 that consists of the isolated vertex @xmath210 .
since @xmath101 is 2-connected , we can extend the forest @xmath220 into an sst of @xmath221 by adding @xmath222 edges of @xmath101 .
+ let @xmath223 be such an extension .
@xmath223 includes a unique simple path @xmath224 from @xmath213 to @xmath215 .
this path starts in the component @xmath116 of @xmath31 , and ends in @xmath214 .
it visits some of the intermediate components @xmath225 in a particular order ( see figure [ fig8p6 ] ) . by appropriately labelling these components
, we may assume that @xmath224 visits @xmath226 in this order ( @xmath227 ) .
let @xmath228 be the edge of @xmath224 that passes from @xmath116 to @xmath229 , @xmath230 ( @xmath231 ) the edge that passes from @xmath225 to @xmath232 , and @xmath233 be the edge that passes from @xmath234 to @xmath214 .
( if @xmath235 , then there is only one edge @xmath236 that passes directly from @xmath116 to @xmath214 . )
+ now we can describe the passage from @xmath164 $ ] to @xmath165 $ ] .
let us start with the simple case @xmath235 .
put @xmath237 , @xmath238 . then @xmath239 , @xmath240 , and @xmath241 .
thus , @xmath242 and @xmath243 are included in i - cliques , while @xmath244 is included in a u - clique .
hence , @xmath245 is the required sequence .
+ when @xmath246 , define @xmath247 then each of the triples @xmath248 ( for @xmath249 ) , and @xmath250 is included in an i - clique , and each of the triples @xmath251 ( for @xmath252 ) is included in a u - clique .
therefore , @xmath253 is the required sequence .
this completes the proof of the theorem . as explained after the statement of theorem [ thm : d_t ]
, the theorem and lemma [ lemma : exist - degenerate ] imply the following corollary .
assume @xmath153 . given @xmath3
, we can determine for each max - clique in it whether it is a u - clique or an i - clique .
in this section we use the results of section [ sec : max - cliques ] to identify the vertices of @xmath3 that represent stars ( i.e. , ssts of diameter 2 ) and brushes ( i.e. , ssts of diameter 3 ) .
the stars , considered in section [ sec : sub : stars ] , are determined only up to an automorphism of @xmath1 as a geometric graph .
the brushes , considered in section [ sec : sub : brushes ] , are determined uniquely given a determination of the stars .
a _ star _ is a tree of diameter 2 . for @xmath254 , we call the spanning star whose center is @xmath213 an _ @xmath213-star _ , and denote it by @xmath255 .
[ thm : stars ] a vertex @xmath150 is a star if and only if all u - cliques that include @xmath115 are of size 3
. recall that by the geometric interpretation of max - cliques presented in section [ sec : sub : geom - meaning ] , if @xmath256 is a non - degenerate u - clique then all its elements are obtained from the graph @xmath144 by removing an edge from its unique cycle . in particular , @xmath257 is the length of the unique cycle of @xmath144 .
if the unique cycle of @xmath144 is self - crossing ( which can occur only if its length is @xmath258 ) then @xmath256 is degenerate .
assume that @xmath150 is an @xmath213-star , and let @xmath256 be a u - clique that includes @xmath115 . since @xmath115 is a star , @xmath144 is obtained from @xmath115 by adding an edge that connects two leaves @xmath259 of @xmath115 .
hence , the unique cycle of @xmath144 , @xmath260,[v , w],[w , x])$ ] , is of length 3 .
thus , @xmath261 , as asserted .
on the other hand , we show that if @xmath150 is an sst of diameter @xmath262 , then @xmath115 belongs to a u - clique of size @xmath263 .
since @xmath264 , @xmath115 contains an internal edge @xmath2 $ ] ( i.e. , both @xmath120 and @xmath121 are not leaves of @xmath115 ) . consider the graph @xmath265\}$ ] , that is obviously a forest with two connected components .
color the vertices of the connected component that includes @xmath120 black and the vertices of the component that includes @xmath121 white .
we would like to show that there exists a colorful edge @xmath92 that uses neither @xmath120 nor @xmath121 and does not cross any edge of @xmath265\}$ ] ( see figure [ fig9p7 ] ) .
this will conclude the proof , since in such a case , denoting @xmath266\}$ ] , we find that @xmath149 is an sst , @xmath141 \in g(p)$ ] , and the unique cycle of the graph @xmath267 is of length @xmath258 .
then , by the geometric interpretation above , if @xmath92 crosses @xmath2 $ ] then @xmath256 is a degenerate u - clique , and otherwise , @xmath257 is equal to the length of the unique cycle of @xmath144 , that is @xmath258 .
hence , in any case , @xmath115 lies in a u - clique of size @xmath263 .
we extend @xmath115 to a triangulation @xmath101 of @xmath82 , and consider three cases : 1 .
* case 1 : @xmath2 $ ] is a boundary edge of @xmath82 .
* the edge @xmath2 $ ] , being colorful , is contained in a colorful triangle @xmath268 .
the other colorful edge of @xmath125 is not a boundary edge of @xmath82 , as otherwise , one of the two connected components of @xmath265\}$ ] consists of a single vertex , which contradicts the assumption that @xmath2 $ ] is an internal edge of @xmath115 . denote that colorful edge @xmath132 $ ] .
the neighbors of @xmath120 in @xmath101 constitute a path @xmath269 $ ] .
since the connected component of @xmath120 in the graph @xmath265\}$ ] ( i.e. , the `` black '' component ) includes more than one vertex , at least one of the @xmath270 s is black .
thus , the path contains a colorful edge @xmath92 that uses neither @xmath120 nor @xmath121 ( see figure [ fig10p8 ] ) .
furthermore , as @xmath92 belongs to @xmath101 , it does not cross any edge of @xmath115 .
hence , @xmath92 is the edge whose existence was claimed .
* case 2 : @xmath121 is an internal vertex of @xmath82 . * in this case , @xmath2 $ ] is an internal edge of @xmath101 , and thus , it is contained in two triangles @xmath271 .
we further divide this case into two sub - cases : 1 .
* case 2a : either @xmath127 or @xmath128 ( or both ) are black . *
assume w.l.o.g .
that @xmath127 is black .
since @xmath121 is an internal vertex of @xmath82 , the neighbors of @xmath121 in @xmath101 constitute a cycle @xmath272 $ ] .
since the connected component of @xmath121 in the graph @xmath265\}$ ] contains more than one vertex , at least one of the @xmath270 s or @xmath128 is white . since @xmath273 are black , this implies that the cycle includes at least two colorful edges , and at least one of them uses neither @xmath120 nor @xmath121 ( see left part of figure [ fig11p8 ] ) .
as in case 1 , this is the desired edge @xmath92 .
* case 2b : both @xmath127 and @xmath128 are white .
* by the same argument as above , @xmath120 has a black neighbor in @xmath101 .
as the neighbors of @xmath120 in @xmath101 form a ( possibly closed ) path , at least one edge of this path is colorful ( see right part of figure [ fig11p8 ] ) .
this edge uses neither @xmath120 nor @xmath121 ( as both edges that use @xmath121 in this path , @xmath274 $ ] and @xmath131 $ ] , are not colorful ) . as in the previous cases ,
this is the desired edge @xmath92 .
* case 3 : both @xmath120 and @xmath121 are boundary vertices of @xmath82 , and @xmath2 $ ] is a diagonal of @xmath82 . * as in case 2 , @xmath2 $ ] lies in two triangles @xmath271 .
we further divide this case to two sub - cases : 1 .
* case 3a : @xmath127 and @xmath128 are of the same color .
* w.l.o.g .
, @xmath127 and @xmath128 are white . by the same arguments as above
, @xmath120 has a black neighbor , and thus , the path of @xmath120 s neighbors includes a colorful edge ( see left part of figure [ fig13p9 ] ) .
this edge does not use @xmath121 , as both edges that use @xmath121 ( that are @xmath274,[b , d]$ ] ) are not colorful , and it clearly does not use @xmath120 .
* case 3b : @xmath127 and @xmath128 are of different colors .
* in this case , the edge @xmath134 $ ] is as desired , since it is colorful , does not use @xmath275 , and does not cross edges of @xmath265\}$ ] ( see right part of figure [ fig13p9 ] ) .
note that since @xmath134 $ ] crosses @xmath2 $ ] , the u - clique that includes @xmath115 in this case is degenerate .
this completes the proof of the theorem .
it may happen that @xmath150 is not a star , but all u - cliques that contain @xmath115 are either degenerate or of size 3 .
let @xmath0 be a set of points in general position in the plane , and let @xmath276 .
@xmath277-brush _ is an sst of diameter 3 whose only internal edge is @xmath278 $ ] .
figure [ fig15p9 ] shows an example of a @xmath277-brush .
in this subsection we aim at identifying the vertices of @xmath3 that represent brushes .
the identification uses distances in the graph @xmath3 , defined ( as usual ) as the length of the shortest path between two vertices , and denoted by @xmath279 .
note that by the structure of @xmath3 , it is clear that for any pair of vertices @xmath280 , we have @xmath281 where @xmath282 is the symmetric difference between the edge sets of the graphs @xmath149 and @xmath115 .
[ thm : brush ] an sst @xmath150 is a @xmath277-brush if and only if it is not a star and @xmath283 for sake of convenience , we divide the theorem into two propositions .
[ prop : brush1 ] assume that @xmath150 satisfies @xmath284 then @xmath115 is a @xmath277-brush , or @xmath285 , or @xmath286 .
first , we note that @xmath287 and thus , by the triangle inequality , @xmath288 for any @xmath150 .
assume that @xmath115 satisfies ( [ eq : brush1 ] ) .
let @xmath289 be the numbers of edges of @xmath115 that emanate from @xmath290 ( respectively ) , and let @xmath291 be the number of edges of @xmath115 that use neither @xmath292 nor @xmath22 .
we consider two cases : 1 . * @xmath278 \not \in t$ ] . * in this case , we have @xmath293 hence , @xmath294 2 .
* @xmath278 \in t$ ] .
* in this case , @xmath295 hence , @xmath296 and equality can hold only if @xmath297 , which means that all edges of @xmath115 emanate either from @xmath292 or from @xmath22 and @xmath278 \in t$ ] , i.e. , @xmath115 is a @xmath277-brush , or @xmath285 or @xmath286 . [
prop : brush2 ] let @xmath150 be a @xmath277-brush . then @xmath284 in order to prove proposition [ prop : brush2 ] , we need a lemma .
let @xmath115 be a @xmath277-brush ( or @xmath285 or @xmath286 ) .
we say that @xmath115 is _ of type @xmath298 _ if @xmath299 and @xmath300 ( i.e. , the numbers of edges of @xmath115 that emanate from @xmath301 are @xmath302 , respectively ) . [ lemma : brush1 ] if @xmath115 is a @xmath277-brush ( or a star ) of type @xmath298 , @xmath303 , then it is adjacent in @xmath3 to some @xmath277-brush ( or star ) @xmath154 of type @xmath304 .
( by symmetry , if @xmath305 then @xmath115 is adjacent in @xmath3 to some @xmath277-brush ( or star ) @xmath306 of type @xmath307 . )
assume w.l.o.g .
that @xmath278 $ ] is placed horizontally , and consider the half - plane above it .
let @xmath308 $ ] be an edge such that the angle @xmath309 between @xmath278 $ ] and @xmath308 $ ] is minimal amongst all edges of @xmath115 that emanate from @xmath292 ( see figure [ fig16p10 ] ) .
let @xmath310\ } ) \cup \{[q , x]\}$ ] .
due to the minimality of the angle @xmath309 , @xmath311 $ ] does not cross any of the edges of @xmath115 that emanate from @xmath292 .
thus , @xmath154 is a @xmath277-brush of type @xmath304 ( or @xmath312 , if @xmath313 ) , and @xmath155 \in g(p)$ ] since @xmath314 .
let @xmath150 be a @xmath277-brush .
assume w.l.o.g .
that @xmath115 is of type @xmath298 . repeated use of lemma [ lemma : brush1 ] enables us to construct a path @xmath315 in @xmath3 such that @xmath316 , @xmath317 , and @xmath318
this implies @xmath319 and @xmath320 , hence , @xmath321 .
since we have shown above that @xmath322 for any @xmath115 , this completes the proof .
in this section we achieve a complete reconstruction of the geometric structure of @xmath1 , based on the identification of stars and brushes presented in section [ sec : stars - and - brushes ] . most of the effort is devoted to obtaining a complete identification of the brushes , in the sense that given a @xmath277-brush @xmath115 and a vertex @xmath323 , we determine whether @xmath324 \in t$ ] or @xmath325 \in t$ ] .
this step is presented in section [ sec : sub : brushes - further ] .
the finalization of the proof of theorem [ thm : main ] , presented in section [ sec : sub : completion ] , is easy .
so far , we know which vertices of @xmath3 are brushes . furthermore ,
if we identify the points of @xmath0 with the stars in @xmath3 ( an identification that is determined only up to an automorphism of @xmath1 ) , we can say for each brush @xmath115 , what are the vertices @xmath301 that are its `` centers '' , and how many edges of @xmath115 emanate from each of the central vertices @xmath301 .
our goal now is to gain full information on the brushes .
namely , for a @xmath277-brush @xmath115 and a vertex @xmath323 , we would like to determine whether @xmath308 \in t$ ] or @xmath311 \in t$ ] . as an intermediate step
, we would like to determine , for given @xmath326 , whether both @xmath213 and @xmath210 are connected in @xmath115 to the same vertex ( either @xmath292 or @xmath22 ) , or one of them is connected to @xmath292 and the other to @xmath22 .
[ prop : brush3 ] let @xmath115 be a @xmath277-brush , and let @xmath327 be different from each other and from @xmath292 and @xmath22 .
the leaf edges of @xmath115 whose endpoints are @xmath213 and @xmath210 emanate from the same internal vertex of @xmath115 if and only if for any @xmath328-brush @xmath149 , we have @xmath329 . one direction of the proposition is immediate . if the leaf edges emanate from the same vertex , e.g. , @xmath308,[p , y]$ ] , then at most one ( and actually , exactly one ) of these edges can belong to an @xmath328-brush ( as @xmath330,[y , p],[p , x])$ ] form a cycle ) .
since every edge of an @xmath328-brush @xmath149 emanates from either @xmath213 or @xmath210 , we have @xmath331 ( as these trees have exactly one edge in common ) .
hence , @xmath332 for any @xmath328-brush @xmath149 .
on the other hand , if the leaf edges emanate from different vertices , e.g. , @xmath308,[q , y]$ ] , it is possible that an @xmath328-brush @xmath149 include both these edges , and then @xmath333 .
we will construct an @xmath328-brush that satisfies this condition , and furthermore , satisfies the stronger condition @xmath334 .
before we show this construction , we need a few preparations .
let @xmath10 be a geometric graph , and let @xmath335 be a point in the plane .
we say that @xmath335 _ sees _ a point @xmath0 if the open segment @xmath336 does not meet any edge or vertex of @xmath10 .
we say that @xmath335 sees an edge @xmath337 if it sees every point @xmath338 , including the endpoints .
[ lemma : sees ] let @xmath339 be a crossing - free geometric graph , with no isolated vertices .
suppose @xmath340 is a disjoint union @xmath341 , where @xmath342 ( say , @xmath343 ) , and each edge of @xmath10 connects a vertex of @xmath344 with a vertex of @xmath345 .
suppose @xmath346 .
then @xmath335 sees some vertex @xmath347 .
we note that a similar lemma was proved in @xcite .
the assumption on @xmath335 in @xcite is @xmath348 , and the assertion is the same as in our lemma .
draw a ray @xmath349 that emanates from @xmath335 , crosses some edge of @xmath10 , and does not meet any vertex of @xmath10 .
( it is clear that such rays exist . )
denote by @xmath34 the first crossing point of @xmath349 with an edge of @xmath10 .
then @xmath34 is an interior point of an edge , say @xmath350 $ ] , of @xmath10 , and @xmath335 sees @xmath34 .
now rotate @xmath349 around @xmath335 towards @xmath351 , until it hits @xmath351 .
if the triangle @xmath352 does not contain any vertex of @xmath10 , except @xmath351 , then @xmath335 sees @xmath351 .
otherwise , there is a first position @xmath353 of the rotated ray that meets @xmath340 .
let @xmath93 be the point of @xmath354 closest to @xmath335 .
then @xmath335 sees @xmath93 . if @xmath355 , we are done .
assume , therefore , that @xmath356 .
clearly , @xmath357 since @xmath358 . hence , @xmath359 . among the edges that emanate from @xmath22 ,
let @xmath360 $ ] be the edge such that @xmath361 is minimal .
as before , rotate @xmath353 around @xmath335 towards @xmath362 , until it hits @xmath362 .
if the triangle @xmath363 does not contain any vertex of @xmath10 , except @xmath22 and @xmath362 , then @xmath335 sees @xmath362 .
otherwise , there is a first position @xmath364 of the rotated ray that meets @xmath340 .
let @xmath365 be the point of @xmath366 closest to @xmath335 .
then @xmath335 sees @xmath365 .
now , we observe that @xmath367 .
indeed , @xmath368 since @xmath369 , and @xmath370 since @xmath371 .
therefore , @xmath372 , which completes the proof .
now we are ready to prove proposition [ prop : brush3 ] .
we already proved above that if the two leaf edges of @xmath115 whose endpoints are @xmath16 emanate from the same vertex , then for any @xmath328-brush @xmath149 , @xmath329 .
assume now that these leaf edges emanate from different vertices .
, these edges are @xmath308,[q , y]$ ] .
we consider two cases , according to the placement of @xmath373 in the plane . in each case , we show that we can pass from @xmath115 to a suitable @xmath328-brush @xmath149 in @xmath374 steps , where in each step we remove one edge and add another edge , while maintaining the simplicity .
this will show that @xmath375 , and thus complete the proof of the proposition .
note that in all the steps of the path connecting @xmath115 to @xmath149 , the edges @xmath308,[q , y]$ ] remain untouched .
* case 1 : @xmath16 are on the same side of @xmath376 .
* in this case , at least one of the edges @xmath308,[q , y]$ ] is included in a line that supports the set @xmath377 ( which means that all points in the set are on the same side of the line ) .
we assume w.l.o.g .
that @xmath308 $ ] has this property .
the passage from @xmath115 to an appropriate @xmath149 is performed by a 4-phase procedure , illustrated in figure [ fig18p12new ] . in each phase
( except for phase 3 that will be described below ) , we consider one of the regions of the plane denoted in the figure : @xmath378 , and deal with all points of @xmath0 that belong to that region . 1
. * region 1 ( reg1 ) .
* this region is the open half - plane to the left of the line @xmath379 .
assume that @xmath380 .
we are going to perform @xmath381 steps : in each step , we take one of these points , remove the edge that connects it to either @xmath292 or @xmath22 , and add an edge that connects it to @xmath213 .
of course , we must maintain the simplicity during all steps , and this is achieved using lemma [ lemma : sees ] .
+ let @xmath382 be the geometric graph whose edges are all edges of @xmath115 of the form @xmath350 $ ] or @xmath383 $ ] , where @xmath351 lies in @xmath384 .
the graph @xmath382 and the point @xmath385 satisfy the assumptions of lemma [ lemma : sees ] , and thus , by the lemma , @xmath213 sees one of its vertices @xmath386 , say @xmath387 .
assume , for example , that @xmath388 \in e(g_0)$ ] .
define @xmath389\ } \cup \{[x , w_1]\}$ ] .
since @xmath213 sees @xmath387 , the edge @xmath390 $ ] does not cross any other edge of @xmath115 .
thus , @xmath71 is an sst and @xmath391 , which implies that @xmath392 \in g(p)$ ] .
+ now , we repeat the first step with the sst @xmath71 in place of @xmath115 .
that is , we define @xmath393 whose edges are all edges of @xmath115 of the form @xmath350 $ ] or @xmath383 $ ] where @xmath351 lies in @xmath384 , except for @xmath388 $ ] . as before
, we apply lemma [ lemma : sees ] with @xmath393 and @xmath385 and obtain a vertex @xmath394 that is seen from @xmath213 .
then , we define @xmath72 by removing from @xmath71 the edge that connects @xmath394 to either @xmath292 or @xmath22 and adding the edge @xmath395 $ ] .
note that the edge @xmath390 $ ] that was not included in @xmath396 can not cross @xmath395 $ ] , as they both emanate from @xmath213 .
+ by continuing in the same fashion , we obtain a sequence @xmath397 such that @xmath398 , @xmath399 \in g(p)$ ] for all @xmath400 , and in @xmath401 , all points in @xmath402 are connected to @xmath213 .
+ it should be noted that the parts of @xmath403 that are not included in the auxiliary graph @xmath404 , i.e. , the edge @xmath278 $ ] and the edges @xmath350,[q , w]$ ] , @xmath405 , are all disjoint from the convex set @xmath384 , and thus can not cross the new edge @xmath406 $ ] ( as @xmath407 ) .
* region 2 ( reg2 ) .
* this region contains all points that lie above @xmath376 and on the right side of @xmath379 .
assume that @xmath408 .
we start with @xmath401 and perform @xmath409 steps : in each step , we consider one of these points , remove the edge that connects it to either @xmath292 or @xmath22 , and add an edge that connects it to @xmath210 .
as before , the simplicity is maintained during all steps , by using lemma [ lemma : sees ] .
+ let @xmath410 be the geometric graph whose edges are all edges of @xmath401 of the form @xmath350 $ ] or @xmath383 $ ] , where @xmath351 lies in @xmath411 .
the graph @xmath410 and the point @xmath412 satisfy the assumptions of lemma [ lemma : sees ] , and thus , by the lemma , @xmath210 sees one of the vertices @xmath413 , call it @xmath414 . without loss of generality , @xmath415 \in g_{k_1}$ ] .
define @xmath416\ } \cup \{[y , w_{k_1 + 1}]\}$ ] .
since @xmath210 sees @xmath414 , the edge @xmath417 $ ] does not cross any other edge of @xmath401 .
thus , @xmath418 \in g(p)$ ] . + by continuing in the same fashion ,
we obtain a sequence @xmath419 such that @xmath399 \in g(p)$ ] for all @xmath400 , and in @xmath420 , all points in @xmath402 are connected to @xmath213 and all points in @xmath421 are connected to @xmath210 .
* phase 3 . * in this phase , we add the edge @xmath89 $ ] and remove the edge @xmath278 $ ] ( that otherwise closes a cycle @xmath330,[y , q],[q , p],[p , x])$ ] ) . formally , we define @xmath422\ } \cup \{[x , y]\}$ ] .
note that the edge @xmath89 $ ] does not cross any edge of @xmath420 , as in @xmath420 , all points of @xmath421 are connected to @xmath210 .
* region 4 ( reg4 ) .
* this region contains all points that lie below @xmath376 and on the right of @xmath379 .
assume @xmath423 .
as all points of @xmath0 except for @xmath373 belong to one of the regions : @xmath424 , we have @xmath425 .
we construct a sequence of ssts @xmath426 such that in @xmath427 , all points in @xmath402 are connected to @xmath213 and all points in @xmath428 are connected to @xmath210 .
hence , @xmath429 is an @xmath328-brush that satisfies @xmath430 , as desired .
+ let @xmath431 be the geometric graph whose edges are all edges of @xmath432 of the form @xmath350 $ ] or @xmath383 $ ] , where @xmath351 lies in @xmath433 .
the graph @xmath431 and the point @xmath412 satisfy the assumptions of lemma [ lemma : sees ] , and thus , by the lemma , @xmath210 sees one of the vertices @xmath434 , say @xmath435 . without loss of generality ,
@xmath436 \in g_{k_1+k_2 + 1}$ ] .
define @xmath437\ } \cup \{[y , w_{k_1+k_2 + 1}]\}$ ] .
since @xmath210 sees @xmath435 , the edge @xmath438 $ ] does not cross any other edge of @xmath401 .
( it should be noted that the fact that @xmath210 lies outside @xmath433 does not disturb us , as all points in region 2 ( that is the region @xmath210 sees @xmath433 through ) are already connected to @xmath210 , and @xmath0 is in general position . )
thus , @xmath439 \in g(p)$ ] . + by continuing in the same fashion
, we obtain a sequence @xmath440 such that @xmath399 \in g(p)$ ] and @xmath429 is the desired @xmath328-brush .
* case 2 : @xmath16 are on different sides of @xmath376 . *
this case is treated in a fashion similar to case 1 .
we divide all points of @xmath441 into two regions , where region 1 ( reg1 ) consists of the points above @xmath376 and region 2 ( reg2 ) consists of the points below @xmath376 ( see figure [ fig21p14new ] ) . in the first phase
, we consider the points of @xmath384 ( excluding @xmath213 ) , disconnect them from @xmath292 or @xmath22 and connect them to @xmath213 instead .
the procedure is identical to the procedure of the first phase of case 1 . in the second phase ,
we consider the points of @xmath411 ( excluding @xmath210 ) , disconnect them from @xmath292 or @xmath22 and connect them to @xmath210 instead .
the procedure is , again , similar .
finally , in the third phase we remove the edge @xmath278 $ ] and insert the edge @xmath89 $ ] instead .
( as at this stage , all points in @xmath442 are connected to either @xmath213 or @xmath210 , this step does not create crossings . ) as a result , we obtain a sequence @xmath443 , such that @xmath399 \in g(p)$ ] for all @xmath400 and @xmath444 is an @xmath328-brush , as desired . as cases
1,2 include all possible placements of @xmath445 , the proof is complete .
now we are ready to identify every brush completely .
[ cor : brush ] let @xmath446 be a @xmath277-brush and let @xmath254 , @xmath323 .
given @xmath3 , we can determine whether @xmath308 \in t$ ] or @xmath311 \in t$ ] .
it follows from the proof of proposition [ prop : brush2 ] that @xmath115 belongs to a path @xmath447 in @xmath3 such that in @xmath403 , @xmath448 and @xmath449 .
consider @xmath71 . since it has only one vertex @xmath450 that is connected to @xmath22 , we can use proposition [ prop : brush3 ] to determine it .
( here we use the assumption that @xmath451 . )
we can then move to @xmath72 and use proposition [ prop : brush3 ] again to determine the additional vertex @xmath452 connected to @xmath22 in @xmath72 .
( note that @xmath453 \in e(t_2)$ ] , and thus , @xmath452 is identified as the unique vertex @xmath213 such that the leaf edge of @xmath72 that emanates from it has the same second endpoint as the leaf edge that emanates from @xmath454 . ) we can continue in the same fashion and get a complete identification of @xmath455 , including @xmath115 .
our last step toward the identification of the geometric structure of @xmath1 is the following easy proposition .
[ prop : brush4 ] let @xmath373 be four different points in @xmath0 . the segments @xmath308 $ ] and @xmath456 $ ] do not cross if and only if there exists a @xmath277-brush that includes the edges @xmath308 $ ] and @xmath456 $ ] .
it is clear that if @xmath308 $ ] and @xmath456 $ ] cross then no @xmath277-brush can contain both edges @xmath308 $ ] and @xmath456 $ ] , as a brush is a _ simple _ tree . if @xmath308 $ ] and @xmath456 $ ] do not cross , then they are strictly separated by some line @xmath204 . in such a case , we can define a @xmath277-brush in which all vertices that lie on the same side of @xmath204 as @xmath292 are connected to @xmath292 , and all other vertices are connected to @xmath22 ( see figure [ fig22p14 ] ) .
this @xmath277-brush includes both @xmath308 $ ] and @xmath456 $ ] .
now we are ready to prove our main theorem .
consider the geometric tree graph @xmath3 .
the vertices @xmath445 are identified with the stars @xmath457 . by theorem [ thm :
brush ] , we can identify all @xmath277-brushes in @xmath3 . by corollary [ cor :
brush ] , we can check for each of them whether it includes both @xmath308 $ ] and @xmath456 $ ] or not . by proposition [ prop : brush4 ] , if none of the @xmath277-brushes contains both @xmath308 $ ] and @xmath456 $ ] , then these segments cross , and otherwise , they do not cross .
this completes the proof of the theorem .
in this section we consider the abstract ( i.e. , non - geometric ) graph @xmath12 .
recall that , as defined in the introduction , the vertices of the tree graph @xmath458 are all the spanning trees of @xmath12 , and two spanning trees are adjacent if they differ in exactly two edges .
we prove theorem [ thm : main - general ] , stating that the automorphism group of @xmath458 is isomorphic to @xmath459 .
it turns out that the theorem can be proved by roughly the same methodology as the proof of theorem [ thm : main ] , as shown below .
altogether , the proof in the abstract setting turns out considerably simpler than its geometric counterpart .
* identification of stars in @xmath458 . * denote @xmath460 , and let @xmath461 . as in the geometric case , our first step is identification of the vertices of @xmath10 that represent stars .
unlike the geometric case , here the identification is immediate . on the other hand ,
if @xmath115 is not a star then it is easy to see that the graph @xmath468 is connected , and thus , there exists @xmath462 that does not share an edge with @xmath115 .
hence , @xmath469 .
we note that since the distance in @xmath10 between any pair of stars in @xmath470 , it follows that the quantity @xmath471 equals @xmath470 if @xmath115 represents a star and @xmath472 otherwise .
consequently , the set of vertices that represent stars is exactly the _ center _ of the graph @xmath10 .
these vertices can be identified with the vertices of @xmath12 in an arbitrary way ( as any automorphism of @xmath12 clearly induces an automorphism of @xmath10 ) .
so , we call the @xmath80 vertices in @xmath10 that represent stars @xmath473 ( in some arbitrary order ) .
* valences of vertices in @xmath458 . *
the next simple step is identifying , for any @xmath162 and any vertex @xmath93 , what is the valence of @xmath93 in @xmath115 .
note that we were not able to obtain such an identification in the geometric setting . * max - cliques in @xmath458 .
* our next step is examination of max - cliques in @xmath10 . as in the geometric case
, we would like to determine whether a given max - clique is a @xmath38-clique or an @xmath35-clique . as the discussion in section [ sec :
sub : max - cliques - basics ] is purely combinatorial , it applies without change to the abstract setting . in particular , all vertices in a @xmath38-clique @xmath256 are obtained from @xmath144 by removing an edge from its unique cycle , and all vertices in an @xmath35-clique @xmath481 are obtained from the two - component forest @xmath143 by adding an edge that connects its two components . as there are no geometric restrictions in our case
, it follows that @xmath257 is equal to the size of the unique cycle in @xmath144 ( and , in particular , is between @xmath482 and @xmath80 ) , and @xmath483 , where @xmath8 is the number of vertices in one of the connected components of @xmath143 .
hence , determination whether @xmath480 is a @xmath38-clique or an @xmath35-clique is non - trivial only if @xmath484 .
a @xmath38-clique @xmath256 is of size @xmath472 if the unique cycle @xmath34 of @xmath144 is of size @xmath472 , which means that @xmath144 consists of @xmath34 plus a single additional edge .
each element of @xmath256 is obtained from @xmath144 by removing one edge from @xmath34 .
assume w.l.o.g . that @xmath485 , and the additional edge is @xmath486 $ ] .
it is clear that @xmath487 is never a leaf in a tree of @xmath256 , @xmath488 is a leaf in all @xmath472 trees of @xmath256 , and each of the vertices @xmath489 is a leaf in exactly two trees of @xmath256 . in addition , @xmath256 has two trees that are paths .
( these are the trees obtained by removing @xmath490 $ ] and @xmath491 $ ] . )
these two trees can be recognized by checking that their sequence of valences is @xmath492 .
an @xmath35-clique @xmath481 is of size @xmath472 if in the two - component forest @xmath143 , one component consists of a single vertex @xmath213 .
assume , in addition , that @xmath481 has two elements that are paths .
( otherwise , we can determine that @xmath481 is an @xmath35-clique by the previous paragraph . )
this is possible only if the second component of @xmath143 is a path @xmath0 .
in such a case , each endpoint of @xmath0 is a leaf in @xmath470 ( of the @xmath472 ) trees of @xmath481 .
as in @xmath256 , all vertices except one are leaves in at most two trees of @xmath256 , this property allows to determine that @xmath481 is indeed an @xmath35-clique . * the automorphism group of @xmath458 . *
our last step is to show that the information on @xmath10 obtained so far is sufficient for determining uniquely the spanning tree represented by each vertex of @xmath10 .
namely , given @xmath162 and two vertices @xmath493 , we would like to determine whether @xmath278 \in e(t)$ ] or not .
if this is possible , it implies that @xmath494 , since our determination is unique up to the arbitrary identification of the vertices of @xmath10 that represent stars with the vertices of @xmath12 .
hence , this will complete the proof of theorem [ thm : main - general ] .
let @xmath162 and suppose @xmath493 , @xmath495 , @xmath496
. then @xmath278 \in e(t)$ ] if and only if @xmath115 has a neighbor @xmath154 in @xmath10 in which the valences of both @xmath292 and @xmath22 are smaller by @xmath497 than in @xmath115 .
if @xmath278 \not \in e(t)$ ] then no removal of an edge from @xmath498 can reduce the valences of both @xmath292 and @xmath22 , and thus , @xmath154 as described in the claim does not exist . on the other hand , if @xmath278 \in e(t)$ ] then the graph @xmath499\}$ ] is a two - component forest in which both components are of size @xmath500 .
hence , there exists an edge @xmath501 $ ] that connects the two components of @xmath77 and uses neither @xmath292 nor @xmath22 .
the tree @xmath502\ } \cup \{[p',q']\}$ ] is a neighbor of @xmath115 as described in the claim . 1 .
find a leaf @xmath503 of @xmath115 such that @xmath504 .
2 . find a neighbor @xmath154 of @xmath115 in @xmath458 such that @xmath505 .
3 . consider the @xmath38-clique @xmath506 , and find a tree @xmath507 such that @xmath508 and @xmath509 .
we find a vertex @xmath510 such that @xmath511 .
we claim that if @xmath512 then @xmath278 \in e(t)$ ] , and otherwise , @xmath278 \not \in e(t)$ ] .
* step 1 . *
first , we note that if @xmath513 for all leaves @xmath503 of @xmath115 , then @xmath115 is a star , and thus , @xmath278 \in e(t)$ ] for the unique @xmath22 whose valence in @xmath115 is greater than @xmath497 and @xmath278 \not \in e(t)$ ] for any other @xmath22 .
hence , we may assume that there exists a leaf @xmath503 such that @xmath514 , and we only have to detect it .
consider the set of leaves of @xmath115 other than @xmath292 : @xmath515 .
( note that we can recognize this set , as we are able to determine valences of vertices . )
we claim that @xmath516 if and only if there exists a neighbor @xmath154 of @xmath115 in @xmath458 such that @xmath517 .
this allows to detect the desired @xmath503 by going over the elements of @xmath30 , and for each of them , going over the neighbors of @xmath115 in @xmath458 and checking whether the claimed neighbor exists . to see that the claim holds , note that a neighbor @xmath154 of @xmath115 satisfies @xmath517 , if and only if it is of the form @xmath518\ } \setminus \{e\}$ ] , for an edge @xmath92 that belongs to the unique cycle @xmath34 of @xmath519\}$ ] and uses neither @xmath292 nor @xmath503 .
if @xmath520 , then @xmath34 is of length @xmath482 , and thus , it has no edges that use neither @xmath292 nor @xmath503 .
thus , no such neighbor @xmath154 exists . if @xmath521 , then @xmath34 is of length @xmath522 , and thus , it includes an edge @xmath92 that uses neither @xmath292 nor @xmath503 .
the tree @xmath523\}$ ] is the desired neighbor of @xmath115 .
* the required neighbor @xmath149 is the tree obtained from @xmath519\}$ ] by removing the unique edge of the cycle @xmath34 that uses @xmath292 but not @xmath503 ( call it @xmath524 $ ] ) .
the @xmath38-clique @xmath506 can be recognized using proposition [ lemma : ui - cliques ] , since there exist only two max - cliques of @xmath458 that include both @xmath115 and @xmath154 a @xmath38-clique and an @xmath35-clique and proposition [ lemma : ui - cliques ] allows us to determine , which of them is the @xmath38-clique .
then , @xmath149 can be recognized as the unique element of @xmath506 in which the valences of @xmath525 are @xmath497 and @xmath526 , respectively .
* step 4 . *
it is clear that the unique vertex whose valence in @xmath149 is smaller by one than its valence in @xmath115 is @xmath510 , as defined in step 3 . by the construction of @xmath34 ,
@xmath524 $ ] is the unique edge of @xmath498 that emanates from @xmath292 , i.e. , @xmath510 is the unique neighbor of @xmath292 in @xmath115 . hence ,
@xmath278 \in e(t)$ ] if and only if @xmath527 , as asserted .
the vertex @xmath510 is detected by comparing the valences of the vertices in @xmath149 with their respective valences in @xmath115 .
m. c. hernando , complejidad de estructuras geom@xmath529tricas y combinatorias , ph.d .
thesis , universitat polit@xmath529ctnica de catalunya , 1999 ( in spanish ) .
available online at : http://www.tdx.cat/tdx-0402108-120036/ | let @xmath0 be a finite set of points in general position in the plane .
the structure of the complete graph @xmath1 as a geometric graph includes , for any pair @xmath2,[c , d]$ ] of vertex - disjoint edges , the information whether they cross or not .
the simple ( i.e. , non - crossing ) spanning trees ( ssts ) of @xmath1 are the vertices of the so - called geometric tree graph of @xmath0 , @xmath3 .
two such vertices are adjacent in @xmath3 if they differ in exactly two edges , i.e. , if one can be obtained from the other by deleting an edge and adding another edge . in this paper
we show how to reconstruct from @xmath3 ( regarded as an abstract graph ) the structure of @xmath1 as a geometric graph .
we first identify within @xmath3 the vertices that correspond to spanning stars .
then we regard each star @xmath4 with center @xmath5 as the representative in @xmath3 of the vertex @xmath5 of @xmath1 .
( this correspondence is determined only up to an automorphism of @xmath1 as a geometric graph . )
finally we determine for any four distinct stars @xmath6 and @xmath7 , by looking at their relative positions in @xmath3 , whether the corresponding segments cross . |
the interplay of band theoretic issues of hybridization and chemical bonding with the quantum chemical issue of strong local correlations is basic to the physics of many important materials . in an important paper @xcite zaanen , sawatzky and allen classified insulating transition metal oxides as `` charge - transfer '' or `` mott '' insulators according to whether the physics could be discussed solely in terms of strong correlated transition metal @xmath1 states or whether transitions to o @xmath2 states
were important to the low energy physics .
the issue arises with particular force in the high temperature superconductors where the cu @xmath3 and @xmath4 states are not far in energy from o @xmath2 states , but the cu @xmath5 state is very far away in energy . in this circumstance
a strong particle - hole asymmetry is expected @xcite , with doped electrons residing on cu sites whereas doped holes reside mainly on the o , but may be bound to cu spins creating
zhang - rice singlets".@xcite quantifying this appealing physical picture requires solving an electronic structure problem with multiple scales , including a correlation energy on the cu site @xmath6ev , a cu - o energy level difference of 2 - 4ev,@xcite and a cu - o hybridization @xmath7ev.@xcite in this paper we use single - site dynamical mean - field theory ( dmft ) to solve a model involving both copper and oxygen orbitals , developing a comprehensive theoretical picture of the electronic structure and optical conductivity of undoped and doped cuprate materials across the charge - transfer - insulator to charge - transfer metal phase diagram , including the effect of antiferromagnetism on the spectra and optics .
we use newly improved exact diagonalization ( ed)@xcite and continuous - time quantum monte carlo ( ct - qmc ) @xcite impurity solvers .
these methods have different sources of error and we find consistent results with the two methods .
our work is related to previous work of dopf _
et al._,@xcite georges _
et al._,@xcite and zlfl _ et al .
_ @xcite who each studied one particular parameter value .
modern developments in computers and solvers mean that we are able to obtain much more information .
our work also has some overlap with more recent work of macridin _ et al .
_ @xcite who used the dynamical cluster approximation on a four site cluster to study momentum dependence and the onset of superconductivity .
we compare our findings to very recent work of weber _
et al._@xcite who used similar methods to study a similar model .
et al . _ focused on specific parameters ; we focus on the spectral functions and conductivities over a wider energy range , varying the charge - transfer gap to explore all regions of the theoretical phase diagram and present a comparison of the low energy behavior of the copper - oxygen model to that of an effective one - band model .
we present evidence that a one - band model provides a reliable picture of the spectral functions and conductivity for frequencies less than about @xmath0ev ( note that this range extends about a factor of two above the charge - transfer gap in frequency ) .
we find an electron - hole asymmetry in the self - energy .
the asymmetry is much more pronounced for parameters such that the undoped material is a charge - transfer insulator ( insulating even in the paramagnetic phase ) .
the asymmetry , however , is not reflected in the fermi velocity renormalization or the low frequency optical matrix oscillator strength , where changes in the electronic structure compensate for the differences in correlation strength . comparison of our results to data suggests that the cuprates are on the metallic side of the single - site dmft phase diagram , with antiferromagnetism being essential to produce the gap in the undoped material . the rest of this paper is organized as following : in section [ mode ] we describe the model , in section [ resu ] we present the calculated phase diagram , spectral functions , self - energies and optical conductivities , in section [ comparison ] we compare the copper - oxygen model results to those obtained from computations performed on the one - band hubbard model .
section [ conc ] contains a summary of our results and a conclusion .
we analyze the canonical two dimensional copper - oxygen " hamiltonian @xcite retaining the cu @xmath8 and o @xmath9 orbitals ( whose momentum ( @xmath10 ) components are created by the operators @xmath11 ) . we allow for the possibility of two - sublattice antiferromagnetism by doubling the unit cell .
the hamiltonian is therefore a 6-band model @xmath12 in the magnetic brillouin zone . to write the band theoretical part we divide the lattice into two sublattices , @xmath13 and @xmath14 , distinguish the oxygen sites displaced from the cu in the @xmath15 and @xmath16 directions , adopt the basis @xmath17 and write @xmath18 where the @xmath19 matrices are @xmath20 and @xmath21 .
here we neglect oxygen - oxygen hopping .
we use the value @xmath22ev suggested by band theory calculations.@xcite because only the cu site is interacting , we may integrate out the oxygen band to obtain an effective one - orbital model , which we solve in the single - site dmft @xcite using ed@xcite and ct - qmc@xcite methods described in the literature . in the ed calculations typically @xmath23 bath states were used and the results were verified by occasional large calculations ; for ct - qmc temperatures studied were typically @xmath24 for the phase diagram and @xmath25 for spectral functions and conductivities .
the main result of the dmft calculation is a self - energy which , in the antiferromagnetic phase , is spin dependent and takes the form @xmath26 with @xmath27 . in the single - site dmft method
the self - energy is momentum independent so the conductivity may be computed from @xcite @xmath28 , \nonumber \label{sigmamatrix}\end{aligned}\ ] ] where the integral is over the magnetic brillouin zone , the spectral function @xmath29 is @xmath30 the matrix green function @xmath31 at frequency @xmath32 and chemical potential @xmath33 is @xmath34 , the current operator @xmath35 is @xmath36 and @xmath37 is the fermi function . for comparison
we have studied the one - band model described by @xmath38 with @xmath39 @xmath40ev is chosen to reproduce the non - interacting bandwidth of the cu - o antibonding band passing through the fermi level and @xmath41 fixed so as to reproduce the correlation gap .
fig . [ phasediagram ] shows the boundary between metallic and paramagnetic insulating solutions , calculated for one hole per cuo@xmath42 unit as a function of interaction strength @xmath43 and cu - o energy level splitting @xmath44(@xmath45 ) in the paramagnetic phase . as in the single - site dmft of the one - band hubbard model @xcite a coexistence region
is observed where both metallic and insulating solutions exist .
we define the insulating and metallic boundaries of the coexistence region at fixed large @xmath43 to be @xmath46 and @xmath47 respectively .
the phase diagram is obtained at @xmath48 using the ed solver ; the results were verified using ct - qmc by scanning @xmath44 at selected @xmath43 values .
we find almost perfect agreement for @xmath47 ; the ct - qmc calculation of @xmath46 line represents the smallest @xmath44 value at which a metallic phase can be found at temperature @xmath49ev .
the known t - dependence of the single - site dmft boundary for the single - band hubbard model @xcite suggests that the discrepancy may simply be a finite - temperature effect related to the different temperatures used in the two calculations .
[ dos ] presents the many body density of states ( dos ) ( electron removal spectrum for energy @xmath50 and electron addition spectrum for energy @xmath51 ) as well as the projections onto the cu-@xmath1 and o-@xmath9 states calculated in the paramagnetic phase using the ed method .
the zero of energy is the chemical potential .
the upper panel presents results for @xmath52 ( so within single - site dmft the paramagnetic phase of the undoped material is metallic ) while the lower panel presents results for @xmath53 such that the undoped material is a charge - transfer insulator .
the main difference between the two spectra is the presence or absence of a gap in the undoped material . at lowest energy
( binding energy @xmath54ev ) a peak is seen , of mainly @xmath1-character .
this peak corresponds to removing one electron and leaving the cu in the @xmath5 state ; it is pushed down from the bare @xmath5 energy by a level repulsion due to hybridization with the o states .
thus it is not correct to identify the position of this peak directly with @xmath43 , as is sometimes done in literature . in the binding energy range @xmath55 a mainly oxygen - like band
the very sharp peak corresponds to the non - bonding oxygen state ; it would be broadened if oxygen - oxygen hopping were taken into account ( although the vanishing of the cu - o hybridization at the @xmath56 point means that a singularity would remain at the bare oxygen energy ) .
the broad structure of mixed oxygen and copper character lying below the non - bonding state may be thought of as the bonding " linear combination of cu and o states pushed down below the non - bonding o level by hybridization to the cu @xmath1 level .
this feature was identified in ref . as the copper upper hubbard band corresponding to the @xmath5 state , but ref . did not present results over a wide enough range to determine if the @xmath54ev peak ( which we find to correspond to the @xmath5 state ) was present in their calculations .
the structure of mixed cu - o character at binding energies in the range @xmath57 corresponds to the zhang - rice singlet states .
calculations ( not shown here ) in which the cu orbital is forced to be fully spin polarized show that these states correspond to holes with the same spin as the deep - lying removal state .
the states in the electron addition spectrum are mainly of cu @xmath4 character , and play the role of the upper hubbard band . the effect of antiferromagnetism on the magnitude of the gap in the charge - transfer insulator state has been the subject of debate , with ref . arguing on the basis of hubbard model calculations that antiferromagnetism increases the gap significantly while the conductivity calculations presented in ref . were interpreted as indicating no significant effect of antiferromagnetism on the gap .
we have used the `` quasiparticle equation '' method of ref . to determine the gap values at @xmath58ev , @xmath59ev finding that the gap in the paramagnetic insulating phase is @xmath60ev while the addition of antiferromagnetism shifts the gap to @xmath61ev .
some of the difference between our results and those of ref . may arise from the extremely small value of the calculated paramagnetic - state conductivity in the near gap region , especially for @xmath62 , which may have led those authors to overestimate the gap in the paramagnetic state . upon doping ,
two changes occur .
first , the chemical potential moves into the zhang - rice band ( hole doping ) or the upper hubbard band ( electron doping ) . for @xmath63
the associated changes in chemical potential are substantial : the fermi level , measured relative to the non - bonding oxygen peak , shifts by almost 2ev ; for the @xmath64 the fermi level changes rather less .
second , as can be seen by inspection of the figure [ dos ] and from table [ areas ] , the relative strengths of the different spectral features evolve .
the issue of the number of states created by doping has received some attention in the literature as a signature of
mottness".@xcite we find that the changes in the spectrum do not have a universal doping or interaction - strength dependence ; however in the charge - transfer insulating " regime @xmath65 , each doped hole adds roughly two states to the zhang - rice band and one to the upper hubbard band , while doping with electrons does essentially the opposite . in the paramagnetic metal case the changes in electronic structure are larger . .integrated density of states ( both spins ) for the @xmath5 ( lowest - lying ) ,
zhang - rice " ( zr ) and upper hubbard band " ( uhb ) spectral features discussed in the text for @xmath58ev , @xmath66ev ( paramagnetic metal in undoped case ) and @xmath59ev ( paramagnetic insulator in undoped case ) at dopings indicated .
[ cols="<,^,<,<,^ , < " , ] to further probe the particle - hole asymmetry we show in the upper panel of fig .
[ selffig ] the self - energy calculated on the matsubara axis for @xmath67 electron and hole doping at @xmath68 .
the near - perfect agreement between the results of ed and ct - qmc calculations serves as a test of the reliability of our results .
the lower panel focuses on the low frequency behavior , presenting the doping dependence of @xmath69 estimated from the values of @xmath70 at the lowest two matsubara points of the ed calculation for both @xmath71 ( paramagnetic insulator ) and @xmath72 ( paramagnetic metal ) . at very low doping
the estimate becomes unreliable because the ( very small ) fermi liquid scale is not easily resolved so we do not present results .
we see again that the self - energy is systematically larger for hole doping than for electron doping , with the difference being more pronounced for the paramagnetic insulator case .
however , the self - energy is not necessarily the most relevant measure of correlation strength . in a multiband model such as the one studied here , the fermi surface @xmath73 is given by the solution of @xmath74 at @xmath75 . the degree to which @xmath1 states participate in the fermi surface states @xmath76 depends on the self - energy , which of course changes across the charge - transfer gap and with doping . by expanding for @xmath77 near @xmath78 and @xmath10 near @xmath79 and defining @xmath80 and @xmath81
, we find that the physical quasiparticle velocity @xmath82 is given by @xmath83 and the bare fermi velocity by the same equation but with @xmath84 .
the bare velocities ( defined here in physical units by multiplying the result above by the lattice constant @xmath85 ) have only about @xmath86 doping dependence on either the electron on the hole doped sides , but change substantially as one goes from electron to hole doping : at @xmath66ev we have @xmath87 for the @xmath67 hole doped and @xmath88 for the @xmath67 electron doped calculation while at @xmath59ev we have @xmath89 for @xmath67 hole doping and @xmath90 for @xmath67 electron doping .
we see from fig .
[ velocityfig ] that for the strong coupling ( @xmath59ev ) case the particle - hole asymmetry in @xmath91 is slightly overcompensated by a difference in @xmath1 character of the ground state wave function , so that the particle - hole asymmetry in the ratio @xmath92 is rather smaller in magnitude and of opposite sign , compared to that in @xmath93 , while in the more weakly correlated case the wave function changes induce an asymmetry in velocity renormalization which , while small , is larger than that in @xmath94 . and inverse of self - energy derivative @xmath95 .
upper panel : @xmath59ev ; lower panel @xmath66ev . ] and inverse of self - energy derivative @xmath95 .
upper panel : @xmath59ev ; lower panel @xmath66ev . ]
we have calculated the optical conductivity and have verified the results via comparison of the integral of our calculated conductivity to the independently calculated `` kinetic energy '' and to the average of the renormalized fermi velocity over the fermi surface .
we also compared results obtained directly on the real axis from the ed calculation to results obtained by analytic continuation of the matsubara axis self - energy obtained from ct - qmc calculations using the methods of ref . .
the upper panel of fig .
[ condfig ] shows the conductivity in the near gap region for a doping of one hole per cu - o@xmath42 unit in the antiferromagnetic phase at @xmath59ev , 4ev and 4.5ev .
as noted above , for parameters @xmath96 the gap in the antiferromagnetic phase is rather larger than the experimentally measured value @xmath97ev.@xcite on the other hand , the gap value determined from the antiferromagnetic phase of the @xmath98 calculation is in reasonable agreement with data .
this comparison places the materials clearly on the metallic side of the paramagnetic - metal / charge - transfer - insulator phase diagram , in agreement with previous analysis based on the one - band hubbard model.@xcite however , it is important to note that the calculated conductivities in this frequency range are about a factor of 2 smaller in magnitude relative to experimental data ( note that a normalization error means that the conductivity results of ref .
are too large by a factor of two )
. some of the difference may arise from transitions to bands not included in the present calculation , but it is possible also that the peierls phase arguments omit important interband matrix elements even within the space of states we consider .
this is an important issue for further study .
the lower panels of fig . [ condfig ] display the doping dependence of the conductivity , calculated in the paramagnetic phase for @xmath66ev .
electron or hole doping adds optical absorption strength ( associated with motion of doped holes ) at frequencies @xmath99ev . for both values of @xmath44 the integrated absorption strength ( up to 2ev )
is found to be comparable for electron and hole doping , despite the differences in self - energy displayed in fig .
[ selffig ] . in this frequency range ,
the optical matrix element is proportional to the fermi velocity which depends on the renormalized cu - o energy difference set by the parameter @xmath100 .
the shifts in @xmath101 as the chemical potential is tuned from hole to electron doping leads to a bare fermi velocity which is larger on the hole - doped side than on the electron - doped side as noted in the previous sub - section to a change in the d - character of the wave function ; these effects lead to changes in the optical matrix elements which compensate to a considerable degree for the change in self - energy .
one of us previously argued @xcite that the experimentally observed similarity in low frequency optical absorption between electron and hole - doped materials implied that the correlation strength was about the same for two systems ; we see that this argument must be treated with caution .
a more striking change is that hole - doping but not electron doping activates a strong transition at about 4ev between the non - bonding oxygen band and the near - fermi - surface states .
this feature is not observed experimentally ; indeed a comparison of figures 7 and 8 of ref .
shows that in the range @xmath102ev the optical absorption in electron doped compounds is slightly larger than in hole - doped compounds .
we note that in a more realistic three - band model with oxygen - oxygen hopping included the spectral weight in this peak would be spread over a wider energy range .
further experimental and theoretical investigation of this issue is important .
a basic issue in the physics of the cuprates is the reduction of the three - band model to an effective one - band model.@xcite we test the reduction by comparing our results at @xmath103 to results obtained by applying the qmc solver to a one - band square - lattice hubbard model at @xmath104 and @xmath40ev chosen to reproduce the splitting between the centroids of the upper and lower hubbard bands .
[ compto1band ] shows the comparison of spectral functions and optical conductivities . to obtain spectra for the hubbard model we used the analytic continuation procedure described in ref . , while the three - band results were obtained using the ed solver at parameters @xmath58ev , @xmath105ev and @xmath48 .
all calculations are done in the paramagnetic phase .
the upper panel of fig .
[ compto1band ] compares the spectral functions .
one sees immediately that the one - band and three - band models provide a reasonably consistent account of the spectra within a few ev of the fermi energy
. some differences are evident . from the upper panels one sees that the one - band model has a slightly larger bandwidth than the three - band model ( the difference is most evident for the half - filled calculation ) .
this difference may be interpreted as indicating that the three - band model at @xmath103 is equivalent to a one - band model at a @xmath41 a bit greater than @xmath106 .
the lower panel of fig [ compto1band ] compares the conductivities . in the half filled case
the two models give a quite consistent account of the absorption above the mott - hubbard / charge - transfer gap edge .
note that because the lowest - lying gap excitation is not optically active , the difference in gap values noted in the previous paragraph is not easy to see in this panel .
the one - band and three - band calculations are done with different methods but as can be seen , the conductivities are similar and we have verified that the spectral weight ( integral of @xmath107 up to say @xmath108ev ) are similar for the two models . however , for hole doping a new feature in the three - band model appears at @xmath109ev .
this is associated with transitions from the non - bonding oxygen band , which is of course not present in the one - band model .
in this paper we have used the single site dynamical mean field method to solve the three - band model believed to be relevant to the copper - oxide materials .
the method includes the physics of zhang - rice singlets ( doped holes reside largely on oxygen sites and have spin opposite to the copper spin ) , reproduces the characteristic features of the spectrum , and reveals ( especially for parameters in the `` charge - transfer insulator '' regime ) a particle - hole asymmetry in the self - energy which however is largely canceled by the difference in @xmath1-content of the near - fermi surface states , leading to very similar velocity renormalizations between electron and hole - doped compounds .
one important implication of this finding is that ( at least for the correlation strengths we have studied ) the fact that doped holes are `` zhang - rice singlets '' while doped electrons are just conventional doubly occupied sites does not imply a significant difference in the physics and suggesting that a reduction to an effective one - band model may be reasonable . the issue of reduction to a one - band model has been considered by many previous authors ; however the discussion has been largely been couched in terms related to reduction to t - j - like models .
we prefer to directly compare spectral functions and conductivity .
we found that for @xmath96 ( paramagnetic insulator ) a one - band model gave a quantitatively accurate description of the physics at scales below about @xmath110ev ( above this scale the effects of non - bonding oxygen states which are not included at all in the one - band model become visible ) .
similar results ( not shown ) are found for @xmath111 .
it is likely that for parameters much deeper in the insulating phase ( which we have not investigated ) the `` zhang - rice '' effects may be more important and the reduction to a one - band model may be more problematic .
a basic question in cuprate physics is the effective correlation strength governing the physics of the low - energy particles responsible for superconductivity .
we addressed this question by calculating the gap in the insulating phase .
the gap is of a `` charge - transfer '' rather than `` mott - hubbard '' nature , as stressed by many previous workers .
our comparison to the one - band model shows that the charge - transfer gap can be used to extract an effective @xmath43 which , when used in a one - band model , reproduces the low energy physics reasonably well .
in contrast to ref .
we find that antiferromagnetism has a pronounced effect on the magnitude of the gap in the insulating state .
we find that there is no reasonable way to obtain a gap of the physical scale ( between @xmath112 and @xmath113ev ) if the materials are assumed to be paramagnetic insulating side of the phase diagram ; rather they must be taken to be more moderately correlated .
this conclusion is in agreement with previous results.@xcite three features of the calculation suggest potentially interesting directions for future research .
first , the quasiparticle band structure depends on the renormalized @xmath1 level energy @xmath114 .
this quantity has ( especially for stronger correlations ) a noticeable dependence on doping and on which side of the charge - transfer gap the materials are on .
it also changes dramatically between antiferromagnetic and paramagnetic states .
further investigation of the physical consequences of these changes would be useful .
this might produce additional insight into the fundamental electronic structure dichotomy in many transition metal oxides between the dramatic evidence for strongly correlated behavior in high energy spectroscopies and the more nearly band - like behavior of the lower energy excitations . concerning the conductivity , as is seen most clearly in fig
. [ compto1band ] , the calculated insulating state conductivity is only a few hundred inverse - ohm - inverse - centimeters , rather smaller than the experimentally measured values @xmath115 .
some part of the discrepancy may arise from other bands , not considered in our calculation , but it may also indicate a failure of the peierls - phase approximation to the conductivity .
a further issue in the comparison of the high - frequency calculated conductivity to data is the appearance , for hole doping but not electron doping , of a strong feature relating to transitions from the non - bonding oxygen bands to the near - fermi - surface states .
this feature implies that the high energy conductivity in the 4 - 6ev range should be greater for hole doped than for electron doped compounds .
this is not seen experimentally .
inclusion of oxygen - oxygen hopping will spread the excess spectral weight over a wider frequency range , perhaps mitigating the discrepancy with experiment .
further , one must bear in mind that the electron and hole doped materials studied experimentally have different crystal structures , introducing further uncertainties in the comparison . however , the issue warrants further study .
this collaboration was begun with support from the columbia / polytechnique / science po /sorbonne alliance program .
ajm and xw are supported by nsf - dmr-0705847 and mc by miur prin 2005 , prot .
200522492 and miur prin 2007 , prot . 2007fw3mjx003 , and ldm by the rutgers center for materials theory , nsf - dmr-0528969 and rtra triangle de la physique .
we acknowledge helpful discussions with c. weber which led to physics insights and to correction of an error in the normalization of the conductivity in an earlier version of this manuscript .
we also thank n. lin for helpful discussions and for aid in verifying the conductivity calculations . | the three - band model relevant to high temperature copper - oxide superconductors is solved using single - site dynamical mean field theory and a tight - binding parametrization of the copper and oxygen bands . for a band filling of one hole per unit cell
the metal / charge - transfer - insulator phase diagram is determined . the electron spectral function , optical conductivity and quasiparticle mass enhancement are computed as functions of electron and hole doping for parameters such that the corresponding to the paramagnetic metal and charge - transfer insulator sides of the one hole per cell phase diagram .
the optical conductivity is computed using the peierls phase approximation for the optical matrix elements .
the calculation includes the physics of `` zhang - rice singlets '' .
the effects of antiferromagnetism on the magnitude of the gap and the relation between correlation strength and doping - induced changes in state density are determined .
three band and one band models are compared .
the two models are found to yield quantitatively consistent results for all energies less than about @xmath0ev , including energies in the vicinity of the charge - transfer gap .
parameters on the insulating side of the metal / charge - transfer insulator phase boundary lead to gaps which are too large and near - gap conductivities which are too small relative to data .
the results place the cuprates clearly in the intermediate correlation regime , on the paramagnetic metal side of the metal / charge - transfer insulator phase boundary . |
pulsars are unique objects , with which all types of elemental interaction could be tested extremely . however , the most elementary question relevant is still open : _ what is the nature of pulsars ? _ it is conventionally thought that pulsars are simply a kind of boring big `` nuclei '' neutron stars , but more and more attention is paid to the quark star model for pulsars @xcite since _ no _ convincing work , neither in theory from first principles nor in observation , has confirmed baade - zwicky s original idea that supernovae produce neutron stars . the bare quark surface is suggested to be a new probe for identifying quark stars with strangeness , and possible observational evidence for bare strange stars appears : the drifting subpulses of radio pulsars , ultra - high luminosity of soft @xmath3-ray repeaters , non - atomic thermal spectra of isolated `` neutron '' stars @xcite . however , can the bare strange star model reproduce most of the general features of pulsars ( especially glitches ) ?
the observation of free precession in psr b1828 - 11 @xcite and psr b1642 - 03 @xcite challenges astrophysicists today to re - consider the internal structure of radio pulsars @xcite .
the current model for glitches involves neutron superfluid vertex pinning and the consequent fluid dynamics
. however , the pinning should be much weaker than predicted in the glitch models , otherwise the vortex pinning will damp out the precession on timescales being much smaller than observed .
in addition the picture , that a neutron star core containing coexisting neutron vertices and proton flux tubes , is also inconsistent with observations of freely precessing pulsars @xcite .
it is then supposed that the hydrodynamic forces presented in a precessing star are probably sufficient to unpin all of the vortices of the inner crust @xcite since a definitive conclusion on the nature of vertex pinning has not been reached yet due to various uncertainties in the microscopic physics .
but recently , levin & dangelo @xcite studied the magnetohydrodynamic ( mhd ) coupling between the crust and the core of a rotating neutron star , and found that the precession of psr b1828 - 11 should decay over a human lifetime .
this well - defined mhd dissipation should certainly be important in order to test the stellar models . an alternative way to understand both glitch and free - precession could be through the suggestion that radio pulsars are solid quark stars @xcite .
a solid quark star is just a rigid - like body , no damping occurs , and the solid pulsar model may survive future observational tests if the free precession keeps the same over several tens of years . a neutron star could not be in a solid state , whereas a cold quark star could be
. such a solid state of quark matter could be very probably skyrme - likequark clusters might also be described as solitons in a similar way . ]
@xcite , the study of which may help us to understand dense quark matter with low temperature .
fluid strange - star ( even with possible crusts ) models were noted to be inconsistent with the observations of pulsar glitches more than one decade ago @xcite . modifications with the inclusion of possible stable particles to form a differentiated structure of so - called strange pulsars was also suggested @xcite , but is not popular because of a disbelief in the employed physics @xcite .
however , can a fully solidified quark star proposed @xcite really reproduce the glitch behaviors observed ?
one negative issue is that giant glitches are generally not able to occur at an observed rate in a solid neutron star @xcite .
nonetheless , more strain energy could be stored in a solid quark star due to an almost uniform distribution of density ( the density near the surface of a bare strange star is @xmath4 g/@xmath1 ) and high shear modulus introduced phenomenologically for solid quark matter with strangeness .
more energy is then released , and this may enable a solid pulsar to glitch frequently with large amplitudes .
furthermore , the post - glitch behaviors may represent damped vibrations . in this paper
, we try to model glitches in a starquake scenario of solid quark stars .
a quake model for a star to be mostly solid was generally discussed by baym & pines @xcite and others @xcite , who parameterized the dynamics for solid crusts , and possible solid cores , of neutron stars .
strain energy develops when a solid star spins down until a quake occurs when stellar stresses reach a critical value .
two possibilities were considered previously .
ruderman @xcite assumed that the entire strain energy is relieved in the quake , while baym & pines @xcite suggested only part of the stress is released and that the plastic flow is negligible . we present a third possibility that , during a quake , the entire stress is almost relieved at first when the quake cracks the star in pieces of small size ( the total released energy @xmath5 may be converted into thermal energy @xmath6 and kinematic energy @xmath7 of plastic flow , @xmath8 , but the part of @xmath7 might be re - stored by stress due to the anelastic flow ( i.e.
, the kinetic energy is converted to strain energy again ) .
as shown in fig.1 , a quark star may solidify with an initial oblateness @xmath9 ; stress ( to be relative to @xmath9 ) increases as the star losses its rotation energy , until the star reaches an oblateness @xmath10 .
a quake occurs then , and the reference point of strain energy should be changed to @xmath11 ( the oblateness of a star without shear energy ) at this moment .
damped vibrations in the potential field ( see the inserted up - right part of fig.1 ) occur therefore .
after a glitch , the star solidifies and becomes elastic body again . .
the horizontal solid lines are for the reference points of the strain .
the steps , drawn very exaggeratedly , are starquakes .
no starquake occurs anymore when @xmath12 falls below @xmath13 ( i.e. , the stress can not be large enough for a quake ) .
, width=340 ] no memory of the first glitch affects the second one .
the density of quark stars with mass @xmath14 can be well approximated to be uniform @xcite . as a star , with an initial value @xmath15 , slows down , the expected @xmath12 decreases with increasing period .
however , the rigidity of the solid star causes it to remain more oblate than it would be had it no resistance to shear .
the strain energy in the star reads @xcite @xmath16 and the mean stress @xmath17 in the star is @xmath18 where @xmath12 is stellar oblateness is defined by @xmath19 ( @xmath20 and @xmath21 are the semimajor and semiminor axes , respectively ) . the oblateness ( or ellipticity ) @xmath22 ( @xmath23 is the nonrotating , spherical moment of inertia ) is related to @xmath24 through @xmath25 . ] , @xmath26 is the volume of the star , and @xmath27 is the mean shear modulus of the star .
we note that , in the following calculations , the stress - increase developed by the decrease of oblateness ( due to the spindown ) is only included . however , azimuthal stress due to the general relativistic effect @xcite ( maybe similar to the frame - dragging effect in vacuum ) of rotating solid stars may contribute significantly , though , unfortunately , the theoretical answers to elastic relativistic - stars with rotation are very difficult to be worked out . the total energy of such a rotating star with mass @xmath28 and radius @xmath29 is the sum of the gravitational energy @xmath30 , the rotation energy @xmath31 , the strain energy @xmath32 , the bulk energy @xmath33 , and the surface energy @xmath34 , @xmath35 where @xmath36 is the reference oblateness before the @xmath37-th glitch occurs , @xmath38 , @xmath39 is the moment of inertia , @xmath40 is the stellar angular momentum , @xmath41 ( @xmath42 the rotation period ) , and the coefficients @xmath43 and @xmath44 measure the gravitational and strain energies @xcite , respectively , @xmath45 @xmath46 the changes of @xmath33 and @xmath34 are much smaller than that of @xmath30 , @xmath47 , or @xmath48 when a star spins down , according to a mass formula for strange quark matter to be analogous to the bethe - weizsacher semi - empirical mass function @xcite .
we therefore neglect @xmath33 and @xmath34 in the following calculations . by minimizing the total energy @xmath49 , a real state satisfies ( note that @xmath50 ) , @xmath51
the reference oblateness is assumed , by setting @xmath52 in eq.([epsilon ] ) , to be @xmath53 a star with oblateness of eq.([epsiloni ] ) is actually a maclaurin sphere .
when the star spins down to @xmath54 , the stress develops to @xmath55,% \label{sigmamu}\ ] ] according to eq.([sigma ] ) .
a glitch takes place if the stress is greater than a critical one , @xmath56 .
the first quake is characterized by a sudden shift of @xmath12 , the amount of which is @xmath57 where the reference point is changed from @xmath9 to @xmath11 . due to the conservation of stellar angular momentum , the sudden change in the oblateness results in an increase of spin frequency
, @xmath58 starquakes would also produce a thermal energy dissipation during glitches that could be expected to be observable as an increase of x - ray luminosity soon after glitch . according to the conservation of energy , one obtains from eq.([e ] ) , @xmath59 where @xmath60 , which can be obtained by @xmath61 from eq.([sigmamu ] ) , is the spin frequency when the first quake occurs , @xmath62 is the released energy in a starquake .
the observed glitch size @xmath63 can be calculated from eq.([epsilon ] ) to eq.([epsinon-1 ] ) .
the calculations , based on these equations , are parameterized by five input quantities : @xmath28 , @xmath64 , @xmath65 , @xmath66 and @xmath67 , and two of them are fixed to be @xmath68 and @xmath69 g/@xmath1 .
2 shows the results for various @xmath65 and @xmath66 .
, vs. the rotation period , @xmath42
. the lines labelled i , ii , and iii are for different shear modulus , @xmath65 .
i : @xmath70 erg/@xmath71 , ii : @xmath72 erg/@xmath73 , and iii : @xmath74 erg/@xmath71 .
the dash - dot , dotted , solid and dashed lines are for @xmath75 , @xmath76 , @xmath77 and @xmath78 erg/@xmath71 , respectively .
we observe that the curves tend towards the asymptotic line when @xmath79 erg/@xmath71 .
the stellar mass and mean density are chosen to be mass @xmath68 and density @xmath69 g/@xmath1 in the calculations .
, width=340 ] we see that , for @xmath80 erg/@xmath71 , @xmath63 decays to be a constant for pulsars with large rotation periods .
small @xmath65 may generally result also in a small jump of @xmath63 for a certain period @xmath42 .
however , for @xmath81 erg/@xmath71 , the asymptotic line is reached , which means that @xmath82 .
we choose @xmath83 erg/@xmath71 in computation , but the curves do not depend sensitively on @xmath84 .
it is found that the glitch amplitude , @xmath63 , could be as high as observed if the shear modulus @xmath85 erg/@xmath1 .
though the quark matter with low temperature and high baryon density is focused recent years , it is still impossible to determine the properties of such qcd phase by first principles ( most of the calculations are started from qcd phenomenological models ) .
one of the proposed states of strange stars at low temperature could be a solid state @xcite ( and also the 2nd paragraph of 4 ) , but the calculations on solid state is much more difficult than that of fluid one .
nevertheless , we may estimate the shear modulus @xmath65 originated only by electric interaction between charged @xmath86quark clusters ( @xmath87 : the quark number in a cluster ) and a uniform background of electrons , through a similar quantum mechanical calculation of metals @xcite .
the effective shear modulus , averaged over polarizations and directions , can be well fitted by @xmath88 in asymptotic case , where @xmath89 is the charge of quark - clusters , @xmath90 the cluster number density and @xmath20 the separation between clusters @xcite .
for strange quark matter with baryon number density @xmath91 and electron number density to be @xmath92 that of quarks , one comes to @xmath93 where @xmath94 baryons per @xmath95 is the nuclear saturation density .
this result can be regarded as a low limit modulus of solid strange quark matter since the van der waals - type color interaction , with a high coupling constant ( to be corresponding to the electric charge @xmath24 in electromagnetic interaction ) , may result in a larger shear modulus .
we might then conclude that @xmath65 in the range of @xmath96 erg/@xmath1 is not impossible .
if the kilohertz quasi - periodic oscillations are relevant to the global oscillation behavior of solid quark stars , the shear modulus could be @xcite @xmath97 erg/@xmath1 , which is much larger than the value ( @xmath98 erg/@xmath1 ) of neutron star crust
. this could be unsurprise due to much high density .
after the first quake , from eq.([sigmamu ] ) , the stress in the star builds up again , at a rate of @xmath99 which is almost a constant for a certain pulsar with @xmath42 and @xmath100 during a period when the effect of @xmath101 is negligible .
another quake occurs after a time of @xmath102 the calculation for certain @xmath42 and @xmath103 is presented in fig.3 .
for a certain set of \{@xmath104 } , the interval @xmath105 become longer for a higher value of @xmath66 .
it is , however , not necessary to expect that a radio pulsar should jump its rotation periodically , because of no reason to show that the critical value @xmath66 ( and maybe @xmath65 ) keeps a constant during its life .
the time separation @xmath106 may present a quasi - periodic nature if @xmath66 has a gaussian distribution .
when @xmath107th quake occurs in a pulsar , the oblateness decreases to @xmath108 at first , while the reference point is changed to @xmath36 from @xmath109 .
soon after the rotational jump ( @xmath110 ) , the star has a tendency to reach the equilibrium oblateness @xmath36 .
certainly , this recovery process depends on the difference , @xmath111 , which is shown in fig.4 as a function of @xmath6 .
increasing oblateness can obviously contribute an additional spindown role besides that of magnetodipole radiation , which results in larger @xmath103 .
such a post - starquake could be observed as a recover behavior during post - glitch . from fig.4
, one can see the values of @xmath111 are generally not sensitive to both @xmath6 and @xmath66 . as a function of @xmath112 .
the lines are grouped for different periods ; `` i '' : @xmath113 s , `` ii '' : @xmath114 s , and `` iii '' : @xmath115 s. the solid and dashed lines are for @xmath75 and @xmath77 erg/@xmath71 , respectively , which are almost the same for certain period @xmath42 .
other parameters chosen are : @xmath68 , @xmath69 g/@xmath1 , and @xmath72 erg/@xmath71 .
, width=340 ] the recovery of @xmath116 is actually a complex process , in which both elastic revert and plastic flow could not be negligible .
nevertheless , this recovery process might be an analog of damped oscillations , with a mathematical description of @xmath117 with @xmath118 . the second term in eq.([ddotepsinon ] )
arise from the anelastic effect during the recovery .
the solution of eq.([ddotepsinon ] ) depends on the relative strength of the elastic spring and that of the damping .
it is well known that three possible solutions exist ; the so - called underdamped case : @xmath119 , the critical case : @xmath120 , and the overdamped case : @xmath121 ; where @xmath122 is the intrinsic frequency .
one comes then to @xmath123 by dimensional analysis ( force @xmath124 and displacement @xmath125 ) . in all the three cases , the oscillation damps by a factor of @xmath126 , with a typical time of @xmath127 .
we are just to discuss the critical damped motion below for an indication . in this case ,
@xmath128 , the time evolution of @xmath12 after the occurrence of starquake can be solved by eq.([ddotepsinon ] ) , @xmath129
a starquake model for pulsar glitches is developed in the regime of solid quark stars , and it is found that the general glitch behaviors ( i.e. , the glitch amplitude @xmath63 and the time interval @xmath106 ) could be reproduced if solid quark matter has properties of shear modulus @xmath130 erg/@xmath1 and critical stress @xmath131 erg/@xmath1 . it is suggested that the post - glitch process could be described as damped oscillations , especially in the critical and the overdamped cases .
anyway , this is only a primary and simplified study of quakes in solid quark stars , more elaborate work , with possible modifications , on both quake and postquake processes is necessary in order to understand the nature of solid quark matter through glitching pulsars .
we are dealing with solid quark stars in this paper .
the quark cooper pairing of the bcs type is suggested in quark matter of low - temperature but high baryon density , which may result in a color superconducting state @xcite , with a large pairing gap on the order of 100 mev .
this kind of condensation in momentum space takes place in case of same fermi momenta ; whereas `` loff''-like state may occur if the fermi momenta of two ( or more ) species of quarks are different @xcite . for 3 flavors of massless quarks , all nine quarks pair in a pattern which locks color and flavor symmetries , as called color - flavor locking ( cfl ) state @xcite
however , for such quark matter , there exists a _ competition _ between color superconductivity and solidification , just like the case of laboratory low - temperature physics .
one needs weak - interaction and low - mass in order to obtain a quantum fluid before solidification .
this is why only helium , of all the elements , shows superfluid phenomenon though other noble elements have similar weak strength of interaction due to filled crusts of electrons .
the strong color interaction ( and the coulomb interaction in the system with strangeness ) may be responsible for a possible solidification of dense quark matter with low temperatures .
further experiments ( in low - energy heavy ion colliders ) may answer whether quark matter is in a state of solid or color - superconductivity .
can a solid neutron star be possible ?
the answer might be _ no _ , because at least the part of neutron matter with approximate nuclear saturation density should be in a fluid state . in this sense , only solid quark matter is possible , and a quark star is identified if one convinces that a pulsar is in a solid state .
we have assumed that the entire strain energy @xmath48 is relieved in a quake , which results in the reference oblateness of eq.([epsiloni ] ) .
however , it is possible that not entire , but most of , the energy @xmath48 is released in a real situation , and the actual reference points are near but larger than @xmath36 . of course , there is a tendency of @xmath132 after the @xmath107th quake , but the effective shear modulus , @xmath133 , of matter broken could be much smaller than that of perfect elastic solid , @xmath65 .
the recovery timescale could be @xmath134 days if @xmath135 is order of @xmath136 erg/@xmath1 .
a large vela glitch on 2000 january 16.319 was noted @xcite , and _
observations were carried out @xmath137 and @xmath138 days after the glitch @xcite , but no temperature change expected in conventional models with released thermal energy of @xmath139 ergs is detected .
this could be understood in this starquake model since ( 1 ) the thermal conductivity of quark matter is much larger than that of hadron matter and ( 2 ) the thermal energy released , @xmath84 , to be much smaller than @xmath140 ergs is possible .
_ acknowledgments _ : this work is supported by national nature sciences foundation of china ( 10273001 ) , and by the special funds for major state basic research projects of china ( g2000077602 ) . the valuable suggestions from an anonymous referee are sincerely acknowledged . | a starquake mechanism for pulsar glitches is developed in the solid quark star model .
it is found that the general glitch natures ( i.e. , the glitch amplitudes and the time intervals ) could be reproduced if solid quark matter , with high baryon density but low temperature , has properties of shear modulus @xmath0 erg/@xmath1 and critical stress @xmath2 erg/@xmath1 .
the post - glitch behavior may represent a kind of damped oscillations .
_ pacs codes : _
97.60.g , 97.60.j , 11.80.f , , , , pulsars , neutron stars , elementary particles |
in most countries , learning microscopic origin of nature is considered to be an important topic in science education . however , there are not so many student experiments or demonstrations which exhibit the existence of individual atom in direct and intuitive manner .
the cloud chamber experiment is a rare exception ; it provides most direct and intuitive way to convince students existence of microscopic particles .
beautiful tracks of particles draw audience s attention and interest .
various chambers have been developed and used in classroom @xcite .
most of simple chambers seem to be based on the results of needles and nielsen @xcite and cowan @xcite , which use a block of dry ice and a beaker filled with ethanol vapour .
the great simplicity of their chamber enables students to make diy chamber at home or classroom .
such a chamber works well if one uses a radioactive source . from spring 2008 , one of the authors ( s.z . )
has been working on classroom experiment program using cloud chamber .
an one - day experiment course had been given for local junior high school students . from spring 2011 , some students ( rest of the authors ) have joined this project as a part of super science high school ( ssh ) activity .
the typical setting of a chamber is shown in figure [ typicalchamber ] .
a chamber is not so large ; smallest example is a glass laboratory dish with radius 5 cm and height 1 cm .
a standard chamber we have used is a round glass container with radius 8.5 cm and 8.0 cm high @xcite .
such chambers require a _ radioactive source . _
it is very hard to find particle tracks of background radiation in this kind of chamber .
one will find few tracks in a minute only in very dark room . in japan
, people s attention to radioactivity is increased much after the fukushima daiichi nuclear disaster following the tohoku earthquake and tsunami on 11 march 2011 .
large part of public unease or fear about radioactivity is due to insufficient education about radioactivity , since curricula of primary and secondary schools lack serious and extensive study of radioactivity .
after the tohoku earthquake , people have been nervous about radioactivity . in particular , parents do not wish their children to join scientific activity using radioactive source , even if it is relatively safe one such as a gas mantle with thorium .
therefore , we think it is very important to develop a sensitive cloud chamber which _ does not require any radioactive source . _ the chamber should be simple , small , portable and cheap since our aim is educational applications such as student experiment course or a workshop for citizens in tohoku area . in this article
, we would like to present a construction of a cloud chamber works well without radioactive sources .
there is only minor modification from a typical chamber shown in this section , but its performance is remarkable .
we also present a result of a performance test .
the construction of our cloud chamber is shown in figure [ fig : chamber ] . a major difference from a common chamber is use of a black anodised aluminium heat
sink as a bottom plate instead of a metal plate .
its dimension is 134 mm @xmath0 134 mm @xmath0 20 mm , weight is 622 g , and a side with fins is placed downward . as we see later ,
the heat sink greatly improves performance of our chamber .
our method has an additional advantage that the heat sink plays a role of supporting base of the chamber which prevents direct contact of side wall with refrigerant .
clear walls without a bottom and a top are placed on the heat sink .
walls and top are made of square , 5 mm thick acrylic plates .
top plate is removable . a piece of black felt e is placed along three sides of a box . upper half of it is soaked with ethanol .
total amount of ethanol is about 10 g for each experiment .
light from pc projector is introduced from a face without a piece of black felt .
all of the above instruments are placed on a shallow styrofoam tray b. liquid nitrogen can be directory poured into this box to this tray .
powder of dry ice can be used also . in this case , a heat sink is placed above powdered dry ice . in the placement
mentioned above , tracks of particles can be observed from above . figure [ fig : working ] is a photo of our cloud chamber in operation .
a pc projector is slanted to obtain fine view of particle tracks . in closing this section
, we would like to mention that our result is not new .
an advantage of our chamber is good performance in spite of its smallness and simpleness .
in fact , various examples of sensitive chambers have been known in japan .
very large and expensive chambers for display are available in some science museums @xcite .
a glass chamber presented in the beginning of this section is also available as a product by rado ltd .
a use of a small heat sink with liquid nitrogen is first presented in @xcite .
another example of a sensitive chamber is given in @xcite .
surprisingly , the very simple construction presented in the previous section is enough to see many particle tracks without any radioactive sources . external electric field is also unnecessary .
we would like to describe operation of a cloud chamber . to begin with
, liquid nitrogen was poured to the styrofoam box .
our lab was at room temperature , so the temperature of the bottom heat sink rises as liquid nitrogen evaporates . during the observation ,
this change of the temperature was measured by a thermocouple .
a result is shown in figure [ fig : bottomtemp ] .
obtained data fits well to a quadratic curve .
particle tracks begin to appear in few minutes . here
we present our result with a video @xcite uploaded to youtube .
figure [ fig : alphabeta ] is captured image from the video .
most thin and wiggy tracks , such as the left picture of figure [ fig : alphabeta ] frequently seen in the chamber , can be considered as those of beta particles .
we can see long tracks since the bottom plate is cooled uniformly .
some of long tracks are slightly curved even though no external magnetic field is imposed .
we think it is due to multiple collision with atoms in the air .
an alpha particle also can be observed as a short , thick line such as the right picture in figure [ fig : alphabeta ] .
one can see that many droplets are formed along a track and fall down toward bottom of the chamber .
although many tracks are seen in the video , there should be no influence of fukushima daiichi nuclear disaster , since in our city ( yokote , akita , japan ) has never observed apparent rise of air dose rate since 11th march .
obtaining of high quality video is important both for scientific analysis and publication on the web .
the later is especially important in for education , since lower quality videos found in the web are not enough to tell the beauty and excitement that people must feel in this experiment .
we find that a digital slr camera ( in our case , eos kiss x3 ) is very useful for this purpose .
it has larger ccd and better lens required for high quality video . in our experiment , a camera is directly put on the top plate as in figure [ fig : camera ] , but use of rigid is better if available .
a qualitative evaluation of the performance of a chamber will be useful for development of new cloud chamber and finding better working condition .
in fact , some test runs tells us that numbers of particle tracks seems to differ for each run . using a recorded video @xcite , we counted all tracks seen in the chamber regardless of its length , width and shape . therefore alpha particles , beta particles and other particles are not distinguished .
figure [ fig : countpermin ] shows a relation between count per 10 seconds and elapsed time .
we observe that , although fluctuation is large , the count is independent of the bottom temperature until it reaches to ` critical ' value .
our observation shows the fall of the count begins around @xmath1 degrees . ,
@xmath2 and @xmath3 degrees respectively .
black and white are converted .
, scaledwidth=90.0% ] while a number of particle tracks does not change below critical temperature , a number of background droplets affects visibility of particle tracks .
many droplets are seen at lower temperature as shown in the left of figure [ fig : droplets ] .
contrast of whole image is reduced at such low temperature .
we also found that alpha particle tracks become thinner at lower temperature , as mentioned in @xcite . on the other hand
, it becomes hard to identify particle tracks at high temperature .
therefore , the best temperature range for observation should be middle region in figure [ fig : countpermin ] .
we expect our cloud chamber will be useful for various application to education , for example , a student research project or demonstrations in public . in particular , it will be important to measure numbers of particle tracks in a city where dose rate is still high . also not examined in this paper , the high quality video enables us various computer analysis such as qualitative evaluation of amount of condensation or automatic counting of particle tracks
. such kind of analysis will be useful for a student s research projects .
we would like to thank hiroki kanda of tohoku university for useful comments at the beginning of our project , ichiro itoh of yokote seiryo gakuin h.s . for providing a voltage multiplier , ryoko yamaishi for support in lab .
we also thank kenichiro aoki of keio university for careful reading of the manuscript .
this work is supported by saito kenzo honour fund and the super science high school funding from japan science and technology agency .
99 andy foland s cloud chamber page http://w4.lns.cornell.edu/~adf4/cloud.html , how to build a cloud chamber !
http://www.youtube.com/watch?v=pewtysxftqk , how to make a diffusion cloud chamber http://www.youtube.com/watch?v=pvcdaa_vvvg needels t s and neilsen c e 1950 a continuously sensitive cloud chamber rev .
21 976 cowan e w 1950 a continuously sensitive cloud chamber rev .
21 991 rado ltd .
http://www.kiribako-rado.co.jp/goods-m.html mori y 1994 let s see positorn in a cloud chamber using liquid nitorogen toray science education award 26 18 , hayashi h 2007 a sensitive cloud chamber in magnetic field as a teaching material of atomic physics j. phys .
55 297 sensitive cloud chamber http://www.youtube.com/watch?v=hcv3fdz1rfk y mori http://www.youtube.com/watch?v=mwm-wo7dokw http://www.youtube.com/watch?v=r9swciiqwvm slatis h 1957 on diffusion cloud chambers nucl . instrum . 1 213 | we present a sensitive diffusion cloud chamber which does not require any radioactive sources . a major difference from a commonly used chamber
is use of a heat sink as its bottom plate .
a result of a performance test of the chamber is given . |
we demonstrate an alternative way to perform compressive cgi ( ccgi ) without any computational overhead once all measurements have been performed by using an adaptive measurement scheme .
we follow a promising strategy for adaptive compressive sensing that suggests replacing the random speckle patterns by directly using the patterns that form the sparse basis @xcite .
we start the discussion of our strategy by recalling the properties of the 2d haar wavelet transform of square images consisting of @xmath2 pixels .
the wavelet decomposition procedure is schematically depicted in figure [ monkey ] .
the decomposition of the image @xmath3 is performed seperately for rows and columns . at first each row is divided into @xmath4 pairs of adjacent pixels .
the partial wavelet transform @xmath5 now consists of the sum and the difference of these adjacent pixels according to the following rules for @xmath6 : @xmath7 repeating that procedure for each column in @xmath8 according to similar rules for @xmath9 gives the full transform @xmath10 : @xmath11 the resulting transform now consists of four quadrants .
the upper left quadrant represents a coarse version of the original image , while the other three quadrants contain information about horizontal , vertical and diagonal edges .
one may now continue and perform another wavelet transform on the upper left quadrant and iteratively repeat this procedure until the pixel in the upper left corner contains the mean intensity of the picture and all other pixels contain information about edges .
now each additional transform performed corresponds to a coarser scale @xmath12 with wavelet coefficients spanning over larger regions , but carrying information over a smaller range of frequencies .
such wavelet representations are efficient in terms of image compression .
only a small portion of natural images consists of edges and only wavelet coefficients corresponding to regions with sharp edges are large , therefore only few large coefficients are sufficient to approximate the full image . as can be seen in figure [ monkey ]
, the number of large wavelet coefficients ( shown in white ) is rather small . 512 pixel baboon test image ( left panel ) and its one - step ( middle panel ) and complete ( right panel ) wavelet transform .
for the transform absolute values of the wavelet coefficients are shown .
white regions correspond to large wavelet values and mark regions with strong edges .
every wavelet coefficient at scale @xmath12 contains information about four pixels of the coarse image of size @xmath13 . also , every wavelet coefficient has four children wavelet coefficients at scale @xmath14 which contain information about the same range of the image . ]
this strategy becomes interesting as the wavelet transformation is hierarchic .
every parent coefficient at some coarse scale has four children coefficients at the next finer scale covering the same spatial region . as it is very likely that the children wavelet coefficients belonging to parent coefficients which are small
will also be small , this offers a great opportunity for image compression in terms of wavelet trees @xcite by cutting of these trees at an adequate scale .
we follow a similar strategy and first take a coarse image of size @xmath15 .
experimentally , this is realized by inserting a phase - only slm ( holoeye - pluto ) in the path of a laser beam polarized such that the slm only introduces a phase shift to it .
the phase pattern imprinted on the slm is the fourier transform of a square superposed with the phase map of a lens . as a consequence , in the focal plane behind the slm
the square is recovered in the spatial intensity pattern of the light beam .
we precomputed 87040 of such phase patterns using an iterative numerical technique based on the adaptive - additive algorithm @xcite .
65536 of these form the pixels of a 256@xmath16256 ( @xmath12=1 ) pixel square .
the other patterns form the pixels of squares of the same size , but consisting of fewer ( 128@xmath16128 ( @xmath12=2 ) , 64@xmath1664 ( @xmath12=3 ) and 32@xmath1632 ( @xmath12=4 ) ) , but larger pixels of size @xmath17 , respectively .
the object to be imaged is placed at the focal plane of the slm ( @xmath18=36 cm ) and the transmission through that object is measured . under the conditions used throughout the manuscript , the whole square has a side length of 32 mm .
when the coarse image is taken , we perform a one - step wavelet transform on it .
now we check the absolute values of the wavelet coefficients corresponding to edges against a predefined threshold @xmath19 .
if the values are larger than @xmath19 , the four children wavelet values at the next finer scale @xmath12 - 1 are measured too . as each wavelet coefficient spans over exactly four pixels at its scale
, it is never necessary to perform more than four measurements in real space to determine any wavelet value .
once all the measurements at the finer scale have been performed , a new finer image can be constructed .
it consists of the newly measured transmission values for regions containing sharp edges and of the transmission values already measured at a coarser scale for regions without edges .
now another one - step wavelet transform is performed on this finer image and again all wavelet values are checked against a new threshold @xmath20 .
this process is repeated until the finest picture at scale @xmath12=1 is constructed .
a summary of the imaging process is presented in fig .
[ example ] .
= 1 is reached or no wavelet coefficient is larger than the threshold value ( lower left panel ) .
the result is then converted back to a real space image using the inverse wavelet transform ( lower right image ) .
for this example the number of measurements needed is roughly 40@xmath21 of the number of pixels present in the image .
note that the upper right , lower left and lower right sector of the wavelet transform correspond to horizontal , vertical and diagonal edges , respectively .
wavelet coefficients have been multiplied by 8 to enhance contrast . ]
we tested the ccgi algorithm using a metal plate containing twelve holes as a test target .
we chose to use a threshold which becomes sharper at finer scales ( @xmath20=2@xmath19 ) and varied the initial threshold @xmath22 , resulting in images of differing quality .
the results are shown in figure [ lochblende ] . here
the transmission maps quantized to 256 greyscales are shown in terms of the decreased acquisition rate @xmath23 , which is the total number of measurements performed on all scales divided by the total number of pixels present on the finest scale ( @xmath0=65536 ) .
the transmission is normalized to the empty space transmission to account for possible inhomogeneities introduced by the slm . gives the decreased acquisition rate .
a faithful image of the plate is already possible with approximately 5 - 7@xmath21 of the measurements required to record every single pixel in full resolution . ]
as can be seen , the image is reproduced quite well at relatively small @xmath23 . at around 2@xmath21
the quality is already sufficient for distinguishing the holes and counting their number .
for @xmath23 around 4@xmath21 the image already looks reasonable . for @xmath23 around 7@xmath21
the recorded image shows good quality . for larger
@xmath23 only small improvements are seen . however , to get a more quantitative measure of the recorded image quality , we calculated the mean squared error @xmath24 ^ 2\ ] ] of the measured image as compared to the reference construction drawing @xmath25 of the metal plate containing the holes . 7@xmath21 . ]
the impression that the image quality does not improve significantly for @xmath267@xmath21 is verified .
the mean squared error roughly follows an exponential decay and saturates approximately at a value of 0.055 for large @xmath23 .
a closer examination of this residual error shows that it is mainly caused by the edges of the holes .
in contrast to the construction drawing , the edges between full transmission and zero transmission are not positioned at pixel borders .
therefore the pixels at the edges show some intermediate transmission and introduce some deviations from the reference image .
= 0.369 . ]
the number of necessary measurements needed for near optimal reconstruction of an image obviously depends on the number of large wavelet coefficients that image carries . in order to demonstrate the adaptivity of our technique
, we kept the threshold setting used for measuring the metal plate at @xmath23=0.131 in figure [ lochblende ] and imaged a more complex object - a 1951 usaf resolution test chart .
the recorded image is shown in figure [ usaf ] .
the image quality is still good , but the algorithm automatically took almost three times more measurements than were needed for the metal plate , resulting in @xmath23=36.9@xmath21 .
the image resolution is reasonable .
one pixel has a length of about 125@xmath27 m in real space which is roughly the size of the lines on the test plate which can still be resolved .
nevertheless the image shows some artifacts .
these are a consequence of a weakness of the algorithm used .
strictly periodic structures like several parallel lines placed next to each other may look like having no edge at all at a coarser scale .
however , such problems may be overcome by more advanced algorithms relying on taking more than just the parent wavelet value into account when deciding on whether a certain wavelet value should be measured or not @xcite .
+ our technique offers a wide range of advantages .
as it is adaptive , one has control over the quality of the image in advance by choosing the thresholds .
the algorithm does not require any additional computationally intensive recovery algorithms needed for standard compressive sensing techniques using pseudorandom illumination patterns .
our technique works reasonable fast .
the slm can be operated at up to 60hz .
photodiodes and readout circuitry working on the same timescale are common today , opening up the possibility to record images of reasonable resolution and quality within few minutes . in particular experimental techniques requiring single pixel detectors like lock - in detection for sensitive pump - probe measurements may benefit from our results .
spatially resolved measurements are to the best of our knowledge not carried out with high resolution using such techniques due to the long measurement durations that would be needed . reducing the number of necessary measurements by a factor of at least fifteen opens up the way to perform such measurements with high spatial resolution .
if the duration of the measurement is more crucial than the image quality , our approach also allows one to perform a fixed number of finer measurements for a preset number of largest wavelet coefficients at each scale instead of using thresholds .
in that way it is possible to take an image in a fixed amount of time .
in conclusion we have developed an adaptive ccgi technique that allows us to record images using a single pixel camera at an acquisition rate fifteen to twenty times below the shannon limit by recording the image directly in a sparse basis .
a number of further research directions arise from our work .
compressive imaging techniques are not limited to recording image information , but have also found usage far beyond simple imaging applications in fields like quantum process tomography @xcite , optical encryption @xcite , fluorescence microscopy and hyperspectral imaging @xcite . also our approach is quite generic . optimized approaches which also take the magnitude of neighboring wavelet coefficients into account @xcite instead of just the parent wavelet coefficient may lead to increased image quality or smaller values of @xmath23 .
also , it is well known that especially designed measurement matrices can drastically reduce the number of needed measurements for exact image reconstruction using seeded belief propagation techniques@xcite .
finally , it is not strictly necessary to use precomputed phase patterns , but one could compute them on the fly , thereby allowing one to even choose an adaptive wavelet basis .
yet , the greatest strength of our approach lies in drastically reducing the needed measurement time for high - resolution images using single - pixel detectors without having any need for computational image reconstruction .
+ we would like to conclude this paper by a comparison between ccgi and standard random gaussian matrix based compressive sensing techniques ( rmcs ) to identify the strengths and weaknesses of our approach in more detail .
obviously , having the image available once all the measurements are taken , is an advantage , but it is also introduces a drawback : ccgi needs to use a predefined sparse basis , while rmcs will automatically find an adequate sparse basis during reconstruction .
accordingly , it is typically possible to achieve a near - ideal reconstruction with fewer measurements using rmcs .
however , it should be noted that the exact number of measurements needed for near - ideal reconstruction is usually not known a priori as it depends on the sparsity of the image .
the needed number of measurements is also unknown in ccgi , but specifying the desired image quality in terms of the threshold @xmath19 ensures that not too many measurements are made .
the exact number of measurements needed for ccgi is hard to predict as it depends on how sparse an image is in the wavelength basis .
for very sparse images , the penalty can be as much as 50% . for less sparse images ,
the differences are less drastic . however , for a comparable number of measurements , ccgi - based methods tend to achieve a better signal to noise ration than rmcs methods do .
see @xcite for a detailed comparison of a slightly modified version of ccgi with state of the art rmcs techniques .
another important benchmark is the performance of compressive sensing techniques in the presence of noise . in ccgi noise
can become a severe problem if the noise magnitude becomes comparable to the threshold chosen .
ccgi is therefore not the method of choice for measuring images containing strong noise or weak signals .
another issue is scalabilty .
going to larger images , increases the necessary number of measurements a computations during data acquisition in ccgi and the complexity of the minimization problem in rmcs .
for all the image sizes we examined , the time needed for performing the measurements was always so much longer than the time needed for performing the wavelet transforms and building the sampling queue that no delay was noticeable . in summary , although other compressive sensing methods based on gaussian random matrices approaches typically need fewer measurements than most ( but not all @xcite ) techniques using deterministic matrices , having the result immediately renders adaptive spatially resolved pump - probe spectroscopy and other delicate spectroscopic techniques with high resolution possible . therefore , we suggest that our technique is well suited for specialized complex problems in physics and spectroscopy which are a priori known to be reasonably sparse in the wavelet basis , while rmcs methods are a much better choice for taking single images , for images where noise is an issue and for images where the sparsest basis is unknown .
the objects to be imaged were placed at the focal plane of the slm ( @xmath18=36 cm ) and the transmission through the objects was measured by a standard commercial photo diode onto which the transmission through the object was focused and a keithley 2000 multimeter was used for measuring the photo diode output .
the hole test plate used consisted of twelve holes .
six of these holes had a diameter of 2 mm , while the other six holes had a diameter of 1 mm .
their average separation was around 1.5 mm .
the laser used for the transmission measurements was a pulsed ti : sapph laser emitting pulses wih a duration of approximately 2ps at a wavelength of 800 nm .
10 url # 1`#1`urlprefix[2]#2 [ 2][]#2 & .
_ _ * * , ( ) . , , & .
_ _ * * , ( ) . , , & .
_ _ * * , ( ) . , & .
_ _ * * , ( ) . , , & .
_ _ * * , ( ) . & .
_ _ * * , ( ) . .
_ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) . , & .
_ _ * * , ( ) . , & .
_ _ * * , ( ) . , , & .
_ _ * * , ( ) . & .
_ _ * * , ( ) . , &
_ _ * * , ( ) .
. _ _ * * , ( ) . , , , & .
_ _ * * , ( ) .
_ et al . _ . _
_ * * , ( ) .
_ _ * * , ( ) . , , , &
_ _ * * , ( )
. _ et al .
_ _ * * , ( ) . , , , & .
_ _ * * , ( ) .
_ _ * * , ( ) .
m.a . designed the experiment and analyzed the data . m.a . and m.b .
wrote the manuscript .
the authors declare no competing financial interests . | compressive sensing is considered a huge breakthrough in signal acquisition .
it allows recording an image consisting of @xmath0 pixels using much fewer than @xmath0 measurements if it can be transformed to a basis where most pixels take on negligibly small values .
standard compressive sensing techniques suffer from the computational overhead needed to reconstruct an image with typical computation times between hours and days and are thus not optimal for applications in physics and spectroscopy .
we demonstrate an adaptive compressive sampling technique that performs measurements directly in a sparse basis .
it needs much fewer than @xmath0 measurements without any computational overhead , so the result is available instantly .
computational ghost imaging ( cgi ) is a novel imaging technique that has received significant attention during the last few years@xcite .
it is a consequent further development of conventional ghost imaging @xcite which allows to record spatially resolved images using a detector without spatial resolution . in conventional ghost imaging
the image is recorded using two spatially correlated light beams , one object and one reference beam .
the object beam illuminates the object to be imaged and is then collected using a bucket detector .
the reference beam never interacts with the object and is recorded using a pixelated device offering spatial resolution .
as both beams are spatially correlated the coincidence count signal allows one to retrieve a ghost image of the object .
ghost imaging using both entangled photons @xcite or classical light @xcite as the spatially correlated twin beam source has been demonstrated . a seminal paper by shapiro @xcite clarified that the sole purpose of the reference beam lies in determining the illumination pattern at the object position at each instant , while the object beam gives data about the transmission of this pattern through the object .
therefore , if one can create a deterministic illumination pattern at the object position , the reference beam becomes obsolete and cgi using just a single beam and a single pixel detector @xcite becomes possible .
this approach has been realized using deterministic speckle patterns created using a spatial light modulator ( slm ) @xcite .
it has also been demonstrated that this technique also offers the possibility to perform compressive sensing @xcite , that is it is possible to reconstruct an image consisting of @xmath0 pixels using much less than @xmath0 measurements by utilizing the fact that natural images are typically sparse@xcite : when transformed to an appropriate basis , they contain many coefficients that are zero or close to it . in practice
, the transmission measured for each speckle pattern constitutes a projection of the object image and compressive sensing is performed by utilizing an algorithm which checks all the possible images which are consistent with the projections performed and finds the image which is the sparsest one .
usually the @xmath1-norm serves as a measure of sparsity and the image which minimizes it , is the optimal reconstruction of the object .
however , this method still has some drawbacks .
the time taken by the reconstruction algorithm can become very long for large images and one needs to compute the speckle pattern at the object position by using the fresnel - huygens propagator on the phase pattern imprinted on the slm .
while the latter is not a big problem - one can calculate the speckle pattern once and reuse the phase pattern masks - the computational overhead , given by the computational effort once all measurements have been made , is a huge problem .
the overhead becomes especially problematic considering typical problems in spectroscopy ( e.g. pump - probe spectroscopy ) , where many similar images need to be taken , while one experimental parameter is changed .
here , it is desirable to have the reconstructed image directly , so one can use this information when taking the next image . for example one could adaptively scan the previous image for regions of large values or strongly varying values and scan these areas with higher resolution in the next image . |
today , bianchi type ix enjoys an almost mythical status in general relativity and cosmology , which is due to two commonly held beliefs : ( i ) type ix dynamics is believed to be essentially understood ; ( ii ) bianchi type ix is believed to be a role model that captures the generic features of generic spacelike singularities .
however , we will illustrate in this paper that there are reasons to question these beliefs .
the idea that type ix is essentially understood is a misconception .
in actuality , surprisingly little is known , i.e. , proved , about type ix asymptotic dynamics ; at the same time there exist widely held , but rather vague , beliefs about mixmaster dynamics , oscillations , and chaos , which are frequently mistaken to be facts . there is thus a need for clarification : what are the known facts and what is merely believed about type ix asymptotics
? we will address this issue in two ways : on the one hand , we will discuss the main rigorous results on mixmaster dynamics , the ` bianchi type ix attractor theorem ' , and its consequences ; in particular , we will point out the limitations of these results . on the other hand
, we will provide the infrastructure that makes it possible to sharpen commonly held beliefs ; based on this framework we will formulate explicit refutable conjectures . historically , bianchi type ix vacuum and orthogonal perfect fluid models entered the scene in the late sixties through the work of belinskii , khalatnikov and lifshitz @xcite and misner and chitr @xcite .
bkl attempted to understand the detailed nature of singularities and were led to the type ix models via a rather convoluted route , while misner was interested in mechanisms that could explain why the universe today is almost isotropic .
bkl and misner independently , by means of quite different methods , reached the conclusion that the temporal behavior of the type ix models towards the initial singularity can be described by sequences of anisotropic kasner states , i.e. , bianchi type i vacuum solutions .
these sequences are determined by a discrete map that leads to an oscillatory anisotropic behavior , which motivated misner to refer to the type ix models as mixmaster models @xcite .
this discrete map , the kasner map , was later shown to be associated with stochasticity and chaos @xcite , a property that has generated considerable interest and confusion ,
see , e.g. , @xcite and references therein . a sobering thought : all claims about chaos in einstein s equations rest on the ( plausible ) belief that the kasner map actually describes the asymptotic dynamics of einstein s equations ; as will be discussed below , this is far from evident ( despite being plausible ) and has not been proved so far .
more than a decade after bkl s and misner s investigations a new development took place : einstein s field equations in the spatially homogeneous ( sh ) case were reformulated in a manner that allowed one to apply powerful dynamical systems techniques @xcite ; gradually a picture of a hierarchy of invariant subsets emerged where monotone functions restricted the asymptotic dynamics to boundaries of boundaries , see @xcite and references therein .
based on work reviewed and developed in @xcite and by rendall @xcite , ringstrm eventually produced the first major proofs about asymptotic type ix dynamics @xcite .
this achievement is remarkable , but it does not follow that all questions are settled . on the contrary , so far nothing is rigorously known , e.g. , about dynamical chaotic properties ( although there are good grounds for beliefs ) , nor has the role of type ix models in the context of generic singularities been established @xcite .
the outline of the paper is as follows . in section
[ basic ] we briefly describe the hubble - normalized dynamical systems approach and establish the connection with the metric approach . for simplicity
we restrict ourselves to the vacuum case and the so - called orthogonal perfect fluid case , i.e. , the fluid flow is orthogonal w.r.t . the sh symmetry surfaces
; furthermore , we assume a linear equation of state . in section [ subsets ]
we discuss the levels of the bianchi type ix so - called lie contraction hierarchy of subsets , where we focus on the bianchi type i and type ii subsets . in section [ nongeneric ]
we present the results of the local analysis of the fixed points of the dynamical system and discuss the stable and unstable subsets of these points which are associated with non - generic asymptotically self - similar behavior .
section [ maps ] is devoted to a study of the network of sequences of heteroclinic orbits ( heteroclinic chains ) that is induced by the dynamical system on the closure of the bianchi type ii vacuum boundary of the type ix state space ( which we refer to as the mixmaster attractor subset ) .
these sequences of orbits are associated with the mixmaster map , which in turn induces the kasner map and thus the kasner sequences .
we analyze the properties of non - generic kasner sequences and discuss the stochastic properties of generic sequences . in section [ furthermix ]
we discuss the main ` mixmaster facts ' : ringstrm s ` bianchi type ix attractor theorem ' @xcite , theorem [ rinthm ] , and a number of consequences that follow from theorem [ rinthm ] and from the results on the mixmaster / kasner map .
in addition , we introduce and discuss the concept of ` finite mixmaster shadowing ' . in the subsection
` attractor beliefs ' of section [ stochasticbeliefs ] we formulate two conjectures that reflect commonly held beliefs about type ix asymptotic dynamics and list some open issues that are directly connected with these conjectures . in the subsection ` stochastic beliefs ' we address the open question of which role the mixmaster / kasner map and its stochastic properties actually play in type ix asymptotic dynamics .
this culminates in the formulation , and discussion , of two ` stochastic ' conjectures . in section [ billiard ]
we present the hamiltonian billiard formulation , see @xcite or @xcite ; we demonstrate that this approach yields a ` dual ' formulation of the asymptotic dynamics .
we point out that the billiard approach is a formidable heuristic picture , but fails to turn beliefs into facts .
we conclude in section [ concl ] with a discussion of the main themes of this paper . throughout this paper
we use units so that @xmath0 and @xmath1 , where @xmath2 is the speed of light and @xmath3 the gravitational constant .
we consider vacuum or orthogonal perfect fluid sh bianchi type ix models ( i.e. , the fluid 4-velocity is assumed to be orthogonal to the sh symmetry surfaces ) with a linear equation of state ; we require the energy conditions ( weak / strong / dominant ) to hold , i.e. , @xmath4 and @xmath5 where @xmath6 , and where @xmath7 and @xmath8 are the energy density and pressure of the fluid , respectively . by
we exclude the special cases @xmath9 and @xmath10 , where the energy conditions are only marginally satisfied . , see @xcite . the case @xmath10 is known as the stiff fluid case , for which the speed of sound is equal to the speed of light .
the asymptotic dynamics of stiff fluid solutions is simpler than the oscillatory behavior characterizing the models with range @xmath11 , and well understood @xcite .
( in the terminology introduced below , the stiff fluid models are asymptotically self - similar . )
we will therefore refrain from discussing the stiff fluid case in this paper .
] as is well known , see , e.g. , @xcite and references therein , for these models there exists a symmetry - adapted frame @xmath12 , [ bixform ] @xmath13 hence , the type ix models naturally belong to the so - called class a bianchi models , see table [ classamodels ] . let @xmath14 where @xmath15 .
furthermore , define @xmath16 where @xmath17 denotes the second fundamental form associated with of the sh hypersurfaces @xmath18 .
the quantities @xmath19 and @xmath20 can be interpreted as the expansion and the shear , respectively , of the normal congruence of the sh hypersurfaces . in a cosmological context
it is customary to replace @xmath19 by the hubble variable @xmath21 ; this variable is related to changes of the spatial volume density according to @xmath22 .
evidently , in bianchi type ix ( and type viii ) there is a one - to - one correspondence between the ` orthonormal frame variables ' @xmath23 ( with @xmath24 ) and @xmath25 ; in particular , the metric @xmath26 is obtained from @xmath27 via .
( for the lower bianchi types i@xmath28 , some of the variables @xmath27 are zero , cf .
; in this case , the other frame variables , i.e. , @xmath29 , are needed as well to reconstruct the metric ; see @xcite for a group theoretical approach . )
.the class a bianchi types are characterized by different signs of the structure constants @xmath30 , where @xmath31 is any permutation of @xmath32 .
in addition to the above representations there exist equivalent representations associated with an overall change of sign of the structure constants ; e.g. , another type ix representation is @xmath33 . [ cols="^,^,^,^",options="header " , ] for all class a models except type ix the constraint implies that @xmath34 ; in bianchi type ix , however , @xmath35 is possible . for type ix ,
define @xmath36 employing and and using that @xmath37 + \delta \geq 0\:,\ ] ] where equality holds iff @xmath38 , we find @xmath39 the function @xmath40 is strictly monotonically increasing along orbits of bianchi type ix . to see this we use and compute @xmath41 \delta\:,\ ] ] where we note that @xmath42 because of the constraints .
in combination with it follows that @xmath43 and @xmath44 are _ bounded towards the past_. the right hand side of the reduced system ( [ ixeq ] ) consists of polynomials of the state space variables and is thus a regular dynamical system .
solutions of of bianchi types i viii are global in @xmath45 , since implies the bounds @xmath34 and @xmath46 which control the evolution of @xmath47 in .
solutions of of bianchi type ix are global towards the past , since @xmath48 is bounded from below ; this follows from , which yields @xmath49 , so that @xmath50 . the decoupled equation for @xmath51 , cf . ,
yields that @xmath52 as @xmath53 , because @xmath54 is non - negative .
since @xmath54 is bounded as @xmath55 , the asymptotics of @xmath51 can be bounded by exponential functions from above and below .
it follows that the equation @xmath56 can be integrated to yield @xmath57 as a function of @xmath45 such that @xmath58 as @xmath55 .
in addition to it is useful to also consider an auxiliary equation for the matter quantity @xmath44 , @xmath59\omega\:.\ ] ] making use of and we conclude that for all orthogonal perfect fluid models with a linear equation of state we have @xmath60\tau\right)$ ] , and hence @xmath61 as @xmath53 , which yields a past singularity .
the divergence of @xmath7 can also be directly read off from the matter equation @xmath62 .
the bianchi type ix state space @xmath63 is characterized by the conditions @xmath64 , @xmath65 , @xmath66 .
we write @xmath67 .
the notation is such that the subscript denotes the non - zero variables among @xmath68 .
setting one or more of these variables to zero ( which corresponds to lie contractions @xcite ) yields invariant boundary subsets which describe more special bianchi types .
since the type ix models exhibit discrete symmetries associated with axes permutations , the contractions generate all possible representations of the more special ( lie contracted ) bianchi types ( which are associated with such permutations ) : the bianchi type @xmath28 subspace @xmath69 is given by the disjoint union of three equivalent sets , @xmath70 , where , e.g. , @xmath71 denotes the type @xmath28 subset with @xmath72 , @xmath73 and @xmath74 ; the bianchi type @xmath75 subspace @xmath76 by the union @xmath77 ; the bianchi type @xmath78 subspace @xmath79 by @xmath80 .
note that the bianchi type @xmath81 subspace does not appear as a boundary subset of @xmath63 .
a bianchi subset contraction diagram for type ix is given in figure [ contraction ] .
[ cc][cc]@xmath82 [ cc][cc]@xmath71 [ cc][cc]@xmath83 [ cc][cc]@xmath84 [ cc][cc]@xmath85 [ cc][cc]@xmath86 [ cc][cc]@xmath87 [ cc][cc]@xmath88 [ rc][cc]@xmath63 [ rc][cc]@xmath69 [ rc][cc]@xmath76 [ rc][cc]@xmath79 each set of the lie contraction hierarchy is the union of an invariant vacuum subset , i.e. , @xmath89 , and an invariant fluid subset , i.e. , @xmath90 . to refer to a vacuum [ fluid ] subset of a bianchi set @xmath91 we use the notation @xmath92 [ @xmath93 . in this spirit , e.g.
, the type ii subset decomposes as @xmath94 . in the following
we analyze the boundary subsets to the extent needed in order to understand the asymptotic type ix dynamics .
the bianchi type i subset is given by @xmath95 , @xmath96 , @xmath74 and @xmath97 ; since @xmath98 , @xmath99 , @xmath100 vanish , we denote this subset by @xmath88 , cf .
figure [ contraction ] .
the _ vacuum subset _ consists of a circle of fixed points_the kasner circle _ @xmath101 , which is characterized by @xmath102 .
it is common to represent different points on @xmath101 in terms of the kasner exponents @xmath103 , @xmath104 each fixed point on @xmath101 represents a kasner solution ( kasner metric ) with the corresponding exponents .
the kasner circle is divided into six equivalent sectors , denoted by permutations of the triple @xmath32 , where sector @xmath31 is characterized by @xmath105 , see figure [ kasnercirc ] .
the boundaries of the sectors are six special points that are associated with solutions that are locally rotationally symmetric ( lrs ) : @xmath106 are given by @xmath107 or @xmath108 and yield the three equivalent lrs solutions whose intrinsic geometry is non - flat ; the _ taub points _
@xmath109 are given by @xmath110 or @xmath111 and correspond to the flat lrs solutions the taub representation of minkowski spacetime .
[ cc][cc]@xmath112 [ cc][cc]@xmath113 [ cc][cc]@xmath114 [ cc][cc]@xmath32 [ cc][cc]@xmath115 [ cc][cc]@xmath116 [ cc][cc]@xmath117 [ cc][cc]@xmath118 [ cc][cc]@xmath119 [ cc][cc]@xmath120 [ cc][cc]@xmath121 [ cc][cc]@xmath122 [ cc][cc]@xmath123 [ cc][cc]@xmath124 [ cc][cc]@xmath125 [ cc][cc]@xmath126 [ cc][cc]@xmath127 [ cc][cc]@xmath128 [ cc][cc]@xmath129 the quantity @xmath130 ( or , equivalently , @xmath131 ) is invariant under changes of the axes and thus naturally captures the ` physical essence ' of a solution independent of the chosen frame ; however , it is typically replaced by the _ kasner parameter _ @xmath132 through @xmath133\:.\ ] ] the kasner parameter @xmath132 parameterizes the kasner exponents uniquely ( up to the permutation symmetry ) ; we have @xmath134 for sector @xmath31 of @xmath101 , where @xmath135 .
therefore , each point on sector @xmath31 is represented by a unique value of @xmath136 . at the boundary points of sector @xmath31 , which are @xmath106 and @xmath137 , the kasner parameter is @xmath138 and @xmath139 , respectively .
permuting @xmath31 yields a physically equivalent state on a different sector ; accordingly , each @xmath135 represents an equivalence class of six points on @xmath140 .
in contrast , @xmath138 describes the three points @xmath141 ; @xmath139 yields @xmath142 . while the bianchi type i vacuum subset coincides with the kasner circle @xmath140 , which is given by @xmath143 , the bianchi type i perfect fluid subset is the set @xmath144 . from it
is straightforward to deduce that there exists a central fixed point , the _
friedmann fixed point _
@xmath145 , given by @xmath146 @xmath147 , which corresponds to the isotropic friedmann - robertson - walker ( frw ) solution . solutions with @xmath148 are given by radial straight lines originating from @xmath140 and ending at @xmath145 .
these results rely on the assumption @xmath149 , see .
let us consider the bianchi type ii subset @xmath150 given by @xmath151 , @xmath152 . in this case ,
the @xmath153-direction is singled out , while there exists a discrete symmetry associated with the interchange of the @xmath154- and @xmath155-direction .
the lrs subset @xmath156 is a subset of codimension one which divides the state space into two equivalent parts , the subsets @xmath157 and @xmath158 ( related by permuting the @xmath154- and @xmath155-axes ) . since the past singularity is of particular interest in our considerations , it is convenient to choose the time direction towards the past , i.e. , to use a reversed time variable @xmath159 according to @xmath160 which we do in the remainder of this section ; accordingly , approach to the past singularity means @xmath161 .
( in section [ maps ] we will see that vacuum type ii orbits are the building blocks for the mixmaster map and the closely related kasner map ( bkl map ) .
it is a well established convention that forward iterations of these maps are directed towards the singularity . to agree with this convention
the use of a past - directed time variable in a discussion of bianchi type ii models thus suggests itself . ) on @xmath150 , the gauss constraint @xmath162 can be used to replace @xmath163 by @xmath44 as a dependent variable .
the system thus becomes @xmath164\,,\ ] ] where @xmath165 and @xmath166 , @xmath167 ; we have @xmath168 .
let us first consider the vacuum subset @xmath169 , i.e. , @xmath89 .
there do not exist any fixed points in the type ii vacuum subset @xmath169 , but the boundary of the vacuum subset coincides with the kasner circle @xmath140 .
the orbits of form a family of straight lines in @xmath169 , where each orbit connects one fixed point on @xmath140 with another fixed point on @xmath140 ; hence each orbit is heteroclinic @xcite .
following the nomenclature of @xcite we call these heteroclinic orbits bianchi type ii _ transitions _ , because each orbit can be viewed as representing a transition from one kasner state to another .
we denote these transitions by @xmath170 , where each @xmath170 emanates from @xmath171 .
if the initial point is a point of sector @xmath172 , then the final point is a point of @xmath173 ; interchanging @xmath154 and @xmath155 yields the transitions emanating from @xmath174 ; if the initial point is @xmath175 , the final point is @xmath137 ; the points @xmath109 and @xmath176 are not connected with any other fixed point ( they are ` fixed points ' under the present ` type ii map ' ) , see figure [ typeii ] .
[ cc][cc]@xmath112 [ cc][cc]@xmath113 [ cc][cc]@xmath114 [ cc][cc]@xmath120 [ cc][cc]@xmath121 [ cc][cc]@xmath122 [ cc][cc]@xmath123 [ cc][cc]@xmath124 [ cc][cc]@xmath125 [ cc][cc]@xmath126 let @xmath177 denote the initial fixed point on @xmath140 ; this point can be represented in terms of the kasner parameter @xmath178 by using .
the orbit ( transition ) emanating from this fixed point is given in terms of an auxiliary function @xmath179 by [ zeqs ] @xmath180 + \eta\ , \sigma^{\mathrm{i}}_{\alpha/\beta } \ : , \qquad \sigma_{\gamma } = -4 [ 1 - \eta ] + \eta \,\sigma^{\mathrm{i}}_{\gamma}\:,\ ] ] where @xmath181 is determined by the equation @xmath182 and the conditions that @xmath183 and @xmath184 .
the quantity @xmath185 is in the interval @xmath186 for @xmath187 ; @xmath188 corresponds to @xmath189 and describes the orbit @xmath190 ; @xmath139 corresponds to @xmath191 and describes the ` isolated ' points @xmath192 , @xmath193 .
since @xmath181 increases from @xmath194 to @xmath185 as @xmath159 goes from @xmath195 to @xmath196 , @xmath185 is called the growth factor @xcite . in section
[ maps ] , the transitions @xmath170 , as represented by , will appear as the building blocks for the mixmaster / kasner map . while there do not exist any fixed points in the vacuum subset of @xmath150 , there exists one fixed point in @xmath150 with @xmath197 , the _ collins - stewart fixed point _
@xmath198 , which corresponds to one representation of the lrs solutions found by collins and stewart @xcite .
@xmath198 is given by @xmath199 and @xmath200 ( which yields @xmath201 ) .
the fixed point @xmath198 is the source ( w.r.t .
@xmath159 ) for all orbits in @xmath150 with @xmath90 . in the limit @xmath202 all solutions in @xmath203
converge to fixed points on the bianchi type i boundary of @xmath150 : there exists one orbit ( which corresponds to an lrs solution ) that converges to @xmath145 as @xmath204 ; every other orbit converges to a fixed point on @xmath205 or @xmath206 on @xmath140 , or to @xmath137 ( in the lrs case ) . for a detailed discussion of these results
see @xcite . in anticipation of theorem [ rinthm ] which implies that generic orbits of bianchi type ix do not have @xmath154-limit , the @xmath154-limit set @xmath208 of a point @xmath209 is defined as the set of all accumulation points towards the past ( i.e. , as @xmath53 ) of the orbit @xmath210 through @xmath211 .
the simplest examples of @xmath154-limit sets are fixed points and periodic orbits .
] points ( w.r.t.the standard future directed time variable @xmath45 ) on any of the bianchi type @xmath28 subsets @xmath71 , @xmath83 , @xmath212 , we refrain from giving a detailed discussion of these subspaces here .
( however , note that in order to _ prove _ theorem [ rinthm ] , a detailed understanding of solutions of bianchi type @xmath28 is essential ; in fact , in the proof of theorem [ rinthm ] bianchi type @xmath28 is ubiquitous ; we refer to @xcite and @xcite . ) in the present context it suffices to note that on each subset @xmath213 there exists a line of fixed points @xmath214 given by @xmath215 and @xmath216 ( so that @xmath89 ) .
since @xmath214 emanates from the point @xmath217 we call it the ` taub line . ' like @xmath137 itself , each of fixed points on @xmath214 is associated with a representation of minkowski spacetime ( in an lrs type @xmath28 symmetry foliation ) .
in the previous section we have given the fixed points associated with the system on @xmath218 .
a local dynamical systems analysis of the fixed points shows whether or not these points attract type ix orbits in the limit @xmath53 .. ] we merely state the results here and refer to @xcite for details . *
each fixed point @xmath219 on @xmath220 is a transversally hyperbolic saddle ( one stable instead of @xmath45 , we must replace ` stable ' by ` unstable ' . ]
mode , see figure [ unstable ] , and three unstable modes ; the stable manifold of @xmath219 coincides with a vacuum type ii transition orbit , see figure [ typeii ] ) .
the taub points @xmath221 are center saddles with a two - dimensional unstable manifold and a three - dimensional center manifold , where @xmath109 is excluded as an @xmath154-limit set on the center manifold , see @xcite ; consequently there do not exist any type ix solutions that converge to any of the points on @xmath140 as @xmath53 . *
the fixed point @xmath145 on @xmath222 is a hyperbolic saddle ( with @xmath222 as a two - dimensional stable manifold and an additional three - dimensional unstable manifold ) .
accordingly , @xmath145 attracts a two - parametric family of type ix orbits as @xmath53 .
these solutions have a so - called isotropic singularity .
* the fixed points @xmath223 ( @xmath224 ) on @xmath225 are hyperbolic saddles ( with @xmath225 as a three - dimensional stable manifold and an additional two - dimensional unstable manifold ) .
the unstable modes are associated with the equations @xmath226 ( for @xmath227 ) .
therefore , each of the fixed points @xmath223 attracts an ( equivalent ) one - parameter set of type ix orbits in the limit @xmath53 . *
each fixed point on @xmath228 on @xmath229 is a center saddle . on the three - dimensional center manifold ( which coincides with @xmath230 , @xmath231 ) the point acts as a ( non - hyperbolic ) sink , see @xcite . since there is a two - dimensional unstable manifold , there exists , for each fixed point on @xmath228 , a one - parameter family of type ix orbits that converges to it as @xmath53 ; these orbits correspond to lrs solutions .
( conversely , generic lrs type ix solutions converge to @xmath228 , see , e.g.,@xcite . ) [ cc][cc]@xmath112 [ cc][cc]@xmath113 [ cc][cc]@xmath114 [ cc][cc]@xmath98 [ cc][cc]@xmath99 [ cc][cc]@xmath99 [ cc][cc]@xmath100 [ cc][cc]@xmath100 [ cc][cc]@xmath98 [ cc][cc]@xmath120 [ cc][cc]@xmath121 [ cc][cc]@xmath122 [ cc][cc]@xmath123 [ cc][cc]@xmath124 [ cc][cc]@xmath125 [ cc][cc]@xmath126 [ cc][cc]@xmath127 [ cc][cc]@xmath128 [ cc][cc]@xmath129 collecting the results we see that the solutions whose @xmath154-limit is one of the fixed points form a subfamily of measure zero of the ( four - parameter ) family of bianchi type ix solutions . following the nomenclature of
@xcite we thus refer to these solutions as _ non - generic _ solutions of bianchi type ix .
alternatively , to capture the asymptotic behavior of these solution , we use the term _ past asymptotically self - similar _ solutions .
( since a fixed point in the hubble - normalized dynamical systems formulation corresponds to a self - similar solution , see e.g. @xcite , solutions that converge to a fixed point are asymptotically self - similar . )
the past asymptotically self - similar solutions comprise the lrs bianchi type ix solutions .
as seen above , generic lrs solutions converge to @xmath228 towards the past ( and each solution that converges to @xmath228 is lrs ) , but there exist exceptional lrs solutions that converge to @xmath145 or @xmath223 .
the remaining orbits whose limit point is either @xmath145 or @xmath223 correspond to past asymptotically self - similar solutions that are non - lrs . clearly , every solution that converges to @xmath145 or @xmath223 is a non - vacuum solution , since @xmath232 at @xmath145 and @xmath223 .
it is natural to ask how the non - generic orbits are embedded in the state space @xmath218 .
the lrs orbits form the three lrs subsets @xmath233 , which are the hyperplanes given by the conditions @xmath234 , @xmath235 , where @xmath236 .
the orbits whose @xmath154-limit set is the fixed point @xmath223 ( for some @xmath154 ) form the set @xmath237 in @xmath63 ; we call @xmath237 the collins - stewart manifold . the local analysis of the fixed point @xmath223 and the regularity of the dynamical system imply that the collins - stewart manifold @xmath237 is a two - dimensional surface ; it can be viewed as a two - dimensional manifold with boundary embedded in @xmath218 ( where this boundary corresponds to an orbit in @xmath238 ) .
analogously , the orbits whose @xmath154-limit set is the fixed point @xmath145 form the set @xmath239 in @xmath63 , which we call the isotropic singularity manifold , since solutions converging to @xmath145 are those with an isotropic singularity .
the isotropic singularity manifold @xmath239 is a three - dimensional hypersurface ; it can be viewed as a three - dimensional manifold with boundary .
generic bianchi type ix models are those that are not asymptotic self - similar and thus constitute examples for asymptotic self - similarity breaking ; for other such examples , see @xcite .
the central theme in this paper is the past asymptotic behavior of the generic models .
the _ mixmaster attractor _
@xmath240 ( alternatively referred to as the bianchi type ix attractor ) is defined to be the subset of @xmath218 given by the union of the bianchi type i and ii vacuum subsets , i.e. , @xmath241 .
since the type ii vacuum subset consists of three equivalent representations we obtain @xmath242 in this section we investigate the structures that the flow of dynamical system induces on @xmath240 .
in particular we discuss the mixmaster map , the kasner map , and the era map . to agree with the well - established convention for these maps , the direction of time will be taken towards the past .
in section [ subsets ] we have seen that the vacuum type ii orbits , i.e. , the orbits on @xmath243 , @xmath224 , are heteroclinic orbits that emanate from and converge to fixed points on the kasner circle @xmath140 , see figures [ typeii ] and [ alltypeii ] . in accord with section [ subsets ] we refer to these orbits as _ transitions _ and denote them by @xmath244 , @xmath245 .
[ rc][cc][0.9][0]@xmath120 [ tl][tl][0.9][0]@xmath121 [ bl][bl][0.9][0]@xmath122 [ lc][cc][0.9][0]@xmath123 [ br][br][0.9][0]@xmath124 [ tr][tr][0.9][0]@xmath125 @xmath246 , @xmath247 on the type ii subsets @xmath248 , @xmath249 , @xmath250 onto @xmath251-space .
concatenation of these transition orbits yields sequences of transitions .
the arrows indicate the direction of time towards the past.,title="fig:",scaledwidth=25.0% ] @xmath246 , @xmath247 on the type ii subsets @xmath248 , @xmath249 , @xmath250 onto @xmath251-space .
concatenation of these transition orbits yields sequences of transitions .
the arrows indicate the direction of time towards the past.,title="fig:",scaledwidth=25.0% ] @xmath246 , @xmath247 on the type ii subsets @xmath248 , @xmath249 , @xmath250 onto @xmath251-space .
concatenation of these transition orbits yields sequences of transitions .
the arrows indicate the direction of time towards the past.,title="fig:",scaledwidth=25.0% ] the type ii transitions are the building blocks for the analysis of the mixmaster attractor . by concatenating transitions we obtain a _ sequence of transitions _ ,
also known as a _
heteroclinic chain_. since each fixed point on @xmath140 ( except for the taub points ) is the initial value for one single transition , see figures [ unstable ] and [ alltypeii ] , the concatenation of transitions is unique : each fixed point ( except for the taub points ) generates a unique sequence of transitions .
note , however , that the ` direction of time ' is relevant . for each fixed point
@xmath252 on @xmath140 , which is not one of the taub points , there exists one single transition emanating from @xmath252 at @xmath253 , but there are two transitions converging to @xmath252 as @xmath202 .
therefore , concatenating transitions in the reversed direction of time leads to ambiguities .
( in terms of the standard future - directed time variable @xmath45 we have the converse statement : it is possible to make unambiguous retrodictions , but not predictions . )
let @xmath254 and let @xmath255 denote the initial point of the @xmath256 transition ( @xmath257 is also the end point of the @xmath258 transition ) .
we refer to the sequence @xmath259 of kasner fixed points , which is induced by the sequence of transitions , as being generated by the _ mixmaster map_. the mixmaster map can be visualized by a map in @xmath251-space , obtained by inscribing @xmath101 in a triangle with corners at @xmath260 and cyclic permutations , from which the ( projections of the ) transition orbits ` originate ' as straight lines ; see figure [ mixmaster ] .
[ cc][cc]@xmath112 [ cc][cc]@xmath113 [ cc][cc]@xmath114 [ cc][cc]@xmath120 [ cc][cc]@xmath121 [ cc][cc]@xmath122 [ cc][cc]@xmath123 [ cc][cc]@xmath124 [ cc][cc]@xmath125 [ cc][cc]@xmath126 [ cc][cc]@xmath127 [ cc][cc]@xmath128 [ cc][cc]@xmath129 let the initial kasner state of a transition be represented by the kasner parameter @xmath261 , where we assume @xmath262 , since neither of the taub points @xmath120 , @xmath121 , @xmath122 can be the initial value for a transition . inserting and into
we find that a transition maps the kasner parameter @xmath263 to the parameter @xmath264 , where @xmath265 ( u^{\mathrm{i } } - 1)^{-1 } & \qquad \text{if } \quad u^{\mathrm{i } } \in [ 1,2]\ : .
\end{array}\right.\ ] ] the information contained in this _ kasner map _ suffices to represent the collection of all transition orbits ( as a whole ) .
however , for each particular mapping @xmath266 , there exist six ( equivalent ) associated transitions ; there exist only three ( equivalent ) associated transitions : @xmath267 , @xmath268 , @xmath269 . ]
this is simply because @xmath263 characterizes the initial kasner fixed point on @xmath140 only up to permutations of the axes .
hence , in order to reconstruct a particular transition from , we supplement with information about the initial sector of the transition , which determines the position of the axes . in terms of the kasner parameter , a sequence of transitions corresponds to an iteration of .
let @xmath254 and let @xmath270 denote the initial kasner state of the @xmath256 transition .
this transition maps @xmath270 to @xmath271 , i.e. , @xmath272 ( u_l - 1)^{-1 } & \qquad \text{if } \quad u_l \in [ 1,2]\ , .
\end{array}\right.\ ] ] we refer to this map as the ( iterated ) _ kasner map _
( which is also known as the bkl map @xcite ) .
since each value of the kasner parameter @xmath273 represents an equivalence class of six kasner fixed points , the kasner map can be regarded as the map induced by the mixmaster map on these equivalence classes via the equivalence relation . in a sequence @xmath274 that is generated by the kasner map , each kasner state @xmath270
is called an _ epoch_. every sequence @xmath274 possess a natural partition into pieces ( which contain a finite number of epochs each ) where the kasner parameter is monotonically decreasing according to the simple rule @xmath275 ; these pieces are called _ eras _ @xcite .
an era begins with a maximal value @xmath276 of the kasner parameter ( where @xmath276 is generated from @xmath277 by @xmath278^{-1}$ ] ) , continues with a sequence of kasner parameters obtained via @xmath279 , and ends with a minimal value @xmath280 that satisfies @xmath281 , so that @xmath282^{-1}$ ] begins a new era .
@xmath283 let us denote the initial ( @xmath284 maximal ) value of the kasner parameter @xmath132 in era number @xmath285 ( where @xmath286 ) by @xmath287 .
following @xcite we decompose @xmath287 into its integer part @xmath288 $ ] and its fractional part @xmath289 , i.e. , @xmath290\ : , \quad x_s= \ { { \mathsf{u}}_s\}\:.\ ] ] the number @xmath291 represents the ( discrete ) length of era @xmath285 , which is simply the number of kasner epochs it contains .
the final ( @xmath284 minimal ) value of the kasner parameter in era @xmath285 is given by @xmath292 , which implies that era number @xmath293 begins with @xmath294 the map @xmath295 is ( a variant of ) the so - called _ era map _ ; starting from @xmath296 it recursively determines @xmath287 , @xmath297 , and thereby the complete kasner sequence @xmath298 .
the era map admits a straightforward interpretation in terms of continued fractions .
consider the continued fraction representation of the initial value , i.e. , @xmath299\:.\ ] ] the fractional part of @xmath300 is @xmath301 $ ] ; since @xmath302 is the reciprocal of @xmath303 we find @xmath304\:.\ ] ] therefore , the era map is simply a shift to the left in the continued fraction expansion , @xmath305 \:\mapsto\ : { \mathsf{u}}_{s+1 } = [ k_{s+1 } ; k_{s+2 } , k_{s+3},\dotsc]\:.\ ] ] the properties of the kasner sequence depend on the initial value @xmath296 .
if and only if the initial kasner parameter @xmath306 is a rational number , i.e. , if and only if @xmath307 , then its continued fraction representation is finite , i.e. , [ cfki ] @xmath308\ : , \tag{\ref{cfki}i}\ ] ] where @xmath309 .
therefore , there exists only a finite number of eras ( where the last one begins with @xmath310 ) , and the kasner sequence is finite . at the end of era number @xmath311 , the kasner parameter reaches @xmath138 , which subsequently terminates the recursion . since @xmath312 is a set of measure zero in @xmath313 , this case is non - generic .
a quadratic irrational ( quadratic surd ) is an algebraic number of degree @xmath314 , i.e. , an irrational solution of a quadratic equation with integer coefficients . if and only if the initial kasner parameter @xmath306 is a quadratic irrational , i.e. , if and only if @xmath315 , where @xmath316 and @xmath317 is not a perfect square ( i.e. , @xmath318 ) , then its continued fraction representation is periodic , i.e. , @xmath319\ : ; \tag{\ref{cfki}ii}\ ] ] the notation is such that the part in parenthesis , i.e. , @xmath320 , is repeated ad infinitum . is a quadratic irrational such that @xmath321 , then @xmath322 $ ] , i.e. , the continued fraction is purely periodic without any preperiod . ]
consequently , the era map becomes periodic ( after the @xmath311 era ) , and we thus obtain a periodic sequence of eras and a periodic kasner sequence @xmath274 . it is straightforward to see that while the period of the era sequence is @xmath323 , the period of the kasner sequence is @xmath324 ; see the examples below . since the set of algebraic numbers of degree two ( or equivalently the set of equations with integer coefficients it is a subset of @xmath325 ) is a countable set , case ( ii ) is also non - generic .
an irrational number is called badly approximable if its markov constant , let @xmath326 denote the distance from @xmath211 to the nearest integer , i.e. , @xmath327 .
the markov constant @xmath328 of a number @xmath329 is defined as @xmath330 , see @xcite .
it is known that @xmath331 for all @xmath332 .
] is finite .
if and only if @xmath306 is badly approximable , then the coefficients ( partial quotients ) in its continued fraction representation are bounded , i.e. , @xmath333 \qquad \text{with}\quad k_i \leq k\:\ , \forall i \tag{\ref{cfki}iii}\ ] ] for some positive constant @xmath334 .
consequently , the sequence of eras and the kasner sequence @xmath274 are bounded , i.e. , @xmath335 @xmath336 . obviously , case ( ii ) is a subcase of case ( iii ) .
( note , however , that there is probably no relationship between case ( iii ) and algebraic numbers of degree greater than @xmath314 , since it is expected that these numbers are well approximable . )
the set of badly approximable numbers are a set of lebesgue measure zero , hence this case is non - generic . if and only if the initial kasner parameter @xmath306 is a well approximable irrational number , then the partial quotients @xmath337 in the continued fraction representation @xmath338 \tag{\ref{cfki}iv}\ ] ] are unbounded ( and we can construct a diverging subsequence from the sequence of partial quotients @xmath339 ) .
this is the generic case , and hence generically the kasner sequence @xmath274 is infinite and unbounded . in terms of continued fractions ,
the kasner sequence @xmath340 generated by @xmath341 $ ] is @xmath342 \rightarrow \big [ k_0 -1 ; k_1 , k_2 , \dotsc \big ] \rightarrow \big [ k_0 -2 ; k_1 , k_2 , \dotsc \big ] \rightarrow \ldots \rightarrow \big[1 ; k_1 , k_2 , \dotsc \big ] \\
\rightarrow \mathsf{u}_1 & = \big [ k_1 ; k_2 , k_3 , \dotsc \big ] \rightarrow \big [ k_1 - 1 ; k_2 , k_3 , \dotsc \big ] \rightarrow \big [ k_1 -2 ; k_2 , k_3 , \dotsc \big ] \rightarrow \ldots \rightarrow \big[1 ; k_2 , k_3 , \dotsc \big ] \\
\rightarrow \mathsf{u}_2 & = \big [ k_2 ; k_3 , k_4 , \dotsc \big ] \rightarrow \big [ k_2 - 1 ; k_3 , k_4 , \dotsc \big ] \rightarrow \big [ k_2 -2 ; k_3 , k_4 , \dotsc \big ] \rightarrow \ldots\end{aligned}\ ] ] let us give some examples for periodic _
era sequences _ and _ kasner sequences_. if @xmath343 = ( 1+\sqrt{5})/2 $ ] , which is the golden ratio , then @xmath344 @xmath345 .
it follows that the kasner sequence is also a sequence with period @xmath194 , @xmath346 if @xmath347 = 1 + \sqrt{2}$ ] , then @xmath348 @xmath345 ; hence the era sequence is a sequence of period @xmath194 .
however , the associated kasner sequence has period @xmath314 , @xmath349 analogously , the initial value @xmath350 = ( 3+\sqrt{13})/2 $ ] generates an era sequence of period @xmath194 and an associated kasner sequence of period @xmath351 .
finally let @xmath352 = 1 + \sqrt{3/2 } \simeq 2.2247 $ ] .
then the era sequence has period @xmath314 with @xmath353 \simeq 2.2247 $ ] for even @xmath311 and @xmath354 \simeq 4.4495 $ ] for odd @xmath311 .
the associated kasner sequence @xmath355 has period @xmath356 , @xmath357 in the state space description of sequences ( in terms of the mixmaster map ) , an _ epoch _ is simply a point @xmath257 on the kasner circle .
( it is one of the six points in the equivalence class associated with the kasner parameter @xmath270 . )
transitions connect epochs and thus generate the mixmaster map .
the kasner parameter @xmath132 can be employed to measure the ( angular ) distance of a point @xmath252 on @xmath140 from the taub points or from the non - flat lrs points : if @xmath358 , then @xmath252 is at an angular distance of approximately @xmath359 from one of the taub points . on the other hand , in the vicinity of the non - flat lrs points ( where @xmath360 ) , @xmath361 is a linear measure for the angular distance of @xmath252 from the closest of the non - flat lrs points .
if @xmath362 , then @xmath252 is closer to one of the non - flat lrs points @xmath106 than to any of the taub points @xmath109 .
therefore , in the state space picture , an _ era _ can be described as a ( finite ) sequence of points @xmath363 , @xmath364 , obtained from the mixmaster map , whose angular distance from the taub points is monotonically increasing .
the era begins with a fixed point @xmath365 that is close to one of the taub points @xmath109 ( the preceding point @xmath366 was closer to one of the other two taub points @xmath176 , @xmath367 , than to @xmath109 ) ; then the distance from the taub point @xmath109 monotonically increases until , for @xmath368 , it exceeds the distance to the non - flat lrs points ; this is the terminal point for the era ; it is connected by the following transition with the initial point of the next era ; a good illustration is figure [ period3 ] , where each era consists of three epochs .
note that due to the equivalence of kasner points , there exist several realizations of one and the same kasner sequence as mixmaster sequences in the state space picture ; this is exemplified by the heteroclinic cycles in figure [ period1a ] and [ period1b ] .
the state space description of the cases ( i)(iv ) , which are characterized by the initial kasner parameter @xmath296 , is the following : * iff @xmath307 , see , the mixmaster sequence of kasner fixed points @xmath369 is finite . after a finite number of transitions , at the end of era @xmath311 , the sequence reaches one of the lrs points @xmath106 ( where @xmath138 ) ; a last transition follows , namely the transition @xmath370 , and the sequence terminates in one of the taub points .
* iff @xmath306 is a quadratic surd , i.e. , @xmath315 for some @xmath371 where @xmath372 is not a perfect square , see , then the mixmaster sequence of kasner points @xmath259 is eventually periodic , where the period is a multiple of @xmath324 ; see figure [ cycle ] . viewed as a periodic sequence of transitions ( which are heteroclinic orbits )
we obtain a heteroclinic cycle . in figure
[ cycle ] we give some of the heteroclinic cycles associated with kasner sequences with periods @xmath194 , @xmath314 , and @xmath351 in their projection onto @xmath373-space .
note that due to permutation symmetry there are several cycles associated with a given periodic kasner sequence .
* iff @xmath306 is a badly approximable irrational number , see , there exists a neighborhood of the taub points @xmath109 such that the mixmaster sequence of kasner points @xmath259 does not enter this neighborhood .
this is simply because there exists a maximal value of the sequence @xmath355 .
* iff @xmath306 is a well approximable irrational number , see , then the mixmaster sequence @xmath259 comes arbitrarily close to the taub points .
this is the generic case .
[ cc][cc]@xmath112 [ cc][cc]@xmath113 [ cc][cc]@xmath114 [ cc][cc]@xmath120 [ cc][cc]@xmath121 [ cc][cc]@xmath122 [ cc][cc]@xmath123 [ cc][cc]@xmath124 [ cc][cc]@xmath125 + in the following we analyze the generic case ( iv ) in more detail .
let @xmath306 be a well approximable irrational number , i.e. , a number whose continued fraction expansion @xmath374\ ] ] defines an unbounded sequence @xmath339 . by construction
, era number @xmath375 contains @xmath337 epochs ( which we call its length ) . a natural question to ask concerns the distribution of the partial quotients @xmath337 . for a ` typical ' well approximable irrational number , how often does the number @xmath194 appear in the sequence @xmath339 ?
how often the number @xmath314 ? and what about the number @xmath376 ? the answer is given by khinchin s law @xcite .
let @xmath377 denote the probability that a randomly chosen partial quotient among @xmath378 equals @xmath379 . in the asymptotic limit , i.e. , for @xmath380 we have @xmath381 i.e.
, the partial quotients of the continued fraction representation of are distributed like a random variable whose probability distribution is given by .
( by , the number @xmath194 appears in @xmath382 of the slots , the number @xmath314 in @xmath383 , and the number @xmath376 in @xmath384 of the slots . )
khinchin s law applies for _ almost all _
numbers @xmath306 . for a ( generic ) kasner sequence @xmath355 with initial parameter @xmath385 $ ] and its associated era sequence @xmath386 , where @xmath387 $ ]
, the expression @xmath388 of represents the probability that a randomly chosen era of @xmath386 has length @xmath389 ; this corresponds to the probability that the initial value of an era is contained in the interval @xmath390 . in this manner ,
the probability distribution makes possible a stochastic interpretation of generic kasner sequences .
the probability distribution results in extraordinary properties of the mixmaster / kasner map , which will be of crucial importance in the considerations of section [ beliefs ] .
we do not discuss details here but refer to future work ; however , we can not refrain from giving a teaser : for a generic kasner sequence @xmath355 and its associated era sequence @xmath386 there exist infinitely many eras such that the length ( i.e. , the number of epochs ) of the @xmath311 era is larger than @xmath391 ; however , for sufficiently large @xmath311 , the length is guaranteed to be bounded by @xmath392 .
( for a proof of this result by borel and bernstein see @xcite ; see also @xcite . )
properties of this kind underline the remarkable intricacies of the heteroclinic structures on the mixmaster attractor .
in this section we turn to what is known about the past asymptotic dynamics of generic type ix solutions .
the main mixmaster fact is ringstrm s ` bianchi type ix attractor theorem ' .
consider a solution of bianchi type ix that is either vacuum or associated with a perfect fluid satisfying @xmath393 .
recall that such a solution is called _ generic _ if it is not past asymptotically self - similar , i.e. , if its @xmath154-limit set is neither the point @xmath145 , nor any of the points @xmath223 , nor a point on @xmath228 ; in other words , a generic solution corresponds to an orbit in @xmath63 that is neither contained in @xmath239 , nor in @xmath237 , nor in @xmath233 .
therefore , the set of generic bianchi type ix states is an open set in @xmath63 .
( to conform with @xcite we use the future directed time variable @xmath45 . )
the main results concerning generic bianchi type ix models are due to ringstrm @xcite ; these results rest on earlier work that is reviewed and derived in @xcite , and on @xcite . in the following we state the main theorem in a version adapted to our purposes .
[ rinthm ] let @xmath394 be a generic solution of bianchi type ix , i.e. , a generic solution of in @xmath63 .
then @xmath395 as @xmath396 .
note that this theorem applies to both the fluid and the vacuum case ; in the latter case becomes @xmath397 and @xmath398 .
the proof of theorem [ rinthm ] given in @xcite is delicate .
in the first part it is proved that the @xmath154-limit set of a generic bianchi type ix solution is non - empty and must contain a point on the kasner circle .
the main part of the proof deals with the fact that the function @xmath399 is in general not monotone .
there exist times where @xmath399 increases ( as @xmath53 ) ; the associated growth must therefore be controlled and shown to be negligible compared to the overall decrease in @xmath400 .
this is done by a careful analysis of the equations and ( approximate ) solutions . in @xcite
we give an alternative and relatively short and succinct proof of theorem [ rinthm ] which is based on an in - depth understanding of the hierarchical structure of the dynamical system ( as represented by figure [ contraction ] ) .
it is important to note , however , that either of the proofs fail in the other bianchi types that are conjectured to exhibit an oscillatory approach towards the singularity , i.e. , the proofs fail for types @xmath401 and viii .
this is unfortunate , since there are reasons to believe that these models are more relevant than type ix as regards the dynamics of generic ( inhomogeneous ) cosmologies , see @xcite .
using the concept of the mixmaster attractor , cf . , we obtain an equivalent formulation of theorem [ rinthm ] : let @xmath402 be a generic solution of bianchi type ix . then @xmath403 where the distance @xmath404 is given as @xmath405 .
theorem [ rinthm ] thus states that the attractor of generic type ix solutions resides on @xmath240 ; however , whether the past attractor is in fact @xmath240 or merely a subset thereof remains open .
( the terminology ` mixmaster attractor ' is seductive but might turn out to be quite misleading . ) likewise , the theorem does not provide any direct information about the details of the asymptotic behavior of solutions .
the prerequisite for a deeper understanding of the asymptotic behavior and the oscillatory dynamics of generic type ix solutions is an understanding of the mixmaster attractor . in section [ maps ] we have identified the structures on the mixmaster attractor @xmath240 that are induced by the flow of the dynamical system : heteroclinic cycles [ case ( ii ) ] and finite [ case ( i ) ] and infinite heteroclinic sequences [ cases ( iii ) and ( iv ) ] .
all these structures qualify ( a priori ) as possible @xmath154-limit sets of generic type ix orbits . in conjunction with these results , the main theorem [ rinthm ]
implies a number of further facts about the attractor`mixmaster attractor facts ' , which we give as a list of * corollaries*. ( for proofs see @xcite , and also @xcite . ) 1 .
[ possessesalpha ] a generic , i.e. not past asymptotically self - similar , type ix orbit possesses an @xmath154-limit point on @xmath240 .
[ onethenall ] if @xmath406 is an @xmath154-limit point of a type ix orbit , then the entire heteroclinic cycle / sequence ( mixmaster sequence ) through @xmath252 must be contained in the @xmath154-limit set .
[ taubthenmany ] if one of the taub points @xmath407 is an @xmath154-limit point of a type ix orbit , then the @xmath154-limit set contains kasner fixed points associated with arbitrarily large values of the kasner parameter @xmath132 .
4 . for generic solutions of bianchi type ix the weyl curvature
scalar @xmath408 ( and therefore also the kretschmann scalar ) becomes unbounded towards the past .
5 . taking into account both the expanding and contracting phases of bianchi type ix solutions , generic bianchi type ix
initial data generate an inextendible maximally globally hyperbolic development associated with past and future singularities where the curvature becomes unbounded .
[ uniformcor ] convergence to the mixmaster attractor is uniform on compact sets of generic initial data : let @xmath409 be a compact set in @xmath63 that does not intersect any of the manifolds @xmath239 , @xmath237 , @xmath233 , so that each initial data @xmath410 generates a generic type ix solution .
let @xmath411 denote the type ix solution with @xmath412 .
then @xmath413 uniformly in @xmath410 .
corollary 3 implies that the @xmath154-limit set contains an infinite set of kasner fixed points in a neighborhood of the taub point(s ) , but this set is not necessarily a continuum of fixed points ; cf . the previous discussion about possible @xmath154-limit sets .
if a kasner point with @xmath414 is contained in the @xmath154-limit of an orbit , so is a taub point .
this is an immediate consequences of the results of section [ maps ] , case ( i ) . on the other hand ,
if the @xmath154-limit set of an orbit is a heteroclinic cycle or a heteroclinic sequence ( possibly in combination with a cycle ) associated with cases ( ii ) and ( iii ) of section [ maps ] , then there exists a neighborhood of the taub points whose intersection with the @xmath154-limit set is empty .
note that the oscillatory behavior of asymptotic type ix dynamics , which we unfortunately know no details about , constitutes an example of asymptotic self - similarity breaking @xcite . in order to make progress
as regards the details of the asymptotic oscillatory behavior , it is natural to first establish the connection between the mixmaster / kasner map and dynamics for a finite time interval @xmath415 , discussed next . to make contact between the mixmaster map and bianchi type ix asymptotic dynamics we introduce the concept of _ finite mixmaster shadowing _ which formalizes the following basic idea : given a sequence of transitions we can choose type ix initial data sufficiently close to the initial data of the sequence so that the type ix solution generated by this data remains close to the sequence for some ` time '
let @xmath416 be a kasner fixed point ( but @xmath417 ) and let @xmath306 be the associated value of the kasner parameter .
there exists a unique sequence of transitions @xmath418 and an associated mixmaster sequence @xmath419 with @xmath416 as initial data ; the associated kasner sequence is @xmath355 .
( if @xmath307 , the sequence terminates at one of the taub points after a finite number of transitions . )
we shall make the definition that a type ix solution shadows a finite piece @xmath420 of the sequence of transitions if it is contained in a prescribed ( small ) tubular neighborhood of @xmath421 .
however , a standard @xmath422-neighborhood of the sequence fails to be a reasonable measure of closeness , because in the vicinity of a taub point the transitions lie so dense that consecutive transitions are not separated from each other by their respective @xmath422-neighborhoods .
therefore , the introduction of _ adapted _ tubular neighborhoods is necessary to take into account the sensitivity of the flow at the taub points and to capture more accurately the intuitive idea of shadowing .
let @xmath423 be small .
a ` _ _ taub - adapted neighborhood _ _ ' of the mixmaster sequence @xmath424 is a sequence @xmath425 of open balls , where @xmath426 is centered at the point @xmath257 and has radius @xmath427 , i.e. , @xmath428 .
( the radius @xmath429 is chosen to ensure that the intersection of the ball @xmath426 with the kasner circle induces more or less a standard @xmath422-neighborhood @xmath430 of the kasner parameter @xmath270 ; recall from section [ maps ] that @xmath359 measures the angular distance of a fixed point from a taub point . ) a taub - adapted tubular neighborhood of the sequence of transitions @xmath418 is the sequence of tubes @xmath431 that linearly interpolate between @xmath426 and @xmath432 . based on this definition we say that a type ix solution @xmath433 _ shadows _ a finite piece @xmath420 of the sequence of transitions if it moves in a prescribed taub - adapted tubular neighborhood @xmath421 , i.e. , if there exists a sequence of times @xmath434 such that @xmath433 is contained in @xmath435 for all @xmath436 $ ] for all @xmath437 . making use of these concepts ,
a formulation of finite mixmaster shadowing is the following : let @xmath423 and @xmath438 . consider the sequence of transitions @xmath439 emanating from an initial kasner point @xmath416 and its taub - adapted tubular neighborhood ( associated with @xmath422 ) .
then there exists @xmath440 such that each type ix orbit @xmath433 that is generated by initial data @xmath441 with @xmath442 shadows the finite piece @xmath421 of the sequence of transitions .
( a proof of this statement in a slightly different form has been given by rendall @xcite .
alternatively , one can invoke the regularity of the dynamical system , the center manifold reduction theorem and continuous dependence on initial data . ) evidently , @xmath443 depends on the choice of @xmath422 and @xmath444 .
more importantly , however , @xmath443 depends on the position of @xmath416shadowing is not uniform ; in particular , if we consider a series of initial points that approach one of the taub points , then @xmath443 necessarily converges to zero along this series shadowing is more delicate in the vicinity of a taub point .
we will return to this issue in some detail in the next section .
finite mixmaster shadowing concerns any generic type ix orbit @xmath433 .
let @xmath445 be an @xmath154-limit point of the type ix orbit @xmath433 ; without loss of generality we may assume that @xmath252 is not one of the taub points .
( the existence of such a point is ensured by corollaries [ possessesalpha][taubthenmany ] of section [ furthermix ] . ) for simplicity we assume that @xmath252 is associated with an irrational value of the kasner parameter , which guarantees that the sequence @xmath259 emanating from @xmath446 is an infinite sequence . since @xmath252 is an @xmath154-limit point of @xmath433 , there exists a sequence of times @xmath447 , @xmath448 ( @xmath449 ) , such that @xmath450 ( @xmath449 ) .
therefore , we observe a recurrence of phases , where the orbit @xmath433 shadows @xmath259 with an increasing degree of accuracy , i.e. , shadowing takes place for increasingly longer pieces of the sequence or in ever smaller neighborhoods .
( if the kasner parameter of @xmath252 is rational , then the sequence @xmath259 is finite and terminates at a taub point .
@xmath433 will shadow this finite sequence recurrently with an increasing degree of precision . )
we will return to this issue in some more detail in the subsection ` stochastic mixmaster beliefs ' of section [ beliefs ] .
the concept of shadowing leads directly to the concept of _ approximate sequences _ which we introduce next .
consider a generic type ix orbit @xmath402 and the function @xmath451 , where the distance @xmath452 is given as @xmath453 .
when the orbit @xmath433 traverses a ( sufficiently small ) neighborhood of a fixed point on @xmath454 , the function @xmath455 exhibits a unique local minimum .
this is immediate from the transversal hyperbolic saddle structure of the fixed point .
( however , the flow in the vicinity of the taub points is more intricate , since these points are not transversally hyperbolic . )
it follows that the function @xmath451 can be used to partition @xmath433 into a sequence of segments in a straightforward manner : into segments seems natural ; it is important to note , however , that any definition depends on the formulation of the problem and is to a certain extent arbitrary .
for instance , instead of using the minima of @xmath451 one might prefer to analyze the projection of the orbit onto @xmath373-space and use the extrema of @xmath456 .
however , the conclusions drawn from any construction of segments are quite insensitive to the details of the definition . ] the local minima of @xmath451 form an infinite sequence @xmath457 such that @xmath458 as @xmath459 .
( this follows directly from corollaries [ possessesalpha ] and [ onethenall ] because @xmath433 has @xmath154-limit point(s ) on the kasner circle and @xmath154-limit points on the type ii subset . ) a _ segment _ of @xmath433 is defined to be the solution curve between two consecutive minima , i.e. , the image of the interval @xmath460 $ ] .
( note that @xmath461 in the approach to the singularity , while the discrete ` time ' @xmath256 is past - directed , i.e. , @xmath462 towards the singularity .
this convention is chosen to agree with the standard convention for the mixmaster and the kasner map . ) in the asymptotic regime , i.e. , in the approach to the mixmaster attractor , finite shadowing entails that a finite sequence of segments will resemble a finite sequence of type ii transitions ; this assumes , however , that the type ix orbit does not come too close to any of the taub points .
( in the neighborhood of a taub point the flow of the dynamical system is much more intricate .
this might yield recurring interruptions of the ` standard behavior ' . )
we call the type ix orbit in its segmented form an _ approximate sequence of transitions_. in addition , we define a sequence of ` check points ' that is associated with an approximate sequence of transitions .
each minimum @xmath463 of @xmath451 is associated with a kasner fixed point @xmath464 ( a ` check point ' ) that is defined as the minimizer on @xmath140 of the distance between @xmath465 and @xmath140 .
( note that the check points @xmath466 do not lie on the type ix orbit @xmath433 . ) since each check point @xmath464 is associated with a value @xmath467 of the kasner parameter , the sequence @xmath466 induces a sequence @xmath468 .
the value @xmath469 is in general not generated from @xmath467 by the exact kasner map , but differs from that value by an error of @xmath470 . in our terminology , the sequence @xmath466 is an approximate mixmaster sequence ; @xmath468 is an approximate kasner sequence , see figure [ checkmap ] .
[ rc][rc][0.9][0]@xmath465 [ lc][lc][0.9][0]@xmath471 [ lc][lc][0.9][0]@xmath472 [ rc][rc][0.9][0]@xmath473 [ cc][cc][0.7][0]@xmath474 [ cc][cc][1][0]@xmath433 [ cc][cc][1][0]@xmath140 [ cc][cc][0.5][2]mixmaster map [ tc][tc][0.5][2]kasner map [ rc][rc][0.9][0]@xmath475 .
the points @xmath465 , @xmath476 , are the local minima of the distance between @xmath433 and @xmath140 .
the ` check point ' @xmath472 is the point on @xmath140 that is closest to @xmath433 ( for @xmath45 in a neighborhood of @xmath463 ) . since each check point @xmath472
is associated with a value @xmath467 of the kasner parameter , the ` approximate mixmaster sequence ' @xmath477 induces an ` approximate kasner sequence ' @xmath468 . in general , @xmath473/@xmath469 are not generated from @xmath472/@xmath467 by the exact mixmaster / kasner map , but differ by an error @xmath478/@xmath470.,title="fig:",scaledwidth=35.0% ] the approximate kasner sequence @xmath467 associated with a type ix orbit @xmath433 does not follow the kasner map exactly . a natural question to ask , however , is whether the errors @xmath470 of the approximate kasner sequence converge to zero as @xmath462 or not .
if @xmath479 as @xmath459 for a type ix orbit , this means that its dynamics is completely described by an ` asymptotic mixmaster / kasner map ' , i.e. , by a map that converges to the mixmaster / kasner map towards the singularity .
however , if @xmath480 as @xmath462 , then the evolution is interrupted repeatedly infinitely many times by phases where the dynamics is completely different from the mixmaster dynamics , e.g. , ` eras of small oscillations ' @xcite .
( in the present state space description an era of small oscillations is associated with type @xmath28 behavior in the vicinity of one of the taub lines @xmath228 , where @xmath481 and @xmath163 are small and of the same order ; for details we refer to the discussion of type @xmath28 dynamics in @xcite . ) in the subsection ` stochastic mixmaster beliefs ' of section [ beliefs ] we will investigate the behavior of @xmath470 along type ix orbits in detail .
there remain several important open problems . in the following
we will address the most pressing questions ; in particular we will transform vague beliefs into refutable conjectures , and give arguments in their favor .
an immediate question is the following : what are the actual @xmath154-limit sets of type ix orbits ?
consider a type ix orbit @xmath153 .
the @xmath154-limit set @xmath482 contains a number of kasner fixed points ( and the associated type ii transitions ) , each of which is characterized by a particular value of the kasner parameter @xmath132 .
the set of kasner parameters obtained in this way from @xmath482 we denote by @xmath483 .
the question of which form @xmath483 can take for different orbits @xmath153 is open , the most significant issues being the following : * is it possible that there exists @xmath153 whose limit set @xmath483 consists of rational numbers only ?
it is straightforward to exclude that @xmath483 coincides with @xmath312 itself ( or a dense subset thereof ) , the reason being that @xmath154-limit sets are necessarily closed .
another a priori constraint is that @xmath483 must contain @xmath484 ( which characterizes the taub points ) if @xmath483 contains a rational number ; see section [ maps ] .
corollary [ taubthenmany ] of section [ furthermix ] then implies that @xmath483 is unbounded , i.e. , @xmath483 contains arbitrarily large values of the kasner parameter .
a set that is compatible with these basic requirements is , e.g. , @xmath485 . whether there exist orbits @xmath153 such that @xmath483 takes this ( or a related ) form remains open .
* is it conceivable that there exist orbits @xmath153 such that @xmath482 is a heteroclinic cycle , see figure [ cycle ] ? in this case
the set @xmath483 is generated by a quadratic surd @xmath486 $ ] via the kasner map ; see section [ maps ] ; in particular , @xmath483 is finite .
however , whether orbits @xmath153 with this particular past asymptotic behavior really exist is an open problem .
* can there exist orbits @xmath153 such that @xmath483 is bounded but contains infinitely many @xmath132-values ? is the kasner sequence generated by a badly approximable number a candidate ?
the kasner sequence @xmath355 generated by a badly approximable number is an infinite sequence that is bounded .
however , there must be at least one accumulation point of this sequence ; if @xmath483 contains @xmath355 , then @xmath483 must contain the accumulation point of the sequence as well ( since @xmath154-limit sets are closed ) .
if this accumulation point is a well approximable number , @xmath483 can not be bounded ; however , if the accumulation point is a quadratic surd , no inconsistencies arise .
hence , a priori , there might exist orbits @xmath153 such that @xmath483 is not finite but still bounded .
whether this is indeed the case is doubtful , but hard to exclude a priori .
an open problem that might be quite separate from the questions raised above concerns the behavior of all type ix orbits save a set of measure zero .
the past attractor of a dynamical system given on a state space @xmath487 is defined as the smallest closed invariant set @xmath488 such that @xmath489 for all @xmath490 apart from a set of measure zero @xcite . the past attractor of the type ix dynamical system coincides with the mixmaster attractor @xmath240 ( rather than being a subset thereof ) . why is this a belief and not a fact ?
theorem [ rinthm ] implies that @xmath491 ; however , it is believed that @xmath492 .
( the usage of the terminology ` mixmaster attractor ' for the set @xmath240 reflects the strong belief in the mixmaster conjecture . )
it is difficult to imagine how the mixmaster attractor conjecture could possibly be violated .
for instance , it seems rather absurd that the past attractor consists only of ( a subset of ) heteroclinic cycles but there exist no proofs .
a closely related belief is the following stronger statement . for almost all bianchi type ix solutions @xmath153 the @xmath154-limit set
@xmath482 coincides with the mixmaster attractor @xmath240 .
we use the term ` almost all ' in a noncommittal way without specifying the measure ; recall that the word ` generic ' already has the well - defined meaning of ` not past asymptotically self - similar ' .
( the usage of the word ` generic ' is in accord with @xcite . )
this subsection is concerned with the ( open ) question of which role the mixmaster / kasner map plays in the asymptotic evolution of type ix solutions .
the basis for our considerations are the results of section [ maps ] where we discussed the mixmaster / kasner map and the stochastic aspects of ( generic ) kasner sequences .
the _ mixmaster stochasticity conjecture _ supposes that these stochastic properties carry over to almost every type ix orbit when represented as an approximate kasner sequence .
the approximate kasner sequence @xmath468 associated with a generic type ix orbit admits a stochastic interpretation in terms of the probability distribution associated with the kasner map , cf
. section [ maps ] .
( this holds with probability one , i.e. , for almost every generic type ix orbit . ) the mixmaster stochasticity conjecture is based on a rather suggestive simple idea : type ix evolution is like trying to follow a path of an ( infinite ) network of paths while the ground is shaking randomly , and where the shaking subsides with time but never stops .
a type ix orbit tries to follow a sequence of transitions on the mixmaster attractor while random errors cause the orbit to lose track . due to the errors , the orbit is incapable of following one particular sequence forever ;
after a finite time , the type ix orbit has deviated too much and it leaves the vicinity of the sequence .
although , temporarily , the orbit is contained in a neighborhood of a different sequence , it is bound to lose track of that sequence as well eventually . accordingly
the type ix orbit is thrown around in the space of mixmaster sequence with the effect that the type ix orbit inherits the stochastic properties of generic mixmaster sequences . in
the following we paint a heuristic picture that makes this idea a little more concrete .
consider a type ix orbit ( approximate sequence of transitions ) and the associated approximate mixmaster sequence @xmath477 .
let @xmath493 be the exact mixmaster sequence with @xmath494 and consider the taub - adapted neighborhood of this sequence associated with some prescribed value @xmath423 .
finite mixmaster shadowing entails that there exists a finite piece @xmath495 of the approximate sequence that is contained in the prescribed taub - adapted neighborhood of the exact sequence @xmath493 .
however , at @xmath496 , the approximate mixmaster sequence @xmath477 leaves the prescribed tolerance interval due to the accumulation of errors .
hence , at @xmath496 we reset the system and consider the exact mixmaster sequence @xmath497 with initial data @xmath498 .
the approximate mixmaster sequence @xmath499 is contained in the taub - adapted neighborhood of the exact sequence @xmath497 up to @xmath500 . at @xmath501
another readjustment becomes necessary .
iterating this procedure and concatenating the finite pieces @xmath502 we are able to construct a sequence @xmath503 with the property that the approximate sequence @xmath477 is contained within the taub - adapted neighborhood ( associated with @xmath422 ) of @xmath503 for all @xmath504 .
the sequence @xmath503 is a piecewise exact mixmaster sequence ; it is exact in intervals @xmath505 .
the length @xmath506 of these intervals grows beyond all bounds as @xmath507 , because shadowing takes place with an increasing degree of accuracy .
the error ( of the order @xmath422 ) between the approximate sequence @xmath477 and the piecewise exact sequence @xmath508 results from the accumulation of numerous small errors .
this obliterates the deterministic origin of the problem and generates a ` randomness ' that leads to stochastic properties .
accordingly , we expect that the exact sequences @xmath509 from which @xmath503 is built constitute a random sample of mixmaster sequences and thus truly reflect the stochastic properties of the mixmaster map . as a consequence , although the sequence @xmath508 is only a piecewise exact mixmaster sequence , it possesses the same stochastic properties as a generic mixmaster sequence . extrapolating this line of reasoning
we are able to complete the argument and find that the approximate sequence @xmath466 itself reflects the stochastic properties of the mixmaster map . to emphasize this aspect of stochasticity of @xmath466
we use the term ` randomized approximate sequence ' .
some comments are in order . in our discussion
we have assumed implicitly that the approximate sequence @xmath466 we consider can shadow any exact mixmaster sequence for a finite number of transitions only .
this is not necessarily the case .
a type ix orbit whose @xmath154-limit set is one of the heteroclinic cycles ( if such an orbit exists ! ) is an obvious counterexample : for every @xmath423 there exists @xmath510 such that @xmath464 is contained in the taub - adapted @xmath422-neighborhood of the mixmaster sequence associated with the cycle .
however , we expect that this type of ` infinite shadowing ' holds at most for orbits of a set of measure zero
. a more serious limitation of the intuitive picture that we have sketched is illustrated by the following related example . consider a type ix orbit whose @xmath154-limit set is the heteroclinic cycle depicted in figure [ period1a ] ( where we note again that the existence of such an orbit is not proven ) .
for the associated approximate kasner sequence @xmath468 we have @xmath511 as @xmath459 .
consider the piecewise exact kasner sequence @xmath512 that is associated with the piecewise exact mixmaster sequence .
each piece @xmath513 is generated from a value @xmath514 = [ 1;1,1,1,1,\ldots,1,k^{(i)}_n , k^{(i)}_{n+1},\ldots]$ ] ; we have @xmath449 as @xmath507 .
it is evident that these pieces do _ not _ form a random sample of kasner sequences .
at best one could conjecture ( and it is probably safe to do so ) that the collection of the remainders @xmath515 $ ] is such a random sample .
even so , the stochastic aspect does not carry over to the approximate sequence @xmath468 , since the approximate sequence leaves the neighborhood of each sequence @xmath516 before that sequence has entered its stochastic regime ( which is characterized by the remainder @xmath515 $ ] ) .
again we invoke ` genericity ' to save the day : we conjecture that , for almost all approximate sequences , the sequences @xmath517 are a true random sample of kasner sequences and that the associated stochastic properties indeed carry over to the approximate sequence @xmath468 itself .
the piecewise exact sequence @xmath503 consists of pieces of length @xmath506 , where @xmath518 grows beyond all bounds as @xmath507 , because shadowing takes place with an increasing degree of accuracy .
however , this does not necessarily imply that @xmath519 as @xmath507 .
the latter property is directly connected with the _ kasner map convergence conjecture_. almost every generic type ix orbit is associated with an approximate kasner sequence @xmath468 such that @xmath520 as @xmath459 , where @xmath470 describes the error between @xmath469 and the value of the kasner parameter generated from @xmath467 by the exact kasner map .
the relevance of the kasner map for the asymptotic evolution of type ix solutions rests on the validity of the kasner map convergence conjecture .
let us thus give a more detailed discussion and present the line of arguments that leads to this conjecture .
an orbit @xmath433 ( with segments @xmath521 , @xmath522 ) generates a sequence of check points @xmath523 , which in turn yields a map @xmath524 .
the parameter @xmath469 ( associated with @xmath525 ) is generated from @xmath467 ( associated with @xmath526 ) by the kasner map plus an error @xmath470 , see figure [ checkmap ] .
the magnitude of the error @xmath470 depends on the initial data of the segment , i.e. , on @xmath465 ; equivalently , we may view @xmath470 as a function depending on ( i ) the position of @xmath472 on @xmath140 , and ( ii ) the vector @xmath527 , which is orthogonal to @xmath140 at @xmath472 . to obtain an estimate for @xmath470 we introduce the _ order of magnitude _ of the error , which we denote by @xmath528 .
we define @xmath528 to be the average of @xmath529 over all vectors @xmath527 of equal length that are orthogonal to @xmath140 ; alternatively , we use the somewhat ` safer ' definition of @xmath528 as the maximum of @xmath529 . by design ,
the order of magnitude @xmath528 of the error is a function of two variables : ( i ) the position of @xmath472 on @xmath140 , which is invariantly represented by @xmath467 , and , instead of @xmath527 itself , ( ii ) the ( orthogonal ) distance of @xmath465 from @xmath472 ( or , equivalently , from @xmath140 ) , i.e. , @xmath530 ( @xmath531 ) , which we denote by @xmath474 , see figure [ checkmap ] ; in brief , @xmath532 . in the following we investigate the behavior of the function @xmath533 under variations of the two arguments .
\(i ) @xmath534 .
keep @xmath535 fixed ( and assume that @xmath535 lies in the interval @xmath536 so that its image under the kasner map is finite ) .
then @xmath537 is a function of the distance @xmath538 such that @xmath539 as @xmath540 ( which is a simple consequence of the regularity of the dynamical system and continuous dependence on initial data ) .
it is reasonable to conjecture that the relation is in fact monotone .
\(ii ) @xmath541 .
keep the distance @xmath538 fixed , so that @xmath537 is a function of @xmath535 .
the fundamental observation is that @xmath537 becomes unbounded as ( a ) @xmath542 , and ( b ) @xmath543 .
case ( a ) is due to the intricacies of the flow in the vicinity of the non - transversally hyperbolic taub points ( where @xmath484 ) ; recall that @xmath544 measures the angular distance of the check point from the nearest taub point . for ( b )
we consider the orbit @xmath545 ( for some @xmath154 ) which corresponds to @xmath546 .
suppose that @xmath472 coincides with @xmath106 ( i.e. , @xmath547 ) .
in general , @xmath473 will not coincide with @xmath109 ( independently of the choice of @xmath548 ) ; therefore @xmath549 .
if @xmath472 is very close to @xmath106 , which means that @xmath467 is close to @xmath194 , a small deviation of @xmath433 from the type ii transition emanating from @xmath472 will still be small at the end point @xmath471 ; however , in the vicinity of the taub point @xmath109 even a small deviation can translate to a large error @xmath550 between @xmath469 and @xmath551 , which in turn results in the asserted blow - up of the order of magnitude of the error .
the qualitative properties of @xmath537 as a function of @xmath538 and @xmath535 are depicted in figure [ uerror ] .
[ cc][cc][0.7][0]@xmath552 [ cc][cc]@xmath535 [ cc][cc]@xmath537 [ cc][cc][1][0]@xmath538 consider a generic type ix orbit @xmath433 and its representation as an approximate sequence of transitions . by theorem [ rinthm ] each orbit converges to the mixmaster attractor . as a consequence
the distance @xmath553 ( @xmath554 ) converges to zero as @xmath459 .
therefore , as @xmath555 ( @xmath556 ) , the sequence of functions @xmath557 of figure [ uerror ] converges to zero .
however , this convergence is merely pointwise and not uniform ; this behavior is crucial to understand the behavior of @xmath470 as @xmath459 .
if a type ix orbit @xmath433 converges to one of the heteroclinic cycles on the mixmaster attractor , then there exists @xmath558 such that @xmath559 for all sufficiently large @xmath256 . on this interval
the function @xmath537 of figure [ uerror ] converges to zero uniformly .
we therefore obtain that @xmath560 and thus @xmath479 as @xmath459 for this type ix orbit .
however , the mixmaster attractor conjecture suggests that almost every type ix orbit @xmath433 has an associated approximate kasner sequence @xmath468 that is unbounded , hence the general case is not so clear . since @xmath468 enters the intervals @xmath561 and @xmath562 for any @xmath558 , @xmath563 ( and thus @xmath470 ) need not necessarily converge to zero .
let us elaborate .
let @xmath564 be fixed . by the implicit function theorem , the inequality @xmath565 is satisfied if and only if @xmath566 or @xmath567 for some functions @xmath568 , @xmath569 , which satisfy @xmath570 and @xmath571 as @xmath572 .
accordingly , for a given type ix orbit ( and its associated approximated kasner sequence @xmath467 ) , we obtain @xmath573 or @xmath574 .
we call the union of the two intervals @xmath575 and @xmath576 the ` hazard zone ' ( associated with epoch number @xmath256 ) ; obviously , the hazard zone is decreasing with @xmath256 .
hence , the order of magnitude of the error , @xmath528 , is larger than @xmath577 at epoch number @xmath256 if and only if the approximate kasner parameter @xmath467 falls into the hazard zone ( associated with @xmath256 ) . in connection with the kasner map convergence conjecture
the question is how often @xmath468 enters the hazard zone , i.e. , how often occurs : finitely many times or infinitely many times ?
suppose that the mixmaster stochasticity conjecture is correct .
then almost every sequence @xmath468 admits a probabilistic interpretation in terms of the probability distribution .
accordingly , we expect the question to turn into an example of a ` 01 law ' : there exists two alternatives : ( i ) type ix orbits ( i.e. , sequences @xmath468 ) that satisfy infinitely many times are generic ( i.e. , of measure @xmath194 in the state space ) ; type ix orbits that do not are non - generic ( i.e. , of measure @xmath578 ) ; ( ii ) type ix orbits that satisfy infinitely many times are non - generic ; orbits that do not are generic .
( in both cases , the two sets , the ` generic set ' and the ` non - generic set ' , are probably non - empty . )
which of these alternatives is actually realized depends on the rate of decay of @xmath579 and the decay and growth of @xmath568 and @xmath569 , respectively , as @xmath572 .
case ( i ) is associated with small ( subcritical ) rates of decay ; case ( ii ) with large ( overcritical ) decay .
we conjecture that case ( ii ) applies , i.e. , is satisfied only finitely many times for generic type ix orbits .
the reason for this conjecture is the fast rate of convergence of the orbit to the mixmaster attractor which follows from the decay of the functions @xmath580 , cf . , and @xmath581 , cf . @xcite ; this leads to the expectation that the r.h .
side of represents a rapidly decaying function . to conclude our line of arguments in favor of the kasner map convergence conjecture , we note that the actual error @xmath470 can be estimated by @xmath528
therefore , if @xmath582 for all @xmath256 except for a finite set , then also @xmath583 for all @xmath256 except for a finite set . since @xmath577 is an arbitrary positive number , the statement of the kasner map convergence conjecture ensues .
we round off this section with some further remarks on the conjectures .
first , we note that one might be led to suppose that there could in fact exist type ix solutions for which the statement of the kasner map convergence conjecture is violated , i.e. , @xmath480 as @xmath462 . in any case
, the class of these special solutions is at most of measure zero .
second , we note that the statement of the kasner map convergence conjecture , i.e. , @xmath479 ( @xmath459 ) , and the statement @xmath519 ( @xmath507 ) , cf . the discussion on piecewise exact sequences , are expressions of one and the same stochastic property .
this emphasizes that the two ` stochastic ' conjectures tightly intertwine .
third , we remark that it is conceivable that ` almost every ' in the two conjectures might be in fact ` every ' . in that case , the non - generic examples of kasner / mixmaster sequences , see section [ maps ] , would not have any counterparts among the type ix solutions .
of particular interest in this context would be to have an answer to question ( ii ) of the second subsection of section [ beliefs ] .
if there do not exist type ix orbits whose @xmath154-limit set is one of the heteroclinic cycles , this would be a strong indication in favor of ` every ' and against ` almost every ' .
if there exist type ix orbits that converge to a cycle , ` almost every ' is the best one can aim for in the mixmaster stochasticity conjecture . finally , let us briefly comment on claims of stochastic and chaotic properties of bianchi type ix asymptotic dynamics .
chaotic aspects of the kasner map and related maps have been studied under various aspects @xcite . however , the relevance of these maps for type ix asymptotics rests on the two conjectures in this section .
if the conjectures turn out to be wrong ( e.g. , if there exists a generic set of solutions that converge to heteroclinic cycles ) , then none of the results on the kasner map carry over to full type ix dynamics .
numerical investigations , see @xcite and references therein , reflect theorem [ rinthm ] and finite shadowing , but it is implausible that numerics can possibly shed light on the actual asymptotic limit of type ix solutions .
numerical errors are unavoidable and random in nature ; these errors generate precisely the type of stochasticity the simulation is looking for . accordingly
, numerical studies will necessarily find type ix solutions to exhibit the same stochastic behavior as generic kasner sequences and are thus not suited to explore the validity of the conjectures .
in this section we briefly review the hamiltonian billiard approach to type ix dynamics in the spirit of @xcite and make contact with the dynamical systems approach .
the metric can be written as @xmath584 where @xmath585 is an arbitrary time variable that is related to proper time @xmath57 by a densitized lapse function @xmath586 according to @xmath587 ; evidently , @xmath588 $ ] .
the hamiltonian for the orthogonal bianchi type ix perfect fluid models is given by @xmath589 where @xmath590 is three - curvature and @xmath591 , cf .
the remark following ( where we recall that @xmath592 ) .
@xmath593 is the inverse of the so - called minisuperspace metric @xmath594 , @xmath595 i.e. , @xmath596 is a @xmath597-dimensional lorentzian metric .
the gravitational potential @xmath598 is given by @xmath599 where we have set the structure constants @xmath600 to one ; the potential for the fluid is given by @xmath601 $ ] .
for further details see , e.g. , @xcite or ( * ? ? ?
* chapter 10 ) ; note that @xmath602 , which is used in these references ; see also @xcite and @xcite .
as argued in @xcite , @xmath603 is expected to be timelike in the asymptotic regime , i.e. , @xmath604 . assuming that @xmath603 is timelike allows to introduce new metric variables instead of @xmath605 .
defining @xmath606 and orthogonal angular metric variables , collectively denoted by @xmath153 , leads to @xmath607 where @xmath608 is the standard metric on hyperbolic space . making a further change of variables , @xmath609 yields @xmath610\:.\ ] ] choosing the lapse according to @xmath611 leads to a hamiltonian of the form @xmath612 + \bar{\rho}^2 \sum_a c_a\,\exp(-2 \bar{\rho}\ , w_a(\gamma))\:,\ ] ] where @xmath613 denotes certain linear forms of the variables @xmath614 , i.e. , @xmath615 ; see @xcite for details .
the essential point is that one expects that @xmath616 towards the singularity and hence that each term @xmath617 $ ] becomes an infinitely high sharp wall described by an infinite step function @xmath618 that vanishes for @xmath619 and is infinite for @xmath620 .
accordingly , only ` dominant ' terms in the potential are assumed to be of importance for the generic asymptotic dynamics , while ` subdominant ' terms , i.e. , terms whose exponential functions can be obtained by multiplying dominant wall terms , are neglected . in the present case
there are three dominant terms in @xmath621 ( which is the minimal set of terms required to define the billiard table ) , the three exponentials @xmath622 .
dropping the subdominant terms in the limit @xmath623 leads to an asymptotic hamiltonian of the form @xmath624 + \sum_{a=1}^3 \theta_\infty(-2 w_a(\gamma))\:,\ ] ] where only the three dominant terms appear in the sum .
the correspondence between the dynamical systems picture and the hamiltonian picture is easily obtained by noting that the hamiltonian constraint is proportional to the gauss constraint .
the dominant terms correspond to the terms @xmath625 , @xmath626 , in ; the subdominant terms are collected in @xmath400 .
the non - trivial dynamics described by resides in the variables @xmath153 , i.e. , in the hyperbolic space .
it can be described asymptotically as geodesic motion in hyperbolic space constrained by the existence of sharp reflective walls , i.e. , the asymptotic dynamics is determined by the type ix ` billiard ' given in figure [ billiardix ] .
based on the heuristic considerations the limiting hamiltonian is believed to describe generic asymptotic dynamics .
[ cc][cc]@xmath627 [ cc][cc]@xmath628 [ cc][cc]@xmath629 the hamiltonian picture ( as represented by the billiard ) and the dynamical systems picture ( as represented by the mixmaster attractor ) are ` dual ' to each other . in the hamiltonian picture
the asymptotic dynamics is described as the evolution of a point in the billiard .
straight lines ( geodesics ) in hyperbolic space correspond to kasner states .
wall bounces correspond to bianchi type ii solutions ; the bounces change the kasner states according to the kasner map .
since the billiard picture emphasizes the dynamics of the configuration space variables , one may say that the hamiltonian billiard approach yields a ` configuration space ' representations of the essential asymptotic dynamics . in the dynamical systems state space picture
the motion in the hamiltonian billiard picture becomes a ` wall ' of fixed points
the kasner circle @xmath101 .
the walls in the hamiltonian billiard are translated to motion in the dynamical systems picture bianchi type ii heteroclinic orbits ; these type ii transitions yield exactly the same rule for changing kasner states as the wall bounces : the kasner map .
since the variables
@xmath630 are proportional to @xmath631 , it is natural to refer to the projected dynamical systems picture as a ` momentum space ' representation of the asymptotic dynamics . summing up , ` walls ' and ` straight line motion ' switch places between the hamiltonian formulation and dynamical systems description of asymptotic dynamics , and the two pictures
give equivalent complementary asymptotic pictures .
the above heuristic derivation of the limiting hamiltonian rests on two basic assumptions : it is assumed that @xmath603 is timelike in the asymptotic regime and that one can drop the subdominant terms .
these assumptions correspond to assuming that @xmath44 and @xmath400 can be set to zero asymptotically , i.e. , the procedure precisely assumes theorem [ rinthm ] .
( the hamiltonian analysis in ( * ? ?
? * chapter 2 ) uses the same assumptions , and hence the present discussion is of direct relevance for that work as well . ) that these assumptions are highly non - trivial is indicated by the difficulties that the proof of theorem [ rinthm ] has presented ; cf .
an alarming example is bianchi type viii ; in this case the heuristic procedure in this respect leads to exactly the same asymptotic results , but so far there exist no proof that @xmath44 and @xmath400 tend to zero towards the singularity for generic solutions .
we elaborate on the differences between type viii and type ix in @xcite .
moreover , like in the state space picture there exists no proof that all of the possible billiard trajectories are of relevance for the asymptotic dynamics of type ix solutions .
for example , there exists a correspondence between periodic orbits in the dynamical systems picture and the hamiltonian billiard picture , and it is not excluded a priori that solutions are forced to one , or several , of these .
we emphasize again that all proposed measures of chaos that take billiards as the starting point , see @xcite and references therein , rely the conjectured connection between the mixmaster map and asymptotic type ix dynamics .
the above discussion shows that there are non - trivial assumptions and subtle phenomena that are being glossed over in heuristic billiard ` derivations . '
nevertheless , we do believe that the billiard procedure elegantly uncovers the main generic asymptotic features , and it might be fruitful to attempt to combine the dynamical systems and hamiltonian picture in order to prove the mixmaster conjectures .
a tantalizing hint is that in the billiard picture @xmath632 becomes an asymptotic constant of the motion . rewriting this dimensional constant of the motion in terms of the dynamical systems variables yields @xmath633\:,\ ] ] where @xmath634 and @xmath635 is a constant .
the purpose of this paper is two - fold . on the one hand
, we analyze the main known results on the asymptotic dynamics of bianchi type ix vacuum and orthogonal perfect fluid models towards the initial singularity . the setting for
our discussion is the hubble - normalized dynamical systems approach , since this is essentially the set - up which has led to the first rigorous statements on bianchi type ix asymptotics @xcite .
( we choose slightly different variables to emphasize the permutation symmetry that underlies the problem . ) the main result ( ` mixmaster fact ' ) is theorem [ rinthm ] which is due to ringstrm @xcite ; for an alternative proof see @xcite .
this theorem , in conjunction with an analysis of the mixmaster attractor , leads to a number of further rigorous results which we list as consequences @xcite .
on the other hand , we draw a clear line between rigorous results ( ` facts ' ) and heuristic considerations ( ` beliefs ' ) .
we make explicit that the implications of theorem [ rinthm ] and its consequences are rather limited , in particular , the rigorous results do not give any information on the details of the oscillatory nature of mixmaster asymptotics .
the mathematical methods required to obtain proofs about the actual asymptotic mixmaster oscillations are yet to be developed ; it is likely that radically new ideas are needed .
this paper provides the infrastructure that might yield the basis for further developments .
our framework enables us to transform vague beliefs to a number of specific conjectures that describe the expected ` complete picture ' of bianchi type ix asymptotics .
the arguments we give in their favor are based on an in - depth analysis of the mixmaster map and its stochastic aspects in combination with the dynamical systems concept of shadowing .
we conclude with a few pertinent comments .
first , as elaborated in @xcite there exists no corresponding theorem to theorem [ rinthm ] in other oscillatory bianchi models such as bianchi type vi@xmath636 or viii ; this suggests that the situation in the general inhomogeneous case is even more complicated than expected .
furthermore , numerical studies are incapable of shedding light on the asymptotic limit .
this is mainly due to the accumulation of inevitable random numerical errors that make it a priori impossible to track a particular type ix orbit .
finally , although the hamiltonian methods are a formidable heuristic tool , so far this approach has not yielded any proofs about asymptotics .
nevertheless , it might prove to be beneficial to explore the possible synergies between dynamical systems and hamiltonian methods . in this paper and in @xcite we have encountered remarkable subtleties as regards the asymptotic dynamics of oscillatory singularities ; this emphasizes the importance of a clear distinction between facts and beliefs .
we thank alan rendall , hans ringstrm , and in particular lars andersson for useful discussions .
we gratefully acknowledge the hospitality of the mittag - leffler institute , where part of this work was completed .
cu is supported by the swedish research council .
motter and p.s .
mixmaster chaos .
127 ( 2001 ) .
more qualitative cosmology 137 ( 1971 ) .
o.i . bogoyavlensky and s.p .
novikov singularities of the cosmological model of the bianchi type ix type according to the qualitative theory of differential equations 747 ( 1973 ) .
o.i . bogoyavlensky .
( springer - verlag , 1985 ) .
global dynamics of the mixmaster model .
2341 ( 1997 ) . h. ringstrm .
curvature blow up in bianchi viii and ix vacuum spacetimes . 713
h. ringstrm .
the bianchi ix attractor . 405
heinzle and c. uggla . a new proof of the bianchi type ix attractor theorem . preprint ( 2009 ) .
t. damour , m. henneaux , and h. nicolai .
cosmological billiards . r145 ( 2003 ) .
h. friedrich , a.d .
the cauchy problem for the einstein equations .
127 ( 2000 ) .
l. andersson , a.d .
quiescent cosmological singularities 479 ( 2001 ) .
jantzen and c. uggla . the kinematical role of automorphisms in the orthonormal frame approach to bianchi cosmology . 353
( 1999 ) .
lin and r.m .
proof of the closed - universe - recollapse conjecture for diagonal bianchi type ix cosmologies .
3280 ( 2003 ) .
heinzle , n. rohr , and c. uggla .
matter and dynamics in closed cosmologies .
083506 ( 2005 ) .
spatially homogeneous dynamics : a unified picture . in _ proc .
int . sch .
e. fermi " course lxxxvi on gamov cosmology " _ , r. ruffini , f. melchiorri , eds .
( north holland , amsterdam , 1987 ) and in _ cosmology of the early universe _ , r. ruffini , l.z .
fang , eds .
( world scientific , singapore , 1984 ) .
arxiv : gr - qc/0102035 .
collins and j.m .
qualitative cosmology .
419 ( 1971 ) .
lim , c. uggla , and j. wainwright .
asymptotic silence - breaking singularities .
2607 ( 2006 ) .
j. wainwright , m. j. hancock , and c. uggla .
asymptotic self - similarity breaking at late times in cosmology .
2577 ( 1999 ) . | we consider the dynamics towards the initial singularity of bianchi type ix vacuum and orthogonal perfect fluid models with a linear equation of state .
surprisingly few facts are known about the ` mixmaster ' dynamics of these models , while at the same time most of the commonly held beliefs are rather vague . in this paper
, we use mixmaster facts as a base to build an infrastructure that makes it possible to sharpen the main mixmaster beliefs .
we formulate explicit conjectures concerning ( i ) the past asymptotic states of type ix solutions and ( ii ) the relevance of the mixmaster / kasner map for generic past asymptotic dynamics .
the evidence for the conjectures is based on a study of the stochastic properties of this map in conjunction with dynamical systems techniques .
we use a dynamical systems formulation , since this approach has so far been the only successful path to obtain theorems , but we also make comparisons with the ` metric ' and hamiltonian ` billiard ' approaches . |
stars more massive than @xmath4 25 @xmath5 form a black hole at the end of their evolution .
stars with non - rotating black holes are likely to collapse `` quietly '' ejecting a small amount of heavy elements ( faint supernovae ) . in contrast , stars with rotating black holes are likely to give rise to very energetic supernovae ( hypernovae )
. we present distinct nucleosynthesis features of these two classes of `` black - hole - forming '' supernovae .
nucleosynthesis in hypernovae is characterized by larger abundance ratios ( zn , co , v , ti)/fe and smaller ( mn , cr)/fe than normal supernovae , which can explain the observed trend of these ratios in extremely metal - poor stars .
nucleosynthesis in faint supernovae is characterized by a large amount of fall - back .
we show that the abundance pattern of the recently discovered most fe - poor star , he0107 - 5240 , and other extremely metal - poor stars are in good accord with those of black - hole - forming supernovae , but not pair - instability supernovae .
this suggests that black - hole - forming supernovae made important contributions to the early galactic ( and cosmic ) chemical evolution .
finally we discuss the nature of first ( pop iii ) stars .
type ic hypernovae 1998bw and 2003dh were clearly linked to the gamma - ray bursts grb 980425 ( galama et al .
1998 ) and grb 030329 ( stanek et al . 2003 ; hjorth et al . 2003 ) , thus establishing the connection between long grbs and core - collapse supernovae ( sne ) .
sne 1998bw and 2003dh were exceptional for sne ic : they were as luminous at peak as a sn ia , indicating that they synthesized 0.3 - 0.5 @xmath5 of @xmath6ni , and their kinetic energy ( ke ) were estimated as @xmath7erg @xmath4 30 ( iwamoto , mazzali , nomoto , et al .
1998 ; woosley , eastman , & schmidt 1999 ; nakamura et al . 2001a ; mazzali et al . 2003 ) .
other `` hypernovae '' have been recognized , such as sn 1997ef ( iwamoto et al .
2000 ; mazzali , iwamoto , & nomoto 2000 ) , sn 1999as ( knop et al .
1999 ; hatano et al .
2001 ) , and sn 2002ap ( mazzali et al .
these hypernovae span a wide range of properties , although they all appear to be highly energetic compared to normal core - collapse sne .
the mass estimates , obtained from fitting the optical light curves and spectra , place hypernovae at the high - mass end of sn progenitors .
in contrast , sne ii 1997d and 1999br were very faint sne with very low ke ( turatto et al . 1998 ; hamuy 2003 ; zampieri et al . 2003 ) . in the diagram that shows @xmath8 and the mass of @xmath6ni ejected @xmath9ni ) as a function of the main - sequence mass @xmath10 of the progenitor star ( figure [ fig2 ] ) , therefore , we propose that sne from stars with @xmath11 20 - 25 @xmath5 have different @xmath8 and @xmath9ni ) , with a bright , energetic `` hypernova branch '' at one extreme and a faint , low - energy sn branch at the other ( nomoto et al .
for the faint sne , the explosion energy was so small that most @xmath6ni fell back onto the compact remnant .
thus the faint sn branch may become a `` failed '' sn branch at larger @xmath10 . between the two branches , there may be a variety of sne ( hamuy 2003 ) .
this trend might be interpreted as follows .
stars with @xmath12 20 - 25 @xmath5 form a neutron star , producing @xmath4 0.08 @xmath13 0.03 @xmath5 @xmath6ni as in sne 1993j , 1994i , and 1987a .
stars with @xmath11 20 - 25 @xmath5 form a black hole ; whether they become hypernovae or faint sne may depend on the angular momentum in the collapsing core , which in turn depends on the stellar winds , metallicity , magnetic fields , and binarity .
hypernovae might have rapidly rotating cores owing possibly to the spiraling - in of a companion star in a binary system .
the light curve modeling of the unusually bright hypernova sn1999as suggests that the progenitor is a core - collapse supernova and the ejected @xmath6ni mass is as large as @xmath15 .
motivated by sn 1990as , umeda & nomoto ( 2004 ) have investigated how much @xmath6ni can be synthesized in core - collapse massive supernovae .
the evolutions of several very massive stars with initial masses of @xmath16 and low metallicity ( @xmath17 ) have been calculated from the main - sequence to `` hypernova '' explosions .
the synthesized @xmath6ni mass increases with the increasing explosion energy and the progenitor mass .
umeda & nomoto ( 2004 ) found that for the explosion energy of 3@xmath18 ergs , for example , the @xmath6ni mass of up to 2.2 , 2.3 , 5.0 , and 6.6 @xmath5 can be produced for the progenitors with masses of 30 , 50 , 80 and 100 @xmath5 , that are sufficiently large to explain sn 1999as .
figure [ rhot ] shows the evolution of the central density and temperature for the 30 and 90@xmath5 models .
more massive stars have larger specific entropy at the center , thus having higher temperature for the same density . for 90@xmath5
, the evolutinary track is very close to ( but outside of ) the `` e@xmath19e@xmath20 pair - instabillity region '' of @xmath21 where @xmath22 denotes the adiabatic index .
the evolution of the central temperature and density is significantly different between the 30 and 90@xmath5 models during si - burning at @xmath23k = @xmath24 .
the central temperature and density of the 90@xmath5 model oscillate several times .
this is because in such massive stars radiation pressure is so dominant that @xmath22 is close to 4/3 , and thus the inner core of the stars easily expands with the nuclear energy released by si - burning .
once it expands , the temperature drops suddenly , the central si - burning stops , and the stellar core turns into shrink .
since only small amount of si is burnt for each cycle , this pulsations occur many times .
umeda & nomoto ( 2004 ) found from the study of 80@xmath25 stars that the number of the oscillations depends on the convective parameter @xmath26 : larger @xmath26 increases the number of the oscillation .
this is because for a larger @xmath26 fresh si is mixed more efficiently into the center , which increase the lifetime of this stage .
the situation is similar to the breathing pulse phase at the end of he - burning , while the variation of the temperature and density is much larger in si burning than in the breathing pulse phase .
the amplitude of the temperature and density variation is larger for more massive stars , which suggests more and more drastic oscillations would occur for larger mass stars .
in core - collapse supernovae / hypernovae , stellar material undergoes shock heating and subsequent explosive nucleosynthesis .
iron - peak elements are produced in two distinct regions , which are characterized by the peak temperature , @xmath27 , of the shocked material . for @xmath28k
, material undergoes complete si burning whose products include co , zn , v , and some cr after radioactive decays . for @xmath29k @xmath30k ,
incomplete si burning takes place and its after decay products include cr and mn ( nakamura et al .
1999 ) .
the right panel of figure [ fig3 ] shows the composition in the ejecta of a 25 @xmath5 hypernova model ( @xmath31 ) .
the nucleosynthesis in a normal 25 @xmath5 sn model ( @xmath32 ) is also shown for comparison in the left panel of figure [ fig3 ] ( umeda & nomoto 2002 ) .
we note the following characteristics of nucleosynthesis with very large explosion energies ( nakamura et al . 2001b ; nomoto et al . 2001 ; umeda & nomoto 2005 ) : \(1 ) both complete and incomplete si - burning regions shift outward in mass compared with normal supernovae , so that the mass ratio between the complete and incomplete si - burning regions becomes larger . as a result ,
higher energy explosions tend to produce larger [ ( zn , co , v)/fe ] and smaller [ ( mn , cr)/fe ] , which can explain the trend observed in very metal - poor stars ( umeda & nomoto 2005 ) .
\(2 ) in the complete si - burning region of hypernovae , elements produced by @xmath33-rich freezeout are enhanced .
hence , elements synthesized through capturing of @xmath33-particles , such as @xmath34ti , @xmath35cr , and @xmath36ge ( decaying into @xmath34ca , @xmath35ti , and @xmath36zn , respectively ) are more abundant .
\(3 ) oxygen burning takes place in more extended regions for the larger ke .
then more o , c , al are burned to produce a larger amount of burning products such as si , s , and ar .
therefore , hypernova nucleosynthesis is characterized by large abundance ratios of [ si , s / o ] , which can explain the abundance feature of m82 ( umeda et al . 2002 ) .
hypernova nucleosynthesis may have made an important contribution to galactic chemical evolution . in the early galactic epoch when the galaxy was not yet chemically well - mixed
, [ fe / h ] may well be determined by mostly a single sn event ( audouze & silk 1995 ) .
the formation of metal - poor stars is supposed to be driven by a supernova shock , so that [ fe / h ] is determined by the ejected fe mass and the amount of circumstellar hydrogen swept - up by the shock wave ( ryan , norris , & beers 1996 ) .
then , hypernovae with larger @xmath8 are likely to induce the formation of stars with smaller [ fe / h ] , because the mass of interstellar hydrogen swept up by a hypernova is roughly proportional to @xmath8 ( ryan et al .
1996 ; shigeyama & tsujimoto 1998 ) and the ratio of the ejected iron mass to @xmath8 is smaller for hypernovae than for normal supernovae . in the observed abundances of halo stars ,
there are significant differences between the abundance patterns in the iron - peak elements below and above [ fe / h]@xmath37 - @xmath38 .
\(1 ) for [ fe / h]@xmath39 , the mean values of [ cr / fe ] and [ mn / fe ] decrease toward smaller metallicity , while [ co / fe ] increases ( mcwilliam et al . 1995 ; ryan et al .
1996 ) .
\(2 ) [ zn / fe]@xmath40 for [ fe / h ] @xmath41 to @xmath42 ( sneden , gratton , & crocker 1991 ) , while at [ fe / h ] @xmath43 , [ zn / fe ] increases toward smaller metallicity ( cayrel et al .
2004 ) .
the larger [ ( zn , co)/fe ] and smaller [ ( mn , cr)/fe ] in the supernova ejecta can be realized if the mass ratio between the complete si burning region and the incomplete si burning region is larger , or equivalently if deep material from the complete si - burning region is ejected by mixing or aspherical effects .
this can be realized if ( 1 ) the mass cut between the ejecta and the compact remnant is located at smaller @xmath44 ( nakamura et al .
1999 ) , ( 2 ) @xmath8 is larger to move the outer edge of the complete si burning region to larger @xmath44 ( nakamura et al .
2001b ) , or ( 3 ) asphericity in the explosion is larger . among these possibilities ,
a large explosion energy @xmath8 enhances @xmath33-rich freezeout , which results in an increase of the local mass fractions of zn and co , while cr and mn are not enhanced ( umeda & nomoto 2002 , 2005 ) . models with @xmath45 do not produce sufficiently large [ zn / fe ] . to be compatible with the observations of [ zn / fe ] @xmath46 , the explosion energy must be much larger , i.e. , @xmath47 for @xmath48 , i.e. , hypernova - like explosions of massive stars ( @xmath49 ) with @xmath50 are responsible for the production of zn . in the hypernova models ,
the overproduction of ni , as found in the simple `` deep '' mass - cut model , can be avoided ( umeda & nomoto 2005 ) . therefore ,
if hypernovae made significant contributions to the early galactic chemical evolution , it could explain the large zn and co abundances and the small mn and cr abundances observed in very metal - poor stars ( fig . [ fig4 ] : tominaga et al .
recently the most fe deficient and c - rich low mass star , he0107 - 5240 , was discovered ( christlieb et al .
2002 , 2004 ) .
this star has [ fe / h ] @xmath51 but its mass is as low as 0.8 @xmath5 .
this would challenge the recent theoretical arguments that the formation of low mass stars , which should survive until today , is suppressed below [ fe / h ] @xmath52 ( schneider et al .
2002 ) .
the important clue to this problem is the observed abundance pattern of this star .
this star is characterized by a very large ratios of [ c / fe ] = 4.0 and [ n / fe ] = 2.3 , while the abundances of elements heavier than mg are as low as fe ( christlieb et al .
interestingly , this is not the only extremely metal poor ( emp ) stars that have the large c / fe and n / fe ratios , but several other such stars have been discovered ( aoki et al .
therefore the reasonable explanation of the abundance pattern should explain other emp stars as well .
we show that the abundance pattern of c - rich emp stars can be reasonably explained by the nucleosynthesis of 20 - 130 @xmath5 supernovae with various explosion energies and the degree of mixing and fallback of the ejecta .
we consider a model that c - rich emp stars are produced in the ejecta of ( almost ) metal - free supernova mixed with extremely metal - poor interstellar matter .
we use pop iii pre - supernova progenitor models , simulate the supernova explosion and calculate detailed nucleosynthesis ( umeda & nomoto 2003 ) . in figure [ fig5 ] ( right )
we show that the elemental abundances of one of our models are in good agreement with he0107 - 5240 , where the progenitor mass is 25 @xmath5 and the explosion energy @xmath53 0.3 ( umeda & nomoto 2003 ) . in this model ,
explosive nucleosynthesis takes place behind the shock wave that is generated at @xmath54 1.8 @xmath5 and propagates outward . the resultant abundance distribution is seen in figure [ fig5 ] ( left ) , where @xmath44 denotes the lagrangian mass coordinate measured from the center of the pre - supernova model .
the processed material is assumed to mix uniformly in the region from @xmath55 1.8 @xmath5 and 6.0 @xmath5 .
such a large scale mixing was found to take place in sn1987a and various explosion models ( hachisu et al . 1990 ) .
almost all materials below @xmath54 6.0 @xmath5 fall back to the central remnant and only a small fraction ( @xmath56 10@xmath57 ) is ejected from this region .
the ejected fe mass is 8 @xmath2 10@xmath58 @xmath5 .
the cno elements in the ejecta were produced by pre - collapse he shell burning in the he - layer , which contains 0.2 @xmath5 @xmath59c .
mixing of h into the he shell - burning region produces 4 @xmath2 10@xmath60 @xmath5 @xmath61n . on the other hand , only a small amount of heavier elements ( mg , ca , and fe - peak elements ) are ejected and their abundance ratios are the average in the mixed region
. the sub - solar ratios of [ ti / fe ] @xmath62 and [ ni / fe ] @xmath62 are the results of the relatively small explosion energy ( @xmath53 0.3 ) . with this `` mixing and fallback '' , the large c / fe and c / mg ratios observed in he0107 - 5240 are well reproduced . in this model
, n / fe appears to be underproduced .
however , n can be produced inside the emp stars through the c - n cycle , and brought up to the surface during the first dredge up stage while becoming a red - giant star ( boothroyd & sackmann 1999 ) .
the `` mixing and fall back '' process can reproduce the abundance pattern of the typical emp stars without enhancement of c and n. figure [ fig7 ] shows that the averaged abundances of [ fe / h ] @xmath63 stars in cayrel et al .
( 2003 ) can be fitted well with the hypernova model of 25 @xmath5 and @xmath53 20 ( left ) but not with the normal sn model of 15 @xmath5 and @xmath53 1 ( right ) ( tominaga et al .
it is of vital importance to identify the first generation stars in the universe , i.e. , totally metal - free , pop iii stars .
the impact of the formation of pop iii stars on the evolution of the universe depends on their typical masses .
recent numerical models have shown that , the first stars are as massive as @xmath4 100 @xmath5 ( abel et al . 2002 ) .
the formation of long - lived low mass pop iii stars may be inefficient because of slow cooling of metal free gas cloud , which is consistent with the failure of attempts to find pop iii stars .
if the most fe deficient star , he0107 - 5240 , is a pop iii low mass star that has gained its metal from a companion star or interstellar matter ( yoshii 1981 ) , would it mean that the above theoretical arguments are incorrect and that such low mass pop iii stars have not been discovered only because of the difficulty in the observations ? based on the results in the earlier section , we propose that the first generation supernovae were the explosion of @xmath4 20 - 130 @xmath5 stars and some of them produced c - rich , fe - poor ejecta .
then the low mass star with even [ fe / h ] @xmath64 can form from the gases of mixture of such a supernova ejecta and the ( almost ) metal - free interstellar matter , because the gases can be efficiently cooled by enhanced c and o ( [ c / h ] @xmath65 ) .
in contrast to the core - collapse supernovae of 20 - 130 @xmath5 stars , the observed abundance patterns can not be explained by the explosions of more massive , 130 - 300 @xmath5 stars .
these stars undergo pair - instability supernovae ( pisne ) and are disrupted completely ( e.g. , umeda & nomoto 2002 ; heger & woosley 2002 ) , which can not be consistent with the large c / fe observed in he0107 - 5240 and other c - rich emp stars .
the abundance ratios of iron - peak elements ( [ zn / fe ] @xmath66 and [ co / fe ] @xmath67 ) in the pisn ejecta ( fig .
[ fig7 ] ; umeda & nomoto 2002 ; heger & woosley 2002 ) can not explain the large zn / fe and co / fe in the typical emp stars ( mcwilliam et al .
1995 ; norris et al . 2001 ; cayrel et al .
2003 ) and cs22949 - 037 either .
therefore the supernova progenitors that are responsible for the formation of emp stars are most likely in the range of @xmath68 @xmath5 , but not more massive than 130 @xmath5 .
this upper limit depends on the stability of massive stars as will be discussed below .
.the results of the stability analysis for pop iii and pop i stars . @xmath69 and
@xmath2 represent that the star is stable and unstable , respectively .
the @xmath70-folding time for the fundamental mode is shown after @xmath2 in units of @xmath71yr ( nomoto et al .
[ cols="^,^,^,^,^,^,^",options="header " , ] [ tab : pop3 ] to determine the upper limit mass of the zero age main sequence ( zams ) , nomoto et al .
( 2003 ) analyzed a linear non - adiabatic stability of massive ( @xmath72 - @xmath73 ) pop iii stars using a radial pulsation code .
because cno elements are absent during the early stage of their evolution , the cno cycle does not operate and the star contracts until temperature rises sufficiently high for the @xmath74 reaction to produce @xmath59c .
we calculate that these stars have @xmath75 , and the central temperature @xmath76 on their zams .
we also examine the models of pop i stars for comparison .
table [ tab : pop3 ] shows the results for our analysis .
the critical mass of zams pop iii star is @xmath77 while that of pop i star is @xmath78 .
this difference comes from very compact structures ( with high @xmath79 ) of pop iii stars .
stars more massive than the critical mass will undergo pulsation and mass loss .
we note that the @xmath70-folding time of instability is much longer for pop iii stars than pop i stars with the same mass , and thus the mass loss rate is much lower . these results are consistent with ibrahim et al .
( 1981 ) and baraffe et al .
however , the absence of the indication of pisne in emp stars might imply that these massive stars above 130@xmath80 undergo significant mass loss , thus evolving into fe core - collapse rather than pisne .
sne 1997cy and 1999e were initially classified as type iin because they showed h@xmath33 emission .
sn 2002ic would also have been so classified , had it not been discovered at an early epoch .
sn 1997cy ( @xmath81 ) is among the most luminous sne discovered so far ( @xmath82 about maximum light ) , and sn 1999e is also bright ( @xmath83 ) . both sne
1997cy and 1999e have been suspected to be spatially and temporally related to a grb ( germany et al .
2000 ; rigon et al .
2003 ) . however
, both the classification and the associations with a grb must now be seen as highly questionable in view of the fact that their replica , sn 2002ic , appears to have been a genuine sn ia at an earlier phase ( hamuy et al .
2003 ; deng et al . 2004 ) .
nomoto et al .
( 2005 ) calculated the interaction between the expanding ejecta and csm ( circumstellar matter ) . for the supernova ejecta ,
the carbon deflagration model w7 was used . after day @xmath4 350 ,
the light curve starts declining . to reproduce the declining part of the light curve , we add the outer csm of 0.2 @xmath5 where the density declines sharply as @xmath84 .
this implies that the total mass of csm is @xmath4 1.3 @xmath5 .
nomoto , k. , suzuki , t. , deng , j. , uenishi , t. , & hachisu , i. 2005 , 1604 - 2004 supernovae as cosmological lighthouse , ed .
l. zampieri et al .
( asp ) , in press norris , j.e . , ryan , s.g . , & beers , t.c .
2001 , , 561 , 1034 | recent studies of core - collapse supernovae have revealed the existence of two distinct classes of massive supernovae ( sne ) : 1 ) very energetic sne ( hypernovae ) , whose kinetic energy ( ke ) exceeds @xmath0erg , about 10 times the ke of normal core - collapse sne , and 2 ) very faint and low energy sne ( @xmath1 0.5 @xmath2 @xmath3erg ; faint supernovae ) .
these two classes of supernovae are likely to be `` black - hole - forming '' supernovae with rotating or non - rotating black holes .
we compare their nucleosynthesis yields with the abundances of extremely metal - poor ( emp ) stars to identify the pop iii ( or first ) supernovae .
we show that the emp stars , especially the c - rich type , are likely to be enriched by black - hole - forming supernovae . |
harmonics generation and wave - mixing are one of the basic phenomena of nonlinear optics which have been extensively studied both theoretically and experimentally with the advent of lasers @xcite .
recent advance in laser technologies has provided ultrahigh intensities for supershort laser pulses that makes achievable non - perturbative regime of harmonic generation , which significantly extends the spectral region accessible by lasers , in particular , for short wavelengths towards vuv / xuv or even x - ray radiation @xcite .
such short wavelength radiation is of great interest due to numerous significant applications , e.g. in quantum control , spectroscopy , sensing and imaging etc .. depending on the laser - atom interaction parameters , harmonic generation may arise from bound - bound @xcite and bound - free - bound transitions via continuum spectrum @xcite .
bound - bound mechanism of harmonic generation without ionization is more efficient for generation of moderately high harmonics @xcite . for this mechanism resonant interaction
is of importance .
besides pure theoretical interest as a simple model , resonant interaction regime exhibits significant enhancement of frequency conversion efficiencies @xcite .
however , to access highly excited states of atoms / molecules by optical lasers the multiphoton excitation problem arises .
required resonantly - driven multiphoton transition is effective for the systems with the mean dipole moments in the stationary states , or three - level atomic systems with close enough to each other two states and nonzero transition dipole moment between them @xcite . as a candidate
, we have studied the hydrogenlike atomic and ionic systems where the atom has a mean dipole moment in the excited stationary states , because of accidental degeneracy for the orbital momentum @xcite
. other interesting examples of efficient direct multiphoton excitation are molecules with a permanent dipole moments @xcite , evenly charged molecular ions at large internuclear distances @xcite , and artificial atoms @xcite realized in circuit quantum electrodynamics ( qed ) setups @xcite . in the work @xcite we have shown that the multiphoton resonant excitation of a three - level atomic / molecular system is efficient by the two bichromatic laser fields .
hence , having efficient two - color multiphoton resonant excitation scheme it is of interest to consider multicolor harmonic generation and wave - mixing processes by an atomic or molecular system under the such circumstances when only a bound states are involved in the interaction process , which is the purpose of the current paper .
the presence of the second laser provides additional flexibility for the implementation of multiphoton resonance expanding the spectrum of possible combinations .
moreover , two - color excitation extends accessible scattering frequencies with sum- and difference components . in the current paper , we employ analytical approach for high - order multiphoton resonant excitation of quantum systems which has been previously developed by us @xcite .
expression for the time - dependent mean dipole moment describing coherent part of scattering spectrum is obtained .
the results based on this expression are applied to hydrogen atom and evenly charged homonuclear diatomic molecular ion .
the main spectral characteristics of the considered process are in good agreement with the results of the performed numerical calculations .
estimations show that one can achieve enhanced generation of moderately high harmonics / wave - mixing via multiphoton resonant excitation by appropriate laser pulses .
our interest is also motivated by the advent of circuit qed setups @xcite where one can realize artificial atoms of desired configuration .
thus , the obtained results may also be is of interest for artificial atoms , and the presented results can be scaled to other systems and diverse domains of the electromagnetic spectrum .
the paper is organized as follows . in section
ii , we present the analytical model and derive the coherent contribution to the multicolor harmonic spectra . in section iii , we present some results of numerical calculations of the considered issue without a multiphoton resonant approximation and compare the obtained spectra with the analytical results . here , we consider concrete systems , such as hydrogenlike atom and evenly charged molecular ion .
finally , conclusions are given in section iv .
we consider a three - level quantum system interacting with the two laser fields of frequencies @xmath0 and @xmath1 as shown in fig.(1a ) .
it is assumed that the system is in a @xmath2 configuration in which a pair of upper levels @xmath3 and @xmath4 with permanent dipole moments are coupled to a lower level @xmath5 .
another possible three - level scheme is @xmath6 configuration shown in fig.(1b ) . in this case
the lower level @xmath5 is coupled to an upper level @xmath3 which has a strong dipole coupling to an adjacent level @xmath7 .
if the separation of the energy levels of the excited states is smaller than laser - atom interaction energy then by a unitary transformation @xcite the problem can be reduced to the @xmath2 configuration fig.(1a ) . as an illustrative example may serve hydrogenlike atom considered in parabolic @xcite and more conventional spherical coordinates
. in parabolic coordinates , the atom has a mean dipole moment in the excited states , while in the second case because of the random degeneracy of the orbital moment there is a dipole coupling between the degenerate states , but the mean dipole moment is zero for the stationary states .
the inverse with respect to the @xmath2 configuration is the polar @xmath8 configuration , which can be realized for artificial atoms @xcite .
hence , as a general model we will consider the scheme of the @xmath2 configuration .
type with mean dipole moments in the excited states and
( b ) @xmath6 configuration with the coupling transition between the excited states .
the considered configurations are unitary equivalent to each other . ]
the hamiltonian for the system within semiclassical dipole approximation is given in form @xmath9@xmath10where , @xmath11 , @xmath12 and @xmath13 are the energies of the stationary states @xmath14 , @xmath15 and @xmath4 respectively , and@xmath16is the interaction part of the hamiltonian with a real matrix element of the electric dipole moment projection @xmath17 @xmath18 , and @xmath19 are slowly varying amplitudes of linearly polarized laser fields , with unit polarization vector @xmath20 and constant relative phase @xmath21
. the diagonal terms in ( interact ) describe the interaction due to the mean dipole moments and are crucial for effective multiphoton coupling .
we consider schrdinger equation @xmath22with hamiltonian ( [ ham ] ) at the resonance @xmath23 for efficient multiphoton coupling .
the resonance detunings are given by relations @xmath24for @xmath25 pair of photon numbers . here and below ,
unless otherwise stated , atomic units ( @xmath26 ) are employed .
our method of solving the schrdinger equation with hamiltonian ( ham ) has been described in detail in @xcite , and will be excluded here .
the time - dependent wave function can be expressed as @xmath27@xmath28@xmath29where @xmath30 are the time - averaged probability amplitudes and @xmath31 are rapidly changing functions on the scale of waves periods . depending on
the ratio of frequencies @xmath32 the resonant condition ( [ detuning ] ) can hold for a single pair of photon numbers -normal resonance or can also be satisfied by diverse pairs of photons numbers ( in principle infinity ) -degenerate resonance .
let us first consider the case of normal resonance .
hence , if resonant condition holds for a pair ( @xmath33 ) then assuming the smooth turn - on of the pump waves , the relation between rapidly oscillating and slow oscillating parts of the probability amplitudes can be written@xmath34@xmath35@xmath36@xmath37where @xmath38@xmath39and@xmath40@xmath41 in deriving these equations we have applied well - known jacobi
anger expansion via bessel functions:@xmath42 now let us proceed to the case of degenerate resonance .
particularly if @xmath43 , where @xmath44 is an integer number , then there are many channels of resonance transitions and one should take into account all possible transitions and the relation between rapidly oscillating and slow oscillating parts of the probability amplitudes can be written @xmath45 @xmath46 \frac{s_{2}+n}{s_{2}}e^{is_{1}\varphi } e^{is_{2}\omega _ { 2}t } , \label{srpd1}\ ] ] @xmath47 @xmath48 where @xmath49 is given by resonance condition@xmath50 in the schrdinger picture the coherent part of the dipole spectrum is expressed as follows @xcite:@xmath51where@xmath52is the time - dependent expectation value of dipole operator . with the help of wave function ( [ wf ] ) the expectation value of the dipole operator ( dev1 ) can be written as @xmath53 @xmath54 @xmath55 combining the solution for slow oscillating parts of the probability amplitudes with ( [ z1 ] ) , ( [ z2 ] ) and ( [ dev2 ] ) one can calculate analytically the expectation value of the dipole operator for an arbitrary initial atomic state .
the fourier transform of @xmath56 gives the coherent part of the dipole spectrum .
the solution for slow oscillating parts of the probability amplitudes analytically is very complicated and in order to reveal the physics of multiphoton resonant excitation process let us consider systems with inversion symmetry : @xmath57 and @xmath58 . as is seen from ( [ dev2 ] ) , for effective harmonic generation in certain resonance conditions one should provide considerable population transfer between the atomic states @xmath14 and upper levels @xmath3 and @xmath4 .
dynamic stark shifts can take the states off resonance , so appropriate detunings for compensation are chosen .
let us first consider the case of normal resonance . for the system initially situated in the ground state ,
the solution for slow oscillating parts of the probability amplitudes is @xcite : @xmath59@xmath60@xmath61where @xmath62 describes dynamic stark shifts@xmath63@xmath64and @xmath65 expresses the frequency of rabi oscillations @xmath66 replacing the probability amplitudes in ( [ dev2 ] ) by the corresponding expressions ( [ avg1])-([avg3 ] ) and taking into account the relations ( [ rpd1])-([rpd3 ] ) one can derive the final analytical expression for @xmath67 . as is seen from eq .
( [ stark ] ) , the dynamic stark shift is proportional to the ratio @xmath68 , while the multiphoton coupling is proportional to @xmath69 .
since large dynamic stark shifts are detrimental for maintenance of considerable population transfer , here we consider systems with @xmath70 .
taking into account the smallness of the parameter @xmath71 , from ( [ dev2 ] ) at the first approximation we derive the following compact analytic formula:@xmath72where@xmath73@xmath74 as we can see from ( [ main ] ) , ( [ amp ] ) , the spectrum consists of doublets @xmath75 . at
that only harmonics with odd sum of harmonic numbers @xmath76 are existed , as it was expected because of inversion symmetry of the considered problem .
now let us consider the degenerate resonance .
particularly if @xmath77 , where @xmath44 is odd , the dipole expectation value can be obtained from eqs .
( [ srpd1])-([srpd3 ] ) , ( [ dev2 ] ) and is given as : @xmath78@xmath79where rabi frequenc is given as:@xmath80 the spectrum is noticeably different for even @xmath44 . in particular
, for bichromatic field with frequencies @xmath81 and @xmath82 dipole spectrum contains also even harmonics of the main frequency , tunable low - frequency ( much smaller than @xmath81 ) part and six hyper - raman lines per harmonic .
dipole moment expectation value is expressed by very bulky formula and will not be given here .
in this section we present numerical calculations for hydrogen atom and homonuclear diatomic molecular ion @xmath83 @xcite with the specific parameters of available laboratory lasers .
the numerical results will be compared with exact results of the dipole spectrum to estimate accuracy and applicability of generalized rotating wave approximation . the time - dependent schrdinger equation for three - level model with hamiltonian ( [ ham ] )
are considered .
the solution for the probability amplitudes has been obtained using runge - kutta algorithm and scattering spectrum is estimated by applying the fast fourier transform method @xcite . for smooth turn - on of the laser fields ,
we consider envelopes with hyperbolic tangent @xmath84 temporal shape , where @xmath85 characterizes the turn - on time and chosen to be @xmath86 .
it is assumed that quantum systems are initially in the ground state @xmath5 . for both systems the main transitions fall in the vacuum ultraviolet range .
thus , @xmath87 @xmath88 . and @xmath89 @xmath90 .
for hydrogen atom and ion @xmath83 , respectively .
this is a spectral domain where strong coherent radiation is difficult to generate and two or higher photon multicolor resonant excitation is of interest . at four - photon two - color resonance ( @xmath91 ) of hydrogen atom with @xmath92 @xmath90@xmath93 @xmath90 , and @xmath94
@xmath95 the solid ( red ) line corresponds to numerical calculations ; the dashed ( green ) line corresponds to the approximate solution ( for better visibility the latter has been slightly shifted to the right ) . ] at five - photon two - color resonance ( @xmath96 ) of hydrogen atom with @xmath92 @xmath90@xmath97 @xmath90 , and @xmath98 @xmath95 the solid ( red ) line corresponds to numerical calculations ; the dashed ( green ) line corresponds to the approximate solution . ]
figure 2 displays the multicolor harmonic and wave - mixing emission rate ( coherent part ) as a function of the ratio @xmath99 ( we assume @xmath100 and @xmath101 ) at the four - photon two - color resonant excitation of hydrogen atom with xef excimer ( 351 nm , @xmath102 ) and ti : sapphire ( 780 nm , @xmath103 ) laser systems with @xmath92 @xmath104 .. for the hydrogen atom @xmath105 @xmath90 . and
@xmath106 @xmath88 . here and below ,
for the chosen parameters the dynamic stark shift is compensated .
the latter provides almost complete population transfer .
the solid ( red ) line corresponds to numerical calculations , while the dashed ( green ) line corresponds to the approximate expression ( [ main ] ) .
note that the numerical and analytical calculations coincide with high accuracy , so for visual convenience to distinguish these curves the spectrum corresponding to analytical calculations ( [ main ] ) has been slightly shifted to the right for figs . 2 - 5 .
figure 3 displays the multicolor harmonic and wave - mixing emission rate at the five - photon two - color resonant excitation of hydrogen atom with ar+ ( 488 nm , @xmath102 ) and ti : sapphire ( 724 nm , @xmath107 ) laser systems with @xmath108 @xmath90 . at seven - photon two - color resonance ( @xmath109 ) of n@xmath110 molecular ion with @xmath111 @xmath90@xmath112 @xmath90 , and @xmath113
@xmath114 the solid ( red ) line corresponds to numerical calculations ; the dashed ( green ) line corresponds to the approximate solution . ]
@xmath115@xmath116 , and @xmath117 @xmath118 in fig .
4 we plot harmonic / wave - mixing emission rate at the seven - photon two - color resonant excitation of n@xmath119 molecular ion with xef excimer ( 351 nm , @xmath120 ) and ti : sapphire ( 820 nm , @xmath107 ) laser systems with @xmath121 @xmath90 . for this system
we take @xmath122 @xmath88 . and
@xmath106 @xmath90 .. for a degenerate case of resonance in fig .
5 we plot harmonic / wave - mixing emission rate at @xmath123 with @xmath124 @xmath88 .
( 849 nm ) at waves electric fields
@xmath92 @xmath104 .. for the case @xmath125 the radiation spectrum is similar to non - degenerate case . however , there are a few differences .
scattering spectrum is richer in satellites of multicolor harmonics , in addition there is low - frequency radiation on rabi frequencies . in fig .
6 we plot low - frequency part of radiation spectrum .
here , the presented triplet which can be tuned by laser parameters lies in thz / ir region .
@xmath90@xmath126 , and @xmath127 @xmath128 ] as is seen from these figures in the coherent spectrum there are as harmonics of the individual waves as well as frequencies with sum / difference components and its harmonics , in accordance with the analytical ansatz ( main_d ) . from eq .
( [ amp ] ) it is clear that for effective harmonic generation one should provide large dipole interaction energy @xmath129 , since the bessel function exponentially decreases with increasing index at the given argument .
the bessel function @xmath130 at large argument values reaches its maximum at @xmath131 .
thus , the cutoff frequency depends linearly on the amplitudes of the laser fields .
let us make some estimations for reasonable interaction parameters .
the average number of photons at the frequency @xmath81 emitted at each lasers shots of duration @xmath85 , rayleigh length @xmath132 on the atomic / molecular ensemble of density @xmath133 can be estimated as @xcite : @xmath134the incident pulse duration is assumed to be @xmath135 @xmath136 , the rayleigh length is taken to be @xmath137 , and for emitters density we assume @xmath138@xmath139 .
for the setup of figs . 3 and 4 with the chosen parameters , the average number of radiated photons at frequencies up to @xmath140 per shot is @xmath141 , which is two orders of magnitude larger than what one expects to achieve with tunneling harmonics generation huillier
we have presented a theoretical treatment of the multicolor harmonics generation and wave - mixing in a three - level atomic - molecular system under two - color multiphoton resonant excitation .
the coherent part of the dipole spectrum was investigated . with the help of an approximate analytical expression for the dynamic wave function of a three - level atom driven by intense laser fields
, we obtained an analytical expression for the time - dependent expectation value of the dipole operator .
then the results obtained were applied to the hydrogen atom and homonuclear diatomic molecular ion .
the spectrum shows harmonics of the individual waves as well as frequencies with sum / difference components and its harmonics .
these peaks have quite large amplitudes .
the latter is the result of multiphoton resonant interaction of the system with the driving bichromatic laser radiation due to the mean dipole moment in the stationary states .
the cutoff frequency depends linearly on the amplitudes of laser fields .
the presence of the second laser field can make easier the implementation of efficient population transfer and harmonic generation as well as allows generation of new frequencies .
analytical calculations in the generalized rotating wave approximation @xcite allow an explanation of the obtained spectrum .
the numerical simulations are in good agreement with the analytical results .
the considered scheme may serve as a promising method for efficient production of multicolor high harmonics .
it should be noted that the obtained results can be applied to other systems in diverse domains of the electromagnetic spectrum .
a. mcpherson , g. gibson , h. jara , u. johann , t. s. luk , i. a. mcintyre , k. boyer , and c. k. rhodes , j. opt .
b * 4 * , 595 ( 1987 ) ; .
m. ferray , a. lhuillier , x. f. li , l. a. lompre , g. mainfray and c. manus , j. phys .
b * * 21 * * , l31 ( 1988 ) .
m. protopapas , c. h. keitel , and p. l. knight , rep .
* 60 * , 389 ( 1997 ) ; p. salires a. lhuillier , p. antoine , and m. lewenstein , adv . at .
, opt . phys . * 41 * , 83 ( 1999 ) ; t. brabec and f. krausz , rev .
phys . * 72 * , 545 ( 2000 ) ; p. agostini and l. f. dimauro , rep .
. phys . * 67 * , 813 ( 2004 ) ; m. c. kohler , t. pfeifer , k. z. hatsagortsyan , and c. h. keitel , adv . at .
* 61 * , 159 ( 2012 ) . z.
h. kafafi , j. r. lindle , r. g. s. pong , f. j. bartoli , l. j. lingg , and j. milliken , chem .
. lett . * 188 * , 492 ( 1992 ) ; p. haljan , t. fortier , p. hawrylak , p. b. corkum , and m. yu .
ivanov , laser phys .
* 13 * , 452 ( 2003 ) ; g. p. zhang , phys .
lett . * 95 * , 047401 ( 2005 ) ; d. golde , t. meier , and s. w. koch , phys .
b * 77 * , 075330 ( 2008 ) .
p. b. corkum , phys .
lett . * 71 * , 1994 ( 1993 ) ; k. j. schafer , b. yang , l. f. dimauro , and k. c. kulander , phys .
lett . * 70 * , 1599 ( 1993 ) ; m. lewenstein , ph .
balcou , m. yu .
ivanov , anne lhuillier , and p. b. corkum , phys .
rev . a * 49 * , 2117 ( 1994 ) ; v. averbukh , o. e. alon , and n. moiseyev , phys .
a * 64 * , 033411 ( 2001 ) ; c. figueira de morisson faria and i. rotter , phys . rev .
a * 66 * , 013402 ( 2002 ) . | we consider harmonics generation and wave - mixing by two - color multiphoton resonant excitation of three - level atoms / molecules in strong laser fields .
the coherent part of the spectra corresponding to multicolor harmonics generation is investigated .
the obtained analytical results on the basis of generalized rotating wave approximation are in a good agreement with numerical calculations .
the results applied to the hydrogen atom and homonuclear diatomic molecular ion show that one can achieve efficient generation of moderately high multicolor harmonics via multiphoton resonant excitation by appropriate laser pulses . |
the grnwald formula approximation and the @xmath5 approximation of the caputo derivative have been regularly used for numerical solution of fractional differential equations @xcite .
the grnwald formula approximation has weights @xmath6 and accuracy @xmath7 .
the @xmath5 approximation has order @xmath4 and weights @xmath8 .
the weights of the @xmath5 approximation are linear combinations of terms which have power @xmath9 . in the present paper
we construct approximations of the caputo derivative whose weights consist of terms which have power @xmath10 and @xmath11 .
the accuracy of the numerical solution of order @xmath4 is influenced by the coefficient of the term @xmath12 in the expansion of the approximation . in table 1 ,
table 3 and table 9 we compute the error and the order of the numerical solutions of the fractional relaxation equation which use approximations , and of the caputo derivative . in all experiments
the accuracy of the numerical solutions using approximation is higher than the accuracy of the numerical solutions using approximations and .
the caputo derivative of order @xmath13 , where @xmath14 is defined as @xmath15 when the first derivative of the function @xmath16 is a bounded function the caputo derivative at the initial point @xmath17 is equal to zero .
the exponential function has caputo derivative @xmath18 and @xmath19 where @xmath20 is the mittag - leffler function ( @xmath21 ) @xmath22 the analytical solutions of ordinary and partial differential equations are often expressed with the mittag - leffler function .
the ordinary fractional differential equation @xmath23 has exact solution @xmath24 the finite - difference schemes for numerical solution of ordinary and partial fractional differential equations use an approximation for the fractional derivative .
let @xmath25 and @xmath26 be the value of the function @xmath16 at the point @xmath27 .
the @xmath5 approximation of the caputo derivative is constructed by dividing the interval @xmath28 $ ] to subintervals of equal length @xmath29 and approximating the first derivative on each subinterval using a second - order central difference approximation .
@xmath30 where @xmath31 , @xmath32 and @xmath33 approximation has accuracy @xmath34 when @xmath35 $ ] ( @xcite ) .
the numbers @xmath36 have properties @xmath37 where @xmath38 and @xmath39 .
.error and order of numerical solution @xmath40 $ ] of equation i with @xmath41 , equation ii with @xmath42 and equation iii with @xmath43 . [ cols="<,^,^,^,^,^,^ " , ] from , approximation has asymptotic expansion @xmath45 in order to construct approximations for the caputo derivative we express expansion formula in the following form @xmath46 in the previous section we showed that @xmath47-s_n[{\alpha}]+{\dfrac{n^{1-{\alpha}}}{{\alpha}(1-{\alpha})}}.\ ] ] in this section we obtain the formula for the coefficient @xmath48 and an approximation for the caputo derivative of order @xmath1 .
by changing the order of summation in @xmath49 we have that @xmath50 . then @xmath51 when @xmath52 we obtain @xmath53 from the identity @xmath54 we obtain @xmath55 @xmath56 @xmath57}-n{\left [ } { \dfrac{n^{1-{\alpha}}}{1-{\alpha } } } \sum_{m=1}^{\infty}\binom{1-{\alpha}}{2m}{\dfrac{b_{2m}}{n^{2m}}}{\right ] }
\\ & + { \dfrac{1}{2}}{\left [ } { \dfrac{n^{2-{\alpha}}}{2-{\alpha } } } \sum_{m=1}^{\infty}\binom{2-{\alpha}}{2m}{\dfrac{b_{2m}}{n^{2m}}}{\right ] } .\end{aligned}\ ] ] from @xmath58+{\dfrac{n^{-{\alpha}}}{{\alpha}}}+{\dfrac{1}{2n^{1+{\alpha}}}},\ ] ] @xmath59-{\dfrac{n^{2-{\alpha}}}{2-{\alpha}}}+{\dfrac{1}{2n^{{\alpha}-1}}}.\ ] ] then @xmath60+{\dfrac{n^{-{\alpha}}}{{\alpha}}}+{\dfrac{1}{2n^{1+{\alpha}}}}{\right ] } -n{\left [ } s_n[{\alpha}]-{\dfrac{n^{1-{\alpha}}}{1-{\alpha}}}+{\dfrac{1}{2n^{\alpha}}}{\right ] } \\ & + { \dfrac{1}{2}}{\left [ } s_n[{\alpha}-1]-{\dfrac{n^{2-{\alpha}}}{2-{\alpha}}}+{\dfrac{1}{2n^{{\alpha}-1}}}{\right ] } , \end{aligned}\ ] ] @xmath61-n s_n[{\alpha}]+{\dfrac{1}{2}}s_n[{\alpha}-1]+{\dfrac{n^{2-{\alpha}}}{{\alpha}({\alpha}-1)({\alpha}-2)}}.\end{aligned}\ ] ] approximation has expansion of order @xmath1 @xmath62 by approximating @xmath63 in using the approximations @xmath64 @xmath65 we obtain the approximation for the caputo derivative of order @xmath1 @xmath66 where @xmath67 @xmath68 @xmath69 @xmath70 when @xmath71 , approximation has weights @xmath72 when @xmath73 : @xmath74 @xmath75 @xmath76 @xmath77 in the previous sections computed the numerical solutions of the fractional relaxation equation of order @xmath78 and @xmath79 using the second - order approximation for the value @xmath80 . in order to obtain a third - order approximation for @xmath80 we need to compute the values of @xmath81 and @xmath82 .
by applying fractional differentiation of order @xmath9 to equation we obtain @xmath83 then @xmath84 . in @xcite
we showed that @xmath85 by differentiating we obtain @xmath86 @xmath87 when the solution of equation has a bounded second derivative , the value of @xmath82 is computed with the formula @xmath88 now we compute the values of the derivatives @xmath89 of the exact solutions of equation i , equation ii and equation iii . @xmath90 @xmath91 then @xmath92 .
@xmath93 then @xmath94 the exact solution of equation i satisfies @xmath95 and @xmath96 .
@xmath97 @xmath98 then @xmath99 .
@xmath100 @xmath101 the exact solution of equation ii satisfies @xmath102 .
@xmath103 @xmath104 @xmath105 then @xmath106 . @xmath107 @xmath108 the exact solution of equation iii satisfies @xmath109 .
let @xmath110 from taylor s formula @xmath111 is an approximation for @xmath80 with accuracy @xmath112 .
the accuracy of numerical solution @xmath113 $ ] with initial values @xmath114 is @xmath115 . in table 10
we compute the maximum error and the order of numerical solution @xmath113 $ ] of equation i , equation ii and equation iii .
in figure 3 we compare numerical solutions @xmath116,ns[40]$ ] and @xmath113 $ ] of equation iii and @xmath117 .
$ ] -red , and @xmath118$]-blue , and @xmath116$]-green for @xmath119 and @xmath120.,scaledwidth=55.0% ]
in the present paper we obtained approximations of the caputo derivative of order @xmath121 and @xmath1 whose weights consist of terms which have power @xmath10 and @xmath11 . in all experiments
the accuracy of the numerical solutions which use approximation for the caputo derivative is higher than the accuracy of the numerical solutions using approximations and . a question for future work is to construct an approximation of order @xmath4 whose weights consist of terms which have power @xmath10 , where the accuracy of the numerical solutions of equation i , equation ii and equation iii is higher than the accuracy of the numerical solutions using the @xmath5 approximation . 99 m. abramowitz , i. a. stegun : _ handbook of mathematical functions with formulas , graphs , and mathematical tables , _ dover , new york , 1964 .
d. baleanu , k. diethelm , e. scalas and j. j. trujillo : _ fractional calculus : models and numerical methods , _ world scientific , ( 2012 ) .
k. diethelm : _ the analysis of fractional differential equations : an application - oriented exposition using differential operators of caputo type , _ springer ; 2010 .
y. dimitrov : _ numerical approximations for fractional differential equations , _ journal of fractional calculus and applications , * 5(3s ) * ( 2014 ) , no .
22 , 145 .
y. dimitrov : _ a second order approximation for the caputo fractional derivative , _ * arxiv:1502.00719 * ( 2015 ) .
y. dimitrov : _ higher - order numerical solutions of the fractional relaxation - oscillation equation using fractional integration , _ * arxiv:1603.08733 * ( 2016 ) .
h. ding , c. li : _ high - order numerical algorithms for riesz derivatives via constructing new generating functions , _ * arxiv:1505.03335 * , ( 2015 ) .
h. ding , c. li : _ high - order algorithms for riesz derivative and their applications ( iii ) _ , fract .
, * 19(1 ) * , ( 2016 ) , 1955 . g. h. gao and z. z. sun , h. w. zhang : _ a new fractional numerical differ - entiation formula to approximate the caputo fractional derivative and its applications , _ j. comput .
* 259 * ( 2014 ) , 3350 .
r. gorenflo , f. mainardi : _ fractional calculus : integral and differential equations of fractional order _ , cism lecture notes ( 2008 ) . m. glsu , y. ztrk and a. anapali : _ numerical approach for solving fractional relaxation - oscillation equation , _ applied mathematical modelling , * 37(8 ) * ( 2013 ) , 59275937 . m. ishteva : _ properties and applications of the caputo fractional operator , _
master thesis , university of karlsurhe ( 2005 ) .
b. jin , r. lazarov and z. zhou : _ an analysis of the l1 scheme for the subdiffusion equation with nonsmooth data , _
i m a j. numer .
, * 36 ( 1 ) * ( 2016 ) , 197221 . c. li , r. wu , h. ding : _ high - order approximation to caputo derivatives and caputo - type advection - diffusion equations , _ communications in applied and industrial mathematics , * 6(2 ) * ( 2014 ) , e-536 . c. li , f. zeng : _ numerical methods for fractional calculus , _ chapman and hall / crc ( 2015 ) .
y. lin , c. xu : _ finite difference / spectral approximations for the time - fractional diffusion equation , _ journal of computational physics , * 225(2 ) * ( 2007 ) , 15331552 .
c. lubich : _ discretized fractional calculus _ , siam j. math .
, * 17 * ( 1986 ) , 704719 .
y. ma : _ two implicit finite difference methods for time fractional diffusion equation with source term , _ journal of applied mathematics and bioinformatics , * 4(2 ) * ( 2014 ) , 125145 .
f , mainardi : _ fractional relaxation - oscillation and fractional diffusion - wave phenomena , _ chaos , solitons @xmath122 fractals , * 7(9 ) * ( 1996 ) , 14611477 .
f. mainardi , r. gorenflo : _ time - fractional derivatives in relaxation processes : a tutorial survey _ , fractional calculus and applied analysis , * 10(3 ) * ( 2007 ) , 269308 .
d. a. murio , implicit finite difference approximation for time fractional diffusion equations , computers & mathematics with applications , * 56(4 ) * ( 2008 ) , 11381145 .
z. odibat : _ approximations of fractional integrals and caputo fractional derivatives , _ applied mathematics and computation , * 178(2 ) * ( 2006 ) , 527533 .
i. podlubny : _ fractional differential equations , _ academic press , san diego ( 1999 )
. h. m. srivastava , j. choi : _ series associated with the zeta and related functions _ , kluwer academic publishers ( 2001 ) . c. tadjeran , m. m. meerschaert and h. p. scheffer : _ a second - order accurate numerical approximation for the fractional diffusion equation , _ journal of computational physics , * 213 * ( 2006 ) , 205213 .
tian , h. zhou and w. deng : _ a class of second order difference approximation for solving space fractional diffusion equations , _ mathematics of computation , * 84(294 ) * ( 2012 ) , 17031727 .
r. wu , h. ding and c. li : _ determination of coefficients of high - order schemes for riemann - liouville derivative , _ the scientific world journal ( 2014 ) , article i d 402373
. y. yan , k. pal and n. j. ford : _ higher order numerical methods for solving fractional differential equations , _ bit numerical mathematics , * 54(2 ) * ( 2014 ) , 555584 . | in this paper we construct approximations for the caputo derivative of order @xmath0 and @xmath1 .
the approximations have weights @xmath2 and @xmath3 , and the higher accuracy is achieved by modifying the initial and last weights using the expansion formulas for the left and right endpoints .
the approximations are applied for computing the numerical solution of ordinary fractional differential equations .
the properties of the weights of the approximations of order @xmath4 are similar to the properties of the @xmath5 approximation . in all experiments presented in the paper the accuracy of the numerical solutions using the approximation of order @xmath4 which has weights
@xmath3 is higher than the accuracy of the numerical solutions using the @xmath5 approximation for the caputo derivative . *
2010 math subject classification : * 26a33 , 34a08 , 34e05 , 41a25 + * key words and phrases : * fractional derivative , approximation , fourier transform , fractional differential equation . |
asteroids are rotating aspherical reflective bodies which thus exhibit brightness variations . as recognized long ago ( russell 1906 , metcalf 1907 ) , studies of their lightcurves provide important constraints on their physical properties , and processes that affect their evolution .
for example , well - sampled and accurate lightcurves can be used to determine asteroid asphericity , spin vector , and even the albedo inhomogeneity across the surface ( magnusson 1991 ) . the current knowledge about asteroid rotation rates and lightcurve properties is well summarized by pravec & harris ( 2000 ) .
the rotational periods range from @xmath02 hours to @xmath0 15 hours .
the lightcurve amplitudes for main - belt asteroids and near - earth objects are typically of the order 0.1 - 0.2 mag .
( peak - to - peak ) .
recently , similar variations have been detected for a dozen kuiper belt objects ( sheppard & jewitt 2002 ) .
the largest amplitudes of @xmath02 mag .
( peak - to - peak ) are observed for asteroids 1865 cerberus and 1620 geographos ( wisniewski et al .
1997 , szab et al 2001 ) .
in contrast to appreciable and easily detectable amplitudes of single - band light curves , typical asteroid color variations are much smaller .
indeed , if albedo did nt vary across an asteroid s surface , then the asteroid would not display color variability irrespective of its geometry while the absence of color variability may also be consistent with a gray albedo variation , the strong observed correlation between asteroid albedo and color ( blue c type asteroids have visual albedo of @xmath1 , while for red s type asteroids @xmath2 , zellner 1979 , shoemaker et al .
1979 ) implies that non - uniform albedo distribution should be detectable through color variability .
following magnusson ( 1991 ) , hereafter we will refer to non - uniform albedo distribution across an asteroid surface as to albedo variegation .
the most notable case of albedo variegation is displayed by 4 vesta which apparently has one bright and one dark hemisphere ( blanco & catalano 1979 ; degewij , tedesco & zellner 1979 , binzel et al .
definite color variations have been detected in only a few dozen asteroids .
a color variability at the level of a few percent has been measured directly for eros ( v - r and v - i , wisniewski 1976 ) and for 51 nemausa ( u - b , v - y , kristensen & gammelgaard 1993 ) . in a study that still remains one of the largest monitoring programs for color variability , degewij
, tedesco & zellner ( 1979 ) detected color variations greater than 0.03 mag . in 6 out of 24
monitored asteroids . in another notable study ,
schober & schroll ( 1982 ) detected color modulation in 49 asteroids .
recently , a spectacular confirmation of albedo variegation has been obtained for eros by near multispectral imaging ( murchie et al . 2002 ) . while similar spatially resolved images are available for several other objects ( e.g. zellner et al .
1997 , binzel et al .
1997 , baliunas et al .
2003 ) , the number of asteroids with observational constraints on their albedo variegation remains small .
here we study asteroid color variability by utilizing the sloan digital sky survey moving object catalog ( hereafter sdssmoc , ivezi et al .
sdssmoc currently contains accurate ( 0.02 mag ) 5-band photometric measurements for over 130,000 asteroids .
a fraction of these objects are previously recognized asteroids with available orbits , and 7,531 of them were observed by sdss at least twice .
we use the color differences between the two observations of the same objects to constrain the ensemble properties , as opposed to studying well - sampled light curves for a small number of objects .
the lack of detailed information for individual objects is substituted by the large sample size which allows us to study correlations between color variability and various physical properties in a statistical sense . also , objects in the sample studied here have typical sizes 110 km , about a factor 10 smaller than objects for which color variations have been reported in the literature .
we describe the sdssmoc and data selection in section 2 , and in section 3 we perform various tests to demonstrate that detected color variability of multiply observed objects is not an observational artefact . in section 4
we search for correlations between the color variability and asteroid physical properties , and summarize our results in section 5 .
sdss is a digital photometric and spectroscopic survey which will cover 10,000 deg@xmath3 of the celestial sphere in the north galactic cap and a smaller ( @xmath0 225 deg@xmath3 ) and deeper survey in the southern galactic hemisphere ( azebajian et al .
2003 , and references therein ) .
the survey sky coverage will result in photometric measurements for about 50 million stars and a similar number of galaxies . about 50% of the survey
is currently finished .
the flux densities of detected objects are measured almost simultaneously in five bands ( fukugita et al . 1996 ; @xmath4 , @xmath5 , @xmath6 , @xmath7 , and @xmath8 ) with effective wavelengths of 3551 , 4686 , 6166 , 7480 , and 8932 , 95% complete for point sources to limiting magnitudes of 22.0 , 22.2 , 22.2 , 21.3 , and 20.5 in the north galactic cap .
astrometric positions are accurate to about 0.1 arcsec per coordinate ( rms ) for sources brighter than 20.5@xmath9 ( pier et al .
2002 ) , and the morphological information from the images allows robust star - galaxy separation ( lupton et al .
2001 ) to @xmath0 21.5@xmath9 .
sdss , although primarily designed for observations of extragalactic objects , is significantly contributing to studies of the solar system objects , because asteroids in the imaging survey must be explicitly detected to avoid contamination of the samples of extragalactic objects selected for spectroscopy .
preliminary analysis of sdss commissioning data ( ivezi et al .
2001 , hereafter i01 ) showed that sdss will increase the number of asteroids with accurate five - color photometry by more than two orders of magnitude , and to a limit about five magnitudes fainter ( seven magnitudes when the completeness limits are compared ) than previous multi - color surveys ( e.g. the eight color asteroid survey , zellner , tholen & tedesco 1985 ) .
sdss moving object catalog is a public , value - added catalog of sdss asteroid observations .
in addition to providing sdss astrometric and photometric measurements , all observations are matched to known objects listed in the astorb file ( bowell 2001 ) , and to the database of proper orbital elements ( milani et al .
1999 ) , as described in detail by juri et al .
( 2002 , hereafter j02 ) .
multiple sdss observations of objects with known orbital parameters can be accurately linked , and thus sdssmoc contains rich information about asteroid color variability .
we select 7,531 multiply observed objects from the second sdssmoc edition ( adr2.dat ) by requiring that the number of observations ( @xmath10 ) is at least two , and use the first and second observations to compute photometric changes .
a `` high - quality '' sample of 2,289 asteroids is defined by two additional restrictions : * in order to avoid the increased photometric errors at the faint end , we require @xmath11 . * in order to minimize the effect of variable angle from the opposition on the observed color , we select only objects for which the change of this angle is less than 1.5 degree . we also utilize a subsample of 541 asteroids that were observed at least four times . since sdss observations of asteroids are essentially random , and the time between them ( days to months , 75% of repeated observations are obtained within 3 months , and 86% within a year ) is much longer than typical rotational periods ( @xmath12 1 day ) , the phases of any two repeated observations are practically uncorrelated .
thus , the distribution of magnitude and color changes for a large ensemble of asteroids is a good proxy for random two - epoch sampling of an asteroid s single - band and color lightcurves . of course , this is strictly true only if the distributions of amplitudes and shapes of these light curves are fairly narrow . for wide distributions of amplitudes and shapes , the observed two - epoch magnitude and
color changes represent convolution of the two effects .
[ fig1 ] color change between two epochs for all 7,531 asteroids from sdssmoc that were observed at least twice .
the symbols with error bars connected by the solid line show the distribution of the @xmath13 color change for a subset of 2,289 asteroids brighter than @xmath14 and with the difference in angles from the opposition smaller than 1.5 degree .
its equivalent gaussian width , determined from the interquartile range , is 0.053 mag .
the dash - dotted line shows the distribution of the @xmath13 color change between two epochs for 21,000 stars brighter than @xmath14 .
its width , which indicates the measurement error for @xmath13 color , is 0.023 mag . ]
the colors of asteroids in sdss photometric system are discussed in detail by i01 .
they defined a principal color in the @xmath15 vs. @xmath16 color - color diagram , @xmath13 , as @xmath17 the @xmath13 color distribution is strongly bimodal ( see fig . 9 in i01 ) , with the two modes at @xmath18 and @xmath19 .
the rms scatter around each mode is about 0.05 mag .
the two modes are associated with different taxonomic classes : the `` blue '' mode includes c , e , m and p types , and the `` red '' mode includes s , d , a , v and j types ( see fig . 10 in i01 ) .
the vesta type asteroids ( type v ) can be effectively separated from `` red '' asteroids using the @xmath20 color ( j02 ) .
the @xmath13 and @xmath20 colors are strongly correlated with dynamical family membership ( ivezi et al .
2002c , hereafter i02c ) .
hereafter , we chose the @xmath13 color as the primary quantity to study asteroid color variability .
[ fig2 ] color change between two epochs for the same asteroids as in in fig .
1 , as a function of the object s velocity . the bottom panel compares the width of the distribution of the @xmath13 color change for 511 objects with @xmath21 ( dashed line ) and for 1,065 objects with @xmath22 deg / day ( solid line ) .
note that the two histograms are statistically indistinguishable indicating that the color measurement is not affected by the object s apparent velocity . ]
[ fig3 ] color change between two epochs for the same asteroids as in in fig .
1 , as a function of the object s apparent magnitude . the bottom panel compares the width of the distribution of the @xmath13 color change for 214 objects with @xmath23 ( dashed line ) and for 898 objects with @xmath24 ( solid line ) .
note that the two histograms are statistically indistinguishable indicating that the color measurement is not affected by the object s apparent magnitude . ]
[ fig4 ] color change between two epochs for the same asteroids as in in fig .
1 , as a function of the change in the @xmath6 band apparent magnitude the bottom panel compares the width of the distribution of the @xmath13 color change for 645 objects with @xmath25 ( dashed line ) and for 1,644 objects with @xmath26 ( solid line ) . note that the two histograms are statistically indistinguishable indicating that the color measurement is not correlated with the change of the object s apparent magnitude . ]
[ fig5 ] color change between two epochs for the same asteroids as in in fig .
1 , as a function of the mean angle from the opposition . the bottom panel compares the width of the distribution of the @xmath13 color change for 712 objects with @xmath27 ( dashed line ) and for 664 objects with @xmath28 ( solid line ) .
note that the two histograms are statistically indistinguishable indicating that the color measurement is not correlated with the mean angle from the opposition . ]
[ fig6 ] magnitude for the full sample , as a function of the change of the angle from the opposition .
the bottom panel compares the @xmath29 distributions for two ranges of the angle change : @xmath30 as the solid line and @xmath31 as the dashed line .
the former is well fit by a sum of two gaussians ( see text ) shown as the dash - dotted line . ] the observed distribution of the @xmath13 color change between two epochs , @xmath32 , for 2,289 selected asteroids , is shown by symbols ( with error bars ) in figure 1 .
its equivalent gaussian width determined from the interquartile range ( hereafter `` width '' ) , is 0.053 mag .
the expected width based on the formal errors reported by the sdss photometric pipeline ( `` photo '' , lupton et al .
2001 ) is 0.02 mag . , indicating that the observed @xmath32 distribution reflects intrinsic asteroid color changes .
however , the formal errors may not be correct . in order to determine the measurement accuracy for the @xmath13 color
, we use 21,000 stars with @xmath11 that were observed twice . the dashed line in figure 1 .
shows their @xmath32 distribution .
its width is 0.023 mag .
, in agreement with the expectations based on formal errors . subtracting the error distribution width of 0.023 mag . in quadrature ,
the intrinsic @xmath13 color root - mean - square ( rms ) variation is 0.04 mag . while fig.1 demonstrates that this measurement is statistically highly significant , in the remainder of this section we test for the presence of spurious observational effects that could be responsible for the observed asteroid color variation .
although the formal photometric errors for stars are correct , asteroids move during observations and their motion could in principle affect the photometric accuracy . while this effect should be negligible ( the images are not strongly trailed ) , we test for it by correlating @xmath32 with the object s velocity . if the photometric accuracy is lower for moving objects , the @xmath32 distribution width should increase with magnitude of the object s velocity .
the @xmath32 vs. @xmath33 diagram is shown in the top panel in figure 2 .
the bottom panel compares the width of the @xmath32 distribution for two subsamples selected by velocity .
the dashed line show the @xmath32 distribution for 511 objects with 0.1 deg / day @xmath34 0.18 deg / day , and the solid line for 1,065 ones with @xmath350.22 deg / day . as evident , the two histograms are statistically indistinguishable
. the photometric errors typically increase with apparent magnitude of the measured object . while the adopted faint limit ( @xmath36 ) is sufficiently bright that sdss photometric errors are nearly independent of magnitude ( ivezi et al .
2002c ) , we test this expectation by correlating @xmath32 with the mean apparent magnitude in fig .
the bottom panel compares the width of the @xmath32 distribution for two subsamples selected by @xmath23 ( dashed line , 214 objects ) and by @xmath24 ( solid line , 898 objects ) .
there is no significant difference between the two histograms . relaxing the magnitude cut to @xmath37 yields the same null result with an increased sample size by a factor of 2 .
we have also tested @xmath32 vs. @xmath29 dependence , illustrated in fig .
4 , and did not detect any correlation .
in order to minimize the opposition effect ( see section 6.1 in i01 ) , the maximum change of the angle from the opposition ( between two observations ) in the analyzed sample is constrained to 1.5 deg ( this limit may be slightly too conservative since we find no correlation between @xmath32 and the _ change _ of the angle from the opposition for changes as large as 10@xmath38 ) . to exclude the possibility that the color change is affected by the _
mean _ angle from the opposition , we analyze @xmath32 as a function of this angle in fig .
5 . there is no discernible correlation . unlike the color change which is not correlated with the angle from the opposition
, the change of magnitudes between two epochs is strongly correlated with this angle , as expected .
the top panel in fig .
6 shows the dependence of @xmath29 on the change of the angle from the opposition .
the comparison of @xmath29 distributions for two ranges of the angle change ( @xmath30 as the solid line and @xmath31 as the dashed line ) are shown in the bottom panel . as evident , even when asteroids are observed at practically the same position , @xmath29 has a wide distribution , mostly due to asteroid rotation .
we find that the @xmath29 distribution can be well fit by a sum of two gaussians with the widths of 0.08 mag and 0.35 mag and amplitude ratio 2:1 , shown as the dash - dotted line .
the measured @xmath29 distribution can be used to constrain the distributions of asteroid axes ratios and rotational periods .
such an analysis , while interesting in its own right , is beyond the scope of this paper .
[ fig7 ] the sdss photometric measurements are obtained in the order r - i - u - z - g , and the elapsed time between the first ( r ) and last ( g ) measurement is @xmath39 minutes . the asteroid brightness variation during this time ,
even if achromatic , introduces a bias in the color measurement .
for example , if an asteroid is observed during the rising part of its lightcurve , the @xmath15 color is biased red , and the @xmath16 color is biased blue . also , due to fixed filter order
, the color biases should be correlated .
the expected correlations are @xmath40 , @xmath41 , @xmath42 , @xmath43 , @xmath44 , @xmath45 . with conservative assumptions that typical rotational period is as short as 2 hours , and that the peak - to - peak amplitude is as large as 05
, the expected rms contribution is the largest for @xmath16 color and equal to @xmath0003 , while for @xmath15 color it is less than 001 .
7 shows the six possible correlations of the four sdss colors .
the rms scatter of each color is also displayed in the figure . as evident , the variations in all four colors are too large by a factor of few to be explained by any plausible rotation parameters . furthermore , the slopes of expected correlations , shown by the dashed lines , are not supported by the data .
we conclude that the non - simultaneous nature of sdss color measurements is not significantly contributing to the observed color variation .
the @xmath46 and @xmath20 color changes show the largest width .
we have verified that imposing a magnitude cut ( @xmath47 ) on the @xmath4 and @xmath8 magnitudes does not decrease the scatter , i.e. the large widths are not caused by the lower sensitivity of these two bands .
figure 7 demonstrates that some color pairs ( e.g. @xmath48 vs. @xmath49 show correlations with a slope which can not be explained by the examined instrumental effects .
we characterize the observed correlations by their linear regression , the @xmath50 of the fit and the linear correlation coefficient @xmath51 .
all the fits cross the origin ( the 0th - order coefficients do not differ significantly from 0 ) .
we determined the following correlations : @xmath52 there are some noteworthy aspects of this test .
first , the change of @xmath46 color does not correlate with the change of @xmath15 and @xmath20 colors , although the scatter of @xmath53 is much higher than for the other colors . also ,
the color indices @xmath54 , where @xmath55 , do not correlate with any other color index that does not include @xmath56 .
this is consistent with a hypothesis that the blue part of the spectrum is affected by a process which causes no color variation of the red colors .
the changes of @xmath15 and @xmath16 colors , which are used to define color @xmath13 , seem to be weakly correlated , with the median slope @xmath57 ( 3 ) .
this implies that , on average , the line connecting two observations in this diagram is more aligned with the second principal axis which is perpendicular to @xmath13 , than with the @xmath13 axis . using the definition of the second principal color , hereafter named @xmath58 , @xmath59 we constructed the principal color diagram shown in the top panel in fig .
note that this diagram is simply a rotated version of the @xmath15 vs. @xmath16 diagram . in this panel
we simply plot the mean value of each principal color , while in the middle panel we connect two individual measurements by lines , for a small subset of objects with @xmath60 ( to avoid crowding ) and a change in each color of at least 0.03 mag . as evident , for objects with large color variations ,
the changes of the two principal colors seem to be somewhat correlated , with the slope larger than 1 .
that is , the @xmath58 color varies more than the @xmath13 color . in the bottom panel
we show the change of @xmath58 color vs. the change of @xmath13 color for all the objects in the `` high - quality '' subsample , as well as the rms scatter in each color .
the observed variability of the @xmath58 color appears nearly sufficient to explain its single - measurement distribution width of @xmath61 mag .
( the rms width of the change in that color is 1.33 of its single - epoch width , i.e. only slightly smaller than the expected value of 1.41 if the variability was the only reason for its finite value ) . in order to test this hypothesis
, we use a subsample of 541 asteroids that were observed at least four times . the distribution width for the mean @xmath58 color , obtained by averaging the four measurements ,
is expected to be smaller than that measured for any individual epoch . on the other hand ,
no significant difference in the distribution widths is expected for the @xmath13 color .
the distributions shown in fig .
9 suggest that the variability contributes significantly to the @xmath58 color distribution width , and much less to the width of each individual mode in the @xmath13 color distribution . since the variability induced motion in the @xmath58 vs. @xmath13 color - color diagram
is more aligned with the @xmath58 axis than with the @xmath13 axis , over 90% of asteroids have the same @xmath13-based classification ( @xmath62 vs. @xmath63 ) in different epochs .
the same behavior also indicates that the color variability can not be explained as due to mixing of the two basic materials , corresponding to c and s types , on the asteroid surfaces .
we will return to this point in section 5 .
[ fig8 ] and a change in each color of at least 0.03 mag .
note that for objects with large color variations , the changes of the two principal color seem to be somewhat correlated .
the bottom panel show the change of @xmath58 color vs. the change of @xmath13 color for all the objects in the `` high - quality '' subsample , as well as the rms scatter in each color .
, title="fig : " ] [ fig9 ] ( top panel ) and @xmath13 ( bottom panel ) color distributions obtained by averaging four measurements ( solid lines ) for 541 asteroids , and those for individual measurements ( dashed lines ) .
the @xmath64 values are the distribution widths , subscripted 1 for single epoch and 4 for average measurements .
the widths for the @xmath13 color distribution are determined separately for objects with @xmath65 ( superscripted b ) and @xmath66 ( r ) .
note that the width of the @xmath58 color distribution is 1.46 larger for single epoch measurements , while the @xmath13 color distribution does not change appreciably . ]
the suite of tests in this section suggest that the observed color variations are not an artefact ( either observational , or caused by phenomena such as differential opposition effect ) .
however the lack of any correlation with physical parameters , such as color , size and family membership , discussed in the next section , may be elegantly explained as caused by some `` hidden '' random error contribution , for example a problem introduced by the processing software .
here we present a test which demonstrates that at least in one aspect the observed color variation is not random .
if it is true that asteroids exhibit a varying degree of color variability , as they do , for example , for single - band variability , then the color changes detected in two independent pairs of observations may correlate to some degree . at the same time
, such a correlation would give credence to the measurements reliability .
we use a subsample of 541 asteroids that were observed at least four times to test whether such a correlation exist .
the top panel in fig .
10 plots the change of @xmath13 color in a pair of observations vs. the change in another independent pair of observations .
we have verified that the marginal distributions in each coordinate are indistinguishable . to illustrate this point , in the bottom panel we compare the distributions of the @xmath13 color change in one pair of observations for two subsamples selected by the @xmath13 color change in the other independent pair observations , as marked by the dashed lines in the top panel . if the color changes in two pairs of observations are uncorrelated , then the two histograms should be indistinguishable . however , they are clearly different , indicating that these independent observations `` know '' about each other ! [ fig10 ] color in two independent pairs of observations ( change in the first pair vs. the change in the second pair ) for 541 asteroids observed at least four times .
the bottom panel compares the histograms of @xmath13 color change in one of the two pairs of observations , for two subsets selected by the @xmath13 color change in the other pair of observations , as marked by the dashed lines in the top panel ( see text ) .
note that the two distributions are different , indicating that these independent observations `` know '' about each other . the same conclusion is obtained when the axes are reversed . ] in order to quantify the statistical significance of the difference between the two histogram , we perform two tests . a two - sample kolmogorov - smirnov test ( lupton 1993 ) indicates that the histograms are different at a confidence level of more than 99% .
another test , proposed by efron and petrosian ( 1992 ) , which uses the entire 2-d sample , indicates that the two variables are correlated at a confidence level of 95%
. the somewhat lower confidence level than for the first test is probably due to the contribution of points with small color changes , whose distribution may be randomized by photometric errors .
we conclude that objects with large color variations in one pair of observations tend to show relatively large color variation in the other , independent , pair of observations .
this difference strongly suggests that the observed color variations are real , and also indicates that for some asteroids color variations are stronger than for others .
a series of tests discussed in the preceding section demonstrate that the detection of asteroid color variability is robust . in this section
we attempt to find correlations between color variability and other parameters such as colors ( a good proxy for taxonomic classes ) , absolute magnitude ( i.e. size ) , and family membership . [ fig11 ] color change between two epochs for the same asteroids as in in fig .
1 , as a function of the mean @xmath13 color . the bottom panel compares the distributions of the @xmath13 color change for 689 objects with mean @xmath65 and 1585 objects with mean @xmath66 .
note that the two histograms are statistically indistinguishable indicating that the color measurement is not correlated with the asteroid s @xmath13 color ( which is a good proxy for taxonomic classification ) . ]
[ fig12 ] magnitude of at least 0.2 mag .
the dots in the bottom panel represent objects with the change in the @xmath13 color of at least 0.05 mag .
note that objects with large changes of magnitudes and colors appear to show the same mean principal color distribution as the full sample . ]
[ fig13 ] color change between two epochs for the same asteroids as in in fig .
1 , as a function of the absolute magnitude . the bottom panel compares the width of the distribution of the @xmath13 color change for 97 objects with @xmath67 ( dashed line ) and for 505 objects with @xmath68 ( solid line ) .
note that the two histograms are statistically indistinguishable indicating that the color measurement is not correlated with the asteroid s absolute magnitude ( i.e. size , in the approximate range 110 km ) . ]
the position of an asteroid in the principal color diagram ( the top panel in fig .
8) is a good proxy for its taxonomic classification ( i01 ) .
thus , a dependence of color variability on taxonomic type would show up as a correlation between the change of @xmath13 color and its mean value .
we show the scatter plot of these two quantities in the upper panel in fig .
the bottom panel compares the distributions of the @xmath13 color change for 689 objects with mean @xmath65 ( dominated by the c type objects ) and 1585 objects with mean @xmath66 ( dominated by the s type objects ) .
the two histograms are statistically indistinguishable .
while the mean - color - selected objects ( blue vs. red ) appear to show the same @xmath13 color change distributions , it may be possible that the objects with the largest @xmath29 or @xmath32 would show different principal color distribution .
we repeat the asteroid principal color diagram constructed with the mean colors ( already shown in fig .
8) in the top panel in fig .
the same distribution is shown by linearly spaced isodensity contours in the middle and bottom panels .
the dots in the middle panel represent objects with the change in the @xmath6 magnitude of at least 0.2 mag .
the dots in the bottom panel represent objects with the change in the @xmath13 color of at least 0.05 mag .
there is no discernible difference between the color distribution for the whole sample and that for the highly variable objects .
asteroids of different size may exhibit different color variability .
the closest proxy for the size , in the absence of direct albedo measurements , is the absolute magnitude .
12 shows correlation between @xmath32 and the absolute magnitude @xmath69 .
the displayed range of @xmath69 roughly corresponds to the 110 km size range .
the bottom panel compares the width of the @xmath32 distribution for 97 objects with @xmath67 ( dashed line ) and for 505 objects with @xmath68 ( solid line ) .
the two histograms are statistically indistinguishable indicating that the color change is not correlated with the asteroid s absolute magnitude ( i.e. size ) .
however , we emphasize that the dynamic range of probed sizes is fairly small .
sdss colors are a good proxy for taxonomic classification , and can be efficiently used to recognize at least three color groups ( j02 ) .
i02c showed by correlating asteroid dynamical families and sdss colors that indeed there are more than just three `` shades '' : many families have distinctive and uniform colors .
motivated by their finding , we obtain a more detailed classification of asteroids using dynamical clustering and correlate it with variability properties .
we define families by 3-dimensional boxes in the space spanned by proper orbital elements ( the boundaries are summarized in table 1 ) , using the results from zappala et al .
there are 6 families with more than 50 members in the sample analyzed here .
the comparison of @xmath70 and @xmath71 distributions for individual families with those for the whole sample are shown in figs .
14 and 15 , respectively . in order to quantitatively assess whether there is any family that differs in its variability properties from the rest of the sample , we performed two - sample kolmogorov - smirnov tests .
none of the families listed in table 1 was found to differ from the mean values for the whole sample at a confidence level greater than 95% .
we conclude that all the examined families show similar variability properties .
[ table1 ] .the definitions of asteroid families in the @xmath72-@xmath73-@xmath74 space [ cols="<,<,<,<",options="header " , ] [ fig14 ] [ fig15 ]
the detection of color variability for a large sample of asteroids discussed here represents a significant new constraint on the physical properties and evolution of these bodies .
the random nature of this variability , implied by the lack of apparent correlations with asteroid properties such as mean colors , absolute magnitude and family membership , could be interpreted as due to some hidden random photometric error .
however , several lines of evidence argue that this explanation is unlikely .
first , the magnitude of the observed effect ( 0.060.11 mag . )
is so large that such photometric errors would have to be noticed in numerous other studies and tests based on sdss photometric data .
if such an error only shows up for moving objects , then the color variability should increase with the apparent velocity , an effect which is not observed ( see fig .
second , independent pairs of observations , discussed in section 3.4 , seem to `` know '' about each other : objects with large color variation in one pair of observations tend to show relatively large color variation in other pair of observations .
this fact can not be explained by random photometric errors .
third , the color variation is not entirely random in the principal colors diagram , as discussed in section 3.3 .
there is a preferred direction for variability - induced motion in this diagram , and the scatter in principal colors is significantly different ( @xmath75 ) . were the color variability caused by random photometric errors , it would not be correlated with the distribution of asteroid principal colors .
the observed color variability implies inhomogeneous albedo distribution over an asteroid surface .
although the color variability is fairly small , it suggests that large patches with different color than their surroundings exist on a significant fraction of asteroids .
for example , consider a limiting case of an asteroid with two different hemispheres , one with c type material , and one with s type material . in this case
the peak - to - peak amplitude of its @xmath13 color variability would be only 0.2 mag .
, with [email protected] mag . , even under the most favorable condition of the rotational axis perpendicular to the line of sight .
taking into account a distribution of the angle between rotational axis and the line of sight would decrease these values further . without detailed
modeling it is hard to place a lower limit on the fraction of surface with complementary color to explain the observed color variations .
however , using simple toy models and colors typical for c and s type asteroids , we find that this fraction must be well over 10% .
the features seen in spatially resolved color images of eros obtained by near spacecraft ( murchie et al . 2002 ) support such a conclusion .
a simple explanation for the existence of patches differing in color from their surroundings is the deposition of material ( e.g. silicates on c type asteroids and carbonaceous material on s type asteroids ) by asteroid collisions .
however , such surfaces would exhibit color variations preferentially aligned with the @xmath13 color axis , contrary to the observations .
for a given fraction of asteroid surface affected by the deposition of new material , the large difference in albedos of silicate and carbonaceous surfaces would probably produce different amplitudes of color variability for s and c type asteroids ( due to large difference in their albedos ) , a behavior that is not supported by the data .
thus , the color variability can not be explained as due to patches of s - like and c - like material scattered across an asteroid surface .
a plausible cause for optically inhomogeneous surface is space weathering .
this phenomenon includes the effects of bombardment by micrometeorids , cosmic rays , solar wind and uv radiation , and may alter the chemistry of the surface material ( zeller & rouca , 1967 ) .
recent spacecraft data indicate that these processes may be very effective in the reddening and darkening of asteroid surface ( chapman 1996 ) . in this interpretation
the @xmath46 color should show the largest variation ( hendrix & vilas 2003 ) , and this is indeed supported by the data presented here ( see fig .
it is not clear , however , how could such processes result in fairly large isolated surface inhomogeneities .
an interesting explanation for surface inhomogeneities is the effects of cratering .
using near spacecraft measurements , clark et al .
( 2001 ) find 30 - 40% albedo variations on the psyche crater wall on eros .
such a large effect could perhaps produce color variations consistent with observations .
however , it seems premature to draw conclusions with detailed modeling .
irrespective of the mechanism(s ) responsible for the observed color variations , our results indicate that this is a rather common phenomenon .
we conclude by pointing out that the sample presented here will be enlarged by a factor of few in a year or two because sdss is still collecting data , and the size of known object catalog , needed to link the observations , is also growing .
furthermore , the faint limit of the known object catalog is also improving , and will result in a larger size range probed by the sample .
apart from sdssmoc , the upcoming ( in 510 years ) deep synoptic surveys such as pan - starrs and lsst will produce samples of size and quality that will dwarf the sample discussed here , and provide additional clues about the causes of asteroid color variability .
this work has been supported by the hungarian otka grants t034615 , fkfp grant 0010/2001 , szeged observatory foundation , and pro renovanda cultura hungariae foundation dt 2002/maj.21 .
.i . acknowledges generous support by princeton university .
funding for the creation and distribution of the sdss archive has been proved by the alfred p. sloan foundation , the participating institutions , the national aeronautics and space administration , the national science foundation , the u.s .
department of energy , the japanese monbuhagakusho , and the max planck society .
the sdss web site is www.sdss.org .
the participating institutions are the university of chicago , fermilab , the institute for advanced study , the japan participation group , the johns hopkins university , the max - planck - institute for astronomy ( mpia ) , the max - planck - institute for astrophysics ( mpa ) , new mexico state university , princeton university , the united states naval observatory , and the university of washington .
lupton , r.h . , gunn , j.e .
, ivezi , . , et al . , 2001 , in _ astronomical data analysis software and systems x _ , asp conference proceedings , vol.238 , p. 269 . edited by f. r. harnden , jr . , francis a. primini , and harry e. payne .
san francisco : astronomical society of the pacific , issn : 1080 - 7926 | we report a detection of statistically significant color variations for a sample of 7,531 multiply observed asteroids that are listed in the sloan digital sky survey moving object catalog . using 5-band photometric observations accurate to @xmath00.02 mag
, we detect color variations in the range 0.06 - 0.11 mag ( rms ) .
these variations appear uncorrelated with asteroids physical characteristics such as diameter ( in the probed 1 - 10 km range ) , taxonomic class , and family membership .
despite such a lack of correlation , which implies a random nature for the cause of color variability , a suite of tests suggest that the detected variations are not instrumental effects .
in particular , the observed color variations are incompatible with photometric errors , and , for objects observed at least four times , the color change in the first pair of observations is correlated with the color change in the second pair .
these facts strongly suggest that the observed effect is real , and also indicate that for some asteroids color variations are larger than for others .
the detected color variations can be explained as due to inhomogeneous albedo distribution over an asteroid surface .
although relatively small , these variations suggest that fairly large patches with different color than their surroundings exist on a significant fraction of asteroids .
this conclusion is in agreement with spatially resolved color images of several large asteroids obtained by near spacecraft and hst . |
elliptic flow , @xmath0 , is a crucial observable of heavy - ion collisions : the large magnitude of @xmath0 at rhic @xcite and lhc @xcite provides the strongest evidence that a low - viscosity fluid is formed in these collisions @xcite .
elliptic flow is determined to a good approximation by linear response to the initial eccentricity @xmath1 , which quantifies the spatial azimuthal anisotropy of the fireball created right after the collision @xcite .
this initial eccentricity comes from two effects : first , the overlap area between the colliding nuclei has the shape of an almond in non - central collisions , where the smaller dimension of the almond is parallel to the reaction plane .
this results in an eccentricity which becomes larger as impact parameter increases , and whose magnitude is model dependent @xcite .
second , even in central collisions , there is a sizable eccentricity due to quantum fluctuations in wave functions of incoming nuclei @xcite , and to the probabilistic nature of energy deposition in nucleon - nucleon collisions .
the magnitude of these eccentricity fluctuations is again a model - dependent issue , which involves the dynamics of the collision at early times @xcite .
the goal of this paper is to show that the distribution of the initial eccentricity is to some extent independent of model details .
more precisely , it can be written to a good approximation as a universal function of two parameters , where one of the parameters corresponds to the reaction plane eccentricity , and the other parameter characterizes the magnitude of fluctuations .
information on the initial state is thus encoded in two numbers .
the initial eccentricity @xmath1 is defined in every event from the initial energy density profile ( see below sec . [
sec : ellipticpower ] ) and thus carries information about how energy is deposited in the early stages of the collision .
there are several models of the initial density profile and its fluctuations , which are typically implemented through monte carlo simulations .
the monte carlo glauber model is used in many event - by - event hydrodynamic calculations @xcite : in this model , one assumes that the energy is localized around each wounded nucleon .
other monte carlo models of the initial state are inspired by saturation physics @xcite and have also been used as initial conditions in hydrodynamic calculations @xcite .
another approach is to use an event generator from particle physics @xcite or a transport calculation @xcite to model the initial dynamics .
each monte carlo model returns a probability distribution for @xmath1 at a given centrality .
a simple parametrization of the distribution of @xmath1 , usually referred to as the bessel - gaussian distribution @xcite , was proposed in @xcite .
it works well for nucleus - nucleus collisions at moderate impact parameters , but fails for more peripheral collisions and/or small systems such as proton - nucleus collisions .
the reason why it fails can be traced back to the fact that it does not take into account the fact that , by definition , @xmath2 in every event .
a new power distribution was recently introduced @xcite which well describes eccentricity distributions when there are only flow fluctuations ( see also @xcite ) , and satisfies @xmath2 by construction . in sec .
[ sec : ellipticpower ] , we propose a generalization of this result : we take into account the eccentricity in the reaction plane by distorting the power distribution into an elliptic power distribution .
this new , two - parameter distribution reduces to the power distribution for an azimuthally - symmetric system . in sec .
[ sec : mcmodels ] , we use the elliptic power distribution to fit the distribution of @xmath1 in pb+pb collisions calculated by monte carlo methods for several models and for all centralities .
we also show that the elliptic power distribution reproduces the magnitude of eccentricity fluctuations , and the cumulants of the distribution of @xmath1 .
the initial anisotropy in harmonic @xmath3 is defined in every event by @xcite @xmath4 where @xmath5 is the energy density near midrapidity shortly after the collision , and @xmath6 are polar coordinates in the transverse plane , in a centered coordinate system , where @xmath7 is the orientation of the reaction plane . in most of this paper ,
we focus on the second harmonic @xmath8 .
@xmath1 is often referred to as the `` participant eccentricity '' and @xmath9 as the `` participant plane '' .
this terminology refers to monte carlo glauber models @xcite , in the context of which these concepts were first introduced @xcite .
note that @xmath10 by definition .
the initial anisotropy can also be written in cartesian coordinates : @xmath11 @xmath12 is the anisotropy in the reaction plane . for a symmetric density profile satisfying @xmath13 , the definition eq .
( [ defepsilon ] ) implies @xmath14 , which in turn implies @xmath15 : the participant plane coincides with the reaction plane .
this is no longer true in the presence of fluctuations .
figure [ 2d ] ( a ) displays the distribution of @xmath16 , for @xmath8 , obtained in a monte carlo glauber simulation of pb+pb collisions in the 75 - 80% centrality range . in this simulation , centrality is defined according the to number of participants .
the maximum of the distribution is at a positive value of @xmath12 , reflecting the large reaction plane eccentricity .
fluctuations around the most probable value are large .
they display characteristic features : 1 .
the width around the maximum is larger along the @xmath17 axis than along the @xmath18 axis .
the distribution of @xmath12 is left skewed with a steeper decrease to the right of the maximum toward a cut - off at @xmath19 .
the usual bessel - gaussian parametrization @xcite assumes that fluctuations are gaussian and isotropic and therefore misses both features ( see fig .
[ 2d ] ( c ) ) .
these features can be traced back to the constraint that the support of the distribution is the unit disk @xmath20 .
our goal in this paper is to derive a generic distribution with these features . in ref .
@xcite , an exact expression for the distribution of @xmath16 for @xmath8 was derived under the following assumptions : * the energy profile is a superposition of @xmath21 pointlike , identical sources : @xmath22 , where @xmath23 denotes the transverse position of the sources . *
the positions of the sources @xmath23 are independent . *
the distribution of @xmath23 is a 2-dimensional gaussian , where the widths along @xmath18 and @xmath17 may differ .
here we denote by @xmath24 the _ ellipticity _ parameter , corresponding to the eccentricity of the distribution of sources in the reaction plane .
it satisfies @xmath25 . under these conditions ,
the distribution of @xmath16 is @xmath26 where @xmath27 .
this probability distribution is normalized : @xmath28 , where integration runs over the unit disk @xmath29 . in this paper
, we argue that eq .
( [ ellipticpower2d ] ) , which we name the elliptic power distribution , provides a good fit to _ all _ models of the initial state .
this success can be ascribed to the fact that the natural support of the elliptic power distribution is the unit disk : this is a major advantage over previous parametrizations .
we treat both the _ ellipticity _ @xmath30 and the _ power _ @xmath31 as fit parameters . in particular , we allow for arbitrary real , positive values of @xmath31 ( as opposed to integer or half - integer ) . for @xmath32 , the distribution eq .
( [ ellipticpower2d ] ) is azimuthally symmetric : @xmath33 this is the one - parameter power distribution introduced in ref .
@xcite , which was shown to fit monte carlo results when the eccentricity is solely created by fluctuations , as for instance in p - p collisions or p - pb collisions .
the power parameter @xmath31 quantifies the magnitude of fluctuations : the smaller @xmath31 , the larger the fluctuations . when the ellipticity @xmath30 is positive , the denominator of eq .
( [ ellipticpower2d ] ) breaks azimuthal symmetry and favors larger values of @xmath12 .
the mean eccentricity in the reaction plane @xmath34 is derived in appendix [ sec : appendix ] as a function of @xmath30 and @xmath31 . because of fluctuations , it is not strictly equal to the eccentricity of the underlying distribution , @xmath30 @xcite .
it is in general smaller , and coincides with @xmath30 only in the limit @xmath35 .
a fit to monte carlo glauber results using the elliptic power distribution is displayed in fig .
[ 2d ] ( b ) .
the fit is not perfect .
specifically , the maximum density is slightly overestimated , while the width of the @xmath36 distribution is slightly underestimated .
note that there are several differences between the ideal case considered in @xcite and the actual glauber calculation , specifically : the correlations between the participants , the fact that their distribution in the transverse plane is not a gaussian , and the recentering correction .
we have checked that switching off the recentering correction does not make agreement significantly better . despite these imperfections ,
the elliptic power distribution captures both features pointed out at the end of sec .
[ sec : example ] , namely , a larger width along the @xmath17 axis , and a steeper decrease to the right of the maximum .
the elliptic power distribution can be somewhat simplified in the limit @xmath35 , corresponding to a large system with small fluctuations . to leading order in @xmath37 , eq .
( [ ellipticpower2d ] ) reduces to a two - dimensional elliptic gaussian distribution : @xmath38 the maximum lies on the @xmath18-axis at @xmath39 and the widths are given by @xmath40 in general , the gaussian is more elongated along the @xmath17 axis , that is , @xmath41 , which corresponds to the first of the two properties listed in sec .
[ sec : example ] .
it is symmetric around its maximum and therefore does not possess the second property , namely , the skewness along the @xmath18 axis .
this property only appears as a next - to - leading correction of order @xmath37 , which is derived in appendix [ sec : asymptotic ] . the usual isotropic gaussian distribution introduced in ref .
@xcite is obtained by setting @xmath42 in eq .
( [ bg2d ] ) .
this parametrization misses both properties and is therefore less accurate than our new elliptic power distribution , as can be seen in fig .
[ 2d ] ( c ) .
in particular , it overestimates the density at the maximum by a factor larger than 2 . in 7580% central pb - pb collisions .
( a ) : @xmath1 , ( b ) : @xmath43 , ( c ) : @xmath44 .
histograms are monte carlo glauber simulations ( same as in fig .
[ 2d ] ( a ) ) .
dashed lines are bessel - gaussian fits using eq .
( [ bg ] ) .
full lines are elliptic power fits using eq .
( [ ellipticpower ] ) for @xmath1 ( parameters as in fig .
[ 2d ] ( b ) ) and @xmath44 ( @xmath45 and @xmath46 ) .
the dash - dotted line for @xmath43 is a power fit using eq .
( [ power ] ) ( @xmath47 ) . ]
since the orientation of the reaction plane is not directly accessible experimentally , the magnitude of the eccentricity @xmath48 matters more than its phase @xmath49 .
monte carlo simulations of the initial state typically return a probability distribution @xmath50 for each centrality @xcite .
it is obtained by transforming @xmath51 to polar coordinates and integrating over the azimuthal angle : @xmath52 it is normalized by construction : @xmath53 .
inserting eq .
( [ ellipticpower2d ] ) into eq .
( [ cartesiantopolar ] ) and using the symmetry of the integrand under @xmath54 , one obtains @xmath55 the integral can be carried out analytically to give @xmath56 however , if the hypergeometric function is not available , or not defined everywhere needed , the integral over angles in eq .
( [ ellipticpower ] ) may be carried out numerically .
equally spaced angles @xmath57 , where @xmath58 .
excellent accuracy is obtained with @xmath59 integration points . ] for @xmath32 , eq .
( [ ellipticpower ] ) reduces to @xmath60 which is the `` power '' distribution @xcite . in the limit @xmath35
this becomes a gaussian . in the limit @xmath35 and @xmath61 , eqs .
( [ bg2d ] ) and ( [ cartesiantopolar ] ) give @xmath62 which is the usual bessel - gaussian distribution @xcite , where we have defined @xmath63 .
note that if @xmath64 , the two - dimensional elliptic gaussian distribution eq .
( [ bg2d ] ) does not give the bessel - gaussian distribution upon integration over @xmath65 @xcite .
figure [ 1d ] ( a ) presents the histogram of @xmath1 obtained by integrating the results of fig .
[ 2d ] ( a ) over azimuthal angle .
the fit using the elliptic power distribution is clearly much better than the fit using the bessel - gaussian .
fit gives a large weight to the last bin because the bessel - gaussian does not go to zero at @xmath66 .
the elliptic power distribution gives a much better fit than the bessel - gaussian , irrespective of the details of the fit procedure . ] for sake of completeness , fig .
[ 1d ] ( b ) and ( c ) also display the distributions of higher order fourier harmonics @xmath43 and @xmath44 .
the initial _ triangularity _ @xmath43 acts as a seed for triangular anisotropy @xcite , in the same way as the initial eccentricity is the origin of elliptic anisotropy . since triangularity is solely created by fluctuations , the distribution of @xmath43 is well reproduced by the single - parameter power distribution , eq .
( [ power ] ) @xcite .
if the two - parameter elliptic power distribution is used for @xmath43 , the @xmath30 parameter comes out to be essentially zero .
the one - parameter fit is significantly better than the two - parameter bessel - gaussian fit , eq .
( [ bg ] ) .
note that the values of @xmath31 are not necessarily the same for ellipticity and triangularity .
the distribution of the fourth harmonic @xmath44 is well fitted by the elliptic power distribution .
the resulting value of @xmath30 is significantly smaller than for the distribution of @xmath1 .
note , however , that the @xmath30 for @xmath44 is not the only origin of anisotropy in the corresponding harmonic , due to large nonlinear terms in the hydrodynamic response @xcite .
from elliptic power fits to the eccentricity distribution and power fits to the triangularity distribution [ cf .
[ 1d ] ( b ) ] , as a function of centrality for the monte carlo glauber model @xcite and the ip - glasma model @xcite .
( b ) : symbols correspond to the value of the ellipticity parameter @xmath30 .
lines correspond to the mean eccentricity in the reaction plane @xmath67 for the glauber model , either calculated directly ( full line ) , or estimated using the elliptic power distribution eq .
( [ epsrp ] ) ( dotted line ) . ]
we now argue that the elliptic power distribution always gives good fits to distributions of @xmath1 in nucleus - nucleus collisions .
figure [ histo ] presents the histogram of @xmath1 in pb+pb at 2.76 tev in several centrality bins , obtained using the monte carlo glauber ( panels ( a ) to ( d ) ) @xcite and the ip - glasma ( panels ( e ) to ( h ) ) @xcite models , together with fits using the elliptic power and the bessel - gaussian distributions .
both distributions are able to fit both models in the 5 - 10% centrality bin .
bessel - gaussian fits become worse as the centrality percentile is increased , while elliptic power fits are excellent for both models and for all centralities .
only four centrality bins are shown in fig .
[ histo ] for sake of illustration , but we have checked that the fits are as good for the other centralities . for the most central bin ( 0 - 5% ) , however , the fit parameters are strongly correlated and can not be determined independently . this can be understood as follows : for central collisions , the elliptic power distribution is very close to a bessel - gaussian distribution , eq .
( [ bg ] ) .
now , to order @xmath68 , this distribution is invariant under the transformation @xmath69 , i.e , the dependence on @xmath30 can be absorbed into a redefinition of the width @xmath70 .
therefore one can not fit @xmath30 and @xmath70 independently when @xmath30 is too small and one can actually use the one - parameter power distribution .
the two models plotted in fig . [ histo ] represent two extremes in the landscape of initial - state models .
the phobos monte carlo model is the simplest model including fluctuations : all participant nucleons are treated as identical , pointlike sources of energy . by contrast , in the ip - glasma model , the energy density is treated as a continuous field and contains nontrivial fluctuations at the subnucleonic level .
the elliptic power distribution is able to fit both extremes .
we have explicitly checked that it works well also for the mc - kln model @xcite .
we therefore conjecture that it provides a good fit to all monte carlo models of initial conditions . the elliptic power distribution , eq .
( [ ellipticpower2d ] ) , encodes the information about the eccentricity distribution into two parameters which are plotted in fig .
[ parameters ] as a function of centrality for the ip glasma and monte carlo glauber models .
as explained above , the two parameters can not be disentangled for very central collisions in practice , the fitting procedure returns a very large error on each parameter : therefore we exclude the most central ( @xmath71 ) bin .
panel ( a ) also displays the values of @xmath31 obtained by fitting the distribution of the _ triangularity _ @xmath43 with the power distribution eq .
( [ power ] ) .
the power parameter @xmath31 increases towards central collisions .
this is expected , since @xmath31 is typically proportional to the system size . in the monte
carlo glauber model , @xmath31 is approximately proportional to the number of participant nucleons @xmath72
. the ellipticity @xmath30 , on the other hand , smoothly increases with centrality percentile , and is somewhat larger for the ip - glasma than for the glauber model , in line with the expectations that saturation - inspired models predict a larger eccentricity than glauber models @xcite .
for the monte carlo glauber model , we also show on the same plot the reaction plane eccentricity @xmath67 : we can either calculate it directly in the monte carlo glauber model ( full line ) or estimate it using eq .
( [ epsrp ] ) below derived from the elliptic power distribution ( dotted line ) .
it is close to the glauber @xmath30 up to mid - centrality .
the difference between @xmath30 and @xmath67 is nevertheless much larger than predicted by the elliptic power distribution .
this can be attributed to the fact that the elliptic power distribution does not reproduce all the fine structure of the two - dimensional distribution ( fig .
[ 2d ] ( a ) ) , even though it provides a very good fit to the distribution of @xmath1 ( figs .
[ 1d ] ( a ) and [ histo ] ) . ) , as a function of collision centrality .
lines : monte carlo glauber results for @xmath1 ( full ) and @xmath43 ( dash - dotted ) .
symbols are results using the elliptic power distribution eq .
( [ ellipticpower ] ) for @xmath1 ( circles ) , and the power distribution eq .
( [ power ] ) for @xmath43 ( squares ) , with parameters fitted to the histograms .
the horizontal dashed line indicates the value @xmath73 corresponding to pure gaussian fluctuations ( eq . ( [ bg ] ) with @xmath30 ) . ] a standard measure of eccentricity fluctuations is the ratio of the standard deviation to the mean @xcite : @xmath74 where angular brackets denote an average over events in a centrality class .
we now check that the elliptic power distribution , fitted to the histogram of @xmath1 , correctly reproduces the magnitude of eccentricity fluctuations .
the ratio eq .
( [ sigmaovermean ] ) is presented in fig .
[ sigovermean ] for @xmath1 and @xmath43 . for central collisions ,
it approaches @xmath75 @xcite for both @xmath1 and @xmath43 , which is the value given by eq .
( [ bg ] ) for @xmath32 . for more peripheral collisions ,
relative eccentricity fluctuations decrease very mildly for @xmath43 , and more strongly for @xmath1 .
for @xmath43 , this mild decrease is captured by fitting with the power distribution , eq .
( [ power ] ) .
the mean of the power distribution is given by @xmath76 while the mean square is @xcite @xmath77 .
the mean of the elliptic power distribution , eq .
( [ ellipticpower ] ) , must be calculated numerically as a function of @xmath30 and @xmath31 .
the eccentricity fluctuations from the elliptic power distribution closely match the monte carlo glauber result in fig .
[ sigovermean ] . .
( b ) : cumulant ratios @xmath78 and @xmath79 .
lines are monte carlo glauber results
. symbols are results using the elliptic power distribution eq .
( [ cumulantsep ] ) with parameters fitted to the distribution of @xmath1 . ]
more detailed information about eccentricity fluctuations is contained in moments or cumulants of the distribution .
the moment of order @xmath80 is defined as @xmath81 .
often , one solely uses _ even _ moments of the distribution @xmath82 , because the corresponding moments of the distribution of anisotropic flow are directly accessible through cumulant analyses @xcite .
the first eccentricity cumulants @xcite are defined by : @xmath83 figure [ fig : cumulants ] displays ratios of successive cumulants obtained in the monte carlo glauber calculation and using the elliptic power distribution , eq .
( [ cumulantsep ] ) .
@xmath84 increases from central to peripheral collisions .
it is smaller than unity by definition .
higher order ratios @xmath78 and @xmath79 are exactly equal to 1 for the bessel - gaussian distribution .
the monte carlo glauber calculation gives ratios slightly smaller than unity , with a nontrivial centrality dependence .
these nontrivial features are reproduced by the elliptic power distribution .
we have introduced a new parametrization of the eccentricity distribution in nucleus - nucleus collisions . like the previously used bessel - gaussian parametrization , it is a two - parameter distribution , but it describes peripheral collisions much better .
this is due to the correct implementation of the constraint that the eccentricity must be smaller than unity in all events .
the consequence of our result is that any model of initial - state eccentricities can be characterized by two numbers for each centrality : the ellipticity @xmath30 , which corresponds closely to the reaction - plane eccentricity ; the power parameter @xmath31 , which governs the magnitude of fluctuations and scales like the number of participants in the glauber model .
since elliptic flow is essentially proportional to the initial eccentricity @xcite , our result can be applied @xcite to the distribution of elliptic flow values , which has been measured recently at the lhc @xcite .
the elliptic power distribution could also be used as a kernel in the unfolding procedure which is used to eliminate finite multiplicity fluctuations @xcite .
it could also be used in fitting the distribution of the flow vector @xcite .
we expect it to give a better result than the bessel - gaussian distribution , which has been found to be not precise for peripheral collisions @xcite .
we thank hiroshi masui for carrying out the monte carlo glauber calculations which inspired this work , bjrn schenke for the ip glasma results , c. loizides for the new version of the phobos glauber , m. luzum and s. voloshin for extensive discussions and suggestions . in particular , we thank s. voloshin for suggesting the name elliptic power , and for useful comments on the manuscript .
jyo thanks the mit lns for hospitality .
ly is funded by the european research council under the advanced investigator grant erc - ad-267258 .
amp was supported by the director , office of science , of the u.s .
department of energy .
the two - dimensional elliptic power distribution eq .
( [ ellipticpower2d ] ) is normalized to unity on the unit disk if @xmath85 and @xmath86 .
we choose the convention @xmath87 throughout this paper . for @xmath88 , eq .
( [ ellipticpower2d ] ) has a maximum on the @xmath18-axis for @xmath89 , where @xmath90 but for @xmath91 , the distribution diverges on the unit circle @xmath92 .
the mean eccentricity in the reaction plane is obtained by integrating eq .
( [ ellipticpower2d ] ) : @xmath93 where @xmath94 denotes the hypergeometric function . in the limit @xmath61
, it simplifies to @xmath95 thus @xmath96 is slightly bigger than @xmath67 due to fluctuations .
we now explain how to evaluate the moments of the elliptic power distribution .
multiplying eq .
( [ ellipticpower2d ] ) by @xmath97 and integrating successively over @xmath36 and @xmath12 , one obtains @xmath98 using this equation , one can express analytically the even moments @xmath99 and the cumulants @xcite @xmath100 as a function of @xmath31 and @xmath30 . using the shorthand notation @xmath101 and inserting into eq .
( [ cumulants ] ) , one obtains @xmath102
we now study the elliptic power distribution eq .
( [ ellipticpower2d ] ) in the limit @xmath35 , corresponding to a large system . to leading order ,
the distribution is a gaussian centered around the intrinsic ellipticity @xmath30 , see eq .
( [ bg2d ] ) .
we therefore write @xmath104 and treat @xmath105 and @xmath36 as small parameters of order @xmath106 . expanding the logarithm of eq .
( [ ellipticpower2d ] ) in powers of @xmath107 and exponentiating , one obtains @xmath108 where @xmath109 is the gaussian distribution in eq .
( [ bg2d ] ) , and @xmath110 and @xmath111 are perturbations of order @xmath107 : @xmath112 @xmath110 is linear , while @xmath111 is cubic in @xmath105 and @xmath36 .
the linear term @xmath110 shifts the maximum of the distribution , which is found by setting @xmath113 in eq .
( [ expansion ] ) and differentiating with respect to @xmath105 : @xmath114 alternatively , this result can be recovered by expanding eq .
( [ max ] ) .
the cubic term @xmath111 skews the gaussian and is responsible for the skewness seen in fig .
[ 2d ] ( b ) and fig .
[ 1d ] ( a ) .
the linear term @xmath110 can be absorbed by shifting the maximum of the gaussian according to eq .
( [ maxnlo ] ) , therefore the difference between @xmath67 and @xmath115 is solely due to @xmath111 : @xmath116 comparing with eq .
( [ maxnlo ] ) , one sees that @xmath117 , which is a consequence of the skewness .
alternatively , eq . ( [ avnlo ] ) can be obtained by expanding eq .
( [ epsrp ] ) .
the first moments can also be evaluated to first order in @xmath37 .
the mean square eccentricity is @xmath118 when @xmath119 , one recovers the result obtained with the power distribution @xcite in the limit @xmath35 . when @xmath120 , the correction is negative , so that rms anisotropy is _ smaller _ than @xmath30 .
the fourth moment is given by @xmath121 from eqs .
( [ mom2 ] ) and ( [ mom4 ] ) , one obtains the cumulant @xmath122 ( see eq .
( [ cumulants ] ) ) : @xmath123 note that @xmath124 for all positive @xmath30 in the limit @xmath35 . in the limiting case @xmath32
, @xmath125 is positive and of order @xmath126 @xcite .
higher order cumulants are all equal to @xmath122 to order @xmath37 .
we now consider a different asymptotic expansion introduced in @xcite , where one treats @xmath31 as a large parameter and @xmath30 as a small parameter , keeping the product @xmath127 fixed .
the only difference with the asymptotic expansion carried out in appendix [ sec : asymptotic ] is that we also treat @xmath30 as a small parameter of order @xmath107 .
therefore the perturbations @xmath110 and @xmath111 in eq . (
[ perturbation ] ) are of order @xmath128 . for sake of consistency , one must carry out the whole expansion to that order .
one obtains @xmath129 where @xmath110 and @xmath111 are defined in eq .
( [ perturbation ] ) and @xmath130 is a new quartic term : @xmath131 where we have introduced the shorthand notation @xmath132 and simplified the expressions of @xmath110 and @xmath111 using eq .
( [ sigmaxy ] ) and @xmath61 . in the isotropic case @xmath32 ,
both @xmath110 and @xmath111 vanish and only @xmath130 contributes .
the mean square eccentricity is given by @xmath133 where the first two terms are the leading order terms , of order @xmath37 , and the two next terms are corrections of order @xmath134 .
the first three terms are present in eq .
( [ mom2 ] ) , while the last term is the contribution of the quartic perturbation @xmath130 in eq .
( [ perturbation2 ] ) .
similarly , one can expand the cumulant @xmath122 , to give for the fourth power : @xmath135 where the first term is the leading term , of order @xmath134 , and the next three terms are corrections of order @xmath136 . in the isotropic case @xmath32
, the exact result is @xmath137 $ ] @xcite , which reduces to @xmath138 for @xmath35 , in agreement with the above result .
k. h. ackermann _ et al .
_ [ star collaboration ] , phys .
lett . * 86 * , 402 ( 2001 ) s. s. adler _ et al . _
[ phenix collaboration ] , phys .
lett . * 91 * , 182301 ( 2003 ) kaamodt _ et al . _
[ alice collaboration ] , phys .
lett . * 105 * , 252302 ( 2010 ) g. aad _ et al . _ [ atlas collaboration ] , phys .
b * 707 * , 330 ( 2012 ) s. chatrchyan _ et al . _
[ cms collaboration ] , phys .
c * 87 * , 014902 ( 2013 ) p. romatschke and u. romatschke , phys .
lett . * 99 * , 172301 ( 2007 ) m. luzum and p. romatschke , phys .
c * 78 * , 034915 ( 2008 ) [ erratum - ibid .
c * 79 * , 039903 ( 2009 ) ] b. alver _ et al . _
[ phobos collaboration ] , phys .
lett . * 98 * , 242302 ( 2007 ) t. hirano , u. w. heinz , d. kharzeev , r. lacey and y. nara , phys . lett .
b * 636 * , 299 ( 2006 ) t. lappi and r. venugopalan , phys .
c * 74 * , 054905 ( 2006 ) m. miller and r. snellings , [ nucl - ex/0312008 . ]
m. l. miller , k. reygers , s. j. sanders and p. steinberg , ann .
nucl . part .
sci . * 57 * , 205 ( 2007 ) c. flensburg , arxiv:1108.4862 [ nucl - th ] .
a. dumitru and y. nara , phys .
c * 85 * , 034907 ( 2012 ) b. schenke , p. tribedy and r. venugopalan , phys .
c * 86 * , 034908 ( 2012 ) h. holopainen , h. niemi and k. j. eskola , phys .
c * 83 * , 034901 ( 2011 ) b. schenke , s. jeon and c. gale , phys .
* 106 * , 042301 ( 2011 ) z. qiu and u. w. heinz , phys .
c * 84 * , 024911 ( 2011 ) p. bozek , phys .
c * 85 * , 014911 ( 2012 ) h .- j .
drescher and y. nara , phys .
c * 76 * , 041903 ( 2007 ) b. schenke , p. tribedy and r. venugopalan , phys .
lett . * 108 * , 252301 ( 2012 ) r. andrade , f. grassi , y. hama , t. kodama and o. socolowski , jr .
lett . * 97 * , 202302 ( 2006 ) z. -w .
lin , c. m. ko , b. -a .
li , b. zhang and s. pal , phys .
c * 72 * , 064901 ( 2005 ) h. petersen , j. steinheimer , g. burau , m. bleicher and h. stocker , phys .
c * 78 * , 044901 ( 2008 ) s. a. voloshin , a. m. poskanzer and r. snellings , arxiv:0809.2949 [ nucl - ex ] . s. a. voloshin , a. m. poskanzer , a. tang and g. wang , phys .
b * 659 * , 537 ( 2008 ) l. yan and j. -y .
ollitrault , phys .
* 112 * , 082301 ( 2014 ) a. bzdak , p. bozek and l. mclerran , nucl .
* a927 * , 15 ( 2014 ) a. bzdak and v. skokov , arxiv:1312.7349 [ hep - ph ] .
d. teaney and l. yan , phys .
c * 83 * , 064904 ( 2011 ) b. alver , m. baker , c. loizides and p. steinberg , arxiv:0805.4411 [ nucl - ex ] .
ollitrault , phys .
d * 46 * , 229 ( 1992 ) .
j. casalderrey - solana and u. a. wiedemann , phys .
lett . * 104 * , 102301 ( 2010 ) s. k. prasad , v. roy , s. chattopadhyay and a. k. chaudhuri , phys .
c * 82 * , 024909 ( 2010 ) r. s. bhalerao and j. -y .
ollitrault , phys .
b * 641 * , 260 ( 2006 ) g. aad _ et al . _
[ atlas collaboration ] , jhep * 11 * , 183 ( 2013 ) b. schenke , p. tribedy and r. venugopalan , nucl .
* a926 * , 102 ( 2014 ) b. alver and g. roland , phys .
c * 81 * , 054905 ( 2010 ) [ erratum - ibid .
c * 82 * , 039903 ( 2010 ) ] f. g. gardim , f. grassi , m. luzum and j. -y .
ollitrault , phys .
c * 85 * , 024908 ( 2012 ) d. teaney and l. yan , phys .
c * 86 * , 044908 ( 2012 ) b. schenke , p. tribedy and r. venugopalan , arxiv:1403.2232 [ nucl - th ] . c. gale , s. jeon , b. schenke , p. tribedy and r. venugopalan , phys .
lett . * 110 * , 012302 ( 2013 ) p. sorensen [ star collaboration ] , j. phys .
g * 34 * , s897 ( 2007 ) w. broniowski , p. bozek and m. rybczynski , phys .
c * 76 * , 054905 ( 2007 ) n. borghini , p. m. dinh and j. -y .
ollitrault , phys .
c * 64 * , 054901 ( 2001 ) a. bilandzic , r. snellings and s. voloshin , phys .
c * 83 * , 044913 ( 2011 ) h. niemi , g. s. denicol , h. holopainen and p. huovinen , phys . rev .
c * 87 * , 054901 ( 2013 ) l. yan , j .- y .
ollitrault , and a. m. poskanzer , quark matter presentation , darmstadt ( 2014 ) .
j. jia and s. mohapatra , phys .
c * 88 * , no . 1 , 014907 ( 2013 ) b. alver _ et al .
_ [ phobos collaboration ] , phys .
* 104 * , 142301 ( 2010 ) j. barrette _ et al .
_ [ e877 collaboration ] , phys . rev .
* 73 * , 2532 ( 1994 ) s. voloshin and y. zhang , z. phys .
c * 70 * , 665 ( 1996 ) c. adler _ et al . _ [ star collaboration ] , phys .
c * 66 * , 034904 ( 2002 ) b. alver , b. b. back , m. d. baker , m. ballintijn , d. s. barton , r. r. betts , r. bindel and w. busza _ et al .
_ , phys .
c * 77 * , 014906 ( 2008 ) | we propose a new parametrization of the distribution of the initial eccentricity in a nucleus - nucleus collision at a fixed centrality , which we name the elliptic power distribution .
it is a two - parameter distribution , where one of the parameters corresponds to the intrinsic eccentricity , while the other parameter controls the magnitude of eccentricity fluctuations . unlike the previously used bessel - gaussian distribution , which becomes worse for more peripheral collisions , the new elliptic power distribution fits several monte carlo models of the initial state for all centralities . |
continued advances in commodity processing and networking hardware make pc ( or workstation ) clusters a very attractive alternative for lattice qcd calculations @xcite . indeed , there are quite a few important problems that can be addressed on pc clusters , and many lattice physicists are taking advantage of this opportunity .
however , for the most demanding problems in lattice qcd , e.g. dynamical fermion simulations with realistic quark masses , one would like to distribute the global volume over as many nodes as possible , resulting in a very small local volume per node .
pc clusters are inadequate to deal with this case because the communications latency inherent in their networking hardware implies that the local volume must not be chosen too small if a reasonable sustained performance is to be achieved . in other words , for typical lattice qcd problems
pc clusters do not scale well beyond a few hundred nodes . in custom - designed supercomputers such as qcdoc @xcite and apenext @xcite
, the communications hardware is designed to reduce the latencies and to assist critical operations ( such as global sums ) in hardware . as a result , these machines are significantly more scalable and allow for much smaller local volumes .
in addition , they provide low power consumption , a small footprint , and a very low price / performance ratio per sustained mflops . on the downside ,
the development effort is considerably higher than for pc clusters , but this effort is amortized by the unique strengths of these machines .
the qcdoc hardware has been described in detail in several previous publications , see refs .
@xcite , therefore we only summarize its most important features here . the qcdoc asic , shown schematically in fig .
[ fig : asic ] , was developed in collaboration with ibm research and manufactured by ibm .
it contains a standard powerpc 440 core running at 500 mhz , a 64-bit , 1 gflops fpu , 4 mbytes of embedded memory ( edram ) , and a serial communications interface ( scu ) which has been tailored to the particular requirements of lattice qcd .
the scu provides direct memory access , single - bit error detection with automatic resend , and a low - latency pass - through mode for global sums . also on the chip
are several bus systems , controllers for embedded and external ( ddr ) memory , an ethernet controller , a bootable ethernet - jtag interface , and several auxiliary devices ( interrupt controller , i@xmath0c interface , etc . ) . a picture of one of the first asics , delivered in june of 2003 , is shown in fig .
[ fig : asic_closeup ] .
the physical design of a large machine is as follows .
two asics are mounted on a daughterboard , together with two standard ddr memory modules ( one per asic ) with a capacity of up to 2 gbytes each .
the only other nontrivial components on the daughterboard , apart from a few leds , are four physical layer chips for the mii interfaces ( two per asic ) and a 4:1 ethernet repeater which provides a single 100 mbit / s ethernet connection off the daughterboard . a picture of the very first two - node daughterboard is shown in fig .
[ fig : db ] .
[ fig : db ] a motherboards holds 32 such daughterboards , eight motherboards are mounted in a crate , and a large machine is built from the desired number of crates . a picture of a qcdoc motherboard is shown in fig .
[ fig : mb ] .
the physics communications network of qcdoc is a 6-dimensional torus with nearest - neighbor connections .
the two extra dimensions allow for machine partitioning in software so that recabling is not required .
a 64-node motherboard has a @xmath1 topology , with three open dimensions and three dimensions closed on the motherboard ( one of which is closed on the daughterboard ) .
the scu links run at 500 mbit / s and provide separate send and receive interfaces to the forward and backward neighbors in each dimension , resulting in a total bandwidth of 12 gbit / s per asic ( of which 8 gbit / s will be utilized in a 4-dimensional physics calculation ) .
in addition to the physics network , there is an ethernet based network for booting , i / o , and debugging , as well as a global tree network for three independent interrupts . the ethernet traffic from / to each motherboard proceeds at 800 mbit / s to a commercial gbit - ethernet switch tree , a parallel disk system , and the host machine .
the latter will be a standard unix smp machine with multiple gbit - ethernet cards .
see fig .
[ fig : network ] .
as of the writing of this article ( september 2003 ) , all major subsystems of the qcdoc asic have been tested in single - daughterboard configurations ( 2 asics per daughterboard ) using a temporary test - jig .
this configuration allows non - trivial communications between the two asics in one of the six dimensions ; for the remaining dimensions the asic communicates with itself in a loop - back mode . extensive memory tests with different sized commodity external ddr sdram modules have been done , tests of the 4 mbyte on - chip edram have been performed , and all the dma units have been used .
high - performance dirac kernels have been run for wilson and asqtad fermion formulations , confirming the performance figures given in table [ tab : performance ] and ref .
no problems with the asic have been found to date . with qcdoc motherboards now in hand for use ( fig .
[ fig : mb ] ) , tests of 64 and 128 node machines are imminent . in our test - jig tests ,
the asics appear to be working well close to the target frequency of 500 mhz . with a fully populated motherboard , and the more stable electrical environment provided by the motherboard as compared to our simple test - jigs
, we will soon be able to test a large number of asics at the 500 mhz design frequency . from preliminary measurements ,
the power consumption per node is about 5 w.
one of the major goals of the development team was to make qcdoc accessible and deployable to a large scientific community , so every attempt has been made to allow users to use standard software tools and techniques . in order to reliably boot ,
monitor and diagnose the 10,000 + nodes in a large qcdoc machine , and to allow user code maximum access to the high - performance capabilities of the hardware , we are writing a custom operating system for qcdoc .
this is modeled on the qcdsp operating system , which has successfully allowed the use of a machine with a similar number of processing nodes .
the qcdoc operating system ( qos ) has been improved in myriad ways , continuing to focus on ease - of - use for the end - user .
we now give a more detailed list of the requirements of the operating system and discuss how these features have been implemented .
in addition to booting the machine , the qos must diagnose hardware errors and monitor the hardware during user program execution .
monitoring is accomplished by regular inspection of on - chip registers which monitor memory and communication link status , along with interrupt handlers which are invoked by changes in hardware state deemed crucial . after booting ,
partitions of the machine may be requested for interactive use by a user or the queuing system .
each partition will only run a single user program at a time ; all multitasking will be handled by the host front - end . during program execution ,
the operating system provides host - qcdoc input / output access using commodity ethernet connections , access to special qcdoc features such as the mesh communication network and on - chip memory and access from qcdoc processing nodes to a parallel disk array .
the qos is designed to coordinate the interactions between the host and qcdoc , insulating the user from many qcdoc - specific details while still providing detailed control over their execution environment .
our solution is qos , written largely from scratch in c++ , c and some assembler .
the software runs on the host computer and the qcdoc nodes and consists of several parts , namely : * the qdaemon on the host , which manages and monitors the entire qcdoc machine .
all interactions go through the qdaemon .
* the qclient layer on the host , which provides access to the qdaemon for a variety of planned user interfaces , such as the batch system and web applications , as well as the existing command line interface , the qcsh . * the qcsh on the host , a modified version of tcsh , which is currently in use and allows complete control and use of qcdoc .
* the boot / run kernels on each node .
the qdaemon is the heart of the operating system , and it is designed and implemented using modern c++ techniques , e.g. it is heavily templated , it employs posix threads to allow efficient use of a multi - processor front end , its queue calls and object lists are thread - safe , and it can drive multiple physical network interfaces .
the single qdaemon on the host uses reliable udp to transport packets between qcdoc and the host , starting with basic udp packets in the boot procedure and adding an rpc protocol to this when the run kernels are loaded .
the qdaemon controls the ip addressing scheme to the nodes , does the partitioning that users request , monitors machine status , directs i / o to the host and evokes hardware testing when needed .
the qdaemon can be accessed by a
root " user for system management and monitoring , but most users will only have access to a limited set of qdaemon s capabilities , those appropriate for the partition where they are running a job .
the qdaemon also provides flexible software partitioning of qcdoc .
the qclient is a library of interfaces which communicates with the qdaemon and can accept input from a variety of planned user tools . to access the qclient , and hence the qdaemon , access to the host computer must be gained by conventional authentication tools such as ssh or a secure web connection .
currently , a command - line interface , the qcsh , communicates with the qdaemon via the qclient library . following the successful qcdsp model ,
qcsh is a command - line interface . starting from a standard unix shell , tcsh ,
extra built - in commands to control qcdoc have been added .
users can use normal shell redirection to control i / o from qcdoc , programs can be executed on qcdoc and backgrounded in qcsh allowing further interaction between the qcsh and qdaemon ( but not qcdoc directly since it is executing a program ) , and standard shell scripts can be used .
the special qcdoc commands in the qcsh all begin with the letter q. some examples are : qinit - establish a connection between the qcsh and the qdaemon ; qpartition_create - create a partition of qcdoc ; qboot - boot a qcdoc partition ; qrun - run a user program ; qhelp - display available qcommands .
the qdaemon will only permit connections from the local host , which allows a unix socket to be passed to the qdaemon from a user interface like the qcsh .
the qdaemon then does user i / o from / to this unix socket and can determine unix user and group identities directly from the host os kernel .
this is a major simplification .
while the qdaemon will be handling many users access to qcdoc , it does not have to open / close files and concern itself with protections .
this is handled by the initial user interface , thereby ensuring correct ownership and permissions .
the first software loaded to a qcdoc node is the boot kernel , which is loaded directly to the data and instruction caches of the 440 core via the ethernet - jtag hardware .
only the 440 core must be functional for the boot kernel to execute .
when execution begins , tests of the on - chip and ddr memory can be done .
once the memory is known to be working , the boot kernel enables the standard 100 mbit ethernet port on the qcdoc asic and the run kernel is loaded down .
the run kernel handles the activation of machine - global features such as the nearest - neighbor communications network .
when these steps complete , which should be on the scale of ten minutes even for a large machine , it is available for general use . all of the steps outlined above are implemented and have been extensively used on our first asics .
the run kernel provides support for users , including access to qcdoc features via system calls .
part of the strategy for using rpc for the host run - kernel communication is to avoid having a separate communications protocol for disk access .
we have written an nfs client for the run kernel that uses the standard rpc - based nfs protocol .
the client supports two mount points and open / read / write / close functionality .
this nfs support has already been tested on qcdoc .
while providing a standard user environment for high - performance computing , the run - kernel will not support multi - tasking .
we have chosen to keep the run - kernel lean and compact to ensure reliability and to keep the software task bounded . a major issue for a massively parallel computer is the effectiveness of its message passing .
in hardware , qcdoc is a nearest - neighbor mesh , the majority of qcd based communications are also nearest - neighbor and consequently qos natively supports nearest - neighbor communications calls .
the qos calls are implemented so as not to interfere with the low latency of the qcdoc communications hardware .
the qcdoc hardware supports efficient global sums ( done via our pass - through hardware ) , which are also accessible via qos calls .
we have also implemented the scidac lattice qcd software committee s qcd message passing ( qmp ) protocol on qcdoc .
this protocol supports nearest - neighbor communications , which we efficiently map to the native qos communications calls , as well as arbitrary communications . for the latter , we will implement a manhattan - style routing .
a simple ( but not trivial ) consequence of qcdoc using a standard processor from ibm s powerpc line is the availability of a whole arsenal of open source and commercial software tools , most notably the gnu tools and ibm s tools .
on qcdoc , users will be able to use the gnu toolset , the closest thing to a standard across computer platforms .
additionally , we have access to the high - performance commercial tools from ibm , such as the xlc / xlc compilers which we have seen outperform the gnu compilers on powerpc platforms by as much as a factor of two .
we expect that users may do initial code development with the gnu toolset and only compile with the more restrictively - licensed ibm compilers when final performance is an issue .
an issue for the larger potential user group for qcdoc is performance of existing physics codes , not written for a qcdoc or qcdsp type of computer architecture .
the milc code was an obvious choice to test , since it is one of the major lattice simulation codes and is written in c. the milc code has been run on qcdoc , both the simulator and now the actual asic , concentrating on the asqtad action .
unmodified milc code gives performance in the few percent range for small lattice volumes , which was easily improved by a few standard c - code modifications as described in @xcite .
a summary of the performance for single precision milc code and our double precision assembly code is given in table [ tab : performance ] .
l|c|c|c action & vol . & assem . &
milc + wilson & @xmath2 & 47% + & @xmath3 & 54% + clover & @xmath2 & 56% + & @xmath3 & 59% + staggered & @xmath2 & 36% & 17% + & @xmath3 & & 21% + asqtad & @xmath4 & 43% & 15% + af & @xmath2 & & 14% + & @xmath3 & & 20% + [ tab : performance ]
a 128-node prototype machine is currently being assembled at columbia . assuming that no major problems are found , two machines of 10 tflops ( peak ) each will be available to ukqcd and the riken - bnl research center by the summer of 2004 .
the operating system and user software needed to utilize these machines is progressing in hand with the hardware developments .
approval is pending for a 3 tflops ( peak ) machine at columbia and a @xmath5 20 tflops ( peak ) installation at bnl for the u.s .
lattice community .
* acknowledgments .
* this work was supported in part by the u.s .
department of energy , the institute of physical and chemical research ( riken ) of japan , and the u.k .
particle physics and astronomy research council . | qcdoc is a massively parallel supercomputer whose processing nodes are based on an application - specific integrated circuit ( asic ) .
this asic was custom - designed so that crucial lattice qcd kernels achieve an overall sustained performance of 50% on machines with several 10,000 nodes .
this strong scalability , together with low power consumption and a price / performance ratio of $ 1 per sustained mflops , enable qcdoc to attack the most demanding lattice qcd problems .
the first asics became available in june of 2003 , and the testing performed so far has shown all systems functioning according to specification .
we review the hardware and software status of qcdoc and present performance figures obtained in real hardware as well as in simulation . |
the experiments were performed in ultrahigh vacuum ( base pressure : @xmath42 mbar ) with a radio - frequency low - temperature scanning tunneling microscope @xcite operated at 5k .
it utilizes a sharp tungsten tip ( electrochemically etched and thermally deoxidized above 1070k ) as , both , imaging probe as well as movable ground - electrode against the flat sample .
the latter is a ag(111 ) single - crystal prepared by repeated cycles of ar@xmath43 ion sputtering ( 600ev ) and thermal annealing at 720k .
the ag(111 ) sample is biased from independent rf- and dc - voltage sources for applying ac and dc @xmath0-fields of @xmath44@xmath45v / m at its surface .
the rf - circuit and electronics are described elsewhere @xcite .
after cooling the sample to 5k , the stm chamber was flooded for 1 min with ar gas at a pressure of @xmath46 mbar yielding an ar coverage of @xmath47monolayers on ag(111 ) .
after preparation by this procedure the ag(111 ) surface is covered by 2d - islands of ar with compact shapes and typical sizes ranging from 30 to 100 nm .
we kindly acknowledge financial support of the project i958 by the austrian science fund ( fwf ) .
s.m . , r.k . and g.s .
designed the experiments .
, s.t . , and s.w .- b . conducted the experiments .
g.s . , s.t .
, s.w .- b . and s.m .
analyzed the data .
, r.k . and g.s .
wrote the manuscript .
s.m . and r.k . planned and supervised the project .
all authors discussed the manuscript .
the authors declare no competing financial interests .
correspondence and requests for materials should be addressed to [email protected] .
27ifxundefined [ 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 * * , ( ) link:\doibase 10.1002/9780470034590.emrstm0584 [ _ _ ] ( , ) @noop * * , ( ) @noop * * , ( ) @noop _ _ , edited by and , , vol .
( , ) pp . @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ , ed .
( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) | we present here a powerful method providing simultaneous atomic spatial and nanosecond temporal resolution for investigating dynamics and structure on the atomic scale , in general .
we reveal the dynamic reorganization of surface ( ad)atoms induced by radio frequency alternating charging and decharging of a metal .
our method utilizes taylor - made nano - fabricated two - dimensional islands of physisorbed argon atoms , acting as motion sensors , probed by a radio frequency low - temperature scanning tunneling microscope .
charge oscillations forced by voltages and currents alternating at radio frequency ( rf ) play a dominant role in practically all electronic devices used in our daily life .
such charge oscillations are known to excite plasmons in the quasi - two - dimensional skin layer at the surface of metals @xcite .
in addition , due to electrostriction @xcite , a periodically modulated electric ( @xmath0 ) field gives rise to forced small - amplitude mechanical oscillations of the atomic lattice , @xmath0-fields of @xmath1v / m typically generating strain of 0.1 % @xcite .
such mechanical oscillations are utilized as surface acoustic waves in piezoelectric biosensors @xcite and frequency filters for telecommunication @xcite ; they enable the tuning of electronic and magnetic properties of materials @xcite as well as the operation of artificial muscles for robots and artificial limbs @xcite ; ultrasonic irradiation facilitates to control spin
lattice relaxation times and peak widths in nuclear magnetic resonance @xcite . for measuring strain ,
macroscopic techniques such as x - ray diffraction @xcite or cantilever beam techniques @xcite are well established .
detection of picometer - scale mechanical amplitudes of the surface atomic lattice has been achieved up to 1ghz by an rf - modified scanning tunneling microscope @xcite .
near - field microwave microscopy has demonstrated the electrodynamical response of the material on length scales far shorter than the free - space wavelength of the microwave @xcite .
however , a direct real - space detection and imaging of the surface atomic lattice and its dynamics has remained elusive , to date , due to the lack of simultaneous spatial and temporal resolution of the detection method applied .
we have developed a method that circumvents these difficulties .
here we investigate with nanometer spatial- and nanosecond temporal resolution the impact of radio - frequency alternating electric charging and de - charging of a metal on its atomic surface structure .
in particular , we reveal the dynamic reorganization of surface ( ad)atoms with a time constant of 147 ns induced by applying a 2-ns - periodic rf - voltage ( 530 mhz ) to the metal surface . as origin , we identify the charge - density oscillations in the metal surface skin layer . for detecting such dynamic processes
we utilize nanometer - sized motion sensors ( fig .
[ fig : sensor ] ) consisting of nano - structured two - dimensional ( 2d ) islands of physisorbed noble - gas atoms @xcite .
our motion sensors are shown herein to structurally transform on oscillating substrates like an rf - biased metal surface .
their structural transformations are shown to be powerful analytical probes for characterizing the underlying physical processes of the excitation at the nano - scale .
( a ) nanometer - sized motion sensor : nano - fabricated 2d - island of ar on ag(111 ) imaged by stm at 5k ( @xmath2nm@xmath3 , @xmath4-scale : 200pm , @xmath5v , 70pa ) ; arrows mark lateral channels fabricated by cutting - out from the island by dc - stm manipulation .
( b ) same 2d - island as in ( a ) before the nano - fabrication step ; left inset : magnified view ( @xmath6nm@xmath3 ) revealing atomic resolution on the ar 2d - island as well as single ar vacancies ( labeled 1 ) ; right inset : atomic - resolution image of the ag(111 ) substrate ( @xmath7nm@xmath3 , 1.2 na , @xmath8mv).,width=321 ] figure [ fig : sensor]a shows exemplarily a typical motion sensor imaged by stm at 5k .
it is based on a 2d - island of ar on ag(111 ) that exhibits a strongly non - equilibrium shape obtained after cutting lateral channels ( marked by arrows ) out of the 2d - island .
this is achieved by means of dc - stm manipulation @xcite for the controlled removal of ar atoms at the channels ( see supplementary figures1 ) .
we demonstrate below that these channels are suitable probes for detecting atomic - scale motional dynamics of surface ( ad)atoms . for comparison , fig .
[ fig : sensor]b displays the 2d - island of ( a ) before the nano - structuring step , i.e. exhibiting its natural compact equilibrium shape .
the left inset of fig.[fig : sensor]b displays a magnified view revealing the regular hexagonal ar atomic lattice of the 2d - islands with ar - ar distance of 0.39 nm , in agreement with previous studies @xcite . a single ar vacancy is labeled 1 . for comparison ,
the atomically resolved ag(111 ) lattice ( @xmath9 nm ) is shown in the right inset of fig .
[ fig : sensor]b . similar to a monolayer of ar on ag(111 ) @xcite , the equilibrium - shape islands are stable for > 12h during continuous imaging by dc - stm at sample bias voltage of @xmath5 to @xmath10v and tunneling current of 50200 pa .
more importantly here , also the motion sensors , i.e. the non - equilibrium 2d - islands ( fig .
[ fig : sensor]a ) , are found to be longterm stable against dc - imaging by stm at 5k ( see supplementary figures2 ) .
response of motion sensor to cw rf - excitation .
( a ) sensor island before rf - excitation imaged by stm ( @xmath2 nm@xmath3 , @xmath5v , 70pa ) .
( b)(f ) same sensor island as in ( a ) after successive cw rf - excitation ( 530mhz , @xmath11db ) for accumulating on - time , @xmath12 ( see labels ) .
( g , h ) dependence of the magnitude of sensor response on @xmath12 ; line : numerical fit , function given in red .
( h ) dependence on rf power @xmath13 ; line : numerical linear fit.,width=317 ] for investigating the impact of rf alternating electric charging and de - charging of the metal on its atomic surface structure , we have connected the ag(111 ) sample to the output of an rf generator in parallel to the dc sample - voltage source .
we set a fixed generator frequency to avoid possible effects of frequency - dependent damping of the rf circuitry and to guarantee a constant rf voltage amplitude at the sample surface for all experiments presented herein .
the frequency was @xmath14mhz , which means that the microwave in the ag sample kept at 5k is confined to a surface skin layer @xcite of thickness @xmath15 nm ( resistivity of ag at 5k is @xmath16 @xmath17 m @xcite and permeability @xmath18vs / am ) . in a first step , we have applied a continuous - wave ( cw ) rf - voltage to the sample for varying time spans @xmath12 . before and after each excitation , the motion sensor was imaged by dc - stm .
intriguingly , our motion sensors respond to the cw excitation with characteristic structural changes as evidenced in fig.[fig : response ] .
excitation for @xmath19min causes a gradual closing of the channels .
this is clearly seen by comparing the images of the motion sensors before ( a ) and after ( b ) the rf excitation .
moreover , the large island merges with the isolated small one on the right .
notice the defects labeled a and b being unaffected by the rf excitation .
the total size of the sensor ( area ) has remained approximately constant , indicating that no significant amount of ar atoms is added or subtracted during @xmath12 .
obviously , channel closing proceeds via directed diffusion ( displacement ) of ar atoms across the ag surface .
the perimeter - to - area ratio of the sensor island decreases monotonically , evidencing the non - random nature of the underlying process .
we emphasize that the sensors do not respond to dc tunneling . repeating the 1min cw excitation leads to a further closing of the channels ( fig .
[ fig : response]c ) , which seems to come to rest after a third excitation ( fig .
[ fig : response]d ) .
complete closure of the channels , however , is achieved after applying an additional 5-min ( fig .
[ fig : response]e ) and 10-min ( fig . [ fig : response]f ) cw excitation .
finally , the motion sensor adopts a compact equilibrium - like shape similar to original ar 2d - islands . by determining the total number of ar atoms ( area ) displaced during the on - time of the rf - voltage
, we have quantified the size of the sensor response .
starting from zero , with increasing @xmath12 the response increases and finally saturates at very long times ( fig .
[ fig : response]g ) . within the experimental range of our method
, we have found a nearly linear dependence of the sensor response on the microwave power ( fig .
[ fig : response]h ) .
it exhibits a low - power threshold of @xmath203dbm generator output power , corresponding to an rf - voltage amplitude at the sample of only a few mv zero - to - peak , considering the damping of our rf circuitry @xcite . in the experiments described so far ,
the stm tip was positioned over the sensor 2d - islands in tunnel contact ( typically @xmath5v , 100pa ) during the rf - excitation . to minimize stm - tip effects
, we carefully checked the tip state during the experiments and have repeated them with different tips ( tip formings ) .
it has turned out , however , that the position of the stm tip during rf excitation is irrelevant for the sensor response .
figure[fig : tip]a shows sensor islands with eight fabricated channels in both horizontal and vertical direction , marked by arrows .
after rf - excitation for @xmath21min all of them have responded ( fig.[fig : tip]b ) , although during the excitation the stm tip was placed over the pristine substrate several tens of nanometers away from the sensors ( tip position marked by cross ) .
apparently , the sensor response is based on a `` non - local '' mechanism , i.e. independent of the close - up range of the tip apex , and isotropic in the surface plane ( see fig.[fig : tip]b ) .
we have confirmed the sensor response up to a surface area of @xmath22nm@xmath3 by manual piezo control ( limited by the scan range of our lt - stm instrument ) .
sensor response is independent of stm tip position .
( a , b ) stm images of six sensor islands , marked by arrows , recorded before ( a ) and after ( b ) cw rf - excitation ( 530mhz , @xmath23min ) ; cross marks tunnel position of stm tip during excitation ( + 0.4v ) ; dashed lines mark radial distance from tip position .
( c - f ) difference images of sensor response obtained by subtracting stm images before and after rf - excitation ( @xmath23min ) at different rf - power levels and tunnel conditions ( see labels ) ; red ( blue ) color marks positive ( negative ) response , i.e. accumulation ( removal ) of ar atoms ; unchanged sensor area is plotted in yellow.,width=317 ] even more intriguing are experiments performed at non - tunneling conditions , where the stm tip was perpendicularly retracted by @xmath24200 nm away from the ag(111 ) surface for suppressing electron tunneling ; the local dc @xmath0-field between the stm tip apex and the sample is decreased by a factor of @xmath24200 . for better clarity the respective results , figs.[fig : tip]c f ,
are displayed as difference images , where red ( blue ) color marks sensor area where ar atoms have been accumulated ( removed ) by the rf - excitation .
the sensor response at non - tunneling conditions , fig.[fig : tip]c , is clearly revealed by the channel closing after @xmath23min ( rf power level was 4db above @xmath25 ) .
it is almost indistinguishable from the response at tunneling conditions ( fig.[fig : tip]d ) .
this finding clearly evidences that sensor response is independent of tunneling electrons as well as the magnitude of the dc @xmath0-field between tip and sample .
notice that edge diffusion is observed at the outermost ar atomic row of the sensor 2d - islands in all our experiments independent of rf excitation ( for details see supplementary information ) . repeating the experiments at a decreased power level of 3db below @xmath25 results in zero response ( no channel closing ) at both non - tunneling ( fig.[fig : tip]e ) and tunneling conditions ( fig.[fig : tip]f ) .
the respective decrease of power corresponds to a 50%-decrease of the rf @xmath0-field amplitude .
this result is indeed surprising : there is no response in ( f ) , although it has at least 100 times larger @xmath0-field compared to ( c ) .
obviously , sensor response depends on the rf - power level , but is uncorrelated with the strength of the rf @xmath0-field between stm tip and junction .
our findings therefore contradict a simple @xmath0-field effect .
notice that our argumentation holds independent of the precise value of the rf - voltage amplitude at the tunneling junction , which is not precisely known .
the results of the non - tunneling experiments clearly rule out ( local ) joule heating at the tunnel junction , which relies on the flow of electric current @xcite , as origin of the sensor response .
heating of the sample by microwave radiation is ruled out , because the sensor response happens in the ( reactive ) near - field of the sample surface , where emission of radiant energy is known to be negligible .
heating of the sample by absorption of electromagnetic energy in the cabling and sample crystal is ruled out because we observe negligible warming of the sample ( e.g. , only @xmath26k upon 5min of cw rf - excitation ) .
the presented experiments suggest that 2d - islands of ar on ag(111 ) act as motion sensors responsive to the dynamic reorganization of surface ( ad)atoms . the reorganization is caused by additional elementary diffusion processes of the sensor s atoms induced by the rf alternating electric charging and de - charging of the metal skin layer upon applying an rf - voltage to the sample .
the additional diffusion is absent at dc - voltage conditions at 5k , where only edge - diffusion is observed ( i.e. diffusion along the same atomic row starting off from a kink site ) .
edge diffusion alone can not explain the response , because channel filling requires ar atoms to move out of an edge row to form a new edge , one row in front of the old one .
we estimate the respective energy barrier of this additional diffusion step to be on the order of @xmath27mev , based on our observation that 60min of heating the sensors to 10k ( @xmath28mev ) causes a similarly large sensor response as 5min of 530 mhz rf - excitation with power of 4db above @xmath25 at 5k ( @xmath29mev ) .
still , fundamental questions remain : what causes the additional diffusion process ?
what is the role of the ag substrate atoms ? in the following we discuss the underlying physical mechanism .
sensor response to pulsed rf - excitation .
( a ) schematics illustrating the periodicity @xmath30 of periodic rf pulses .
( b e ) difference images of sensor response to rf excitation ( 530mhz , @xmath31db ) at non - tunneling conditions ; red ( blue ) color marks accumulation ( removal ) of ar atoms ; unchanged sensor area is plotted in yellow ; stm tip is positioned 500 nm away from the island center ; ( b ) after cw - excitation for 1min ; ( c)-(e ) after excitation by pulse train of @xmath32 periodic 50 ns - pulses ( equivalent of 1 min total rf on - time ) with different periodicity of 200 , 600 , and 1000 ns .
( f ) dependence of the magnitude of response on @xmath30 ; red line : numerical fit @xmath33.,width=317 ] to gain further insight , we have investigated the sensor response to pulsed rf - excitation .
we have obtained practically the same results for pulsed experiments at tunneling and non - tunneling conditions ; for brevity we show and discuss herein only the results at non - tunneling conditions .
we have applied pulse trains consisting of periodic 50ns - pulses of frequency 530mhz with different values of pulse period varying between @xmath34 and 1000ns ( fig.[fig : puls]a ) .
each pulse train contained the same total number of @xmath32 pulses , equivalent of a total on - time of 1min of the rf - excitation . compared to cw - excitation for 1min ( fig.[fig : puls]b ) , pulsed excitation yields a significantly smaller response ( fig.[fig : puls]c ) ; increasing the pulse period decreases the response further ( figs.[fig : puls]d , e ) .
this finding indicates that the dynamic processes underlying the sensor response exhibit a time constant @xmath35 with a value similar to the period of pulses applied . a quantitative evaluation of the @xmath30-dependence of the response is displayed in fig.[fig : puls]f .
numerical fitting an exponential function yields a value of @xmath36ns for the time constant of the sensor response .
this value is about six orders of magnitude larger than typical decay times of excited ( surface ) plasmons @xcite , surface - state electrons @xcite , and ( surface ) phonons @xcite on ag(111 ) .
it seems more likely that @xmath35 belongs to the ( collective ) mechanical excitations of ar atoms involved in the restructuring of the sensors .
similar decay times were observed for the collective vibrations of ensembles of small physisorbed molecules on au(111 ) @xcite .
our experimental results obviously rule out several processes as origin of the sensor response : tunnel current , the electric field between stm tip and sample , local joule heating , radiative heating and heating by absorption of electromagnetic energy in the cabling and/or sample crystal .
these findings indicate that sensor response is caused by processes in the ag sample induced by the charging and decharging at radio frequency . in other words ,
we identify the periodic - in - time deviation from charge neutrality in the skin layer of the sample as origin of the observed sensor response .
since the dynamics of electrons and atomic lattice are known to be closely related to each other on time scales of @xmath37ns , studied herein , a manifold of different effects is expected to occur simultaneously , contributing to sensor response : ( i ) electrostriction mechanically strains the surface atomic lattice @xcite affecting the bonding geometry ; reversible mechanical strain as large as 0.15% has been observed in charged nanoporous pt samples @xcite , corresponding to mechanical stress of about 1gpa .
hence , the application of an rf voltage is expected to enforce mechanical vibrations of the surface atomic lattice at the same frequency ( here 530mhz ) with amplitude @xmath38 , contributing to the measured dc tunnel current via the well - known relation @xmath39 .
notice , that the rf - induced change of sample surface height @xmath38 may be misinterpreted as `` rf - induced contribution to dc bias voltage '' , `` apparent shift of work function '' , or `` apparent smearing of @xmath40 '' ( at const-@xmath41 and const-@xmath4 conditions ) .
( ii ) excitation of acoustic surface plasmons ( asps ) , known to exist on ag(111 ) @xcite , is expected to influence adsorbate dynamics ( here : sensor response ) , since the decay of asps can generate ( surface ) phonons @xcite .
our sensors promise to facilitate future experimental studies on the rf - excitation of asps .
( iii ) electric polarization of the surface atoms induces repulsive electrostatic forces between neighboring atoms or even affects the van der waals - london dispersion forces of ar - ar as well as ar - ag atoms @xcite , facilitating enhanced ar diffusion . in conclusion
, we have demonstrated that nano - fabricated monolayer 2d - islands of ar on ag(111 ) at 5k act as experimental probes for detecting surface ( ad)atom mechanical motion and dynamic processes .
to showcase the strength of our method , we reveal a dynamic reorganization of surface ( ad)atoms at the ar / ag(111 ) interface , caused by radio - frequency charge - density oscillations and related electric fields induced by an external rf voltage .
our experimental method is expected to enable quantitative characterization of atomic structural dynamics with unprecedented detail , in general .
this is relevant , in particular , for the study of weakly bound systems with nanometer spatial- and nanosecond temporal resolution , including monolayer solids , surfaces , ( hetero)interfaces and 2d nanostructures . |
the _ systematic _ uncertainties in the determination of the ages , masses , and to a lesser extent extinction and metallicity of young extragalactic star clusters , based on broad - band imaging observations , but using a variety of analysis approaches are poorly understood .
our aim in this paper is to intercompare the results obtained from _ hubble space telescope ( hst ) _ observations of carefully selected young star cluster samples using a variety of commonly used modelling techniques , and characterise the major differences among the techniques .
we will do this specifically with respect to our own extensively tested multi - dimensional approach , which we will use as our main benchmark . at one time or another during its lifetime
, every galaxy will be affected by the external gravitational effects , however minor , exerted by its neighbour galaxies .
irrespective of the precise interaction geometry , the results of such interactions are often striking : depending on the available gas reservoir , violent star formation will ensue , frequently predominantly in the guise of active star _ cluster
_ formation ( e.g. , whitmore et al .
1999 , de grijs et al .
2001 , 2003a , b , and references therein ) .
thus , where the above scenario holds , the age distribution of a galaxy s star cluster population reflects its violent interaction history .
the study of young and intermediate - age star cluster systems in a variety of galactic environments out to @xmath3 mpc has become a major field in extragalactic astrophysics in recent years , significantly stimulated by the superb imaging quality of the _ hst_. one of the key diagnostic tools often utilised to predict the fate of such cluster systems is the cluster luminosity function ( clf ; based on broad - band imaging observations ) .
significant age spreads in young cluster systems which might still be undergoing active cluster formation affect the observed clf ( meurer 1995 , fritze v .
alvensleben 1998 , 1999 , de grijs et al .
2001 , 2003b ) .
this might , in fact , make an intrinsically log - normal clf appear as a power - law ( e.g. , miller et al .
1997 , fritze v .
alvensleben 1998 ) ; the exact shape of the intrinsic clf , whether a power law or a log - normal distribution , is still being debated ( e.g. , vesperini 2000 , 2001 , vs. fall & zhang 2001 ; see also lamers et al .
it is obviously very important to obtain accurate age estimates for the individual clusters within such a system and to correct the observed clf to a common age , before interpreting it as an _ intrinsic _ clf ( fritze v .
alvensleben 1999 , de grijs et al .
2001 , 2003b ) .
the metallicities of star clusters produced in the high - pressure environments of galaxy interactions , mergers and starbursts are an important discriminator against the old milky way - type globular clusters ( gcs ) , thought to be the oldest building blocks of most nearby spiral and elliptical galaxies .
they are expected to correspond to the interstellar medium ( ism ) abundances of the interacting / starburst galaxies , and are therefore most likely significantly more metal - rich than those of halo gcs in the milky way and other galaxies with old gc systems .
however , ism abundances span a considerable range among different galaxy types , from early - type spirals to dwarf irregulars ( e.g. , ferguson et al . 1998 ) , and may also exhibit significant radial gradients ( oey & kennicutt 1993 , zaritsky , kennicutt & huchra 1994 , richer & mccall 1995 ) . hence
, a considerable metallicity range may be expected for star clusters produced in interactions of various types of galaxies and even among the clusters formed within one global galaxy - wide starburst .
a significant increase of the ism abundance in massive gas - rich galaxies may occur during strong bursts ( fritze v .
alvensleben & gerhardt 1994 , their fig .
at the same time , typically within a few @xmath4 yr , some fraction of the gas enriched by dying first - generation burst stars may well be shock - compressed to cool fast enough to be built into later generations of stars or clusters produced in the ongoing burst . the same effect
may occur when multiple bursts occur in a series of close encounters between two or more galaxies before their final merger .
precise ( relative ) metallicity determinations for individual young star clusters are not only important to address these issues , but also for the correct derivation of ages from broad - band colours or spectral energy distributions ( seds ) .
dust extinction is often very important in young cluster systems .
in particular the youngest post - starburst galaxies and galaxies with ongoing starbursts often show strong and patchy dust structures .
for instance , the youngest clusters in the overlap region of the two galactic discs in the antennae galaxies are highly obscured in the optical and are best detected in near or mid - infrared observations ( mirabel et al .
1998 , mengel et al . 2001 ) . similarly , lanon et al .
( 2003 ) discovered one of the most massive young star clusters in m82 based on near - infrared ( nir ) _ hst _ observations ; at optical wavelengths the cluster is completely obscured .
older merger remnants like ngc 7252 or ngc 3921 seem to have blown their inner regions clear of all the gas and dust left over from intense star formation ( e.g. , schweizer et al .
extinction estimates toward individual clusters are therefore as important as individual metallicity estimates in order to obtain reliable ages and masses .
spectroscopy of individual massive young clusters , although feasible today with 8m - class telescopes for the nearest systems , is very time - consuming , since observations of large numbers of clusters are required to obtain statistically significant results .
multi - passband imaging is a very interesting and useful alternative , in particular if it includes coverage of nir and/or ultraviolet ( uv ) wavelengths ( e.g. , de grijs et al .
2003c , anders et al .
there are obviously limitations to the accuracy of the cluster parameters derived from broad - band imaging observations ( e.g. , de grijs et al .
2003b , c , anders et al .
2004b , bastian et al . 2004 ) , but the _ relative _ overall characteristics derived for the cluster _ populations _ as a whole appear to be relatively robust . in this paper
we assess the systematic uncertainties in age and mass determinations , and to a lesser extent also in extinction and metallicity determinations , for young star cluster systems based on the use of broad - band , integrated colours , employing a variety of independently developed methods to analyse extragalactic star clusters as so - called `` simple stellar populations '' ( ssps ) : star clusters are the simplest objects to model , since they approximate single - age , single - metallicity populations with a range of stellar masses .
stochastic sampling effects of the stellar initial mass function ( imf ) also affect star cluster properties , in particular for low - mass objects .
however , since they affect broad - band photometry to a smaller extent than spectroscopy , and because we are dealing here with high - mass clusters only , we will not include these effects in this paper .
our main aim in this paper is to intercompare the results obtained for sets of well - calibrated cluster seds using a variety of commonly used modelling techniques , specifically with respect to our own extensively tested multi - dimensional approach ( see section [ method5.sec ] ) . in order to determine the _ absolute _ systematic uncertainties caused by the intrinsic differences in the models and methods in use in the literature
, we distributed sets of broad - band star cluster photometry ( described in section [ data.sec ] ) to representatives of the various groups active in this field , and requested them to derive the cluster parameters using their specific methodology , wherever the data allowed this to be done .
the models and methods are described in section [ methods.sec ] ; we emphasize that most of the comparisons among methods done in this paper should be considered relative to the results obtained using the analysed method described in section [ method5.sec ] .
we summarise the results from applying our analysed approach to a large grid of artificial cluster data in section [ artdata.sec ] , in order to establish the theoretical benchmark for further comparisons among approaches . in section [ comparison.sec ]
we compare the overall , relative parameter distributions , while in section [ detailed.sec ] one - to - one comparisons between the various method+model combinations for the individual clusters in both of our samples are discussed ; the results from each of the methods used for both cluster samples are included in appendix a. we extend this discussion by considering the effects of converting the cluster photometry to different filter systems ( section [ filters.sec ] ) , and conclude the paper in section [ final.sec ] .
the field of stellar population modelling using extragalactic compact star cluster systems has undergone a major expansion since high - resolution , well - calibrated _ hst _ observations became available to the community .
the application of stellar population synthesis of galactic subcomponents has become almost trivial for galaxies within @xmath5 mpc , while star cluster population modelling is very well feasible out to @xmath6 mpc , at least for the brighter ( and therefore more massive ) clusters within a given cluster population ( see , e.g. , de grijs et al . [ 2003d ] and pasquali , de grijs & gallagher [ 2003 ] for examples toward and close to the distance limit ) .
ongoing and future _ hst _ programmes will continue to provide high - resolution uv - optical - nir imaging of large samples of galaxies out to these distances .
we therefore expect that modelling simple stellar populations and their broad - band seds will remain a key diagnostic tool for both the study of the evolution of extragalactic star cluster systems and their relation to milky way - type gcs , and for the analysis of galactic star formation and interaction histories .
therefore , we decided to focus our comparison of models and methods on _ hst_-based imaging data .
in addition , the calibration of _ hst _ measurements is well - understood and does , therefore , not introduce additional uncertainties as caused by , e.g. , temporal variations in the atmospheric transmission that one would have to deal with if ground - based observations were used .
we selected subsamples from large populations of young star cluster systems extensively studied in the literature , which we required to be among the brighter members of their respective populations ( thus minimising the observational uncertainties ) , as well as spanning a large age range ( based on preliminary analyses , as described below ) .
ideally , we would have preferred to select cluster samples for which both _ hst _
measurements in a minimum of four broad - band passbands could be obtained , as well as independently determined parameters from spectroscopic observations .
unfortunately , such data sets are yet not available , however .
on the other hand , schweizer , seitzer & brodie ( 2004 ) recently showed convincingly that spectroscopic age determinations are not necessarily better or more accurate than photometrically obtained ages , at least in the age range from @xmath7 myr . * ngc 3310 , a nearby spiral galaxy exhibiting dominant star cluster formation in a circumnuclear starburst ring .
this galaxy was covered by _
hst _ by the largest possible wavelenth range .
the full set of eight broad - band stmag magnitudes for the @xmath8 clusters in the galaxy s centre ( located in the starburst ring and outside of it ) , from f300w ( `` mid - uv '' ) to f205w were published and analysed in de grijs et al .
( 2003a , c ) .
we selected 17 of these clusters for the present analysis , all with well - determined , high - quality photometry in the entire set of available passbands ( see table [ ngc3310.tab ] ) . + [ cols="^,^,^,^,^,^,^,^,^ " , ] | high - resolution _ hubble space telescope ( hst ) _ imaging observations of star cluster systems provide a very interesting and useful alternative for stellar population analyses to spectroscopic studies with 8m - class telescopes . here
, we assess the systematic uncertainties in ( young ) cluster age , mass , and to a lesser extent extinction and metallicity determinations , based on broad - band imaging observations with the _
hst_. our aim here is to intercompare the results obtained using a variety of commonly used modelling techniques , specifically with respect to our own extensively tested multi - dimensional approach .
any significant differences among the resulting parameters are due to the details of the various , independently developed modelling techniques used , rather than to the stellar population models themselves . despite the model uncertainties and the selection effects inherent to most methods used , we find that the _ peaks _ in the relative age and mass distributions of a given young ( @xmath0 yr ) cluster system can be derived relatively robustly and consistently , to accuracies of @xmath1 and @xmath2 , respectively , assuming gaussian distributions in cluster ages and masses for reasons of simplicity .
the peaks in the relative mass distributions can be obtained with a higher degree of confidence than those in the relative age distributions , as exemplified by the smaller spread among the peak values of the mass distributions derived .
this implies that mass determinations are mostly insensitive to the approach adopted .
we reiterate that as extensive a wavelength coverage as possible is required to obtain robust and internally consistent age and mass estimates for the individual objects , with reasonable uncertainties .
finally , we conclude that the _ actual _ filter systems used for the observations should be used for constructing model colours , instead of using conversion equations , to achieve more accurate derivations of ages and masses .
methods : data analysis galaxies : spiral galaxies : starburst galaxies : star clusters |
the spatial power spectrum of the hi 21 cm intensity in the small magellanic cloud was obtained by stanimirovic ( 1999 ) .
interestingly , it is a power law over scales as large as that of the smc itself .
similar power laws have been observed by crovisier & dickey ( 1983 ) and by green ( 1993 ) in the galaxy .
the outstanding feature in the case of the smc is the large scale of the observed correlations .
the power laws signal underlying long range correlations in what looks like a field of random fluctuations of the intensity . for an optically thin medium along the line of sight , the 21 cm intensity is proportional to the column density
. therefore , the fluctuations in 21 cm intensity , represent fluctuations in density .
a natural interpretation of the observed power spectra is that the underlying correlations in density fluctuations are due to a turbulence in which velocity fluctuations , that are coupled to density fluctuations , give rise to the observed power laws .
the turbulence interpretation was suggested by goldman ( 2000 ) and stanimirovic & lazarian ( 2001 ) .
goldman ( 2000 ) suggested that this large scale turbulence was generated by instabilities in the bulk flows that resulted from the tidal interaction during the last close passage of the large magellanic cloud ( lmc ) @xmath0 ago ( gardiner & noguchi 1996 ) .
however , since the observations catch a snapshot of the intensity field and since the turbulence timescales are very long ( @xmath1gyr ) one can not rule out the possibility of a _
static _ correlated density field that reflects initial conditions . in the present paper
we propose a test to decide between these two alternatives .
staveley - smith ( 1997 ) observed 501 hi super shells in the smc .
the proposed test relies on the fact that the timescale and age of the turbulence ( if indeed there ) are typically @xmath2 orders of magnitude larger than the lifetimes of the super shells .
therefore , they have formed in the turbulent gas and their observed radial velocities should reflect the turbulent velocity field in the gas in which they where formed .
we wish to look at them as markers registering the ambient gas velocity .
if the radial velocity field exhibits spatial correlations consistent with the those of the turbulence , assumed as responsible for the 21 cm intensity spectra , it will strengthen the case for dynamical turbulence as the source of the hi intensity power spectrum .
we use the data of the 501 super shells reported in table 1 of staveley - smith ( 1997 ) . for each super shell
, the residual radial velocity was found by subtracting from the observed velocity the large scale best fit , up to a shear .
@xmath3 with @xmath4 where @xmath5 the coordinates of each shell @xmath6 are in units of pc and were obtained from the angular coordinates by adapting a distance of 60 kpc to the smc .
the velocities are in units of km / s .
the subtracted large scale velocity field is composed of a mean velocity and a shear .
the magnitude of the shear is consistent with values obtained by gardiner & noguchi ( 1996 ) .
we have computed the second order structure function and the autocorrelation for the residual velocity field along lines parallel to the coordinate axes .
interpolation was used to fit the discrete data along the lines to a continuous function .
the different lines yielded similar results . for simplicity
, homogeneous and isotropic velocity field is assumed . in this case , the structure function and the autocorrelation depend only on the distance between the two points , @xmath7 .
the structure function is @xmath8 similarly , the autocorrelation function is @xmath9 the angular brackets denote ensemble averaging .
assuming ergodicity , in addition to homogeneity and isotropy , ensemble averaging equals space averaging .
as stated above , we use averages over lines so that @xmath10 where @xmath11 is the length of the line . similarly , @xmath12 the results of a typical computation are presented in figures 1 - 2 .
figure 1 shows the structure function @xmath13 . for very small values of @xmath14 @xmath15 , for larger values of @xmath14
it varies as @xmath16 and then it saturates .
the index @xmath17 characterizes the inertial range of the turbulent velocity spectral function : @xmath18 . in kolmogorov turbulence characterizing incompressible fluid @xmath19 . in the case of turbulence in compressible medium @xmath20
this was also the value deduced by goldman ( 2000 ) on the basis of the 21 cm intensity power spectrum .
these two power lows are presented in figure 1 .
the precision of the data is not enough to decide between them , even though the @xmath20 line seems to follow better the slope of the computed structure function .
as function of scale in pc
. the thin lines have slopes @xmath21 and @xmath22 . the upper line has a slope of @xmath21 . ]
the autocorrelation function is shown in figure 2 .
it behaves as an autocorrelation function of a turbulent velocity rather than uncorrelated velocity fluctuations .
as function of scale in pc . ]
figure 3 presents the turbulence spectral function @xmath23 computed from the autocorrelation function .
the curve is noisy but a power law range is clear . also here
the turbulence spectral functions with @xmath19 and @xmath20 are plotted .
the two slopes are compatible with the computed spectral function , although @xmath20 seems preferable .
the wavenumber range shown corresponds to spatial scales between @xmath24 , which is in this case the length of the line @xmath11 , and @xmath25 .
higher wavenumbers correspond to spatial scales that are smaller than the average radius of the shells , and therefore the computed turbulence spectrum is not valid for these scales .
the results of the present work strengthen the case for the turbulence interpretation of the 21 cm power spectra of the smc .
the residual radial velocities of the super shells exhibit statistical spatial correlations expected from turbulence .
the turbulence spectrum and structure function are consistent with a kolmogorov spectrum , @xmath26 , and with that of incompressible turbulence , @xmath20 .
the latter seems preferable .
it equals the value deduced by goldman ( 2000 ) from the hi intensity fluctuations .
stanimirovic , s. , stavely - smith , l. , dickey , j. m. , sault , r. j. , and snowden , s. l. 1999 , mnras , 302 , 417 staveley - smith , l. , sault , r. j. , hatzidimitriou , d. , kesteven , m. j , and mcconnell , d. , 1997 , mnras , 289 , 225 stanimirovic , s. , and lazarian , a. 2001 , apj , 551 , l53 | the spatial power spectrum of the hi 21 cm intensity in the small magellanic cloud ( stanimirovic 1999 ) is a power law over scales as large as those of the smc itself .
it was interpreted as due to turbulence by goldman ( 2000 ) and by stanimirovic & lazarian ( 2001 ) .
the question is whether the power spectrum is indeed the result of a dynamical turbulence or is merely the result of a structured static density . in the turbulence interpretation of goldman ( 2000 )
the turbulence was generated by the tidal effects of the last close passage of the lmc about 0.2 gyr ago .
the turbulence time - scale was estimated by goldman to be 0.4 gyr , so the turbulence has not decayed yet .
staveley - smith ( 1997 ) observed in the smc about five hundreds of hi super shells . their age is more than an order of magnitude smaller than the turbulence age .
therefore , if the turbulence explanation holds , their observed radial velocities should reflect the turbulence in the gas in which they formed . in the present work we analyze the observed radial velocities of the super shells .
we find that the velocities indeed manifest the statistical spatial correlations expected from turbulence .
the turbulence spectrum is consistent with that obtained by goldman(2000 ) . |
the ionized gas in the galaxy has been traditionally associated with bright regions surrounding hot stars , called str@xmath2mgren spheres or classical h ii regions .
most h ii regions , however , are found only at low galactic latitudes with a scale height of @xmath3 pc , which is much smaller than that of the diffuse ionized gas ( dig ) or the warm ionized medium ( wim ) ( e.g. , * ? ? ?
in fact , the classical h ii regions contain only @xmath4 of ionized hydrogen in the galaxy , and the remaining @xmath5 resides in warm ( @xmath6 ) and diffuse ( @xmath7 ) regions .
the wim occupies approximately 20 - 30 % of the volume of @xmath8-thick , plane - parallel layer of our galaxy ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the wisconsin @xmath9 mapper ( wham ) northern sky survey has provided information on the distribution , kinematics , and other physical properties of the wim @xcite . the diffuse wim is now recognized as one of the major components of the interstellar media ( ism ) in both our galaxy and external galaxies ( e.g. , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) ( see also @xcite for a review ) .
the properties of the wim have been revealed through observations of dispersion measure ( dm ) @xmath10 and emission measure ( em ) @xmath11 where @xmath12 is the electron density and @xmath13 is the length along the line of sight ( los ) .
for instance , em / dm and dm@xmath14/em give estimations of the electron density and the extent of the wim , respectively ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in addition , the widths of probability density functions ( pdfs ) of dm and em are known to be related to the root - mean - square ( rms ) sonic mach number , @xmath15 , of the wim ( e.g. , * ? ? ?
* ; * ? ? ?
the wim is most likely to be in a state of turbulence , considering a very high reynolds number inferred from observed velocities and expected viscosity @xcite .
evidences for turbulence include the log - normal distributions of 1 ) ems from the h@xmath16 intensity in the wisconsin h@xmath16 mapper @xcite survey @xcite , 2 ) ems from the h@xmath16 intensity in external galaxies ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , and 3 ) column densities of the dig in our galaxy @xcite . the log - normality can be explained naturally by turbulent distribution of gas ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the best evidence for turbulence comes from the power spectrum presented in @xcite .
it is a composite power spectrum of electron density collected from observations of velocity fluctuations of the interstellar gas , rotation measures ( rms ) , dms , interstellar scintillations , and others .
the spectrum covers a huge range of @xmath17 cm .
the whole range of the spectrum is approximately fitted to the power spectrum of kolmogorov turbulence with slope @xmath18 .
the wim is permeated with magnetic fields .
measuring faraday rotation of polarized radio emissions against background sources has been the most popular method for exploring magnetic fields there .
the rotation of the plane of linearly polarized light is due to the birefringence of magneto - ionic medium .
the angle of rotation is given by @xmath19 where rm is the rotation measure defined by @xmath20 here , @xmath21 is the wavelength of polarized light and @xmath22 is the los magnetic field strength .
the units of rm , @xmath12 , @xmath23 and @xmath13 are rad m@xmath24 , @xmath25 , @xmath26 g , and pc , respectively .
rm / dm gives an estimation of the los magnetic field strength , weighted by the electron density , @xmath27 the method , for instance , has been used to reconstruct the large - scale magnetic field in our galaxy by many authors ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite obtained the distributions of rms along contiguous loss in the constellations of auriga and horologium in multi - frequency polarimetric observations with the westerbork synthesis radio telescope .
while the peak in the frequency distribution of those rms reflects the regular component of magnetic field , @xmath28 , the spread should measure the turbulent component .
this means that if a distribution of rm is observed , its spread provides another way to quantify the magnetic field in turbulent ionized media such as the wim .
recently , using numerical simulations , @xcite found that in turbulent media with @xmath0 , the width of the distribution of rm/@xmath29 ( @xmath29 is the average value of rms ) is rather tightly related to the strength of the regular field along the los , @xmath30 .
they suggested an empirical formula , which can be used to estimate @xmath30 when the distribution of rm/@xmath29 is available .
however , the mach number is not necessarily unity in the wim .
the turbulent velocity dispersion has a range of values @xmath31 km s@xmath32 @xcite , and the temperature also has a range values @xmath33 k ( e.g. , * ? ? ? * ) .
so in the wim , the turbulent flow motions , although not always having @xmath0 , are expected to be somewhere between mildly subsonic and mildly supersonic . as a matter of fact , several groups have suggested that the turbulent wim has @xmath34
( e.g. , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . in this paper
, we extend the work of @xcite by incorporating the mach number dependence .
specifically , we consider turbulent media with @xmath35 and 2 along with @xmath0 . in section 2
, we outline our simulations . in section 3 , we present a new relation among @xmath30 , rm/@xmath29 , and @xmath15 .
we also suggest a relation between @xmath15 and the dispersion of log - normal distribution of em , which can be used to estimate the mach number .
summary follows in section 4 .
we performed three - dimensional simulations using a code based on the total variation diminishing ( tvd ) scheme @xcite , by solving the following set of equations for isothermal , compressible magnetohydrodynamics ( mhds ) @xmath36
@xmath37 @xmath38 @xmath39 where @xmath40 is the isothermal sound speed .
turbulence was driven by imposing a solenoidal ( incompressible ) forcing , .
we followed the recipes of @xcite and @xcite for .
perturbations , satisfying @xmath41@xmath42 , were drawn from a gaussian random field in the fourier space of the wavevector .
the magnitude was determined by the top - hat power distribution in a narrow wave - number range of @xmath43 , where @xmath44 is the computational box size .
the perturbations were converted to quantities in the real space by fourier transform , and then added into the computational domain at every @xmath45 .
in contrast to the original recipes , we used a different seed number for realization of perturbations at every @xmath46 .
the amplitude of the forcing was fixed in such a way that the resulting input rate of the kinetic energy is a constant . initially a static , uniformly magnetized medium with density @xmath47 and magnetic field @xmath28 along the @xmath48-direction was assumed .
a periodic computational box with @xmath49 grid zones was used .
there are two parameters in the problem , the initial plasma beta , @xmath50 , and the rms sound mach number at the saturated stage of turbulence , @xmath51 . to cover the ranges of values expected in the wim of our galaxy
, we included the cases with @xmath52 0.1 , 1 , 10 and @xmath53 , 1 , 2 .
the values of @xmath54 were set by the initial condition . for @xmath15 ,
the amplitude of forcing was set by trial and error , so that @xmath15 became close to the predefined value at the saturated stage .
we present nine simulations in this paper .
the initial magnetic field strength is related to the dimensionless parameter , @xmath54 , by @xmath55 assuming that hydrogen is completely ionized , helium is neutral , and the number ratio of hydrogen to helium is 10 . if we take 8000 k and 0.03 @xmath25 as the representative values of temperature and electron density in the wim , the initial magnetic field strength in our simulations corresponds to 4.1 , 1.3 , 0.41 @xmath26 g for @xmath54 = 0.1 , 1 and 10 , respectively .
it covers the range of the regular magnetic field strength in our galaxy ( e.g. , * ? ? ?
* and references therein ) .
figure 1 shows the evolution of @xmath15 as a function of time in units of @xmath56 .
note that @xmath57 is the turbulent turnover time in our simulations .
the rms flow speed , @xmath58 , and so @xmath15 initially increase and then saturate .
saturation is reached around @xmath59 , as previously shown ( e.g. , * ? ? ?
we ran simulations up to @xmath60 .
we then took 11 snapshots ( black dots in figure 1 ) at the saturated stage and used them to calculating the quantities in following sections .
the errors below are the standard deviations of the 11 snapshot data .
the magnetic field strength estimated with rm using equation ( 5 ) would be unbiased , only if the correlation between @xmath22 and @xmath61 is null or weak , as noted in , e.g. , @xcite . both observations @xcite and numerical simulations ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) have indicated that the correlation would be negative and positive in subsonic and supersonic turbulent flows , respectively .
we calculated the correlation coefficients for our simulation data , @xmath62^{1/2 } \left[\sigma_{i , j , k}(\rho_{i , j , k}-\bar \rho)^2\right]^{1/2}},\ ] ] where @xmath63 and @xmath64 are the average values of @xmath65 and @xmath66 .
table 1 shows the resulting correlation coefficients for nine simulations . as in @xcite
, we have small values , @xmath67 , except for the case of @xmath68 and @xmath69 ; even in the case , we have @xmath70 .
the weak correlation means that the rm field strength in equation ( 5 ) should correctly represent the true magnetic field strength , as further discussed in @xcite .
the correlation in our simulations looks consistent to those of previous works .
for instance , we have @xmath71 for @xmath72 and @xmath73 , and @xcite reported a negative correlation for @xmath74 and @xmath75 .
we note that the correlation coefficients would depend not only on @xmath54 and @xmath15 , but also on the nature of forcing .
it is well known that the properties of turbulence , such as the density pdf and power spectrum , depend on whether the forcing is solenoidal or compressible ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
recently , for instance , @xcite reported a study of supersonic turbulence with solenoidal and compressive drivings where the density pdf was analyzed in details .
in general , compressible forcing results in stronger compression and so larger standard deviation of @xmath76 , @xmath77 , than the solenoidal forcing .
it tells that the correlation coefficient is expected to be different in turbulences driven by solenoidal and compressible forcings , and the results presented in the next section could be affected by the nature of forcing . in this paper , we consider only the turbulence with solenoidal forcing , leaving the effects of the nature of forcing as a future study .
@xcite demonstrated that the frequency distribution of rms in turbulence simulation is well fitted to the gaussian for the case of @xmath79 .
furthermore , @xmath30 is shown to be anti - correlated with the full width at half maximum ( fwhm ) of the frequency distribution of rm/@xmath29 , @xmath78 , as noted in introduction . from a physical point of view
, the broadening of the width of rm distribution is caused by fluctuating gas and magnetic field .
so it is easily expected that @xmath78 would also depend on @xmath15 . here , we further explore the possible relation among @xmath30 , @xmath78 , and @xmath15 based on the new simulations described in the previous section . to obtain the relation among @xmath30 , @xmath78 , and @xmath15
, we took five viewing angles , @xmath80 and @xmath81 , where @xmath82 is the angle between the regular magnetic field , @xmath83 , and the los ( @xmath84 ) .
we first calculated rm for the @xmath82 s in our nine simulations .
table 2 show @xmath85 , normalized with rm due to @xmath30 , @xmath86 , where @xmath87 is the number density of the background medium .
@xmath88 is close to unity indicating that @xmath85 is caused mostly by @xmath30 , except for @xmath89 where the contribution of perturbed @xmath90 is more important .
we then calculated the frequency distribution of rm/@xmath29 , @xmath91 , and @xmath78 .
for the gaussian distribution , the fwhm is related to the standard deviation , @xmath92 , as @xmath93 .
the resulting 45 fwhms are listed in table 3 .
figure 2 shows @xmath78 versus the los regular field strength , @xmath84 , for the 45 fwhms . here
, the magnetic field strength is for the representative values , @xmath94 k and @xmath95 @xmath25 .
@xmath78 is larger for larger @xmath54 and for larger @xmath82 , as already pointed in @xcite .
so @xmath78 is anti - correlated with @xmath30 . for each @xmath15 , we fitted the relation between @xmath30 and @xmath78 with dashed lines .
the broadening of the width of rm distribution is due to the fluctuating magnetic field and electron density .
so as expected , @xmath78 is larger for larger @xmath15 .
we got the relation among @xmath30 , @xmath78 , and @xmath15 by fitting the 45 fwhms in table 3 to a single formula , @xmath96 , where @xmath97 and @xmath98 are fitting parameters .
the best fit we found , @xmath91 , was used to calculate @xmath78 in this paper , while the log of it , @xmath99 , was used in @xcite .
it is because published observations mostly provide @xmath91 ( see the next section ) , and so it is easier to measure the fwhm of @xmath91 .
we found that the fwhm of @xmath91 is @xmath100 times smaller than the fwhm of @xmath99 . with the fwhm of @xmath99
, we would obtain the fitted relation that is consistent with that of ( * ? ? ?
* equation ( 3 ) ) . ] is @xmath101 where the errors indicate the fitting uncertainty .
again , @xmath30 is for @xmath94 k and @xmath95 @xmath25 , and scales as @xmath102 for other values of @xmath103 and @xmath104 ( see equation ( 10 ) ) .
figure 3 shows @xmath30 as a function of @xmath105 for fwhms in figure 2 along with the fitting , demonstrating the goodness of the fitting .
the empirical relation in equation ( 12 ) would provide a handy way to quantify the los regular field strength in regions where the mach number and the rm distribution have been obtained .
it is interesting to see that the relation is applied even to the case of fairly large viewing angle , @xmath89 .
however , it is clear that the relation should break down if @xmath106 and so @xmath107 .
so it can be applied only to regions with @xmath108 not too small .
the above relation can be approximately reproduced from @xmath109 for alfvnic perturbations , where @xmath110 is the alfvn speed . in the case
that the correlation between @xmath65 and @xmath66 is weak ( see section 2 ) , roughly @xmath111 , so @xmath112 .
also , @xmath113 and @xmath114 . combining these
, we have @xmath115 . of course
, the perturbations in our simulations are not totally alfvnic and the correlation between @xmath65 and @xmath66 is not completely null .
so we got the relation with the exponents which are somewhat different from unity .
to derive @xmath30 using equation ( 12 ) , it requires us to know not only the rm distribution , but also the mach number
. it would be handy if we have an independent way to determine the mach number . from numerical simulations for isothermal , hydrodynamic or mhd turbulence
, it is known that the density pdf is approximately fitted to the log - normal distribution , and its standard deviation , @xmath117 , increases as the rms mach number of turbulent flows increases @xcite .
it has been shown that @xmath117 could be related to @xmath15 as @xmath118 with @xmath119 for turbulence with solenoidal forcing and @xmath120 for turbulence with compressible forcing , regardless of the presence of the magnetic field ( see , e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
observationally , however , it is not trivial to get the distribution of volume density ; it is easier to measure the distribution of column density ( dm ) or em .
it has been argued that the column density and em follow the log - normal distribution too ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . here
, we look for a relation between the dispersion of em distribution and the mach number , which may be used to estimate @xmath15 .
we obtained the frequency distribution of @xmath121 for five angles , @xmath122 , in our nine simulations .
the distribution is fitted to the gaussian .
we then calculated the fwhm , @xmath116}$ ] , of the distribution .
table 4 lists the resulting 45 fwhms .
figure 4 shows @xmath116}$ ] versus @xmath15 for the 45 fwhms .
@xmath116}$ ] is most sensitive to @xmath15 , while its dependence on @xmath54 and @xmath82 is weaker .
the average values of @xmath116}$ ] are @xmath123 , @xmath124 , @xmath125 for @xmath15=0.5 , 1 , and 2 , respectively , where the average and standard deviation are taken for 165 data ( 5 @xmath126 11 snapshots ) .
they are also shown in figure 4 too .
our result is roughly consistent with that of @xcite , where the width was from simulations with @xmath127 grid zones .
filled hexagons in figure 4 plot @xmath128 ( converted from @xmath92 ) in their table 5 .
@xcite also argued that the dependence of @xmath116}$ ] on magnetic field is weak .
we fitted @xmath15 versus @xmath116}$ ] to a linear function , and the best fit we got is @xmath129 } - 0.13(\pm0.26).\ ] ] the above can be reproduced from @xmath130 . with @xmath131 for @xmath132 ( @xmath119 for solenoidal forcing ) and @xmath133}$ ] for @xmath134 , we have @xmath135}$ ] , which is close to our fitted relation . as @xmath130 ,
our relation should depend on forcing ; it is applicable only to the case of solenoidal forcing .
we also checked the correlation between rm and em .
both quantities involve the electron density along the los , and so a positive correlation is expected .
we calculated the correlation coefficients between rm and @xmath136 with two - dimensional , spatial distributions , for five viewing angles , @xmath122 , in nine simulations using a formula similar to that in equation ( 11 ) .
the resulting coefficients are listed in table 5 . for small @xmath54 s and @xmath82 s , that is , for the cases with sufficiently large @xmath30 s
, there is a quite strong , positive correlation , as expected .
the correlation , however , is weaker for larger @xmath54 s and @xmath82 s .
our result shows that the correlation is very week and can even become negative for @xmath137 .
@xcite found a relation between the magnetic field strength along the los , @xmath30 , and the fwhm of the frequency distribution of rm/@xmath29 , @xmath78 , for turbulent media of @xmath0 .
but the mach number in the wim is not necessarily unity . in this paper
, we incorporated the mach number dependence in the relation . for it , we performed three - dimensional simulations of isothermal , compressible mhd turbulence for @xmath15 = 0.5 , 1 , 2 and @xmath54 = 0.1 , 1 , 10 .
the parameters were chosen to cover the range of values expected in the wim of our galaxy . - from the frequency distribution of rm/@xmath29 which is well fitted with the gaussian , we calculated @xmath78 .
we suggest a relation among @xmath30 , @xmath78 , and @xmath15 , which is shown in equation ( 12 ) . - the frequency distribution of em is well fitted with the log - normal distribution .
we calculated the fwhm of the frequency distribution of @xmath121 , @xmath116}$ ] .
we suggest a relation between @xmath15 and @xmath116}$ ] , which is shown in equation ( 13 ) .
the relation in equation ( 12 ) would provide a handy way for a quick and rough estimation of @xmath30 in the turbulent wim regions where observations of rm distribution are available and @xmath15 is known . in the case
that @xmath15 is unknown , the relation in equation ( 13 ) could be used for an estimation of @xmath15 , if observations of em distribution are available .
the work of qw was supported by the nsfc ( grants 11143001 , 11103003 , 11133005 ) , the national basic research program of china ( 2009cb824800 ) , the doctoral program of higher education ( 20110142120037 ) , and the fundamental research funds for the central universities ( hust : 2011ts159 ) .
the work of jk was supported by national research foundation of korea through grant k20901001400 - 10b1300 - 07510 .
the work of dr was supported by a research fund of chungnam national university .
numerical simulations were performed by using a high performance computing cluster at the korea astronomy and space science institute .
armstrong , j. w. , rickett , b. j. , & spangler , s. r. 1995 , apj , 443 , 209 balsara , d. s. & kim , j. 2005 , apj , 634 , 390 beck , r. 2007 , a&a , 470 , 539 beck , r. , shukurov , a , sokoloff , d. , & wielebinski , r. 2003 , a&a , 411 , 99 benjamin , r. 1999 , in interstellar turbulence , ed .
j. franco & a. carraminana , 49 burkhart , blakesley , lazarian , a. , & gaensler , b. m. 2012 , apj , 749 , 145 burkhart , b. , falceta - goncalves , d. , kowal , g. , & lazarian , a. 2009 , apj , 693 , 250 berkhuijsen , e. m. & fletcher , a. 2008 , mnras , 390 , l19 berkhuijsen , e. m. , mitra , d. , & mueller , p. 2006
nachrichten , 327 , 82 collins , j. a. & rand , r. j. 2001 , apj , 551 , 57 cho , j. & lazarian , a. 2003 , mnras , 345 , 325 crutcher , r. m. 1999 , apj , 520 , 706 elmegreen , b. g. & scalo , j. 2004 , ara&a , 42 , 211 federrath , c. 2013 , mnras , 436 , 1245 federrath , c. & klessen , r. s. apj , 763 , 51 federrath , c. , klessen , r. s. , & schmidt , w. 2008 , apj , 688 , l79 federrath , c. , klessen , r. s. , & schmidt , w. 2009 , apj , 692 , 364 federrath , c. , roman - duval , j. , klessen , r. s. , schmidt , w. , & mac low , m .- m . 2010 , a&a , 512 , 81 frick , p. , stepanov , r. , shukurov , a. , & sokoloff , d. 2001 , mnras , 325 , 649 gaensler , b. m. , et al .
2011 , nature , 478 , 214 gaensler , b. m. , madsen , g. j. , chatterjee , s. , & mao , s. a. 2008 , pasa , 25 , 184 g@xmath138mez , g. c. , benjamin , r. a. , & cox , d. p. 2001 , aj , 122 908 haffner , l. m. , et al .
2009 , reviews of modern physics , 81 , 969 haffner , l. m. , reynolds , r. j. , tufte , s. l. , madsen , g. j. , jaehnig , k. p. , & percival , j. w. 2003 , apjs , 149 , 405 haffner , l. m. , reynolds , r. j. , & tufte , s. l. 1999 , apj , 523 , 223 han , j. l. , beck , r. , & berkhuijsen , e. m. 1998 , a&a , 335 , 1117 han , j. l. , manchester , r. n. , lyne , a. g. , qiao , g. j. , & van straten , w. 2006 , apj , 642 , 868 haverkorn , m. , katgert , p. , & de bruyn , a. g. 2003 , a&a , 403 , 1031 haverkorn , m. , katgert , p. , & de bruyn , a. g. 2004 , a&a , 427 , 169 hill , a. s. , benjamin , r. a. , kowal , g. , reynolds , r. j. , haffner , l. m. , & lazarian , a. 2008 , apj , 686 , 363 indrani , c. & deshpande , a. a. 1999 , new a , 4 , 33 kim , j. , ryu , d. , jones , t. w. , & hong , s. s. 1999 , apj , 514 , 506 kowal , g. , lazarian , a. , & beresnyak , a. 2007 , apj , 658 , 423 kritsuk , a. g. , norman , m. l. , padoan , p. , & wadner , r. 2007 , apj , 665 , 416 mac low , m .- m . , balsara , d. s. , kim , j. , & de avillez , m. a. 2005 , apj , 626 , 864 mac low , m .- m .
1999 , apj , 524 , 169 miller , s. t. & veilleux , s. 2003 , apj , 592 , 79 nordlund , a. k. & padoan , p. 1999 , in interstellar turbulence , ed .
j. franco & a. carraminana ( cambridge : cambridge university press ) , 218 ostriker , e. c. , gammie , c. f. , & stone , j. m. 1999 , apj , 513 , 259 ostriker , e. c. , stone , j. m. , & gammie , c. f. 2001 , apj , 546 , 980 padoan , p. , nordlund , a. , & jones , b. j. t. 1997 , mnras , 288 , 145 padoan , p. & nordlund , a. 1999 , apj , 526 , 279 passot , t. & vzquez - semadeni , e. 2003 , a&a , 398 , 845 rand , r. j. , kulkarni , s. r. , & hester , j. j. 1990 , apj , 352 , 1 reynolds , r. j. , haffner , l. m. , & lufte , s. l. 1999 , apj , 525 , l21 reynolds , r. j. 1991 , apj , 372 , l17 seon , k .- i .
2009 , apj , 703 , 1159 stone , j. m. , ostriker , e. c. , & gammie , c. f. 1998 , apj , 508 , l99 tabatabaei , f. s. , beck , r. , kr@xmath139gel , e. , krause , m. , berkhuijsen , e. m. , gordon , k. d. , & menten , k. m. 2007 , apjs , 475 , 133 tufte , s. l. , reynolds , r. j. , & haffner , l. m. 1999 , in interstellar turbulence , ed . j. franco & a. carraminana ( cambridge : cambridge univ . press ) , p27 vzquez - semadeni , e. 1994 , apj , 423 , 681 wu , q. , kim , j. , ryu , d. , cho , j. , & alexander , p. 2009 , apj , 705 , l86 c|ccc + & @xmath152=0.1 & @xmath152=1 & @xmath152=10 + & & @xmath35 & + @xmath153 & @xmath154 & @xmath155 & @xmath156 + @xmath157 & @xmath154 & @xmath154 & @xmath158 + @xmath159 & @xmath154 & @xmath160 & @xmath161 + @xmath162 & @xmath154 & @xmath163 & @xmath164 + @xmath165 & @xmath166 & @xmath167 & 0.02@xmath168 + & & @xmath0 & + @xmath153 & @xmath154 & @xmath169 & @xmath170 + @xmath157 & @xmath160 & @xmath169 & @xmath171 + @xmath159 & @xmath160 & @xmath172 & @xmath173 + @xmath162 & @xmath174 & @xmath171 & @xmath175 + @xmath165 & @xmath167 & @xmath176 & @xmath177 + & & @xmath178 & + @xmath153 & @xmath179 & @xmath171 & @xmath180 + @xmath157 & @xmath156 & @xmath173 & @xmath181 + @xmath159 & @xmath182 & @xmath183 & @xmath184 + @xmath162 & @xmath185 & @xmath186 & @xmath187 + @xmath165 & @xmath188 & @xmath189 & @xmath189 + c|ccc + & @xmath152=0.1 & @xmath152=1 & @xmath152=10 + & & @xmath35 & + @xmath153 & @xmath190 & @xmath191 & @xmath192 + @xmath157 & @xmath193 & @xmath194 & @xmath195 + @xmath159 & @xmath196 & @xmath197 & @xmath198 + @xmath162 & @xmath199 & @xmath200 & @xmath201 + @xmath165 & @xmath202 & @xmath203 & @xmath204 + & & @xmath0 & + @xmath153 & @xmath205 & @xmath206 & @xmath207 + @xmath157 & @xmath208 & @xmath209 & @xmath210 + @xmath159 & @xmath211 & @xmath212 & @xmath213 + @xmath162 & @xmath214 & @xmath215 & @xmath216 + @xmath165 & @xmath217 & @xmath218 & @xmath219 + & & @xmath178 & + @xmath153 & @xmath220 & @xmath221 & @xmath222 + @xmath157 & @xmath198 & @xmath223 & @xmath224 + @xmath159 & @xmath225 & @xmath226 & @xmath227 + @xmath162 & @xmath228 & @xmath229 & @xmath230 + @xmath165 & @xmath231 & @xmath232 & @xmath233 + c|ccc + & @xmath152=0.1 & @xmath152=1.0 & @xmath152=10.0 + & & @xmath35 & + @xmath153 & @xmath234 & @xmath235 & @xmath236 + @xmath157 & @xmath193 & @xmath237 & @xmath238 + @xmath159 & @xmath239 & @xmath239 & @xmath240 + @xmath162 & @xmath241 & @xmath238 & @xmath240 + @xmath242 & @xmath243 & @xmath244 & @xmath243 + & & @xmath0 & + @xmath153 & @xmath241 & @xmath245 & @xmath246 + @xmath157 & @xmath247 & @xmath248 & @xmath246 + @xmath159 & @xmath249 & @xmath250 & @xmath251 + @xmath162 & @xmath252 & @xmath211 & @xmath253 + @xmath242 & @xmath254 & @xmath246 & @xmath214 + & & @xmath178 & + @xmath153 & @xmath255 & @xmath256 & @xmath257 + @xmath157 & @xmath258 & @xmath259 & @xmath260 + @xmath159 & @xmath261 & @xmath262 & @xmath263 + @xmath162 & @xmath264 & @xmath265 & @xmath266 + @xmath242 & @xmath267 & @xmath268 & @xmath269 + l|ccc + & @xmath152=0.1 & @xmath152=1.0 & @xmath152=10.0 + & & @xmath35 & + @xmath153 & @xmath270 & @xmath271 & @xmath272 + @xmath157 & @xmath170 & @xmath273 & @xmath274 + @xmath159 & @xmath275 & @xmath276 & @xmath277 + @xmath162 & @xmath278 & @xmath279 & @xmath280 + @xmath242 & @xmath281 & @xmath282 & @xmath283 + & & @xmath0 & + @xmath153 & @xmath273 & @xmath284 & @xmath285 + @xmath157 & @xmath286 & @xmath287 & @xmath288 + @xmath159 & @xmath289 & @xmath290 & @xmath291 + @xmath162 & @xmath292 & @xmath293 & @xmath294 + @xmath242 & @xmath295 & @xmath296 & @xmath297 + & & @xmath178 & + @xmath153 & @xmath298 & @xmath299 & @xmath300 + @xmath157 & @xmath301 & @xmath257 & @xmath302 + @xmath159 & @xmath303 & @xmath304 & @xmath305 + @xmath162 & @xmath306 & @xmath307 & @xmath308 + @xmath242 & @xmath309 & @xmath310 & @xmath311 + -1.2 cm -1.8 cm 0.5 ( top ) , 1 ( middle ) , and 2 ( bottom ) as a function of time in units of @xmath56 .
the solid , dashed and dotted lines are for @xmath152= 0.1 , 1 and 10 , respectively .
black - dots indicate the epochs of 11 snapshots , whose data were used to calculate the quantities and standard deviations presented in this paper.,title="fig : " ] -1.5 cm -2 cm -0.4 cm vs. los regular magnetic field strength .
black , red , and blue symbols are for @xmath35 , 1 , and 2 , respectively .
dashed lines display their best fits .
circles , triangles , and squares are for @xmath312 , @xmath313 , and @xmath314 , respectively , for five viewing angles .
error bars show the standard deviation of 11 snapshot data.,title="fig : " ] -2.5 cm -2 cm -0.6 cm and @xmath15 . black circles , red triangles , and blue squares are for @xmath35 , 1 and 2 , respectively .
error bars show the standard deviation of 11 snapshot data .
solid line displays our best fit ( equation ( 12)).,title="fig : " ] -2.5 cm -2 cm -1 cm vs. rms mach number . open symbols and their error bars are the same as in figure 2 . filled green circles and
their error bars show the average and standard deviation of @xmath116}$ ] for each @xmath15 .
solid and dashed lines display the best fit and 1 @xmath92 envelope ( equation ( 13 ) ) .
filled black hexagons show the results of @xcite.,title="fig : " ] -2.5 cm | @xcite suggested an empirical relation between the magnetic field strength along the line of sight ( los ) and the dispersion of faraday rotation measure ( rm ) distribution in turbulent media with root - mean - square sonic mach number @xmath0 . in this paper , we extend the work by incorporating the mach number dependence .
media with @xmath1 are considered to cover the mach number range of the warm ionized medium ( wim ) of our galaxy .
three - dimensional , magnetohydrodynamic isothermal turbulence simulations with solenoidal forcing are used .
we suggest a new relation among the los magnetic field strength , the dispersion of rm distribution , and the mach number , which approximately represents the relation for alfvnic perturbations .
in addition , we suggest a relation between the mach number and the dispersion of log - normal distribution of emission measure ( em ) , which is basically the relation for the mach number and the density dispersion . the relations could be used for a quick and rough estimation of the los magnetic field strength in the turbulent wim .
ism : magnetic fields - methods : numerical - mhd - turbulence |
the development of optical lattice systems based on laser technology has renewed interest in strongly correlated lattice systems .
one of the most striking phenomena of the optical - lattice systems is the superfluid - mott insulator ( sf - mi ) phase transition ; the sf phase ( i.e. , the coherent - matter - wave phase ) emerges when the kinetic energy is larger enough compared with the on - site repulsive interaction .
otherwise , the mi phase , i.e. , the number - state phase without coherence emerges .
the low - lying excitations of these optical - lattice systems can be described by using the bose hubbard model .
the temperature of trapped - atom systems can be extremely low , and hence , we hereafter assume the ground states of the system .
spin degrees of freedom also play an important role in optical - lattice systems . in theory ,
lots of analytical and numerical studies have been performed for the spin-1 bose hubbard model @xcite , including rigorous results for a finite system @xcite . in the case of antiferromagnetic spin - spin interactions , the perturbative mean - field approximation ( pmfa ) @xcite indicates that when filling with an even number of bosons , the mi phase is considerably more stable against the sf phase than when filling with an odd number of bosons .
this conjecture has been found by density matrix renormalization group ( dmrg ) @xcite and quantum monte carlo ( qmc ) methods @xcite in one dimension ( 1d ) .
recently , qmc methods also confirmed that conjecture in a two - dimensional ( 2d ) square lattice @xcite .
another interesting property of the spin-1 bose hubbard model with antiferromagnetic interactions is the first - order phase transition : the sf - mi phase transition is of the first order in a part of the sf - mi phase diagram .
the first - order transition has also been studied by using the variational monte carlo @xcite and qmc @xcite methods in a 2d square lattice .
the qmc results indicate that the phase transition can be of the first order , which is consistent with mean - field ( mf ) analysis @xcite .
however , the first - order transition disappears for strong antiferromagnetic interactions ; a mf calculation similar to that of ref . @xcite and
the qmc study @xcite show that the first - order sf - mi transition from the mott lobe with two bosons per site disappears when @xmath1 and @xmath2 , respectively .
thus , we assume strong interactions where the sf - mi transition is of the second order . for the second - order sf - mi transition , the strong - coupling expansion of kinetic energy @xcite is excellent for obtaining the phase boundary .
this method has been applied to the spinless @xcite , extended @xcite , hardcore @xcite , and two - species models @xcite , and the results agree well with qmc results @xcite .
thus , in this study , we perform the strong - coupling expansion with the spin-1 bose hubbard model . in another publication @xcite
, we examined the case of hypercubic lattices . in this study
, we examine the triangular lattice and compare the results with those of a square lattice to clarify whether the lattice structure plays a key role for the sf - mi transition .
the triangular lattice is intriguing because it frustrates the spin systems or spinful fermi systems .
the rest of this paper is organized as follows : section ii briefly introduces the spin-1 bose hubbard model and the strong - coupling expansion .
section iii provides our results .
a summary of the results is given in section iv .
some long equations that result from the strong - coupling expansion are summarized in appendix a.
the spin-1 bose hubbard model is given by @xmath3 , where @xmath4.\end{aligned}\ ] ] here , @xmath5 and @xmath6 are the chemical potential and the hopping matrix element , respectively .
@xmath7 ( @xmath8 ) is the spin - independent ( spin - dependent ) interaction between bosons .
we assume repulsive ( @xmath9 ) and antiferromagnetic ( @xmath10 ) interaction .
@xmath11 ( @xmath12 ) annihilates ( creates ) a boson at site @xmath13 with spin - magnetic quantum number @xmath14 . @xmath15 ( @xmath16 ) is the number operator at site @xmath13 .
@xmath17 is the spin operator at site @xmath13 , where @xmath18 represents the spin-1 matrices . in this study
, we assume a tight - binding model with only nearest - neighbor hopping and @xmath19 expresses sets of adjacent sites @xmath13 and @xmath20 . when @xmath21 , the ground state is the mi phase with the lowest interaction energy .
the number @xmath22 of bosons per site is odd when @xmath23 , whereas it is even when @xmath24 .
the mi phase with an even number of bosons is @xmath25 here , @xmath26 implies the boson number @xmath22 , the spin magnitude @xmath27 , and the spin magnetic quantum number @xmath28 at site @xmath29 .
however , for the mi phase with an odd number of bosons per site , we define a nematic state with @xmath28 : @xmath30 because we assume antiferromagnetic interactions .
the dimerized state is degenerate with @xmath31 for @xmath32 and is considered to be the ground state for finite @xmath0 in 1d . therefore , the results based on @xmath31 are basically limited to 2d or larger dimensions .
next , we define the defect states by doping an extra particle or hole into @xmath33 and @xmath31 as follows : @xmath34,\label{he}\\ \psi^{\rm part}_{\rm odd}&=&\frac{1}{\sqrt{n } } \sum_{i}\big [ % f^{\rm part}_{{\rm odd},i } \otimes\prod_{k\ne i } |n_0,1,0\rangle_k\big],\label{po}\\ \psi^{\rm hole}_{\rm odd}&=&\frac{1}{\sqrt{n } } \sum_{i}\big [ % f^{\rm hole}_{{\rm odd},i } \otimes\prod_{k\ne i } |n_0,1,0\rangle_k\big ]
. \label{ho}\end{aligned}\ ] ] here , @xmath35 is the number of lattice sites .
we assume that these defect states can be regarded as the sf states doped with infinitesimal numbers of particles or holes . by applying the rayleigh
schrdinger perturbation theory to these mi and defect states , we obtain the energy of the mi state and that of the defect states up to the third order in @xmath0 .
the results for the energy per site of the mi state and of defect states in the square lattice are summarized in appendix a. the phase can be determined by these energies . specifically , if @xmath36 , then the phase is sf ( mi ) , where @xmath37 is the energy of the mi state and @xmath38 [ @xmath39 is the energy of the defect state with one extra particle ( hole ) . the sf
mi phase boundary is thus determined by @xmath40 or @xmath41 and for ( b ) @xmath42 in the square lattice .
the solid curves show the results up to third order in t. results up to first order ( second order ) in @xmath0 are also shown by the green dashed ( blue dot - dashed ) curve . ,
title="fig:",height=288 ] and for ( b ) @xmath42 in the square lattice .
the solid curves show the results up to third order in t. results up to first order ( second order ) in @xmath0 are also shown by the green dashed ( blue dot - dashed ) curve .
, title="fig:",height=288 ] + but for ( a ) @xmath43 and ( b ) @xmath42 in the triangular lattice . , title="fig:",height=288 ]
but for ( a ) @xmath43 and ( b ) @xmath42 in the triangular lattice . , title="fig:",height=288
] + figures [ ex.sq ] and [ ex.tr ] show the phase diagram obtained from the strong - coupling expansion for the square lattice and for the triangular lattice , respectively . in both lattices ,
the mi phase for even - boson filling is considerably more stable against the sf phase than for odd - boson filling , which is consistent with qmc studies for the square lattice and mf studies .
the area of the mi phase for even - boson ( odd - boson ) filling increases more ( decreases more ) for @xmath44 than for @xmath45 .
the convergence of the strong - coupling expansion from the first to the third order in @xmath0 is fairly good .
we also find that higher - order terms mostly render the sf phase more stable against the mi phase because the area of the mi phase mostly becomes smaller for higher - order expansions , which is similar to the case of spinless bose hubbard model . and for ( b ) @xmath42 obtained by the strong - coupling expansion ( solid curve ) and its chemical - potential fitting ( brown dashed curve ) in the square lattice .
the red dot - dashed curve shows the pmfa results . ,
title="fig:",height=288 ] and for ( b ) @xmath42 obtained by the strong - coupling expansion ( solid curve ) and its chemical - potential fitting ( brown dashed curve ) in the square lattice .
the red dot - dashed curve shows the pmfa results .
, title="fig:",height=288 ] + the phase - boundary curve obtained by the strong - coupling expansion up to the third order in @xmath0 has a cusp at the peak of the mott lobe .
however , the chemical potential in infinite order in @xmath0 should follow a power - law scaling near @xmath46 , which is the value of @xmath0 at the tip of the mott lobe . following ref .
@xcite to the spinless bose hubbard model in 2d
, we perform a chemical - potential fitting method which assumes @xmath47 with a critical exponent @xmath48 . here
@xmath49 and @xmath50 are the regular polynomial functions of @xmath0 . by using our strong - coupling expansion up to the third order in @xmath0
, we immediately determine @xmath51 , @xmath52 , @xmath53 , and @xmath54 by setting @xmath55/2 $ ] . in order to determine @xmath56 , @xmath57 , @xmath58 , and @xmath46 we compare the taylor expansion of @xmath0 in @xmath59 with @xmath60/2 $ ] .
figures [ fit.sq ] and [ fit.tr ] compare the phase - boundary curves obtained by the third - order strong - coupling expansion with those obtained by the chemical - potential fitting in the square lattice and in the triangular lattice , respectively .
the chemical - potential fitting makes the phase - boundary curves smooth and natural , as expected .
the results obtained by the pmfa are also presented in both figures .
but for ( a ) @xmath43 and ( b ) @xmath43 in the triangular lattice . , title="fig:",height=288 ] but for ( a ) @xmath43 and ( b ) @xmath43 in the triangular lattice .
, title="fig:",height=288 ] + for the square lattice , the qmc data show , for the @xmath61 mott lobe , @xmath62 when @xmath63 ( as per our interpretation of fig .
15 of ref .
@xcite ) . our third - order strong - coupling results and the chemical - potential fitting show @xmath64 when @xmath63 and @xmath65 when @xmath63 , respectively .
this shows that the chemical - potential fitting clearly improves the results obtained by the strong - coupling expansion up to third order in @xmath0 . on the other hand , @xmath66 is obtained by the pmfa when @xmath63 .
hence the results obtained by qmc methods are in between those obtained by the chemical - potential fitting and pmfa , at least in the case of the square lattice where qmc has been performed .
this seems to reflect the fact that the pmfa includes a part of the infinite - order terms in @xmath0 but neglects the quantum fluctuations that possibly stabilize the mi phase , whereas the chemical - potential fitting based on the strong - coupling expansion up to third order in @xmath0 includes the quantum fluctuations but does not directly include the higher - order terms in @xmath0 that possibly stabilize the sf phase . comparing the results obtained in the square lattice and those obtained in the triangular lattice , we find that @xmath46 is roughly @xmath67 times greater in the square lattice than that in the triangular lattice , because of the difference between the number of nearest - neighbor sites @xmath68 [ @xmath69 in the square ( triangular ) lattice ]
. however , there seems to be no significant qualitative difference between the results obtained in the square lattice and those obtained in the triangular lattice .
these results suggest that there the sf - mi transition has no significant dependence on the lattice structure .
in this paper , we present the strong - coupling expansion for the spin-1 bose hubbard model for the square and triangular lattices . in both lattices , we confirm that for filling with an even number of bosons , the mi phase is considerably more stable against the sf phase than for filling with an odd number of bosons .
we also fit the phase - boundary curves obtained by the strong - coupling expansion to the scaling form of the curves by using the chemical - potential fitting method .
the fitting curves smooth and natural as expected , and the values of @xmath46 are clearly improved .
overall , there has been no essential difference in the phase boundary curves for the square and triangular lattices .
this might be due to our assumption that the zeroth order in @xmath0 mi state is a nematic state .
however , qmc study @xcite of the square lattice indicates that the magnetic structure factor shows no trace of magnetic order anywhere in the phase diagram .
thus , in contrast to the usual antiferromagnetic spin systems , there might be no significant dependence on the lattice structure for spin-1 bosons in two or more dimensions .
the author acknowledges makoto yamashita and yuta toga for fruitful discussions about the spin-1 bose hubbard model .
by using @xmath33 and @xmath31 [ eqs .
( [ me ] ) and ( [ mo ] ) ] , the energy per site of the mi state in the square lattice is @xmath70 \label{oddenergy}\end{aligned}\ ] ] up to the third order in @xmath0 , and where @xmath71 is the number of nearest - neighbor sites in the square lattice .
the corresponding mi - state energy in the triangular lattice is @xmath72\nonumber\\ & & -\frac{2}{27}z_tt^3\big [ \frac{118}{125}\frac{(n_0 - 1)(n_0 + 4)(2n_0 + 3)}{(u_0 + 4u_2)^2 } + \frac{32}{25}\frac{(n_0 - 1)(n_0 + 4)(2n_0 + 3)}{(u_0 + 4u_2)(u_0+u_2)}\nonumber\\ & & + \frac{4}{25}\frac{(n_0 ^ 2 - 1)(9n_0 + 1)+(n_0 + 2)(n_0 + 4)(9n_0 + 26)}{(u_0+u_2)^2}\nonumber\\ & & + \frac{8}{5}\frac{(n_0 + 1)(n_0 + 2)(2n_0 + 3)}{(u_0+u_2)(u_0 - 2u_2 ) } + \frac{(n_0 + 1)(n_0 + 2)(2n_0 + 3)}{(u_0 - 2u_2)^2}\big ] \label{oddenergytri}\end{aligned}\ ] ] up to the third order in @xmath0 , and where @xmath73 is the number of nearest - neighbor sites in the triangular lattice . up to the second order in @xmath0 ,
the expressions for the mi energy is the same for both lattices except for the difference between @xmath74 and @xmath75 .
however , in contrast to the square lattice , the third - order terms in @xmath0 exist for the triangular lattice . the energies of the defect states can be obtained by using @xmath76 , @xmath77 , @xmath78 , and @xmath79 [ eqs .
( [ pe])([ho ] ) ] up to the third order in @xmath0 . for the square lattice , we obtain @xmath80\nonumber\\ & & -\frac{z_st^3}{27}n_0(n_0 + 3 ) \big\ { ( z_s-1)\big [ \frac{(2n_0 + 3)z_s-3(3n_0 + 8)}{(u_0 + 2u_2)^2}\nonumber\\ & & + \frac{2}{u_0 + 2u_2}\big ( 2\frac{n_0 + 5}{2u_0 + 3u_2}+\frac{n_0 + 2}{2u_0}\big )
+ \frac{4(n_0 + 3)}{3u_2(u_0 + 2u_2)}\big]\nonumber\\ & & -z_s\big [ 2\big(\frac{n_0
+ 5}{(2u_0 + 3u_2)^2}+\frac{n_0 + 3}{(3u_2)^2}\big ) + \frac{n_0 + 2}{(2u_0)^2}\big ] + \frac{4}{3u_2}\big ( \frac{1}{5}\frac{n_0 + 5}{2u_0 + 3u_2}+\frac{n_0 + 2}{2u_0}\big ) \big\ } , \label{evenpart}\end{aligned}\ ] ] @xmath81\nonumber\\ & & -\frac{z_st^3}{27}n_0(n_0 + 3 ) \big\ { ( z_s-1)\big [ \frac{(2n_0 + 3)z_s-3(3n_0 + 1)}{(u_0 + 2u_2)^2}\nonumber\\ & & + \frac{2}{u_0 + 2u_2}\big ( 2\frac{n_0 - 2}{2u_0 + 3u_2}+\frac{n_0 + 1}{2u_0}\big ) + \frac{4n_0}{3u_2(u_0 + 2u_2)}\big]\nonumber\\ & & -z_s\big [ 2\big(\frac{n_0 - 2}{(2u_0 + 3u_2)^2}+\frac{n_0}{(3u_2)^2}\big ) + \frac{n_0 + 1}{(2u_0)^2}\big ] + \frac{4}{3u_2}\big ( \frac{1}{5}\frac{n_0 - 2}{2u_0 + 3u_2}+\frac{n_0 + 1}{2u_0}\big ) \big\},\label{evenhole}\end{aligned}\ ] ] @xmath82\nonumber\\ & & -\frac{z_st^2}{9}(n_0 + 4)\big[2\frac{n_0 - 1}{2u_0 + 3u_2 } + \frac{n_0 + 2}{2u_0 } -\frac{68}{25}\frac{n_0 - 1}{u_0 + 4u_2 } -\frac{8}{5}\frac{n_0 + 2}{u_0+u_2}\big]\nonumber\\ & & -\frac{2}{45}z_s(2z_s+3)t^2\frac{(n_0 + 1)(n_0 + 4)}{3u_2}\nonumber\\ & & -\frac{z_s(z_s-1)t^3}{27}\frac{(n_0 + 1)(n_0 + 2)}{(u_0 - 2u_2)^2 } \big[(2n_0 + 3)z_s-(5n_0 + 6)\big]\nonumber\\ & & -\frac{4}{675}z_s(z_s-1)t^3\frac{n_0 + 1}{(u_0+u_2)^2 } \big[(n_0 - 1)(9n_0 + 1)z_s-2(17n_0 ^ 2 + 26n_0 + 32)\big]\nonumber\\ & & -\frac{z_s(z_s-1)^2t^3}{27}\frac{n_0 + 1}{u_0+u_2 } \big[\frac{32}{25}\frac{(n_0 - 1)(n_0 + 4)}{u_0 + 4u_2 }
+ \frac{8}{5}\frac{(n_0 + 2)(2n_0 + 3)}{u_0 - 2u_2}\big]\nonumber\\ & & -\frac{2}{27}z_s(z_s-1)t^3(n_0 + 1)(n_0 + 4 ) \big\{\frac{n_0 - 1}{2u_0 + 3u_2}\big[\frac{34}{25}\frac{1}{u_0 + 4u_2 } + \frac{4}{5}\frac{1}{u_0+u_2}\big]\nonumber\\ & & + \frac{n_0 + 2}{2u_0}\big[\frac{1}{u_0 - 2u_2 } + \frac{4}{5}\frac{1}{u_0+u_2}\big ] + \frac{1}{3u_2}\big[\frac{2}{25}(8z_s+9)\frac{n_0 - 1}{u_0 + 4u_2 } + \frac{4}{5}z_s\frac{n_0 + 2}{u_0+u_2}\big]\big\}\nonumber\\ & & -\frac{4}{135}z_s(2z_s+3)t^3\frac{(n_0 + 1)(n_0 + 4)}{3u_2 } \big[\frac{1}{5}\frac{n_0 - 1}{2u_0 + 3u_2}+\frac{n_0 + 2}{2u_0}\big]\nonumber\\ & & + \frac{2}{675}z_st^3\frac{(n_0 + 1)(n_0 + 4)}{(3u_2)^2 } \big[2(n_0 - 11)z_s^2 + 9(3n_0 + 7)z_s-9(n_0 + 4)\big]\nonumber\\ & & + \frac{z_st^3}{27}(n_0 + 1)(n_0 + 4)\big [ \frac{68}{25}(z_s-1)\frac{n_0 - 1}{(u_0 + 4u_2)^2}\nonumber\\ & & + 2z_s\frac{n_0 - 1}{(2u_0 + 3u_2)^2}+z_s\frac{n_0 + 2}{(2u_0)^2}\big ] , \label{oddpart}\end{aligned}\ ] ] @xmath83\nonumber\\ & & -\frac{z_st^2}{9}(n_0 - 1)\big[2\frac{n_0 + 4}{2u_0 + 3u_2 } + \frac{n_0 + 1}{2u_0 } -\frac{68}{25}\frac{n_0 + 4}{u_0 + 4u_2 }
-\frac{8}{5}\frac{n_0 + 1}{u_0+u_2}\big]\nonumber\\ & & -\frac{2}{45}z_s(2z_s+3)t^2\frac{(n_0 - 1)(n_0 + 2)}{3u_2}\nonumber\\ & & -\frac{z_s(z_s-1)t^3}{27}\frac{(n_0 + 1)(n_0 + 2)}{(u_0 - 2u_2)^2 } \big[(2n_0 + 3)z_s-(5n_0 + 9)\big]\nonumber\\ & & -\frac{4}{675}z_s(z_s-1)t^3\frac{n_0 + 2}{(u_0+u_2)^2 } \big[(n_0 + 4)(9n_0 + 26)z_s-2(17n_0 ^ 2 + 76n_0 + 107)\big]\nonumber\\ & & -\frac{z_s(z_s-1)^2t^3}{27}\frac{n_0 + 2}{u_0+u_2 } \big[\frac{32}{25}\frac{(n_0 - 1)(n_0 + 4)}{u_0 + 4u_2 } + \frac{8}{5}\frac{(n_0 + 1)(2n_0 + 3)}{u_0 - 2u_2}\big]\nonumber\\ & & -\frac{2}{27}z_s(z_s-1)t^3(n_0 - 1)(n_0 + 2 ) \big\{\frac{n_0 + 4}{2u_0 + 3u_2}\big[\frac{34}{25}\frac{1}{u_0 + 4u_2 } + \frac{4}{5}\frac{1}{u_0+u_2}\big]\nonumber\\ & & + \frac{n_0 + 1}{2u_0}\big[\frac{1}{u_0 - 2u_2 } + \frac{4}{5}\frac{1}{u_0+u_2}\big ] + \frac{1}{3u_2}\big[\frac{2}{25}(8z_s+9)\frac{n_0 + 4}{u_0 + 4u_2 } + \frac{4}{5}z_s\frac{n_0 + 1}{u_0+u_2}\big]\big\}\nonumber\\ & & -\frac{4}{135}z_s(2z_s+3)t^3\frac{(n_0 - 1)(n_0 + 2)}{3u_2 } \big[\frac{1}{5}\frac{n_0 + 4}{2u_0 + 3u_2}+\frac{n_0 + 1}{2u_0}\big]\nonumber\\ & & + \frac{2}{675}z_st^3\frac{(n_0 - 1)(n_0 + 2)}{(3u_2)^2 } \big[2(n_0 + 14)z_s^2 + 9(3n_0 + 2)z_s-9(n_0 - 1)\big]\nonumber\\ & & + \frac{z_st^3}{27}(n_0 - 1)(n_0 + 2)\big [ \frac{68}{25}(z_s-1)\frac{n_0 + 4}{(u_0 + 4u_2)^2}\nonumber\\ & & + 2z_s\frac{n_0 +
4}{(2u_0 + 3u_2)^2}+z_s\frac{n_0 + 1}{(2u_0)^2}\big ] .
\label{oddhole}\end{aligned}\ ] ] for the triangular lattice , we obtain @xmath84\nonumber\\ & & -\frac{z_tt^3}{27}n_0\big\ { \frac{n_0 + 3}{(u_0 + 2u_2)^2 } \big[(2n_0 + 3)z_t^2 - 2(4n_0 + 9)z_t-3(17n_0 + 32)\big]\nonumber\\ & & -2\frac{n_0 + 5}{(2u_0 + 3u_2)^2}\big[(n_0 + 3)z_t-2n_0\big ] -\frac{n_0 + 2}{(2u_0)^2}\big[(n_0 + 3)z_t-2n_0\big ] -2(z_t-2)\frac{(n_0 + 3)^2}{(3u_2)^2}\nonumber\\ & & + 2(z_t+1)\frac{n_0 + 3}{u_0 + 2u_2 } \big[2\frac{n_0 + 5}{2u_0 + 3u_2}+\frac{n_0 + 2}{2u_0}\big]\nonumber\\ & & + 4\frac{n_0 + 3}{3u_2}\big[\frac{(n_0 + 3)z+n_0 - 3}{u_0 + 2u_2 } + \frac{1}{5}\frac{n_0 + 5}{2u_0 + 3u_2}+\frac{n_0 + 2}{2u_0}\big]\big\ } \label{evenparttri}\end{aligned}\ ] ] @xmath85\nonumber\\ & & -\frac{z_tt^3}{27}(n_0
+ 3)\big\ { \frac{n_0}{(u_0 + 2u_2)^2 } \big[(2n_0 + 3)z_t^2 - 2(4n_0 + 3)z_t-3(17n_0 + 19)\big]\nonumber\\ & & -2\frac{n_0 - 2}{(2u_0 + 3u_2)^2}\big[n_0z_t-2(n_0 + 3)\big ] -\frac{n_0 + 1}{(2u_0)^2}\big[n_0z_t-2(n_0 + 3)\big ] -2(z_t-2)\frac{n_0 ^ 2}{(3u_2)^2}\nonumber\\ & & + 2(z_t+1)\frac{n_0}{u_0 + 2u_2 } \big[2\frac{n_0 - 2}{2u_0 + 3u_2}+\frac{n_0 + 1}{2u_0}\big]\nonumber\\ & & + 4\frac{n_0}{3u_2}\big[\frac{n_0z+n_0 +
6}{u_0 + 2u_2 } + \frac{1}{5}\frac{n_0 - 2}{2u_0 + 3u_2}+\frac{n_0 + 1}{2u_0}\big]\big\ } \label{evenholetri}\end{aligned}\ ] ] @xmath86\nonumber\\ & & -\frac{z_tt^2}{9}(n_0 + 4)\big[2\frac{n_0 - 1}{2u_0 + 3u_2 } + \frac{n_0 + 2}{2u_0 } -\frac{68}{25}\frac{n_0 - 1}{u_0 + 4u_2 } -\frac{8}{5}\frac{n_0 + 2}{u_0+u_2}\big]\nonumber\\ & & -\frac{2}{45}z_t(2z_t+3)t^2\frac{(n_0 + 1)(n_0 + 4)}{3u_2}\nonumber\\ & & + \frac{2z_tt^3}{3375}\frac{(n_0 - 1)(n_0 + 4)}{(u_0 + 4u_2)^2 } \big[85(n_0 + 1)z_t+2(524n_0 + 701)\big]\nonumber\\ & & + \frac{4}{675}\frac{z_tt^3}{(u_0+u_2)^2 } \big\{(n_0 ^ 2 - 1)\big[-(9n_0 + 1)z_t^2 + 4(7n_0 + 3)z_t+8(11n_0 + 4)\big]\nonumber\\ & & + ( n_0 + 2)(n_0 + 4)\big[5(n_0 + 1)z_t+2(37n_0 + 88)\big]\big\ } \nonumber\\ & & + \frac{z_tt^3}{27}\frac{(n_0 + 1)(n_0 + 2)}{(u_0 - 2u_2)^2 } \big[-(2n_0 + 3)z_t^2 + 2(3n_0 + 4)z_t+19n_0 + 26\big]\nonumber\\ & & + \frac{2z_tt^3}{675}\frac{(n_0 - 1)(n_0 + 4)}{(2u_0 + 3u_2)^2 }
\big[25(n_0 + 1)z_t-34(n_0 - 1)\big]\nonumber\\ & & + \frac{z_tt^3}{27}\frac{(n_0 + 2)(n_0 + 4)}{(2u_0)^2 } \big[(n_0 + 1)z_t-2(n_0 + 2)\big]\nonumber\\ & & + \frac{2z_tt^3}{675}\frac{(n_0 + 1)(n_0 + 4)}{(3u_2)^2 } \big[2(n_0 - 11)z_t^2 + 9(3n_0 + 7)z_t-27(n_0 + 4)\big]\nonumber\\ & & -\frac{32z_tt^3}{675}\frac{(n_0 - 1)(n_0 + 4)}{(u_0 + 4u_2)(u_0+u_2 ) } \big[(n_0 + 1)z_t^2 - 2(n_0 + 1)z_t-(13n_0 + 19)\big]\nonumber\\ & & -\frac{8}{135}z_t(z_t^2 - 2z_t-7)t^3\frac{(n_0 + 1)(n_0 + 2)(2n_0 + 3 ) } { ( u_0+u_2)(u_0 - 2u_2)}-\frac{16z_tt^3}{135}\frac{(n_0 - 1)(n_0 + 2)(n_0 + 4)}{2u_0(2u_0 + 3u_2)}\nonumber\\ & & -\frac{68z_tt^3}{675}\frac{(n_0 - 1)(n_0 + 4)}{(2u_0 + 3u_2)(u_0 + 4u_2 ) } \big[(n_0 + 1)z_t+n_0 + 7\big]\nonumber\\ & & -\frac{8}{135}z_t(z_t+1)t^3\frac{(n_0 ^ 2 - 1)(n_0 + 4)}{(2u_0 + 3u_2)(u_0+u_2 ) } -\frac{8z_tt^3}{135}\frac{(n_0 + 2)(n_0 + 4)}{2u_0(u_0+u_2 ) } \big[(n_0 + 1)z_t+n_0 + 7\big]\nonumber\\ & & -\frac{2}{27}z_t(z_t+1)t^3\frac{(n_0 + 1)(n_0 + 2)(n_0 + 4)}{2u_0(u_0 - 2u_2 ) } -\frac{4}{675}z_t(8z_t^2+z_t+9)t^3\frac{(n_0 ^ 2 - 1)(n_0 +
4)}{3u_2(u_0 + 4u_2)}\nonumber\\ & & -\frac{8}{135}z_t^2(z_t-1)t^3\frac{(n_0 + 1)(n_0 + 2)(n_0 + 4)}{3u_2(u_0+u_2 ) } -\frac{4}{675}z_t(2z_t+3)t^3\frac{(n_0 ^ 2 - 1)(n_0 + 4)}{3u_2(2u_0 + 3u_2)}\nonumber\\ & & -\frac{4}{135}z_t(2z_t+3)t^3\frac{(n_0 + 1)(n_0 + 2)(n_0 + 4)}{3u_2\cdot2u_0 } \label{oddparttri}\end{aligned}\ ] ] @xmath87\nonumber\\ & & -\frac{z_tt^2}{9}(n_0 - 1)\big[2\frac{n_0 + 4}{2u_0 + 3u_2 } + \frac{n_0 + 1}{2u_0 } -\frac{68}{25}\frac{n_0 + 4}{u_0 + 4u_2 } -\frac{8}{5}\frac{n_0 + 1}{u_0+u_2}\big]\nonumber\\ & & -\frac{2}{45}z_t(2z_t+3)t^2\frac{(n_0 - 1)(n_0 + 2)}{3u_2}\nonumber\\ & & + \frac{2z_tt^3}{3375}\frac{(n_0 - 1)(n_0 + 4)}{(u_0 + 4u_2)^2 }
\big[85(n_0 + 2)z_t+2(524n_0 + 871)\big]\nonumber\\ & & + \frac{4}{675}\frac{z_tt^3}{(u_0+u_2)^2 } \big\{(n_0 + 2)(n_0 + 4)\big[-(9n_0 + 26)z_t^2 + 4(7n_0 + 18)z_t+8(11n_0 + 29)\big]\nonumber\\ & & + ( n_0 ^ 2 - 1)\big[5(n_0 + 2)z_t+2(37n_0 + 23)\big]\big\ } \nonumber\\ & & + \frac{z_tt^3}{27}\frac{(n_0 + 1)(n_0 + 2)}{(u_0 - 2u_2)^2 } \big[-(2n_0 + 3)z_t^2 + 2(3n_0 + 5)z_t+19n_0 + 31\big]\nonumber\\ & & + \frac{2z_tt^3}{675}\frac{(n_0 - 1)(n_0 + 4)}{(2u_0 + 3u_2)^2 } \big[25(n_0 + 2)z_t-34(n_0 + 4)\big]\nonumber\\ & & + \frac{z_tt^3}{27}\frac{n_0 ^ 2 - 1}{(2u_0)^2 } \big[(n_0 + 2)z_t-2(n_0 + 1)\big]\nonumber\\ & & + \frac{2z_tt^3}{675}\frac{(n_0 - 1)(n_0 + 2)}{(3u_2)^2 } \big[2(n_0 + 14)z_t^2 + 9(3n_0 + 2)z_t-27(n_0 - 1)\big]\nonumber\\ & & -\frac{32z_tt^3}{675}\frac{(n_0 - 1)(n_0 + 4)}{(u_0 + 4u_2)(u_0+u_2 ) } \big[(n_0 + 2)z_t^2 - 2(n_0 + 2)z_t-(13n_0 + 20)\big]\nonumber\\ & & -\frac{8}{135}z_t(z_t^2 - 2z_t-7)t^3\frac{(n_0 + 1)(n_0 + 2)(2n_0 + 3 ) } { ( u_0+u_2)(u_0 - 2u_2)}-\frac{16z_tt^3}{135}\frac{(n_0 ^ 2 - 1)(n_0 + 4)}{2u_0(2u_0 + 3u_2)}\nonumber\\ & & -\frac{68z_tt^3}{675}\frac{(n_0 - 1)(n_0 + 4)}{(2u_0 + 3u_2)(u_0 + 4u_2 ) } \big[(n_0 + 2)z_t+n_0 - 4\big]\nonumber\\ & & -\frac{8}{135}z_t(z_t+1)t^3\frac{(n_0 - 1)(n_0 + 2)(n_0 + 4)}{(2u_0 + 3u_2)(u_0+u_2 ) } -\frac{8z_tt^3}{135}\frac{n_0 ^ 2 - 1}{2u_0(u_0+u_2 ) } \big[(n_0 + 2)z_t+n_0 - 4\big]\nonumber\\ & & -\frac{2}{27}z_t(z_t+1)t^3\frac{(n_0 ^ 2 - 1)(n_0 + 2)}{2u_0(u_0 - 2u_2 ) } -\frac{4}{675}z_t(8z_t^2+z_t+9)t^3 \frac{(n_0 - 1)(n_0 + 2)(n_0 + 4)}{3u_2(u_0 + 4u_2)}\nonumber\\ & & -\frac{8}{135}z_t^2(z_t-1)t^3 \frac{(n_0 ^ 2 - 1)(n_0 + 2)}{3u_2(u_0+u_2 ) } -\frac{4}{675}z_t(2z_t+3)t^3 \frac{(n_0 - 1)(n_0 + 2)(n_0 + 4)}{3u_2(2u_0 + 3u_2)}\nonumber\\ & & -\frac{4}{135}z_t(2z_t+3)t^3\frac{(n_0 ^ 2 - 1)(n_0 + 2)}{3u_2\cdot2u_0 } \label{oddholetri}\end{aligned}\ ] ] as for the mi energy , the expressions for the mi energy up to the second order in @xmath0 are the same for both lattices except for the difference between @xmath74 and @xmath75 but the third - order terms in @xmath0 are different from each lattice .
this difference between the third - order terms originates from particles or holes that can ( can not ) return to their original site through three hopping processes in the triangular ( square ) lattice .
10 stamper - kurn d m and ueda m 2013 _ rev .
phys . _ * 85 * 1191 ; lewenstein m , sanpera a and ahufinger v 2012 _ ultracold atoms in optical lattices : simulating quantum many - body systems _
( oxford univ .
katsura h and tasaki h 2013 _ phys .
lett . _ * 110 * 130405 tsuchiya s , kurihara s and kimura t 2004 _ phys .
a _ * 70 * 043628 rizzi m , rossini d , de chiara g , montangero s and fazio r 2005 _ phys .
lett . _ * 95 * 240404 batrouni g g , rousseau v g and scalettar r t 2009 _ phys .
lett . _ * 102 * 140402 apaja v and syljusen o f 2006 _ phys .
a _ * 74 * 035601 toga y , tsuchiura h , yamashita m , inaba k and yokoyama h 2012 _ j. phys .
jpn . _ * 81 * 063001 de forges de parny l , hbert f , rousseau v g and batrouni g g arxiv:1306.5515 kimura t , tsuchiya s and kurihara s 2005 _ phys . rev . lett_. * 94 * 110403 krutitsky k v and graham r 2004 _ phys . rev .
a _ * 70 * 063610 krutitsky k v , timmer m and graham r 2005 _ phys .
rev . a _ * 71 * 033623 freericks j k and monien h 1996 _ phys .
b _ * 53 * 2691 ; freericks j k and monien h 1994 _ europhys . lett . _ * 26 * , 545 freericks j k , krishnamurthy h r , kato y , kawashima n and trivedi n 2009 _ phys .
a _ * 79 * 053631 iskin m and freericks j k 209 _ phys .
a _ * 79 * 053634 hen i , iskin m and rigol m 2010 _ phys .
* 81 * 064503 iskin m 2010 _ phys . rev .
a _ * 82 * 033630 kimura t 2013 _ phys .
a _ * 87 * 043624 | we examine the superfluid - mott insulator ( sf - mi ) transition of antiferromagnetically interacting spin-1 bosons trapped in a square or triangular optical lattice . we perform a strong - coupling expansion up to the third order in the transfer integral @xmath0 between the nearest - neighbor lattices .
as expected from previous studies , an mi phase with an even number of bosons is considerably more stable against the sf phase than it is with an odd number of bosons .
results for the triangular lattice are similar to those for the square lattice , which suggests that the lattice geometry does not strongly affect the stability of the mi phase against the sf phase . |
let us briefly motivate our interest in determining the spectral type of a self - adjoint operator .
let @xmath0 be a probability measure on the real line @xmath1 .
it is well - known that this measure has a lebesgue decomposition @xmath2 , where @xmath3 is absolutely continuous with respect to the lebesgue measure on @xmath1 , @xmath4 is singular with respect to lebesgue measure , and has no atomic part , i.e. @xmath5 for all @xmath6 , and @xmath7 is purely atomic .
this decomposition of a probability measure has important applications in the theory of a self - adjoint operator @xmath8 on a ( separable ) hilbert space @xmath9 . associated with @xmath8 is the spectral measure @xmath10 .
the spectral theorem states that we have @xmath11 if @xmath12 , then @xmath13 is a probability measure , which is supported on the spectrum @xmath14 of @xmath8 .
the lebesgue decomposition of probability measures leads to an orthogonal decomposition of the hilbert space @xmath15 each subspace is the closure of vectors @xmath16 , such that @xmath17 is purely absolutely continuous , etc .
the subspaces reduce the operator @xmath8 , such that @xmath18 is a self - adjoint operator on @xmath19 , etc . in the case of absolutely continuous and singular continuous parts
, one defines the corresponding parts of the spectrum to be those of the restricted operators . in the point spectrum case one
usually takes @xmath20 to be the set of eigenvalues of @xmath8 , in order to handle the case , when the operator has a dense set of eigenvalues .
the spectrum of the operator restricted to @xmath21 is then the closure of this set .
the spectral types of an operator @xmath8 , which is the hamiltonian of a quantum mechanical system , is related to the dynamics of the system , although the relation is by no means simple .
the relation comes from the representation of the time evolution operator@xmath22 as @xmath23 in some quantum mechanical systems ( e.g. atoms and molecules ) the absolutely continuous part is related to the scattering states , since @xmath24 tends to zero for @xmath25 ( a consequence of the riemann
lebesgue lemma ) , and the eigenvalues of @xmath8 are related to the bound states . in many of these systems
one expects that the singular continuous component is absent , and many techniques have been developed to prove this type of result . in solid state physics
the situation is somewhat different , and here one has a wider variety of spectral types .
these applications have motivated us to seek new criteria for identifying the spectral type of a self - adjoint operator .
the components of a probability measure can be identified via a transform of the measure .
two of these are well - known , viz .
the fourier transform and the borel transform . in this paper
we address the question of identifying the components using a more general transform .
we give results using a general approximate identity , and an associated continuous wavelet transform .
concerning the literature , the connection between an approximate identity and the continuous wavelet transform was discussed by holschneider @xcite , while wavelet coefficients of fractal measures were studied by strichartz @xcite . in the theory of self - adjoint operators finer decomposition of spectra with respect to hausdorff measures
was first used by last @xcite and general criteria for recovering a measure from its borel transform was done by simon @xcite .
we need to introduce conditions on our function @xmath26 .
several of these can be relaxed in some of the results .
we use the standard notation @xmath27 .
[ a1.1 ] assume that @xmath28 , @xmath29 , @xmath26 is even , and there exist @xmath30 and @xmath31 , such that @xmath32 and assume that @xmath33 . in the sequel
we always impose this assumption on @xmath26 .
we introduce the notation @xmath34 in particular , the family @xmath35 is an approximate identity .
let @xmath0 be a probability measure on @xmath36 in what follows , with lebesgue decomposition @xmath37 .
let @xmath38 be a function .
we recall that the convolution @xmath39 is defined , when the integral converges . since @xmath26 is bounded , the convolution @xmath40 is defined for all @xmath41 . for @xmath42
we define @xmath43 whenever the limit on the right - hand side exists .
we can now state the results .
we first give results based on @xmath44 and @xmath45 , and then on an associated continuous wavelet transform .
[ thm1 ] let @xmath0 be a probability measure .
then we have as follows : 1 .
let @xmath26 satisfy assumption [ a1.1 ] .
then for every continuous function @xmath38 of compact support , the following is valid .
@xmath46 2 .
3 . assume @xmath48 and @xmath49 finite .
then we have @xmath50 where @xmath51 .
@xmath52 [ thm2 ] let @xmath0 be a probability measure . then for any bounded interval @xmath53 the following are valid . 1 .
let @xmath54 then @xmath55 & \hskip -1.2pc \quad\ , = c \left(\sum_{x \in ( c , d ) } \mu(\{x\})^2 + \frac{1}{2 } [ \mu(\{c\})^2 + \mu(\{d\})^2 ] \right).\end{aligned}\ ] ] 2 . for @xmath56 , we have @xmath57 this theorem has the following corollary .
@xmath52 [ cor1 ] let @xmath0 be a probability measure .
then we have the following results : 1 .
@xmath0 has no point part in @xmath58 $ ] , if and only if @xmath59 2 . if @xmath0 has no absolutely continuous part in @xmath53 , if and only if for some @xmath60 , @xmath61 now to state the results in terms of the continuous wavelet transform , we introduce @xmath62 under assumption [ a1.1 ] we clearly have @xmath63 from the assumption .
integration by parts and eq .
imply that @xmath64 satisfies the admissibility condition for a continuous wavelet , i.e. @xmath65 . thus we can define the continuous wavelet transform of a probability measure @xmath0 as @xmath66 the connection between the approximate identity and this transform is @xmath67 this result follows from @xmath68 and the definitions .
we have the following analogue of theorem [ thm1 ] . [ thm3 ]
let @xmath0 be a probability measure .
then we have the following results : 1 .
we have @xmath69 2 .
let @xmath70 .
assume that @xmath71 exists .
then @xmath72 where @xmath73 was defined in theorem [ thm1 ] .
we note that for @xmath74 we can replace @xmath75 by @xmath76 for any @xmath77 ( see the proof of the theorem ) .
we also have the following analogue of theorem [ thm2](1 ) . [ thm4 ]
let @xmath0 be a probability measure .
then for any bounded interval @xmath53 we have the following result .
let @xmath78 then we have @xmath79 & \quad\,= c_h\left(\sum_{x\in(c , d)}\mu(\{x\})^2 + \frac{1}{2}(\mu(\{c\})^2 + \mu(\{d\})^2)\right).\end{aligned}\ ] ] even when the quantity @xmath49 does not exist ,
it is possible to say something on the wavelet transforms , to cover the cases of measures which are not supported on the sets where such limits exist . set @xmath80 then we have the following theorem . [ thm5 ]
let @xmath0 be a probability measure , and let @xmath26 satisfy assumption [ a1.1 ] .
then @xmath81 is finite for any @xmath82 , whenever @xmath83 is finite for the same @xmath82 , and , if @xmath26 is non - negative , they are both finite or both infinite .
[ rem - new-1 ] the above theorem implies that if @xmath84 , then there is no singular part of @xmath0 supported in @xmath85 . finally as
an application of the above theorems we consider @xmath86 to be a separable hilbert space and @xmath87 a self - adjoint operator .
then we have the following theorem . [ thm6 ]
suppose @xmath87 is a self - adjoint operator on @xmath86 .
consider a function @xmath26 satisfying assumption [ a1.1 ]
. then 1 .
@xmath88 is in the point spectrum of @xmath87 , if for some @xmath89 , @xmath90 2 .
let @xmath91 be a borel set of positive lebesgue measure .
then @xmath92 , if for some @xmath93 , @xmath94 , @xmath95 3 .
the point spectrum of @xmath87 in @xmath85 is empty , if and only if for some orthonormal basis @xmath96 , of @xmath86 , one has for every @xmath97 , @xmath98 4 .
the absolutely continuous spectrum of @xmath87 in @xmath85 is empty , if and only if for some orthonormal basis @xmath96 of @xmath86 , one has for every @xmath97 and some @xmath56 , @xmath99
throughout the computations below the letter @xmath100 denote a constant , whose value may vary from line to line .
part ( 1 ) : since @xmath38 is a continuous function of compact support and @xmath44 is bounded for each @xmath101 , @xmath102 is absolutely integrable and the integral is uniformly bounded in @xmath103 .
therefore , by an application of fubini , a change of variable @xmath104 and dominated convergence theorem , in that order , it follows that @xmath105 & = \lim_{a\rightarrow 0 } \int \hbox{d}\mu(y ) \int f(x ) \tilde{\psi}_a(x - y ) \hbox{d}x \\[.3pc ] & = \lim_{a\rightarrow 0 } \int \hbox{d}\mu(y ) \int f(ax+y ) \psi(x ) \hbox{d}x \\[.3pc ] & = \int \hbox{d}\mu(y ) \int \left(\lim_{a\rightarrow 0 } f(ax+y)\right ) \psi(x ) \hbox{d}x \\[.3pc ] & = \int f(y)\hbox{d}\mu(y ) \cdot \int \psi(x ) \hbox{d}x.\end{aligned}\ ] ] part ( 2 ) : this is a direct consequence of the definition of the integral noting pointwise that we have @xmath106 we also need to use the dominated convergence theorem to interchange the limit and the integral .
part ( 3 ) : let @xmath107 denote the distribution function of @xmath0
. then we have @xmath108 & = \frac{1}{a^\alpha}\int_{\mathbf{r}}\psi^\prime ( y ) \phi_\mu(x - ay ) \hbox{d}y\nonumber\\[.3pc ] & = - \int_0^\infty \!\!\psi^\prime ( y ) ( 2y)^\alpha \frac{\phi_\mu(x+ay ) - \phi_\mu(x - ay)}{(2ay)^\alpha } \hbox{d}y , \label{eqnnt}\end{aligned}\ ] ] where in the first step we used integration by parts , in the next step we used changed variables and in the last step we used the oddness of @xmath109 to split the integral into positive and negative half - lines and multiplied by appropriate powers .
we observe that @xmath110 for each @xmath103 , and is finite by assumption .
furthermore , the function @xmath111 is a bounded measurable function , and due to we can take the limits inside the integral sign in and use the dominated convergence theorem .
part ( 1 ) : we have @xmath112 since the function @xmath44 is bounded , the interval @xmath53 is bounded , and @xmath0 is a probability measure , the right - hand side integral converges absolutely , so we used fubini to interchange integrals to get the equality above . let @xmath113 suppose @xmath114 , then using the bound @xmath115 , we see that the bound @xmath116 & = \frac{c}{a } \left(\int_{|x|\leq|y_1-y_2|/2}+\int_{|x|\geq|y_1-y_2|/2}\right)(\cdots)\hbox{d}x\\[.3pc ] & \leq \frac{ca^{\delta}}{|y_1-y_2|^{\delta } } \int_{-\infty}^{\infty}{\langle{x / a}\rangle}^{-\delta } \hbox{d}(x / a)\\[.3pc ] & \leq \frac{ca^{\delta}}{|y_1-y_2|^{\delta}}\end{aligned}\ ] ] is valid .
it follows that @xmath117 for @xmath118 .
it remains to consider @xmath119 .
this is done by noting that @xmath120 from which taking limits , we obtain the stated value for the coefficient , either @xmath100 or @xmath121 , based on whether @xmath122 or @xmath123 , using the evenness of @xmath26 . now to complete the proof , we note the estimate @xmath124 where the constant @xmath125 is independent of @xmath126 , @xmath127 , and @xmath128 . thus the proof is completed using the dominated convergence theorem .
part ( 2 ) : we adapt the arguments in @xcite to the case at hand .
we split the measure in three components : @xmath129 . here
@xmath130})\hbox{d}\mu$ ] , @xmath131 with @xmath132)$ ] , and @xmath133 is purely singular , and supported on @xmath134 $ ] .
we have for @xmath135 $ ] the estimate @xmath136}a^{-1}{\langle{(x - y)/a}\rangle}^{-\delta}\hbox{d}\mu_1(y ) \leq c a^{\delta-1}.\ ] ] we now look at the @xmath137 part .
we have , for @xmath138 , by the reverse hlder inequality @xmath139 & \quad\ , \leq \left(\int_c^d | ( \tilde{\psi}_a * g)(x ) - a_\psi g(x)| \hbox{d}x\right)^{p } ( d - c)^{1-p},\end{aligned}\ ] ] which implies that @xmath140 in @xmath141 , @xmath142 .
now we will show that the singular part @xmath133 does not contribute to the limit .
so assume that @xmath133 is purely singular and that its support @xmath143 is contained in @xmath144 $ ] .
since @xmath133 is singular , by the definition of support , @xmath143 satisfies @xmath145 and @xmath146 , with @xmath147 denoting the lebesgue measure . by the regularity of the lebesgue measure , given an @xmath148
, there is an open set @xmath149 , such that @xmath150 , with @xmath151 .
we also have latexmath:[$|o| \leq |o\hbox{$\setminus$}s| + the above reverse hlder inequality gives @xmath156 & \quad\,+\int_{(c , d)\setminus o } |(\tilde{\psi}_a*\mu_3)(x)|^p \hbox{d}x\\[.3pc ] & \leq |o|^{1-p } \mu_3((c , d))^p \|\psi\|_1^p \\[.3pc ] & \quad\,+|d - c|^{1-p } \left(\int_{(c , d)\setminus o } then using assumption [ a1.1 ] that @xmath161 , and setting @xmath162 , @xmath163 & \leq c\int_{(c , d ) \setminus o } \frac{1}{a } \int_{\mathbf r}\langle{(x - y)/a}\rangle^{-\delta } \hbox{d}\mu_3(y ) \hbox{d}x\\[.3pc ] & \leq c\int_{(c , d ) \setminus o } h(x)(\tilde{\phi}_a*\mu_3)(x ) \hbox{d}x.\end{aligned}\ ] ] the function @xmath164 satisfies assumption [ a1.1 ] , so theorem [ thm1](1 ) is applicable with @xmath26 replaced by @xmath164
. therefore the last term , which has positive integrand , converges to @xmath165 as @xmath126 goes to zero , which is bounded by @xmath166 , @xmath167 using the facts that @xmath168 and @xmath169 . using the inequality @xmath170 for @xmath171 and non - negative numbers
@xmath126 , @xmath172 , @xmath173 , we have @xmath174 { } & \quad+\int_c^d |(\tilde{\psi}_a * \mu_2 ) ( x)-a_{\psi}g(x)|^p \hbox{d}x\\[.3pc ] { } & \quad+\int_c^d |(\tilde{\psi}_a * \mu_3 ) ( x)|^p \hbox{d}x.\end{aligned}\ ] ] putting the above estimates together and using @xmath175 arbitrary , one gets @xmath176 now the spaces @xmath177 , @xmath171 , are metric spaces with the metric @xmath178 .
it then follows from the triangle inequality for this metric that @xmath179 since @xmath180 , the result follows .
consider the case when @xmath83 is finite for some @xmath82 and for some fixed @xmath184 .
then for any @xmath185 , @xmath186 ) and estimating the right - hand side , one has , by assumption [ a1.1 ] , @xmath187 now taking the @xmath188 of the left - hand side the finiteness of @xmath189 follows .
on the other hand , since @xmath26 is positive continuous with @xmath190 , there is a @xmath191 such that @xmath192 . using this and the evenness of @xmath26 , @xmath193 & \geq \frac{1}{a^\alpha}\int_{-\beta a}^{\beta a } \frac{1}{2 } \hbox{d}\mu(y+x ) \\[.3pc ] & \geq \frac{1}{2a^\alpha } [ \mu(x + a\beta ) - \mu(x - a\beta)],\end{aligned}\ ] ] where @xmath194 is used to get the first inequality .
the above inequalities immediately imply that since @xmath195 is fixed , @xmath196 implies the same for @xmath197 . | in this paper we give some new criteria for identifying the components of a probability measure , in its lebesgue decomposition .
this enables us to give new criteria to identify spectral types of self - adjoint operators on hilbert spaces , especially those of interest .
[ theore]*theorem * [ theore]remark [ theore]proposition [ theore]lemma [ theore]definition [ theore]assumption [ theore]corollary [ theore]example [ theore]proof of theorem |
in the paper @xcite ( hereafter called ff1 ) , the authors investigated the consequences of the assumption that the high - transverse - momentum particles seen in hadron - hadron collisions are produced by a single , hard , large - angle elastic scattering of quarks , one from the target and one from the beam .
the fast outgoing quarks were assumed to fragment into a cascade jet of hadrons .
the distribution of quarks in the incoming hadrons were determined from lepton - hadron inelastic scattering data , together with certain theoretical constraints such as sum rules , etc .
the manner in which quarks cascade into hadrons was determined from particle distributions seen in lepton - hadron and lepton - lepton collisions supplemented by theoretical arguments .
parameterizations for single - hadron fragmentation functions ( sff ) were obtained . in the next work @xcite a new , much simpler , parameterization for these functions was provided .
the answers to many other questions concerning the details of the hadronization process were given .
this hadronization model for the production of mesons in the jets of initial quarks and antiquarks turned out very successful and is used so far in the design of experiments in which quark jets may be observed and for comparison with the available experimental data .
it provides a parameterization of the properties of the jet of mesons generated by a fast outgoing quark .
it is assumed that the meson that contains the original quark leaves momentum and flavor to a remaining jet in which the particles are distributed ( except for scaling of the energy and possible changes of flavor ) like those of the original jet .
one function , the probability @xmath2 that the remaining jet has a fraction @xmath3 of the momentum of the original jet , is chosen ( as a parabola ) so the final distribution of charged hadrons agrees with data from lepton experiments .
all the properties of quark jets are determined from @xmath2 and three parameters : a ) the degree that su(3 ) is broken in the formation of new quark - antiquark pairs ( @xmath4 is taken as half as likely as @xmath5 ) @xmath6 ( as usually , it is adopted the well - known condition following from the isospin symmetry @xmath7 ) and for the sum of probabilities it is taken @xmath8 , because the probabilities of the production of the heavier quarks are neglected ; b ) the spin nature of the primary mesons ( in the basic version of model is assumed to be vector and pseudoscalar with equal probability ) @xmath9 ; c ) the mean transverse momentum given to these primary mesons . in the rest of the paper we will frequently refer to the ref .
@xcite and for this reason it is convenient to denote its as ff2 , and model , which was developed there to denote as field - feynman quark jet model ( ffqjm ) .
+ we would like present here the explanation of the success of the ffqjm , which was given by gottschalk in his lectures @xcite : `` specifically , the ffqjm is not a ' ' theory `` of how hadronization works . rather , it is simply a parameterization of hadron formation in jets which reproduces two important experimental facts : 1 ) approximate scaling of the energy distributions of hadrons within a jet ; 2 ) limited transverse momenta of the produced hadrons with respect to the jet axis .
these qualitative features of hard scattering data are not ' ' explained `` or ' ' derived `` in the ffqjm .
rather , the ffqjm is a parameterization of hadron production from fast quarks which is constructed in such a manner that the above experimental observations are reproduced . ''
+ let us briefly discuss the parameters from points a ) and b ) of ffqjm presented above .
they play important role in the understanding of the hadronization process . in the early model of field and feynman
presented in ff1 , the point a ) coincides with that from ffqjm , but point b ) differs . for its in ff1 is chosen simple version @xmath10 , @xmath11 . in the simple lund model @xcite , which has very large similarities with the ffqjm , the point a ) coincides with latter , but for parameters of point b ) the ratio is chosen in the form @xmath12 as given by statistical spin counting . in the later version of lund model ,
the so called standard lund model @xcite the point a ) is changed to @xmath13 which is followed from the tunneling production mechanism , but point b ) coincides with one from ffqjm , because authors state that the spin - spin forces enhance the production of the lighter pseudoscalars , from the naive @xmath14 vector to pseudoscalar ratio to something closer to @xmath15 .
later these parameters were investigated in detail in the process of @xmath16 annihilation ( see , for instance , refs .
@xcite and @xcite ) . in
@xcite it is shown that the production rates of light - flavor hadrons are set by their masses , spins , strangeness suppression , and strongly influenced by the spin - spin interactions of their quarks .
the vector - to - pseudoscalar meson and decuplet - to - octet baryon suppresions are the same and are explained by the hyperfine mass splitting .
the strangeness suppression factor , @xmath17 , is the same for mesons and baryons .
it is related to the difference in the constituent quark masses , @xmath18 , where @xmath19 , at the temperature , @xmath20 .
comparison of the parameters from ffqjm with the ones obtained in other models showes that their values differ significantly .
this means , that values of parameters in ffqjm were determined with large uncertainties .
inclusion in consideration the new sets of experimental data and new observables can improve description and shift the parameters to their precise values .
+ in ffqjm , using formalism based on the recursive principle and available experimental data , were obtained sff @xmath21 , which are the distributions of hadrons of flavor @xmath22 from a quark @xmath23 , which carry away the fraction @xmath24 of its energy , and dihadron fragmentation functions ( dff ) @xmath25 , which are the distributions of hadrons of flavors @xmath26 and @xmath27 from a quark @xmath23 , which carry away the fractions @xmath0 and @xmath1 of its energy , respectively . using these results , authors proposed several experiments , which can verify model . among others
, they proposed to measure the correlation functions for oppositely and identically charged hadrons .
it was predicted , that correlation function for the oppositely charged hadrons must show the characteristic short - range correlation behavior .
the correlation between two identically charged hadrons , on the other hand , was predicted to be quite small . in ff2
the correlation function of two hadrons was presented as a function of their `` @xmath24 rapidity '' given by @xmath28 .
were considered energies large enough that a plateau was developed .
rapidity of the first hadron was fixed in region of plateau , @xmath29 ( which corresponds @xmath30 ) and @xmath31 was changed in wide region @xmath32 . as it is well known , in this model energy and momentum of the quark jets do not precisely conserved .
the violation of the energy conservation reaches a few per cents . in these conditions
the consideration of hadrons with small enough energy may be doubtful .
+ in this work we study the correlation function in the framework of ffqjm in detail , for the one definite reaction , the semi - inclusive electroproduction of hadrons on proton in deep - inelastic scattering ( dis ) region , at moderate energies ( energy of the virtual photon or , which is the same , the energy of the initial quark in order of 10 - 20 gev ) , where the plateau does not formed and where the small values of @xmath24 in order of @xmath33 do not belong to the current fragmentation region . instead of pairs of oppositely and identically charged hadrons we consider pairs of hadrons having definite charge and flavor ; and instead of rapidity , which has very narrow changing region at moderate energies , we use variable @xmath24 . it is shown that the correlation function has interesting properties .
in particular for some pairs of hadrons there are kinematic regions over variable @xmath1 ( at fixed values of @xmath0 ) where it is positive and unexpectedly large . the experimental investigation of the correlation function in these kinematic regions
has twofold interest , from one side it can help to investigate some delicate properties of quark jets and from the other side can help to verify the physical suppositions of model , and to obtain more precise values of parameters .
+ there are , of course , several obvious defects of the model , which were discussed by various people begun with authors of model , field and feynman . the detailed criticism of model can be found in work @xcite .
there are , also , many attempts to improve the model for the more precise description of the experimental data ( see , for instance , refs .
@xcite and @xcite ) .
+ however , despite on the available defects , the model is able to describe the experimental data in dis region with good enough precision , as the numerous applications are shown . therefore , we think that it can be used for the qualitative study of the two hadron correlation functions at moderate energies . + and last point , which we would like to discuss here , is the place of this work among others , devoted to the investigation of the dff .
at present , many questions connected with the hadron - hadron correlation become the objects of the experimental investigations in the high - energy lepton - lepton , lepton - hadron , lepton - nucleus ,
hadron - hadron , hadron - nucleus and nucleus - nucleus interactions . as a consequence
are arisen theoretical works , which devoted to the explanation and description of the existing experimental data or to the proposals for future experiments .
there is huge amount of such experimental and theoretical works .
naturally it is impossible to present them in one article .
we present only two of them , because they reflect common approach to the investigation of dff on phenomenological level .
they are refs .
@xcite , and @xcite . in these works
authors had concentrated on the definition of the dff and derived the dglap evolution equations for the non - singlet quark dff @xcite and singlet quark and gluon dff @xcite . the more attention was paid to the derivation of dglap evolution equations and less to the derivation of proper initial conditions .
about initial conditions used in @xcite , authors themselves note , that in numerical study of the nonsinglet quark dff they used a simple ansatz for the initial condition as @xmath34 , which at best is just a guess and differs significantly from the inherent hadron correlations in a single jet . in the next work
@xcite the initial conditions for the dff were extracted from jetset at a scale @xmath35 .
although both the sff and dff evolve rapidly with @xmath36 , their ratio , which frequently is used for the comparison with experimental data , has a very weak @xmath36 dependence .
this is especially true for the ratio in the case of the gluon fragmentation function , which shows practically no change with @xmath36 .
the results of evolution are strongly dependent , however , on the initial conditions and thus on the actual values of @xmath0 and @xmath1 . in this work
we study the dis at moderate energies and @xmath36 , i.e. we investigate initial conditions for dff for different pairs of hadrons and different values of @xmath0 and @xmath1 .
we show , that different choice of the types of hadrons and their fractional energies leads to the essentially different initial conditions for dff .
+ the paper is organized as follows . in section 2
we briefly discuss the structure of ffqjm and point out the way of the derivation of the formulae for the calculations of sff and dff , and present expression for correlation function .
section 3 presents results and discussion .
our conclusions are presented in section 4 .
let us consider the semi - inclusive dis process on proton , in which a two hadron system is observed in the final state : @xmath37 we would like to study correlation between two hadrons in the framework of the ffqjm .
before we want briefly remind , that main ingredients of ffqjm are the probability function @xmath38 and three free parameters : 1 ) the ratio @xmath39 ; 2 ) the probability of pseudoscalar meson production @xmath40 and 3 ) the mean transverse momentum of the primary mesons . as we already mentioned in introduction , we do not consider third parameter and suppose that integration over it is performed .
it is easily to see , that parameters from points 1 and 2 are correlated with parameter @xmath41 entering in function @xmath2 . in ffqjm
the ratio @xmath39 is fixed on value @xmath42 , the choice of parameter @xmath43 and correlated with it parameter @xmath41 leads to the two sets of parameters ( two versions of model ) considered in ff2 .
they are : a ) simple version , which takes into account the direct production of pseudoscalar mesons only , with corresponding set of parameters @xmath44 ; and b ) basic version , which takes into account both the direct production of pseudoscalar mesons and their production from the decays of resonances , with set of parameters @xmath45 .
+ now we introduce the notion `` rank '' following ff2 .
let us consider the `` hierarchy '' structure of the final mesons produced when a quark of type `` a '' fragments into hadrons .
new quark pairs @xmath46 , @xmath47 , @xmath48 etc . , are produced and `` primary '' mesons are formed .
the `` primary '' meson @xmath49 that contains the original quark is said to have `` rank '' one , the `` primary '' meson @xmath50 `` rank '' two , the `` primary '' meson @xmath51 `` rank '' three , etc .
finally , some of the `` primary '' mesons decay and it is assigned , that all the decay products to have the `` rank '' of the parent .
the order in `` hierarchy '' is not the same as order in momentum or rapidity . +
as it was mentioned above we use for sff and dff formulae obtained in ff2 .
unfortunately , corresponding expressions are very long and inconvenient for reading .
we do not presented these formulae here , and only remind the numbers of corresponding equations in ff2 . for sff
it is used formula ( 2.57 ) from ff2 .
the structure of dff is more complicate and consists of five terms : ( i ) the probability that the primary meson at @xmath0 is of rank 1 and the primary meson at @xmath1 is of rank 2 , given by ( 2.43a ) ; ( ii ) the probability that @xmath0 is of rank one , but @xmath1 is of rank higher than 2 , given by ( 2.43b ) ; ( iii ) the probability that the primary meson at @xmath0 is not first in rank but higher , but the primary meson at @xmath1 is directly of next rank to the one at @xmath0 , given by ( 2.43c ) ; ( iv ) neither the primary meson at @xmath0 or @xmath1 is first in rank , nor are they adjacent , given by ( 2.43d ) . as it is pointed out in ff2 , the complete dff for producing two hadrons of flavor @xmath52 and @xmath53 is given by symmetrizing ( 2.43a - d ) with respect to @xmath0 and @xmath1 ( see eq .
( 2.44 ) of ff2 ) .
as pointed out in ff2 , inclusion in the consideration the mesons produced from the decays of resonances complicates the structure of equations ( 2.43a - d ) and adds next term ( v ) corresponding to the contribution from the case where both the mesons at @xmath0 and @xmath1 came from the decay of the same resonance ( type @xmath54 ) given by formula ( 2.59 ) of ff2 .
+ there are two sources of correlations in the discussed model . naturally , there is the correlation among secondary particles that are the decay products of the same primary meson .
in addition , however , the primary mesons are not formed at random in @xmath24 .
primary mesons adjacent in rank are correlated in both flavor and @xmath24 since they each contain a quark ( or antiquark ) that came from the same @xmath55 pair .
the two primary mesons of adjacent rank tend to occur near each other in @xmath24 .
all flavor correlations in the quark jets occur between primary mesons of adjacent rank .
the flavor of a meson of rank @xmath56 is independent of the flavor of the meson of rank @xmath57 .
+ the observable which is widely used for the investigation of correlation between two hadrons produced in the any hard process is the two - body correlation function .
it is a function of single and double hadron multiplicities @xmath58 and @xmath59 , respectively . in our case
@xmath60 denote the hadrons , @xmath61 ; @xmath62 denote the fraction of the virtual photon energy @xmath63 carried by @xmath64-th hadron , @xmath65 where @xmath66 is the energy of the @xmath64-th hadron . in this work
we will use , as observable , the `` normalized '' two body correlation function ( see , for instance , ff2 and references therein ) @xmath67 in the quark - parton model the single hadron multiplicity is @xmath68 where @xmath69 and @xmath70 are inclusive and semi - inclusive cross sections , @xmath71 and @xmath72 are the electric charge of the quark in units of the elementary charge and distribution function of the quark with flavor @xmath73 in proton , @xmath74 is the bjorken variable @xmath75 , where @xmath36 is the photon virtuality and @xmath76 is the proton mass .
@xmath77 is the sff for the production of hadron @xmath22 by @xmath64-th quark , @xmath24 denote the fraction of the virtual photon energy @xmath63 carried by hadron .
@xmath78 where @xmath79 is the semi - inclusive cross section in which two hadrons in the final state are observed .
the dff @xmath80 , are the distributions of hadrons of flavors @xmath26 and @xmath27 from a quark @xmath64 , which carry away the fractions @xmath0 and @xmath1 of its energy , respectively .
we perform calculations mainly for basic version of the model . in this case
it is necessary to take into account that a part of the final pseudoscalar mesons are arisen in result of decays of the other hadrons .
we use contributions from 13 hadrons ( @xmath81 , @xmath82 , @xmath83 , @xmath84 , @xmath85 , @xmath3 , @xmath86 , @xmath87 , @xmath88 , @xmath89 , @xmath90 , @xmath91 , @xmath92 ) .
further we will call they `` resonances '' although among them are long living hadrons also .
some of them , in result of decay , give contribution only in pion or kaon production , but there are resonances which decay into @xmath93 pairs . from eqs.(2)-(4 ) we are obtained the correlation functions @xmath94 for the pairs of mesons with different types and/or electric charges ( @xmath95 , @xmath93 and @xmath96 ) . for calculations were used expressions for sff and dff obtained in the framework of the ffqjm and the distribution functions of the quarks in the proton from @xcite . for self - consistency of the consideration we use distribution functions of the quarks in the proton in the leading order ( lo ) approximation . for comparison with the basic version , we will present results of calculations for simple version of model also .
+ = 8.cm
= 8.cm in fig.1
the correlation functions @xmath94 for the pairs of hadrons of the same type ( @xmath95 and @xmath96 ) , with identical and opposite electric charges as a functions of @xmath97 are presented .
calculations were performed for a fixed values of @xmath98 equal to @xmath99 ( circles ) and @xmath42 ( triangles ) . at each fixed value of @xmath98 two curves with fixed values of @xmath74 too small values , because then become large the contributions of two jet events , which connected with the pomeron exchange . in other words , it is need take into account that the part of its time in order of @xmath100 the virtual photon behaves as a @xmath55 pair . if this time is essentially larger than the size of proton then instead of the photon with target will interact quark - antiquark pair .
we can escape this situation imposing the restriction @xmath101 .
] equal to @xmath102 ( filled symbols ) and @xmath42 ( open symbols ) are presented .
the correlation functions are calculated in the basic version of the ffqjm , taking into account hadrons produced both directly and from decays of resonances . only for the pairs of hadrons of the same type and different electric charges they have positive values ( see panels c , d , g , h ) .
the correlation functions for @xmath103 and @xmath104 pairs have similar shapes and become positive only at smaller value of @xmath0 using in calculations ( @xmath105 ) . at bigger value of @xmath0 using in calculations ( @xmath106 ) they are negative . in the cases of @xmath107 and @xmath108
pairs we see more strong positive correlation ( pay attention on the scale of corresponding panels ) .
positive correlation persists practically over all range of the change of variables .
next peculiarity of the correlation functions for these pairs is the strong enough @xmath74 - dependence . in case of @xmath107
more strong @xmath74 - dependence takes place for @xmath105 , while in case of @xmath108 it does for @xmath106 .
pairs of hadrons with identical flavors ( having the same type and identical electric charges ) ( @xmath109 , @xmath110 , @xmath111 and @xmath112 ) have correlation functions which are negative in all region of calculated @xmath1 .
they decrease ( practically linearly ) in all region of @xmath1 .
later we will discuss the reasons of the different behavior of the pairs of hadrons of the same type , for cases , when they have opposite and identical electric charges .
+ in fig.2 the correlation functions for the pairs of hadrons with different types ( @xmath93 - pairs ) , with identical and opposite electric charges are presented .
notations are the same as in fig.1 .
the correlation functions are calculated in basic version of the ffqjm .
they are negative practically everywhere .
their behavior over variable @xmath1 are close to the linear .
we see only small differences in the shapes and values .
the @xmath74 - dependence is practically absent .
+ in fig.3 the correlation functions @xmath94 for the pairs of hadrons of the same type ( @xmath95 and @xmath96 ) , with identical and opposite electric charges as a function of @xmath97 are presented .
notations are the same as in fig.1 .
the correlation functions are calculated in the simple version of the ffqjm , taking into account only hadrons produced directly .
easily to see , that in comparison with basic version , presented in fig.1 , absolute values of @xmath94 here bigger .
the qualitative behavior of the correlation functions in basic and simple versions of model is very close .
the positive correlation for @xmath103 and @xmath104 is more prominent in simple version , while for @xmath107 and @xmath108 pairs the situation is practically the same in both versions .
+ = 8.cm
= 8.cm in fig4 .
the correlation function @xmath94 for @xmath95 pairs with opposite electric charges as a function of @xmath97 is presented .
calculations were performed for a fixed values of @xmath98 equal to @xmath102 , @xmath99 and @xmath42 .
the correlation function is calculated in the basic version of the ffqjm , taking into account hadrons produced both directly and from decays of resonances .
it is presented the contributions of the different mechanisms in correlation function .
the contribution of not adjacent in rank hadrons only ( filled triangles ) ; the sum of contributions of not adjacent and adjacent hadrons ( open circles ) ; and final result , the sum of three contributions : of not adjacent hadrons , adjacent hadrons and hadrons produced from the same resonance ( filled circles ) . + in the fig.5 the same as in fig.4 for @xmath96 pairs is presented ( pay attention on the scale ) .
+ now let us discuss briefly the figs.4 and 5 .
first observation is , that taking into account only contributions from not adjacent mesons we obtain for pairs @xmath103 , @xmath104 , @xmath107 and @xmath108 the negative correlation functions which have the behavior very close to the cases for pairs of mesons of the different types , or for pairs of mesons of the same type and electric charges ( identical flavors ) .
the main source of the positive correlation is the possibility for the mesons entering in the pairs @xmath103 , @xmath104 , @xmath107 and @xmath108 to be produced in the adjacent rank .
second source of the positive correlation , the possibility for the mesons be produced from the decay of the same resonance is very small and can be neglected . from panels
a , b , c of fig.4 for @xmath103 pair , we see that while the contribution of not adjacent mesons is decreased with the increasing of @xmath1 , the contribution of adjacent mesons is practically constant in all region of @xmath1 .
the similar behavior takes place for @xmath104 pair ( see panels d , e , f of fig.4 ) .
situation is more complicated for the @xmath107 and @xmath108 pairs ( see fig.5 ) .
the behavior of the contributions from not adjacent mesons is like the @xmath95 case in fig.4 .
the contribution of the adjacent mesons for @xmath96 pairs has more complicate behavior as a function of @xmath0 and @xmath1 and at some values of @xmath0 and @xmath1 can be essentially larger than the contribution from not adjacent mesons .
it is the source of the big and nontrivial positive correlations for the @xmath107 and @xmath108 pairs .
+ it is very important to verify , that results obtained above do not depend essentially from the choice of the parton distribution functions . for this goal we repeated all calculations with another set of parton distribution functions .
were used lo parton distribution functions from @xcite .
it was obtained , that difference between two results is smaller than 1 per cent .
+ we propose to measure the correlation functions for the @xmath107 and @xmath108 pairs experimentally .
the experimental discovery of the fact of big positive correlation will means , that main mechanism of hadronization is the fragmentation of the initial quark via consecutive production of @xmath55 pairs and their further combination into final mesons .
in the absence of any dynamical correlations , the probability to observe in a single jet one hadron with fraction of energy @xmath0 and second hadron with @xmath1 , together with anything else , would be equal to the product of the probabilities to find hadrons with @xmath0 and @xmath1 in different jets and @xmath94 , as we defined it in eqs.(2)-(4 ) , would be close to zero . really , it is true only in case @xmath113 . when @xmath114 the correlation function for the pair of hadrons , which can not be produced adjacent in rank becomes negative , because of the necessity of the production of the additional particles . even in the case , when hadrons can be produced adjacent in rank part of energy is lost because the additional hadrons are produced in result of decays of resonances , which also leads to the reduction of correlations .
we obtained such a behavior for all considered pairs of mesons , besides @xmath103 , @xmath104 , @xmath107 and @xmath108 . the prominent positive correlations for the @xmath107 and @xmath108 pairs in the framework of ffqjm
were obtained .
the experimental investigation of @xmath94 for these cases can help to verify the basic assumptions of the model and receive the valuable information about hadronization process . + the model does not include quantum effects .
it has dealt solely with probabilities and not amplitudes .
it is supposed that in dis region the interference effects are small .
if the interference effects , nevertheless , do not small , it will felt in selected pairs @xmath103 , @xmath104 because of the interference between the amplitudes for the production of the hadrons @xmath26 and @xmath27 adjacent and not adjacent in rank .
for the pairs @xmath107 and @xmath108 the interference will be practically impossible , because of the necessity of the production of the additional pair of @xmath4 , and construction of specific final state .
+ the advatage of our approach to the derivation of sff and dff in comparison with , for instance , @xcite , and @xcite is that we have , in the framework of ffqjm , the analytic formulae for the sff and dff , which are obtained at moderate values of @xmath36 .
they can serve as a initial state sff and dff for the qcd evolution .
moreover we can have separately the contributions in dff from different mechanisms ( production of hadrons adjacent and not adjacent in rank , production from the decay of the same resonance , etc . ) .
it is very interesting to perform the qcd evolution for the different initial states and to trace the influence of the choice of the initial state , on the final state at high @xmath36 . + as we mentioned above
, the main source of the positive correlation is the possibility for the mesons entering in the pairs @xmath103 , @xmath104 , @xmath107 and @xmath108 to be produced in the adjacent rank .
this mechanism is not the peculiarity of the ffqjm and is available in all hadronization models of the fragmentation type ( for instance @xcite and @xcite )
. it will be very interesting item to receive the correlation functions in other hadronization models .
+ 99 ( ff1 ) r.d.field and r.p.feynman , phys.rev .
* d15 * ( 1977 ) 2590 ( ff2 ) r.d.field and r.p.feynman , nucl.phys . * b136 * ( 1978 ) 1 t.gottschalk , lectures at the 19-th int .
school of elementary particle physics , kupari - dubrovnik , yugoslavia , 1983 .
b.andersson , g.gustafson and c.peterson , nucl.phys . * b135 * ( 1978 ) 273 b.andersson et al .
, phys.rep .
* 97 * ( 1983 ) 31 m.anselmino et al .
, phys.lett . * b427 * ( 1998 ) 356 p.v.chliapnikov , phys.lett . *
b462 * ( 1999 ) 341 r.jengo , a.krzywicki and b.petersson , nucl.phys .
* b65 * ( 1973 ) 319 jing hua , bo - qiang ma , eur.phys.j . *
c30 * ( 2003 ) 207 a.majumder and xin - nian wang , phys.rev . *
d70 * ( 2004 ) 014007 a.majumder and xin - nian wang , phys.rev .
* d72 * ( 2005 ) 034007 m.gl@xmath115ck , e.reya , a.vogt , z.phys . *
c67 * ( 1995 ) 433 t.gehrmann and w.j.stirling , phys.rev .
* d53 * ( 1996 ) 6100 | very successful hadronization model for the production of mesons in the jets of initial quarks was proposed thirty years ago by field and feynman .
the model is used so far in the design of experiments and for comparison with the experimental data . in this work
it is shown that correlation function in that model , as a function of the energies of two final hadrons has interesting properties ( energies are taken in fractions of initial quark energy , and equal @xmath0 and @xmath1 , respectively ) . in particular , for some pairs of mesons there are kinematic regions over variable @xmath1 ( at fixed values of @xmath0 ) , where it is positive and unexpectedly large .
the experimental investigation of the correlation functions in these kinematic regions has twofold interest , from one side it can help to investigate some delicate properties of quark jets and from the other side can help to verify the physical suppositions of model , and to obtain more precise values of parameters . |
fundamental quantum many - body systems involve the interaction of bosonic and/or fermionic particles .
the spin of a particle makes it behave very differently at ultracold temperatures below the degeneracy temperature .
there are thus fundamental differences between the properties of bosons and fermions .
however , as bosons are not subject to the pauli exclusion principle , they can collapse under suitable conditions into the same quantum groundstate the bose - einstein condensate ( bec ) .
remarkably , even a small attraction between two fermions with opposite spin states and momentum can lead to the formation of a bardeen - cooper - schrieffer ( bcs ) pair that has a bosonic nature .
such bcs pairs can undergo the phenomenon of bec as temperature tends to absolute zero . over the past few decades , experimental achievements in trapping and cooling atomic gases
have revealed the beautiful and subtle physics of the quantum world of ultracold atoms , see recent review articles @xcite . in particular , recent experiments on ultracold bosonic and fermionic atoms confined to one dimension ( 1d )
have provided a better understanding of the quantum statistical and dynamical effects in quantum many - body systems @xcite .
these atomic waveguide particles are tightly confined in two transverse directions and weakly confined in the axial direction .
the transverse excitations are fully suppressed by the tight confinement . as a result
the trapped atoms can be effectively characterised by a quasi-1d system , see fig .
[ fig:1d ] .
the effective 1d inter - particle potential can be controlled in the whole interaction regime .
in such a way , the 1d many - body systems ultimately relate to previously considered exactly solved models of interacting bosons and fermions .
this has led to a fascinating interplay between exactly solved models and experimental developments in 1d .
inspired by these developments , the study of integrable models has undergone a renaissance over the past decade .
their study has become crucial to exploring and understanding the physics of quantum many - body systems .
the study of bethe ansatz solvable models began when bethe ( 1931 ) introduced a particular form of wavefunction the bethe ansatz ( ba ) to obtain the energy eigenspectrum of the 1d heisenberg spin chain .
after laying in obscurity for decades , the ba emerged to underpin a diverse range of physical problems , from superconductors to string theory , see , e.g. , batchelor ( 2007 ) . for such exactly solved models ,
the energy eigenspectrum of the model hamiltonian is obtained exactly in terms of the ba equations , from which physical properties can be derived via mathematical analysis . from 1931 to the early 1960s
there were only a handful of papers on the ba , treating the passage to the thermodynamic limit and the extension to the anisotropic xxz heisenberg spin chain @xcite .
yang and yang ( 1966a ) coined the term bethe s hypothesis and proved that bethe s solution was indeed the groundstate of the xxz spin chain @xcite .
the next development was the exact solution of the 1d bose gas with delta - function interaction by lieb and liniger ( 1963 ) , which continues to have a tremendous impact in quantum statistical mechanics @xcite .
they diagonalised the hamiltonian and derived the groundstate energy of the model .
this study was further extended to the excitations above the groundstate @xcite .
mcguire ( 1964 ) considered the model in the context of quantum many - body scattering in which the condition of non - diffractive scattering appeared .
developments for the exact solution of the 1d fermi gas with delta - function interaction @xcite are discussed in the next subsection .
a key point is yang s observation @xcite that a generalised bethe s hypothesis works for the fermion problem , subject to a set of cubic equations being satisfied .
this equation has since been referred to as the yang - baxter equation ( ybe ) after the name was coined by takhtadzhan and faddeev ( 1979 ) .
baxter s contribution was to independently show that such relations also appear as conditions for commuting transfer matrices in two - dimensional lattice models in statistical mechanics @xcite .
moreover , the ybe was seen as a relation which can be solved to obtain new exactly solved models .
the ybe thus became celebrated as the masterkey to integrability @xcite .
the study of yang - baxter integrable models flourished in the 70 s , 80 s and 90 s in the canberra , st petersburg , stony brook and kyoto schools , with far reaching implications in both physics and mathematics . during this period
the ybe emerged as the underlying structure behind the solvability of a number of quantum mechanical models .
in addition to the xxz spin chain , examples include the xyz spin chain @xcite , the @xmath0 model at supersymmetric coupling @xcite and the hubbard model @xcite .
three collections of key papers have been published @xcite .
further examples are strongly correlated electron systems @xcite , spin exchange interaction @xcite , kondo physics of quantum impurities coupled to conduction electrons in equilibrium @xcite and out of equilibrium @xcite , the bcs model @xcite , models with long range interactions @xcite , two josephson coupled becs @xcite , bcs - to - bec crossover @xcite , atomic - molecular becs @xcite and quantum degenerate gases of ultracold atoms @xcite .
a significant development in the theory of quantum integrable systems is the algebraic ba @xcite , essential to the so called quantum inverse scattering method ( qism ) , a quantized version of the classical inverse scattering method .
the qism gives a unified description of the exact solution of quantum integrable models .
it provides a framework to systematically construct and solve quantum many - body systems @xcite ) .
other related threads are the quantum transfer matrix ( qtm ) @xcite and @xmath1-systems @xcite from which one can derive temperature - dependent properties in an exact non - perturbative fashion . applications of this approach include the heisenberg model @xcite , higher - spin chains @xcite , and integrable quantum spin ladders @xcite .
@xmath1-systems and integrability in general also play a fundamental role in the gauge / string theories of high energy physics @xcite .
yang - baxter integrability has also played a crucial role in initiating and inspiring progress in mathematics . particularly to the theory of knots , links and braids @xcite and the development of quantum groups and representation theory @xcite . in the mid-60
s many physicists worked on extending the results obtained by lieb and liniger ( 1963 ) and mcguire ( 1964 ) for 1d bosons with delta - function interaction to the problem of 1d fermions .
mcguire @xcite solved the eigenvalue problem of @xmath2 fermions of the same spin and one fermion of opposite spin and studied the low - lying excited states with repulsive and attractive potentials .
the dynamics of this one spin - down fermi problem has been studied @xcite .
the problem of @xmath3 fermions of the same spin with two fermions of opposite spin was solved by flicker and lieb ( 1967 ) .
a further step came when gaudin ( 1967a,1967b ) and yang ( 1967 ) solved the general problem in terms of a nested ba for arbitrary spin population imbalance .
gaudin derived the groundstate energy for the balanced ( fully paired ) case for attractive interaction , pointing out that the result is equivalent to that for repulsive bosons @xcite . the delta - function interacting two - component fermi gas
is commonly referred to as the gaudin - yang model .
yang s concise solution of the problem had a profound impact .
as already remarked above , a key point in the solution is that the matrix operators describing many - body scattering can be factorised into two - body scattering matrices , provided that a set of cubic equations
the yang - baxter equation are satisfied by the two - body scattering matrices .
this in turn is equivalent to no diffraction in the outgoing waves in three - body scattering processes . in this sense yang
s solution completes mcguire s formulation of the scattering process in the context of the 1d bose gas .
indeed , the @xmath4-matrix obtained for the 1d bose gas is known as the simplest nontrivial solution of the yang - baxter equation @xcite .
the solution of the 1d fermi problem triggered a series of further breakthroughs .
yang ( 1968 ) obtained the @xmath5-matrix of the delta - function interacting many - body problem for boltzmann statistics @xcite .
the exact solution of the 1d fermi gas with higher spin symmetry was obtained by sutherland @xcite .
the 1d hubbard model solved by lieb and wu ( 1968 ) is a fundamental model in the theory of strongly correlated electron systems .
its solution is a significant example of the factorization condition ( the ybe ) in which the quasimomenta of particles @xmath6 is replaced by @xmath7 .
the lieb - wu solution thus gives a similar set of integral equations as yang s fredholm equations for the continuum gas .
this exactly solved model has been extensively studied in the literature . the exact results for the hubbard model not only provide the essential physics of 1d strongly correlated electronic systems @xcite , but also are relevant to phenomena in high @xmath8 superconductivity .
indeed , the 1d hubbard model is an archetypical many - body system featuring fulde - ferrel - larkin - ovchinnikov ( fflo ) pairing , universal tomonaga - luttinger liquid ( tll ) physics , spin - charge separation and quantum entanglement @xcite .
although further study @xcite of the 1d fermi gas was initiated soon after its solution , it was not until much later that this model began to receive more attention @xcite as a result of the brilliant experimental progress in ultracold atom physics .
the fundamental physics of the model is determined by the set of generalised fredholm integral equations obtained in the thermodynamic limit .
takahashi ( 1970a ) discussed the analyticity of the fredholm equations in the vanishing interaction limit .
a thorough study of the fredholm equations for the gaudin - yang model with attractive and repulsive interactions has been carried out @xcite .
the numerical solution of the fredholm equations has also been discussed in the context of harmonic traps @xcite . in particular , the eigenfunction has been obtained explicitly for the fermi gas in the infinitely strong repulsion limit by using the hard - core contact boundary condition @xcite and group theoretical methods @xcite .
the next major advance with implications for the 1d fermi problem was the solution of the finite temperature problem for 1d bosons .
yang and yang ( 1969 ) showed that the thermodynamics of the lieb - liniger bose gas can be determined from the minimisation conditions of the gibbs free energy subject to the ba equations .
takahashi went on to make significant contributions to yang and yang s grand canonical approach to the thermodynamics of 1d integrable models @xcite .
takahashi gave the general name thermodynamic bethe ansatz ( tba ) equations to the yang - yang type of equations for the thermodynamics .
he discovered spin strings patterns to the ba equations in addition to those for the groundstate of the spin chain @xcite . using a similar spin string hypothesis , gaudin ( 1971 ) studied the thermodynamics of the heisenberg - ising chain .
lai ( 1971a , 1973 ) independently derived the tba equations for spin-1/2 fermions in the repulsive regime .
it turns out that takahashi s spin string hypothesis allows one to study the grand canonical ensemble for many 1d many - body systems with internal degrees of freedom , e.g. , the 1d fermi gas @xcite , the 1d hubbard model @xcite , the quantum sine - gordon model @xcite , the kondo problem @xcite among many other integrable models .
building on takahashi s spin string hypothesis , schlottmann derived the tba equations for @xmath9 fermions with repulsive and attractive interactions @xcite .
the yang - yang method has been revealed to be an elegant way to analytically access not only the thermodynamics , but also correlation functions , quantum criticality and tll physics for a wide range of low - dimensional quantum many - body systems @xcite .
the yang - yang thermodynamics of the 1d bose gas has been tested in recent experiments @xcite .
recently , numerical schemes have been developed to solve the tba equations of the 1d two - component spinor bose gas with delta - function interaction @xcite .
the qtm method has also been applied to the thermodynamics of the 1d bose and fermi gases with repulsive delta - function interaction @xcite .
the canberra group and their collaborators @xcite developed an asymptotic method to calculate the thermodynamics of strongly interacting bosons and fermions in an analytic fashion using the polylog function in the framework of the yang - yang and takahashi methods .
this approach does away with the need to numerically solve the tba equations for these systems at quantum criticality , where the temperature is very low and the inter - particle interaction is strong .
the renewed interest over the past decade in 1d fermions has been on a number of related fronts .
here we give a brief introductory outline of these developments .
quantum matter at low temperatures has already been seen to exhibit some remarkable physical properties , such as bec and superfluidity .
fermionic quantum matter with mismatched fermi surfaces has long been expected to exhibit more exotic behaviour than seen in conventional materials .
the two - component attractive fermi gas is particularly interesting due to its connection with the exotic pairing phase the fflo state involving bcs pairs with nonzero centre - of - mass momenta . in this phase , where the system is partially polarized
, the fermi energies of spin - up and spin - down electrons become unequal .
originally , fulde and ferrell ( 1964 ) discovered that under a strong external field , superconducting electron pairs have nonzero pairing momentum and spin polarization .
larkin and ovchinnikov ( 1965 ) found that the formation of pairs of electrons with different momenta , i.e. , @xmath10 and @xmath11 with non zero @xmath12 is energetically favoured over pairs of electrons with opposite momenta , i.e. , @xmath10 and @xmath13 , when the separation between fermi surfaces is sufficiently large .
consequently , the density of spins and the superconducting order parameter become periodic functions of the spatial coordinates .
theoretical study of the fflo state in 1d interacting fermions was initiated by k. yang ( 2001 ) , who used bosonization to study the pairing correlations .
the fflo - like pair correlations and spin correlations for the attractive hubbard model were later investigated numerically by two groups @xcite .
both groups showed the power - law decay of the form @xmath14 for the pair correlation , with spatial oscillations depending solely on the mismatch @xmath15 of the fermi surfaces .
thus the momentum pair distribution has peaks at the mismatch of the fermi surfaces .
the fflo state has since been studied by various methods : density - matrix renormalization group ( dmrg ) @xcite , quantum monte carlo ( qmc ) @xcite , mean field theory and other methods @xcite .
very recently , the asymptotic correlation functions and fflo signature were analytically studied using the dressed charge formalism in the context of the gaudin - yang model @xcite . however , a convincing theoretical proof for the existence of the 1d fflo state in the expansion dynamics of the 1d polarized fermi gas after its sudden release from the longitudinal confining potential is still rather elusive , see recent further developments @xcite .
so far the spatial oscillation nature of fflo pairing has not been experimentally confirmed .
it was shown @xcite that large - spin atomic fermions exhibit rich pairing structures and collective modes in low - energy physics .
further progress towards understanding many - body physics with large - spin fermi gases was made @xcite on spin-3/2 systems which can be realized with @xmath16cs , @xmath17be and @xmath18ba ultracold atoms @xcite .
such systems exhibit a generic @xmath19 symmetry ( isomorphically , @xmath20 symmetry ) .
the spin-3/2 system with @xmath21 symmetry can exhibit a quartet state ( four - body bound state ) .
more generally , ultracold atoms offer an exciting opportunity to investigate spin - liquid behavior via trapped fermionic atoms with large - spin symmetry @xcite .
the trimer state ( trions " ) consisting of fermionic @xmath22li atoms in the three energetically lowest substates has been reported @xcite . on the other hand ,
fermionic alkaline - earth atoms display an exact @xmath9 spin symmetry with @xmath23 where @xmath24 is the nuclear spin @xcite .
such fermionic systems with enlarged @xmath9 spin symmetry are expected to display a remarkable diversity of new quantum phases and quantum critical phenomena due to the existence of multiple charge bound states .
de salvo _ et al . _
( 2010 ) have reported quantum degeneracy in a gas of ultracold fermionic @xmath25sr atoms with @xmath26 in an optical dipole trap .
an experiment by taie _
( 2010 ) dramatically realised the model of fermionic atoms with @xmath27 symmetry where electron spin decouples from its nuclear spin @xmath28 for @xmath29yb together with atoms of its spin-1/2 isotope .
this group also successfully realized the @xmath30 mott - insulator state with ultracold fermions of @xmath29yb atoms in a 3d optical lattice @xcite . in the context of large - spin ultracold atomic fermions , lecheminant _ et al . _
( 2005 ) considered 1d ultracold atomic systems of fermions with general half - integer spins .
the instability of the bcs pairing phase and molecular superfluid phase in these systems have been studied by a low - energy approach . the low - energy physics and competing orders in large - spin fermionic systems in a 1d lattice
were further investigated @xcite . in this scenario , a new class of integrable models of ultracold fermions and bosons with large - spin symmetries
were found @xcite .
they derived the ba solutions for spin-3/2 fermions with @xmath19 symmetry and the @xmath31-invariant model of fermions . from the integrable model perspective
, the study of multi - component attractive fermions was initiated a long time ago by yang ( 1970 ) and takahashi ( 1970b ) . in the light of ultracold higher spin atoms , controzzi and tsvelik ( 2006 ) proposed an exact solution of a model describing the low energy physics of spin-3/2 fermionic atoms in a 1d lattice .
the exact results obtained from 1d many - body systems with higher spin symmetries have provided insight into understanding the few - body physics of trions @xcite ; quartet states ( four - body charge bound states ) @xcite ; and an arbitrary large spin - neutral bound state of different sizes @xcite .
the study of critical phenomena and universal tll physics in low - dimensional ultracold atomic fermi gases with large pseudo - spin symmetries is a rapidly developing frontier in ultracold atom physics .
quantum criticality describes a v - shaped phase of quantum critical matter fanning out to finite temperatures from the quantum critical point ( qcp ) .
it is associated with competition between the two distinct ground states near the qcp . near a qcp
, the quantum critical behaviour is characterized by the energy gap @xmath32 and a diverging length scale @xmath33 , where @xmath34 is the critical value of the driving parameter @xmath35 .
the universality class of quantum criticality is characterized by the dynamic critical exponent @xmath36 and the correlation exponent @xmath37 @xcite .
the many - body system is expected to show universal scaling behaviour in the thermodynamic quantities at quantum criticality due to the collective nature of many - body effects .
thus a universal and scale - invariant description of the system is expected through the power - law scaling of thermodynamic properties . however , understanding the various aspects of quantum criticality in quantum systems represents a major challenge to our knowledge of many - body physics @xcite .
ultracold atoms have become the tool of choice to simulate and test universal quantum critical phenomena .
the study of quantum criticality and finite - size scaling in trapped atomic systems is thus attracting considerable interest @xcite .
the experimental study of critical behaviour in a trapped bose gas was initiated by donner _
in particular , significant experimental progress on quantum criticality and quantum phase transitions in 2d bose atomic gases has been made @xcite .
some of the remarkable features of criticality in general are the notions of universality class and symmetry . using integrable quantum field theory , zamolodchikov ( 1989 ) was able to show that the 2d ising model in a magnetic field , or equivalently the quantum ising chain with a transverse field @xcite displays @xmath38 symmetry close to the critical point .
such exotic quantum symmetry in the excitation spectrum was observed in a recent experiment in the traditional setting of condensed matter physics @xcite .
zhou and ho ( 2010 ) have proposed a precise theoretical scheme for mapping out quantum criticality of ultracold atoms . in this framework , exactly solvable models of ultracold atoms , exhibiting quantum phase transitions , provide a rigorous way to explore quantum criticality in many - body systems .
the equation of state has been obtained for a number of key integrable models , allowing the exploration of tll physics and quantum criticality .
these include the gaudin - yang fermi gas @xcite , the lieb - liniger bose gas @xcite , the fermi - bose mixture @xcite and the spin-1 spinor bose gas with antiferromagnetic spin - spin exchange interaction @xcite .
the exact results for the scaling forms of thermodynamic properties in these systems near the critical point illustrate the physical origin of quantum criticality in many - body systems .
many remarkable 1d quantum phenomena have been experimentally observed due to recent rapid progress in material synthesis and tunable manipulation of ultracold atoms .
these developments have provided a better understanding of significant quantum statistical effects and strong correlation effects in low - dimensional quantum many - body systems .
the observed results to date are seen to be in excellent agreement with results obtained using the mathematical methods and analysis of exactly solved models .
these include the experimental realization of the tonks - girardeau gas @xcite and a quantum newton s cradle , i.e. , a demonstration of out - of - equilibrium physics in arrays of trapped 1d bose gases @xcite and quantum correlations @xcite .
( 2009 ) made a further experimental breakthrough by realising a stable highly excited gas - like phase , called the super tonks - girardeau gas , in the strongly attractive regime of bosonic cesium atoms .
the yang - yang thermodynamics and thermal fluctuations of an ultracold bose gas of @xmath39rb atoms were further tested in a series of publications @xcite .
the universal low - energy physics was demonstrated as hosting a tll @xcite .
the experimental research using ultracold fermi gases to explore pairing phenomena in a 1d fermi gas was first reported in 2005 @xcite . in a major breakthrough towards understanding the exotic pairing signature and quantum phase diagram of the attractive fermi gas , liao _ et al . _
( 2010 ) measured the finite temperature density profiles of trapped fermionic @xmath22li atoms , see fig.[fig:1d ] .
they confirmed the key features of the @xmath40 phase diagram predicted from the exact solution @xcite . in light of these recent developments we review the ba solution of the gaudin - yang model in sec .
ii and discuss the physical understanding of the solution in terms of bcs pairing , the polaron problem , molecule states and the super tonks - girardeau gas . in sec .
iii we further discuss many - body phenomena in the gaudin - yang model .
especially , we discuss 1d fermions in a harmonic trap and review the universal features of 1d interaction , including magnetism , fflo - like pairing , tll physics , spin - charge separation , universal thermodynamics and quantum criticality . in sec .
iv we review recent progress on mixtures of ultracold atoms and the exact solution of the 1d fermi - bose mixture .
v reviews the exotic many - body physics of 1d multi - component interacting fermions , including the three - component fermi gas , the @xmath9 invariant fermi gases , and spin-3/2 fermions with @xmath19 symmetry .
the discussion in this section covers magnetism for systems of large - spin fermions , trions , molecular states of different sizes , multi - component tll phases , universal low - temperature thermodynamics and critical behaviour caused by population imbalance . in sec .
vi , we focus on the asymptotics of various relevant correlation functions for the gaudin - yang model and multi - component fermi gases
. the characteristics of the fflo - like pairing correlations and spin - charge separation correlation functions are discussed in the framework of conformal field theory .
the experimental breakthroughs with quasi-1d ultracold atoms and tests of 1d many - body physics are reviewed in sec .
vii . a brief conclusion and an outlook on future developments
are given in sec .
the hamiltonian @xmath41 describes a 1d @xmath42-function interacting two - component ( spin-@xmath43 ) fermi gas of @xmath44 fermions with mass @xmath45 and an external magnetic field @xmath46 constrained by periodic boundary conditions to a line of length @xmath47 .
the function @xmath48 is the trapping potential .
the field operators @xmath49 and @xmath50 describe the fermionic atoms in the states @xmath51 and @xmath52 , respectively . the @xmath42-type interaction between fermions with opposite hyperfine states preserves the spin states such that the zeeman term in the hamiltonian ( [ ham-1 ] ) is a conserved quantity .
the experimental realization @xcite of this system of interacting fermions is described in sec .
vi . the coupling constant @xmath53 , where @xmath54 can be tuned by feshbach resonance @xcite . for repulsive interaction @xmath55 and for attractive interaction @xmath56 .
where appropriate , we use units of @xmath57 . a dimensionless coupling constant @xmath58 is used to characterize physical regimes , i.e. , @xmath59 for the strong coupling regime and @xmath60 for the weak coupling regime . here
@xmath61 is the linear number density . for a homogeneous gas ,
i.e. , @xmath62 , the eigenvalue problem for hamiltonian ( [ ham-1 ] ) reduces to the 1d @xmath44-body delta - function interaction problem @xmath63 solved by gaudin and yang .
bethe s hypothesis states that the wavefunction of such a many - body system is a superposition of plane waves . in the domain @xmath64 , the wave function
is given by @xmath65 \exp \textrm{i}(k_{p1}x_{q1}+\ldots+k_{pn}x_{qn } ) \label{wave}\end{aligned}\ ] ] where both @xmath66 and @xmath67 are permutations of the integers @xmath68 .
the sum runs over all @xmath69 permutations @xmath70 .
the @xmath71 coefficients @xmath72 $ ] of the exponentials can be arranged as an @xmath71 matrix .
the columns are denoted by @xmath71 dimensional vectors @xmath73 @xcite .
for example , for two fermions with one spin - up and one spin - down , the wave function is written as @xmath74e^{\mathrm{i } ( k_1x_1+k_2x_2)}+[21,12]e^{\mathrm{i } ( k_2x_1+k_1x_2 } \right ) \nonumber\\ & & + \theta_{21}\left([12,21]e^{\mathrm{i } ( k_2x_1+k_1x_2)}+[21,21]e^{\mathrm{i } ( k_2x_2+k_1x_1 } \right)\nonumber\end{aligned}\ ] ] where @xmath75 denotes the step function @xmath76 .
a plane wave repeatedly reflected from the hyperplanes @xmath77 gives a total of @xmath69 plane waves .
the idea in setting up such an ansatz is an attempt at a hypothetical solution followed by demonstrating that it gives the eigenfunction of the many - body problem , rather than solving the problem directly .
the derivative of the wavefunction is discontinuous when two particles are infinitesimally close to one another .
this property can be derived by considering the eigenvalue problem @xmath78 in the center of mass coordinate @xmath79 and the relative coordinate @xmath80 of the two adjacent particles involved .
this discontinuity in the first derivative of the wave function and the continuity of the wave function at @xmath77 give a two - body scattering relation between the adjacent vector coefficients @xmath81 .
the matrix operator @xmath82 is defined by @xmath83 where @xmath24 is the identity and @xmath84 is the permutation operator acting on the vector @xmath85 . due to the fermi statistics , @xmath86 for all @xmath87 and @xmath88 .
yang denoted the unequal indices @xmath89 in the three particle scattering process as the interchanges with coordinates @xmath90 and @xmath91 under the permutation @xmath92 .
the consistency condition for factorizing the many - body scattering matrix into the product of two - body scattering matrices @xmath82 leads to the celebrated ybe @xmath93 where @xmath94 . defining the @xmath4-matrix by @xmath95 and spectral paramters @xmath96 and @xmath97 the ybe
is often written in the form @xmath98 returning to solving the problem of @xmath44 particles in a periodic box of length @xmath47 , the second step is to apply the periodic boundary condition @xmath99 on the wavefunction with period @xmath47 for every @xmath100 .
the two - column young tableau @xmath101 $ ] @xcite encodes the spin symmetry , where @xmath102 and @xmath103 are the numbers of fermions in the hyperfine levels @xmath104 and @xmath51 such that @xmath105 .
this gives the second eigenvalue problem @xmath106 where @xmath107 is an abbreviation of the amplitude of the wave function which provides the eigenvector of the @xmath44 operators @xmath108 with @xmath109 @xmath110 using bethe s hypothesis again , yang solved the eigenvalue problem ( [ ba2 ] ) by the ansatz @xmath111 where @xmath112 are the coordinates of the down - spin fermions and @xmath113 are the spin rapidities within the function @xmath114 . by the symmetry of the young tableau
@xmath115 $ ] , the vector @xmath116 describes a spin system with a number of @xmath103 spins on an @xmath44-site lattice .
the generalized ansatz ( [ func - a ] ) plays an important role in solving multi - component many - body problems @xcite . alternatively , the eigenvalue problem ( [ ba2 ] ) can be worked out in a straightforward way in terms of the qism , where the operator @xmath117 can be written in terms of the quantum transfer matrix @xcite .
this approach was introduced in the study of the 1d hubbard model @xcite .
the energy eigenspectrum is given in terms of the quasimomenta @xmath118 of the fermions via @xmath119 subject to the ba equations which in terms of the function @xmath120 are @xmath121 for @xmath122 and @xmath123 .
all wave numbers @xmath124 are distinct and uniquely define the wave function ( [ wave ] ) @xcite .
the fundamental physics of the model is determined by the ba equations ( [ be ] ) . for repulsive interaction ,
the quasimomenta @xmath125 are real , but the rapidities @xmath126 are real only for the groundstate .
the complex roots @xmath127 are the spin strings for excited states . in the thermodynamic limit
, i.e. , @xmath128 , where @xmath129 is finite , the ba equations ( [ be ] ) can be written as generalized fredholm equations @xmath130 where the integration boundaries @xmath131 and @xmath132 are determined by @xmath133 . in the above equations , the kernel function @xmath134 .
the functions @xmath135 denote the bethe root distributions , with @xmath136 the quasimomenta distribution function and @xmath137 the spin rapidity parameter distribution function .
the groundstate energy per unit length is given by @xmath138 .
for the attractive regime , the quasimomenta @xmath125 of fermions with different spins form two - body bound states , i.e. , the wave numbers are complex with @xmath139 in the thermodynamic limit @xcite . here @xmath140 .
the excess fermions have real quasimomenta @xmath141 with @xmath142 . in the thermodynamic limit ,
the density of unpaired fermions @xmath143 and the density of pairs @xmath144 satisfy the fredholm equations @xmath145 the distribution @xmath144 coincides with the distribution function of the real parts of the bound states .
the linear densities are defined by @xmath146 and @xmath147 .
the groundstate energy per unit length is given by @xmath148 .
in the context of magnetism , the magnetization per unit length is defined by @xmath149 . by definition
, the groundstate energy can be expressed as a function of total particle density @xmath61 and magnetization @xmath150 . in the grand canonical ensemble
, the magnetic field @xmath46 and the chemical potential @xmath35 can be obtained via the relations @xmath151 which have been used to work out the phase diagrams of the attractive fermi gas @xcite and the repulsive fermi gas @xcite .
we now turn to extracting such information from both the discrete and continuum versions of the ba solution .
the 1d fermi gas ( [ ham-1 ] ) with spin population imbalance exhibits an unconventional pairing order that presents a major subtlety of many - body correlations in the gaudin - yang model . starting with the discrete ba equations ( [ be ] )
, we will review how the exact solution enables us to understand precisely such subtle many - body physics driven by the interaction of quantum statistics and dynamics .
in particular , we will see that for the weakly and strongly attractive coupling regimes 1d interacting fermions give significantly different phenomena : weakly bound bcs - like pairs vs tightly bound molecules . for weakly attractive interaction , i.e. ,
@xmath152 , two fermions with spin - up and spin - down form a weakly bound pair with a small binding energy @xmath153 , where the two - body binding energy is less than the kinetic energy @xcite .
the complex conjugate pair leads to an exponential decay of the wave function with a factor @xmath154 .
thus the balanced case has a bcs - like fully paired state where the size of the cooper pairs is much larger than the mean average distance between the fermions . in this weakly attractive regime
, the energy gap separating the first triplet excited state from the groundstate is found to have an asymptotic behaviour @xmath155 as @xmath156 @xcite .
in fact , the ba equations ( [ be ] ) for weak attraction give an explicit relation @xmath157 between the external field and magnetization in the thermodynamic limit .
the lower critical field gives the energy gap at @xmath158 .
this relation indicates a vanishing energy gap @xmath159 for @xmath160 . here
the magnetization is defined by @xmath161 @xcite . for a polarized gas with weak attraction , fermi statistics
lead to segmentation in quasimomentum space , i.e. , the excess fermions are located at the two outer wings in quasimomentum space , see fig .
[ fig : distribution ] . for a finite size system with arbitrary polarization @xmath162 ,
the ba equations ( [ be ] ) determine @xmath103 weakly bound bcs pairs with @xmath163 and @xmath164 unpaired fermions with real @xmath124 @xcite . in this case ,
@xmath165 and @xmath166 are symmetrically distributed around zero in the quasimomentum parameter space , see fig.[fig : distribution ] . assuming that @xmath103 is odd and @xmath44 is even , the first few leading orders of the positive roots @xmath165 and @xmath166 are determined by the equations @xmath167,\nonumber\\ \lambda_{\alpha}&\approx & \frac{2n_{\alpha}\pi}{l}+\frac{3c}{2l\lambda_{\alpha}}+\frac{c}{l}\sum_{\begin{array}{l } \beta = 1\\ \alpha \neq \beta\end{array } } ^{(m_{\downarrow}-1)/2}\left [ \frac{2\lambda_{\alpha } } { \lambda_{\alpha}^2-\lambda_{\beta}^2}\right]\nonumber\\ & & + \frac{c}{2l}\sum_{j=1}^{(n-2m_{\downarrow})/2}\left [ \frac{2\lambda_{\alpha}}{\lambda_{\alpha}^2-k_j^2}\right],\label{bad-2}\end{aligned}\ ] ] where @xmath168 , and @xmath169 .
these case @xmath170 is excluded in eq .
( [ bad-2 ] ) .
the root patterns reveal the cooperative nature of many - body effects , i.e. , an individual quasimomentum depends on that of all the particles . here
the momenta of unpaired fermions and bound pairs depend on the scattering energies between pair and between paired and unpaired fermions .
this indicates that the quantum statistics of the weakly interacting fermions is mutual according to exclusion statistics @xcite . from eq .
( [ bad-2 ] ) , the groundstate energy per unit length is given by @xcite @xmath171 here the groundstate energy ( [ e - dis - w ] ) is also valid for weakly repulsive interaction , i.e. , for @xmath172 .
this leading order correction to the interaction energy indicates a mean field effect . on the other hand , for strong attraction ,
i.e. , @xmath173 ( or @xmath174 ) , the discrete ba equations ( [ be ] ) give the root patterns @xmath175 for bound pairs and real @xmath176 for unpaired fermions @xcite . here
@xmath177 and @xmath178 , the number of unpaired fermions @xmath179 .
the binding energy @xmath180 is the largest energy scale than the kinetic energies of pairs or excess fermions . for a strong attraction ( up to order @xmath181 ) ,
@xmath44 fermions have root patterns @xcite @xmath182 where the effective statistical parameters are given by @xmath183 and @xmath184 with @xmath185 . in this strong attraction limit
, the groundstate energy per unit length is given by @xmath186 , where the energies of excess fermions and pairs are given by @xmath187 the bound states behave like hard - core bosons which can be viewed as ideal particles with fractional exclusion statistics .
however , the bound pairs have tails and they interfere with each other .
it is impossible to separate the intermolecular forces from the interference between molecules and single fermions . from this explicit form of the groundstate energy
, we see that for @xmath188 , the single atoms are repelled by the molecules , i.e. , @xmath189>0.\end{aligned}\ ] ] here @xmath190 is the groundstate energy per unit length of the balanced gas .
this result indicates that the single atoms are repelled by the molecules on the tightly bound dimer limit .
the atom - dimer scattering problem of three fermions in a quasi - one - dimensional trap has been studied @xcite . in higher dimensions ,
a spin - down fermion immersed in a fully polarized spin - up fermi sea gives rise to the quasiparticle phenomenon called fermi polaron @xcite .
the fermi polaron is a spin - down impurity fermion dressed by the surrounding scattered fermions in the spin - up fermi sea . in particular , recent observations of fermi polarons in a 3d or 2d tunable fermi liquid of ultracold atoms @xcite provide insightful understanding of quasiparticle physics in many - body systems . for an attractive polaron , with increasing attraction ,
the single spin - down fermion undergoes a polaron - molecule transition in the fermionic medium @xcite .
for repulsive interaction , theoretical studies have suggested the existence of such novel quasiparticles repulsive polarons @xcite .
the properties of repulsive polarons , such as the energy , life time , and quasiparticle residue , give a fundamental understanding of the coherent nature of the quasiparticle .
the repulsive polaron is metastable and eventually decays to either a molecule state or an attractive polaron with particle - hole excitations in the majority fermi sea .
this quasiparticle phenomenon was experimentally observed by a magnetically tuned feshbach resonance on the bec side with positive scattering length @xcite .
so far , most studies concerning the first order nature of the polaron - molecule transition in a 3d fermionic medium @xcite involve a variational ansatz with some approximations that are ultimately not justified in low dimensions @xcite .
it is generally accepted that there actually do not exist quasiparticle excitations in 1d systems due to the collective nature of the 1d many - body effect .
the elementary excitations in 1d are still eigenstates , where all particles are involved in a low energy nature .
therefore , we can not find a simple operator , acting on the groundstate , to get a quasiparticle excitation , unlike for higher dimensional systems .
but this does not rule out a well - defined quasi - particle like behaviour , e.g. , a polaron , which is a typical example of the collective nature of the 1d many - body effect .
the quantum impurity problem in 1d trapped ultracold atoms has shed new insight on the collective nature of particles @xcite .
the ba solvable models are likely to provide a rigorous treatment of polaron - like phenomena in different mediums @xcite .
in particular , a 1d attractive polaronic phenomenon does occur if one ( or a few ) spin - down fermion ( fermions ) is ( are ) immersed into a large spin - up fermi sea @xcite .
the excitation energy of a system with one spin - down fermion has a certain momentum - dependent relation , which includes a mean field attractive binding energy plus a classical kinetic energy of polaron with effective mass @xmath191 . a cartoon picture is shown in fig .
[ fig : cartoon ] . .
for strong attraction ( lower panel ) , the single impurity fermion binds with one spin - up fermion from the fermi sea to form a tightly bound molecule ( yellow circle ) of a two - atom with a sole molecule binding energy and an effective mass @xmath192 . ]
mcguire ( 1965,1966 ) studied the exact eigenvalue problem of @xmath2 fermions of the same spin and one fermion of the opposite spin .
he calculated the energy shift caused by this extra spin - down fermion by solving the equation @xmath193 for the quasi momentum @xmath194 with @xmath109 and @xmath195 . here
@xmath196 for an attractive interaction strength .
mcguire found a hermitian conjugate pair @xmath197 and @xmath3 real roots @xmath198 .
the energy is given by @xmath199 .
this single impurity problem was recently studied @xcite by means of the ba equations ( [ be ] ) . for an attractive interaction , the quasimomenta @xmath200 of a pair and @xmath3 real roots @xmath118 with @xmath201
are determined by the equations ( [ be ] ) with @xmath202 .
it was found @xcite that the imaginary part @xmath203 in the pair is determined by the equation @xmath204 .
for an excited state with total momentum of the system @xmath205 , the spin - down fermion associated with the weakly bound pair in the fully - polarized fermi sea thus has a nonzero momentum @xmath206 which depends on all individual momenta of the spin - up fermions .
this givens a collective signature of the 1d many - body effect .
thus the energy shift is explicitly given by @xmath207 which behaves like a polaron quasiparticle . here ,
the attractive mean field binding energy of the polaron is given by @xmath208 for weak attraction .
the fermi energy is @xmath209 .
we see that this binding energy depends solely on the fermi energy of the medium and interaction strength in 1d . in eq .
( [ polaron - e2 ] ) , the polaron - like state has an effective mass @xmath210 which is almost the same as the actual mass of the fermions due to the decoupling from the bound pair in the weak coupling limit .
we point out that the polaron - like state only occurs for few impurity fermions immersed into a fully - polarized fermi sea . for a weakly repulsive interaction and in the thermodynamic limit
, the low energy physics of the 1d fermi gas is described by a spin - charge separation theory .
the spin rapidity parameters decouple from the quaismomenta of the fermions .
however , using the ba equations ( [ be ] ) , a single spin - down fermion immersed into the 1d fully - polarized fermi sea with weak repulsion can form a repulsive fermi polaron , with energy of the form ( [ polaron - e2 ] ) and an effective mass @xmath210 . but here the impurity fermion receives a positive mean field shift @xmath211 from the fermionic medium . in the opposite limit , a spin - down fermion immersed into a fully polarized spin - up medium with strong attraction , i.e. , with @xmath212 , the bound pair has @xmath200 and the @xmath3 real roots @xmath118 with @xmath201 are given by @xcite @xmath213 with @xmath214 . for an excited state with total system momentum @xmath205 , the relation between the centre - of - mass pair quasimomentum @xmath215 and the total momentum of the system @xmath205 is given by @xmath216 $ ] , which is independent of the individual quasimomenta of the spin - up fermions .
the energy shift is given by @xmath217 with the chemical potential @xmath218 , where the molecule energy is given by @xmath219 the binding energy of the bound state is @xmath220 which tends to the binding energy of a sole molecule @xmath221 in the strongly attractive regime @xmath222 .
the effective mass of the molecule @xmath223 becomes twice the actual mass of the fermions in this limit . from the shift energies ( [ polaron - e2 ] ) and
( [ polaron - e2-m ] ) , we see clearly that as the attractive interaction grows , the spin - down fermion binds only with one spin - up fermion from the medium to gradually form a tightly bound molecule .
the polaron - molecule crossover is regarded as a change from a mean field attractive binding energy of a polaron with an effective mass @xmath224 to the binding energy of a single molecule with an effective mass @xmath225 as the attraction grows from zero to infinity .
the non - equilibrium dynamics of an impurity in a 1d lattice within a harmonic trap have been studied using numerical methods and the ba solution @xcite .
the numerical simulation of an impurity injected into a 1d quantum liquid has been reported @xcite . in the last section we discussed the solutions to the discrete ba ( [ be ] ) in the limits @xmath226 .
they give rise to different phenomena in the two extreme limits .
usually , the many - body phenomena of interest refers to the physics of the system in the thermodynamic limit , where @xmath227 keeping @xmath129 finite .
this entails considering the solutions to the two sets of fredholm equations ( [ be - r ] ) and ( [ fermi2-a ] ) for the repulsive and attractive regimes . in order to conceive the physical nature of the super tonks - girardeau gas , we first recall the lieb - liniger bose gas with zero - range delta - function interaction , where the tonks - girardeau gas was determined by a fermi - bose mapping to an ideal fermi gas @xcite . for a strong attractive interaction , mcguire ( 1964 )
predicted that the quasimomenta of the bosons are given in terms of a bound state of @xmath44-particles , with @xmath228 $ ] with @xmath229 . in this case , the wave function is approximately given by @xmath230 where @xmath231 is a normalization constant .
the energy of the mcguire cluster state is given by @xmath232 .
however , if the interaction strength is abruptly changed from strongly repulsive to strongly attractive , the highly excited gas - like state may be metastable against this cluster - like state due to the fermi pressure inherited from the repulsive tonks - girardeau gas .
this gas - like state exhibits a more exclusive quantum statistics than the free fermi gas , and is called super tonks - girardeau gas .
the super tonks - girardeau gas - like state was first predicted by astrakharchik _
( 2005 ) and further proved by batchelor _
( 2005b ) using the exact ba solution of the lieb - liniger bose gas .
remarkably , such a highly excited state was realized in a breakthrough experiment @xcite .
the equally populated components in an attractive fermi gas give rise to physics related to the cross - over between a bcs superfluid and a bec @xcite . in this context , for a balanced attractive fermi gas with @xmath233 , the discrete ba equations ( [ be ] ) reduce to @xmath234 with @xmath235 .
this is equivalent to the equations @xmath236 with @xmath237 for the lieb - liniger gas in the super tonks - girardeau phase under the identification @xmath238 , @xmath239 and @xmath240 @xcite .
since the bound pair formed by two fermions with opposite spin has a mass @xmath240 , the @xmath241 bound pairs are equivalently described by the super tonks - girardeau phase of the interacting bose gas with effective 1d scattering length @xmath242 .
this relation is also obtained by an exact mapping based on the two - body scattering problem associated with bec - bcs crossover @xcite . for the balanced fermi gas ,
the binding energy is subtracted from the energy that gives the energy of the bosonic pairs of a two - atom , with result @xmath243 .
the energy eigenvalues of the bosons are given by @xmath244 .
in this regard , the groundstate of the balanced attractive fermi gas can be viewed as the fermionic super tonks - girardeau gas @xcite . the identification between
the balanced attractive fermi gas and the attractive lieb - liniger bose gas suggests an effective attraction between pairs . in the thermodynamic limit , the ba equations ( [ lamda ] ) give a particular fredholm equation which can be deduced from ( [ fermi2-a ] ) , i.e. , @xmath245 where the fermi pair momentum @xmath246 is determined by @xmath247 .
it turns out that @xcite the reduced fredholm eq .
( [ bec - bcs ] ) maps to the lieb - liniger integral equation for 1d spinless bosons on identifying @xmath240 , @xmath248 and @xmath249 . for a weak attraction , the groundstate is the bcs - like pairing state with a pairing correlation length larger than the average interparticle spacing and the energy is given by @xmath250 .
in particular , for strong attraction , the groundstate of bound pairs determined by ( [ bec - bcs ] ) can be regarded as a particular super tonks - girardeau gas of hardcore bosons @xcite .
the distribution function of the pair density plotted in fig .
[ bec - bcs crossover ] provides an understanding of the subtle bec - bcs crossover in the balanced fermi gas . in the weak coupling regime
, the single quasimomentum essentially depends on that of the other particles .
this gives a signature of mutual statistics @xcite .
however , in the limit @xmath251 , the quasimomentum distribution becomes an equally - spaced separation .
this indicates free fermi non - mutual statistics .
further study on the dimer - dimer scattering properties in the confinement induced resonance has been reported @xcite , see also @xcite . for different values of interaction strength @xmath252 @xcite .
the analytic result for the distribution function matches the numerical solution .
the quasimomentum distribution indicates the fermionization from mutual statistics to non - mutual generalized exclusion statistics as @xmath252 increases . from iida and wadati ( 2005 ) . ]
furthermore , it was demonstrated @xcite that another metastable highly excited gas - like state without bound pairs in the strongly attractive regime can be realized through a sudden switch of the interaction from strongly repulsive to strongly attractive . in the limit @xmath253 ,
this gas - like state is still an eigenstate of the system with the energy per particle @xmath254 , but it is a highly excited state . from the experimental point of view , these different quantum states can be tested from measuring the frequencies of the lowest breathing mode from the mean square radius of the 1d trapped gases in a harmonic potential @xcite , e.g. , in the super tonks - girardeau bose gas @xcite . the low breathing mode featuring different states of the strongly repulsive and attractive fermi gas can be analyzed via the local density approximation .
the lowest breathing mode is given by the mean square radius of the trapped fermionic tonks - girardeau gas @xmath255 , see fig .
[ breathering - mode ] . here
the frequency ratio @xmath257 exhibits a peak which is a typical characteristic of the super tonks - girardeau phase .
further evidence for this fermionic super tonks - girardeau gas - like state has been seen in the experimental observation of the fermionization of two distinguishable fermions @xcite .
it is also particularly interesting that a ferromagnetic transition is likely to occur in 1d strongly interacting fermions across the resonance from infinite repulsion to finfinite attraction @xcite . .
the quantum gases are trapped in a 1d harmonic potential @xmath258 . here
@xmath259 and @xmath260 stand for the frequency ratio @xmath261 for the groundstate and fermionic super tonks - girardeau gas , respectively . from l .- m .
guan and chen ( 2010 ) . ]
despite the two sets of fredholm integral equations ( [ be - r ] ) and ( [ fermi2-a ] ) for the homogeneous gas being derived long ago @xcite , their analytical study are still restricted to particular regimes , e.g. , @xmath262 , @xmath263 and @xmath264 .
the first few terms in the asymptotic expansions of the groundstate energy for the 1d attractive fermi gas for both the strong and weak coupling cases has been calculated in terms of power series @xcite and in terms of legendre polynomials @xcite .
the first few terms of the groundstate energy have been derived recently @xcite by an asymptotic expansion for ( a ) strong repulsion , ( b ) weak repulsion , ( c ) weak attraction and ( d ) strong attraction . for strong repulsion ,
the groundstate energy of the gaudin - yang model is given by @xcite @xmath265 , & { \rm for } \,\
, p=0,\\ \frac{1}{3}n^3\pi^2\left [ 1-\frac{8n_{\downarrow}}{c } + \frac{48n_{\downarrow}^2}{c^2 } \right .
-\frac{1}{c^3}\left(256n_{\downarrow}^3-\frac{32}{5}\pi^2n^2n_{\downarrow}\right ) \right ] , & { \rm for } \,\ ,
} \right .
\label{energy - sr}\end{aligned}\ ] ] here @xmath266 is the riemann zeta function .
the leading order ( @xmath267 ) correction was also found in @xcite .
[ fig : energy ] shows that this groundstate energy is a good approximation for the balanced and imbalanced fermi gas with a strongly repulsive interaction . for strong attraction ,
the groundstate energy is given by @xmath268 , where @xmath269,\label{e0u - sa}\\ e_0^b & \approx & \frac{n_{\downarrow}^3\pi^2}{6}\left[1+\frac{2(2n_{\uparrow}-n_{\downarrow})}{|c| } + \frac{3(2n_{\uparrow}-n_{\downarrow})^2}{c^2 } \right.\nonumber\\ & & \left .
-\frac{4}{15|c|^3 } \left ( 180n_{\downarrow}n_{\uparrow}^2 + 20\pi^2n_{\uparrow}^3 - 90n_{\uparrow}n_{\downarrow}^2 - 22\pi^2n_{\downarrow}^3\right.\right.\nonumber\\ & & \left.\left.+15n_{\downarrow}^3 - 120n_{\uparrow}^3 + 63\pi^2n_{\downarrow}^2n_{\uparrow}-60\pi^2n_{\downarrow}n_{\uparrow}^2\right ) \right].\label{e0b - sa}\end{aligned}\ ] ] this energy is highly accurate for arbitrary polarization as can be seen in fig .
[ fig : energy ] . the high precision of expansions for the groundstate energy of the attractive fermi gas were also studied @xcite .
in contrast to the strong coupling case , it is more difficult to proceed with asymptotic expansion for the two sets of fredholm equations ( [ be - r ] ) and ( [ fermi2-a ] ) at vanishing interaction strength . in terms of the polarization @xmath70 ,
the groundstate energy in weak attraction limit was found to be @xcite @xmath270 the coefficient @xmath132 is a complicated function obtained from the power series expansions with respect to @xmath252 .
however , it contains divergent sums and the coefficients are singular as @xmath160 @xcite .
so far , only the leading order correction to the interaction energy is mathematically convincing and consistent with the result ( [ e - dis - w ] ) obtained from the discrete ba equations ( [ be ] ) . beyond the mean field term ,
finding the next leading term in the groundstate energy is still an open problem . for zero polarization , krivnov and ovchinnikov ( 1975 )
found @xmath271 obtained from the 1d hubbard model in dilute limit .
iida and wadati ( 2007 ) found the term @xmath272 .
this difference reveals a subtlety of the vanishing interaction limit , i.e. , the two limits @xmath273 and the thermodynamic limit ( @xmath274 with @xmath129 finite ) do not commute .
in fact , the groundstate energy ( [ energy - iida ] ) only counts the density distributions away from the integration boundaries in the fredholm equations ( [ be - r ] ) and ( [ fermi2-a ] ) , i.e. , @xmath275 and @xmath276 with @xmath277 . at the integration edges , these distribution functions are not analytically extractable as @xmath278 @xcite .
the analyticity of the groundstate energy at @xmath279 is still in question @xcite .
takahashi showed that ( a ) the groundstate energy function @xmath280 is analytic on the real @xmath281 axis when @xmath282 and ( b ) @xmath283 is analytic on the real @xmath281 axis except for @xmath284 when @xmath285 .
however , the fredhom equations for weakly repulsive and attractive interactions are identical as long as @xmath286 and @xmath287 where the integration boundaries match each other between the two sides @xcite . in this identical region ,
the asymptotic expansions of the energies of the repulsive and attractive fermions are identical to all orders as @xmath273 .
but the identity of the asymptotic expansions may not mean that the energy analytically connects due to the divergence of the fredholm equations in the region @xmath288 . in units of @xmath289 .
the figure shows the comparison between the asymptotic solutions and numerical solutions of the fredholm equations for different polarization . in the attractive regime ,
the binding energy @xmath290 was subtracted . from guan and ma ( 2012 ) . ]
so far , we have only discussed the groundstate properties of the gaudin - yang model .
we now survey the wide range of fundamental many - body physics exhibited in the model .
the particularly interesting feature of the attractive fermi gas is the exotic fflo - like pairing , where the system is gapless with mismatched fermi points between the two fermi seas . in the gapped phase
, it is well understood that the correlation function for the single particle green s function decays exponentially @xcite @xmath291 with @xmath292 and @xmath293 , whereas the singlet pair correlation function decays as a power of distance , i.e. , @xmath294 . here
@xmath295 is the energy gap , and the critical exponents @xmath296 and @xmath297 are both greater than zero
. however , once the external field exceeds the lower critical field , the system has a gapless phase where both of these correlation functions decay as a power of distance and the pairs lose their dominance .
the molecule and excess fermions form the polarized fflo - like pairing state , where the spatial oscillations of the pairing correlation are caused by an imbalance in the densities of spin - up and spin - down fermions , i.e. , @xmath298 . in section [ cft - cf ]
, we will further discuss the pair and spin correlations with the spatial oscillation signature in the context of conformal field theory . plane .
the phase boundaries are obtained from eq .
( [ mu - h ] ) in terms of the numerical solution of the ba equations ( [ be - r ] ) . from orso ( 2007 )
. + lower panel : phase diagram of the model in the @xmath299 plane with @xmath300 and density @xmath301 .
the dashed lines denote the two critical lines ( [ hc1 ] ) and ( [ hc2 ] ) .
the coloured phases are obtained by numerical solution of the energy - magnetization ( [ e - mz ] ) . from he _
, title="fig : " ] + in terms of the polarization , the gaudin - yang model with attractive interaction exhibits three quantum phases at zero temperature : the fully paired phase which is a quasicondensate with zero polarization , the fully polarized normal fermi gas with @xmath302 , and the partially polarized fflo - like phase with polarization @xmath303 , see the phase diagram in @xmath304 plane and in @xmath299 plane , see fig .
[ fig : phase - s ] .
the two phase diagrams describe the quantum phases in terms of the grand canonical and canonical ensembles .
the phase boundaries in the two phase diagrams can be mapped onto each other @xcite . in this gapless phase
, the magnetic properties can be exactly described by the external field - magnetization relation @xmath305 where @xmath306 and @xmath307 are given by ( [ mu - h ] ) .
this relation reveals an important energy transfer relation among the binding energy , the variation of fermi surfaces and the external field . for fixed density and strong attraction ,
the paired phase with magnetization @xmath308 is stable when the field @xmath309 , where the lower critical field is given by @xmath310 .
\label{hc1}\ ] ] when the external field exceeds the upper critical field @xmath311 , \label{hc2}\ ] ] a phase transition from the fflo - like phase into the normal gas phase occurs , see fig .
[ fig : phase - s ] .
the lower critical field gives the energy gap in the spin sector .
the magnetization can be obtained from ( [ e - mz ] ) , see fig .
[ fig : mz - s ] .
it was found @xcite that in the vicinity of the critical fields @xmath312 and @xmath313 , the system exhibits a linear field - dependent magnetization @xmath314 \end{array}\right.\label{f2-mz}\end{aligned}\ ] ] with a finite susceptibility . for fixed total number of particles , or say in a canonical ensemble , the magnetic field driven phase transitions in the 1d fermi gases with an attractive interaction are linear - field - dependent , which was also found in the @xmath315 attractive fermi gas @xcite .
the magnetism of the attractive fermi gases has been discussed by schlottmann @xcite .
however , the argument which was made on the initial slope of the magnetization in these papers @xcite does not appear to be correct for a fixed total number of particles .
the reason has been discussed @xcite : the bound pairs which have to be broken up to yield the particles with uncompensated spins form a fermi sea , their density of states is finite at the fermi level , and that keeps the initial susceptibility finite " .
it is also shown @xcite that the curvature of free dispersion at the fermi points couples the spin and change modes and leads to a linear critical behaviour and finite susceptibility for a wide range of models .
they showed that when the magnetic field @xmath316 , the magnetization @xmath317 for a fixed chemical potential .
however , for fixed density , the magnetization @xmath318 as @xmath319 .
this leads to a finite onset susceptibility given by @xmath320 with the pair density stiffness @xmath321 in the strong attraction limit @xmath322 .
here we further remark that for finitely strong attraction the onset susceptibility @xmath323 where @xmath324 is the tomonaga - luttinger liquid parameter at the critical point and @xmath325 is the stiffness of bound pairs in the limit @xmath326 .
the magnetization in the hubbard model with a half - filled band gives rise to the square root dependence on the field @xcite , where low density solitons appearing in the spin sector above the critical field behave like free fermions @xcite .
more rigorously speaking , the linear field - dependent magnetization is clearly seen from the energy transfer relation ( [ e - mz ] ) , where the effective chemical potentials @xmath327 and @xmath328 .
thus the linear term @xmath329 in the relation ( [ e - mz ] ) gives a finite susceptibility at the onset of magnetization . in experiments ,
1d quantum atomic gases are prepared by loading ultracold atoms in an anisotropic harmonic trap with strong transverse confinement and weak longitudinal confinement . in general , interacting many - body systems trapped in a harmonic potential is a rather complicated problem .
the problem of the 1d hamiltonian ( [ ham-1 ] ) trapped in a harmonic potential @xmath330 has been studied by various methods @xcite . for @xmath331 ,
the eigenvalue problem of the trapped gas has been studied analytically in @xcite .
the energy shift for @xmath331 @xcite is given by @xmath332 , where @xmath333 is a scattering length and @xmath334 is the euler gamma function @xcite .
the system of two fermions with arbitrary interaction in a 1d harmonic potential has been experimentally investigated @xcite . in this experiment , the tonks - girardeau state and the metastable super tonks - girardeau state have been observed .
this problem for arbitrary number of particles was studied analytically @xcite , where the limiting cases @xmath335 and @xmath284 have been studied using group theory .
in particular , yang ( 2009 ) gave an analysis of the groundstate energy of fermions in a 1d trap with @xmath42-function interaction .
in the light of yang s argument , for any value of interaction strength , the eigenvalue problems : ( a ) the trapped hamiltonian with symmetry @xmath336 $ ] in full @xmath337 space and ( b ) the hamiltonian in region @xmath338 with the boundary condition that the wave function vanishes on its surface , are equivalent . here
the region @xmath338 is bounded by @xmath339 planes at which the wave function @xmath340 vanishes . for any value of @xmath341
, the groundstate wave function for problem ( b ) has no zeros in the interior of @xmath338 and is not degenerate .
this thus suggests that the groundstate energy of the system with total spin @xmath342 increases monotonically and approaches to the energy @xmath343 .
the lieb and mattis theorem @xcite further suggests @xmath344 if @xmath345 . for @xmath346 ,
the groundstate energy of the trapped gas with total spin @xmath347 is given by @xmath348 , which is independent of the total spin @xmath347 .
for @xmath284 and @xmath342 , the energy is given by @xmath349 ^ 2+[n/2-j]^2 \right)$ ] .
ma and yang ( 2009 ) argued that @xmath350 with @xmath351 where @xmath352 .
in particular , for @xmath346 , the exact wave function of the system @xmath353 where the spatial wave function and symmetric spin wave function have been derived explicitly @xcite @xmath354_{i=1,\ldots , n}^{j=1,\ldots , n}\nonumber \\ \psi_j&= & \sum _ { \alpha = 1}^{n!/((n - m ) ! m!)}\left\{{\cal y } _
\alpha ^{[n - m , m ] } q_\alpha \right\}z_\alpha \nonumber\end{aligned}\ ] ] for the symmetry @xmath355
@xmath356 with @xmath357 .
the basis tensor function @xmath358}$ ] was constructed explicitly from group theory @xcite .
the fermion density distribution for 1d interacting fermions with harmonic trapping has strong oscillations on top of a uniform density cloud @xcite .
these oscillations can be described by an analytical form of the density distribution @xcite @xmath359 for @xmath360 , where the density cloud is given by the thomas - fermi profile , i.e. , @xmath361 with a thomas - fermi radius @xmath362 .
if the longitudinal confinement is weak enough , the atomic density varies smoothly along the longitudinal direction and so the atomic gases can be treated as locally homogeneous systems @xcite .
this type of approximate treatment is known as the local density approximation ( lda ) . in this way
density functional theory has been used to study 1d interacting fermions @xcite . to ensure the validity of the lda
, the correlation length @xmath363 should be much smaller than the characteristic inhomogeneity length @xmath364 , i.e. , @xmath365 .
the two length scales @xmath363 and @xmath366 are determined by the local chemical potential @xmath367 and the local density @xmath368 . obviously , from the definition of @xmath366 , the lda becomes invalid near the edge of an atomic cloud where the density drops rapidly .
however , in real measurements , almost all signal strengths are proportional to the density . therefore , due to the very small density at the edge
, the central region of large density dominates the measurement signals . for a large number of particles , @xmath369 ,
the density profiles of the trapped gas can be precisely analyzed within the lda . in a harmonic trap ,
the equation of state ( [ mu - h ] ) can be reformulated within the lda by the replacement @xmath370 in which @xmath371 is the position and @xmath372 is the frequency within the trap . using the lda for the 1d bose gas in a harmonic trap , its global chemical potential reads @xcite @xmath373-v(z ) = \mu_{0}[n(z ) ] -\frac{1}{2}m\omega_{z}^{2 } z^2,\ ] ] where the local chemical potential @xmath374 $ ] at position @xmath36 is given by the chemical potential for a homogeneous system of an uniform density @xmath375 .
the total number of atoms @xmath44 is given as @xmath376 .
similarly , for a 1d two - component fermi gas in a harmonic trap , the global chemical potential is @xcite @xmath377=\mu_0-\frac{1}{2}m\omega_x^2x^2\nonumber\ ] ] where the chemical potential @xmath378 $ ] can be obtained from the homogenous gas ( [ mu - h ] ) .
@xmath379 is the total linear number density and @xmath380 is the local spin polarization .
they can be determined from restriction on the total particle number @xmath381 and polarization @xmath382 which are rewritten as @xcite @xmath383 here @xmath384 , @xmath385 and @xmath386 . if the trapping potentials are the same for the two spin components , calculations for the integrable homogenous attractive gas confined to a 1d trapping potential thus lead to a two - shell structure composed of a partially polarized 1d fflo - like state in the trapping centre surrounded by wings composed of either a fully paired state or a fully polarized fermi gas @xcite , see fig .
[ fig : phase - attraction ] .
this prediction was verified by liao _
( 2010 ) by the observation of three distinct phases in experimental measurements of ultracold @xmath22li atoms in an array of 1d tubes .
the analytical study of the phase diagram of the 1d attractive fermi gas has been presented in @xcite . and inner shell @xmath387
are theoretically predicted by means of the ba equations ( [ be - r ] ) within the lda .
this figure shows the radii vs polarization @xmath70 for the values of the parameter @xmath388 . from orso ( 2007 ) . ]
the tll @xcite , describing the collective motion of bosons , has played an important role in the novel description of universal low energy physics for low - dimensional many - body physics @xcite . in 1d systems of interacting bosons , fermions or spin systems ,
the effect of quantum fluctuations is strong enough to yield striking anomalous quantum phenomena . in this approach ,
for example , the low energy physics of a 1d interacting fermion system can be described by a bilinear form of bosonic creation and annihilation operators .
the tll is phenomenologically treated by bosonization techniques @xcite based on a linearization of the dispersion relation of the particles in the collective motion , i.e. , @xmath389 , here @xmath390 is the sound velocity of the collective motion .
in contrast to the fermi liquid , this thus leads to a power law density of states for the tll at the fermi energy @xmath391 , i.e. , @xmath392 , where the exponent @xmath393 depending on the so - called tll parameter @xmath394 . in general
, the correlation functions of such 1d systems at zero temperature show a power - law decay determined by the tll parameter @xmath394 and the velocity @xmath390 .
these critical systems not only have global scale invariance but exhibit local conformal invariance . with the help of exact ba solutions , a wide class of 1d interacting systems can be mapped onto tlls in the low - energy limit , including the electronic systems with spin degrees of freedom like spin - charge separation @xcite .
moreover , progress in treating such collective motion of particles beyond the low - energy limit was made by imambekov and glazman ( 2009a,2009b ) and imambekov _ et al . _
this method can be applied to a wide variety of 1d systems with collective motion of particles .
this generalized tll theory could be possibly justified through exact ba results for 1d integrable models in ultracold atoms and correlated electronic systems .
in contrast to the conventional quasiparticles carrying both spin and charge degrees , the elementary excitations form spin and charge waves that propagate with different velocities in 1d @xcite . the relativistic dispersion relation for each one of these excitations
is written as @xmath395 , where @xmath396 is the energy gap and @xmath397 is the velocity . for a gapless excitation with vanishing energy gap @xmath398 . for 1d interacting systems ,
this gives a phonon dispersion that leads to conformal invariance in the excitation spectrum .
however , for a large energy gap , the dispersion can be rewritten as @xmath399 which is the classical dispersion of a free particle with an effective mass @xmath400 . from the ba solution ( [ be ] ) with @xmath401 ,
the charge and spin velocities are @xmath402 for the weak coupling regime @xcite . here the fermi velocity @xmath403 . for strong attraction ,
the charge and spin velocities are given by @xmath404 and @xmath405 @xcite with an energy gap @xmath406 .
this gap increases with increasing interaction strength @xmath252 so that the spin velocity is divergent in the strongly attractive limit .
however , for strong repulsion the charge velocity @xmath407 tends to the fermi velocity and the spin velocity goes to zero @xmath408 @xcite due to suppression of spin transportation due to the strong repulsion , see fig .
[ fig : velocities ] .
we shall discuss tlls and spin - charge separation phenomena in the attractive fermi gas in the following two sections .
for the 1d balanced fermi gas .
the solid lines are the velocities obtained by numerically solving the ba equations ( [ be ] ) .
the dotted lines denote the analytical result for the velocity in the weak coupling regimes . from batchelor _ et al . _
( 2006a ) .
, title="fig : " ] the yang - yang formalism with its generalisation for the study of thermodynamics of ba integrable systems @xcite is a convenient tool for the study of universal thermodynamics and quantum criticality in the presence of external fields . at finite temperatures and in the thermodynamic limit
, the densities in the fredholm equations ( [ fermi2-a ] ) evolve into occupied and unoccupied roots in the whole parameter spaces , namely @xmath409 .
in particular , the roots in the spin sector form complicated string patterns that characterize the spin excitations , i.e. , spin wave bound states .
the density distribution functions of pairs , unpaired fermions and spin strings involve the densities of ` particles ' @xmath410 and ` holes ' @xmath411 ( @xmath277 ) . following the yang - yang grand canonical ensemble method ,
the grand partition function is written as @xmath412 , in terms of the gibbs free energy @xmath413 and the magnetic field @xmath46 , the chemical potential @xmath35 and the entropy @xmath5 @xcite . in terms of the dressed energies
@xmath414 and @xmath415 for paired and unpaired fermions , the equilibrium states are determined by the minimization condition of the gibbs free energy , which gives rise to a set of coupled nonlinear integral equations
the tba equations @xcite . for the attractive gaudin - yang model ,
these equations are @xmath416 @xmath417 the function @xmath418 is the ratio of the string densities . here
@xmath419 denotes the convolution integral @xmath420 . the function @xmath421 is given , e.g. , in @xcite .
the gibbs free energy per unit length is given by @xmath422 where the effective pressures of the bound pairs and unpaired fermions are given by @xmath423 in the grand canonical ensemble , the total number of particles associated with the chemical potential @xmath35 can be changed .
the fermi sea of unpaired fermions can be lifted by the external field .
the entropy @xmath5 is a measure of the thermal disorder .
the spin fluctuations ( spin strings ) are ferromagnetically coupled to the fermi sea of unpaired fermions .
the direct numerical computation of the tba equations was presented by kakashvili and bolech ( 2009 ) .
the tba equations for this model involve an infinite number of coupled nonlinear integral equations that impose a number of challenges to accessing the physics of the model . for zero external field , the lowest excitations split into collective excitations carrying charge and collective excitations carrying spin .
this leads to the phenomenon of spin - charge separation .
the charge excitations are described by sound modes with a linear dispersion . however , for the external field in excesses of the lower critical field the spin gap vanishes .
in contrast to the spin - charge separation formalism , the spin - charge coupling drastically changes the critical behaviour in the attractive regime of the fermi gas .
the tba equations ( [ tba - full ] ) indicate that the spin fluctuations ( the spin wave bound states ) are ferromagnetically coupled to the fermi sea of unpaired fermions @xcite . in contrast to the antiferromagnetic coupling @xmath424 for repulsive regime @xcite , the spin - spin exchange interaction in the spin sector is described by an effective spin-1/2 ferromagnetic chain with a coupling constant @xmath425 in the strong coupling regime @xmath426 .
the ferromagnetic spin wave fluctuations are produced due to the thermal fluctuation in the fermi sea of unpaired fermions .
however , @xmath427 tends to zero for @xmath322 . therefore the spin transportation becomes weaker and weaker until it vanishes as @xmath428 . at zero temperature
all unpaired fermions are polarized and spin strings are fully suppressed . in this gapless phase , excitations involve particle - hole excitations and spin - string excitations . the tba equations ( [ tba - full ] ) can be greatly simplified in the strong coupling regime due to the suppression of spin fluctuations , where @xmath429 as @xmath430 .
thus one can extract the universal tll physics using sommerfeld expansion for temperatures less than chemical potential and magnetic field .
in fact , in this spinless phase , the spin fluctuation is suppressed in the limit @xmath430 and @xmath431 .
thus the bound pairs and unpaired fermions form a two - component tll .
conformal invariance predicts that the energy per unit length has a universal finite - size scaling form that is characterized by the dimensionless number @xmath432 , which is the central charge of the underlying virasoro algebra @xcite .
the finite - size corrections to the groundstate energy have been analytically derived @xcite @xmath433 where @xmath434 with @xmath435 and @xmath436 the velocities of unpaired fermions and bound pairs , respectively . for strong interaction ,
they are given explicitly by @xmath437 where @xmath438 , @xmath439 and @xmath440 .
we will describe universal behaviour of the macroscopic properties of this fermi gas in the following subsections .
although a phase transition in 1d many - body systems at finite temperatures does not exist , the system does exhibit universal crossover from relativistic dispersions to quadratic dispersions .
thus at low temperatures , the bound pairs , normal fermi gas , and the fflo phase become relativistic tlls of bound pairs ( @xmath441 ) , unpaired fermions ( @xmath442 ) , and a two - component tll ( @xmath443 ) , respectively , see fig .
[ fig : entropy1 ] .
a detailed discussion has been given @xcite .
for the temperature @xmath444 , where @xmath391 is the fermi energy , the leading low temperature correction to the free energy of the polarized gas can be calculated explicitly using sommerfeld expansion with the pressures ( [ pressure ] ) , namely @xcite @xmath445 which belongs to the universality class of the gaussian model with central charge @xmath434 . for strong attraction , the velocities are given in ( [ vb - vu ] ) . in the above equation ,
the groundstate energy @xmath446 is as given in sec .
[ ground - state ] .
in fact , from the tba equations ( [ tba - full ] ) , the universal thermodynamics ( [ ff - c ] ) can be shown to be valid for arbitrary interaction strength .
plane showing a contour plot of the entropy in the strong interaction regime .
the dashed lines are determined from the deviation from linear temperature dependent entropy obtained from the result ( [ ff - c ] ) .
the universal crossover temperatures separate the tlls from quantum critical regimes . from yi _
] the two branches of gapless excitations in the 1d fflo - like phase form collective motions of particles .
the low energy ( long wavelength ) physics of the strongly attractive fermi gas is described by an effective hamiltonian @xmath447+\frac{v_b}{2}\left[(\partial _ x \phi_{b})^2+(\partial _
x\theta_{b})^2 \right]\nonumber\\ & & -\frac{h}{2}\frac{\partial _ x \phi_{u}}{\sqrt{\pi } } -\mu \frac{(\partial _ x \phi_u+2\partial_x \phi_b)}{\sqrt{\pi}}\end{aligned}\ ] ] as long as the spin fluctuation is frozen out @xcite . here
the fields @xmath448 with @xmath449 are the density and current fluctuations for the pairs and unpaired fermions
. however , in the spin gapped phase , i.e. , for @xmath309 , the energy gap in the spin sector leads to an exponential decay of spin correlations , whereas the singlet pair correlation and charge density wave correlations have a power - law decay @xcite .
thus the system in the spin gapped phase forms a so - called luther - emery liquid @xcite .
as we have seen , the 1d attractive fermi gas exhibits various phases of strongly correlated quantum liquids and is thus particularly valuable to investigate quantum criticality . near a quantum critical point , the many - body system is expected to show universal scaling behaviour in the thermodynamic quantities due to the collective nature of the many - body effects . in the framework of yang - yang
tba thermodynamics , exactly solvable models of ultracold atoms , exhibiting quantum phase transitions , provide a rigorous way to treat quantum criticality in archetypical quantum many - body systems , such as the gaudin - yang fermi gas @xcite , the lieb - liniger bose gas @xcite , a mixture of bosons and fermions @xcite , and the spin-1 bose gas with both delta - function interaction and antiferromagnetic interaction @xcite . at zero temperature ,
the quantum phase diagram in the grand canonical ensemble can be analytically determined from the so - called dressed energy equations @xcite @xmath450 which are obtained from the tba equations in the limit @xmath430 .
the integration boundaries @xmath451 and @xmath452 characterize the fermi surfaces for bound pairs and unpaired fermions , respectively .
it is convenient to use dimensionless quantities where energy and length are measured in units of binding energy @xmath453 and @xmath454 respectively . in terms of the dimensionless quantities @xmath455 , @xmath456 , @xmath457 , @xmath458 , @xmath459 ,
the phase boundaries have been determined analytically from ( [ tba - f ] ) , see @xcite .
there are four phases denoted by vacuum @xmath460 , fully paired phase @xmath461 , ferromagnetic phase @xmath462 , and partially paired @xmath463 or ( @xmath464-like ) phase presenting the same phase diagram as in fig .
[ fig : phase - attraction ] . the low density and strong coupling limits are particularly important to study quantum criticality .
in fact , the tba equations ( [ tba - full ] ) can be converted into a dimensionless form with the above rescaling . following the notation used in @xcite ,
the phase boundaries between @xmath465 , @xmath466 , @xmath467 and @xmath468 are denoted by @xmath469 to @xmath470 respectively .
the closed forms of the critical fields @xmath471 are needed to determine scaling functions of thermodynamic properties . here
@xmath472 applies to all regimes and @xmath473 are expressions in the strongly interacting regime .
the above critical fields @xmath473 correspond to the upper and lower critical fields in the @xmath474 plane , which were found in @xcite . the tba equations ( [ tba - full-1 ] ) and ( [ tba - full ] ) encode the microscopic roles of each single particle that lead to a global coherent state quantum criticality .
the quantum criticality is manifested by universal scaling of thermodynamic properties near the critical points .
the key input to obtain critical scaling behaviour is to derive the form of the equation of state which takes full thermal and quantum fluctuations at low temperatures into account .
the dimensionless form of the pressure @xcite @xmath475 serves as equation of state , where , to @xmath476 , the pressures of the bound pairs and unpaired fermions are given by @xmath477 + o(c^4),\nonumber\\ \tilde{p}^u&=&-\frac{t^{\frac{3}{2}}}{2\sqrt{2\pi}}f_{3/2}^{u}\left[1 + 2\tilde{p}^b\right ] + o(c^4 ) \end{aligned}\ ] ] with in addition @xmath478 in these equations the functions @xmath479 , @xmath480 , @xmath481 , and @xmath482 are defined by @xmath483 and @xmath484 , with the notation @xmath485 , @xmath486 .
the function @xmath487 is the polylog function and @xmath488 . despite the equation of state ( [ eos ] )
having only a few leading terms in expansions with respect to the interaction strength , it contains thermal fluctuations in contrast to the tll thermodynamics ( [ ff - c ] ) .
the tll thermodynamics ( [ ff - c ] ) has been derived from low temperature expansion along @xmath489 .
this universal thermodynamics is a consequence of the linearly dispersing phonon modes @xcite , i.e. , the long wavelength density fluctuations of two weakly coupled gases or a gas of bound pairs or single fermions . the quantum critical regime lies beyond @xmath490 . in this limit , the equation of state ( [ eos ] ) provides closed forms for the scaling functions of thermodynamic quantities , such as density , magnetization and the compressibility .
near the critical point , the thermodynamic functions can be cast into a universal scaling form @xcite .
the explicit universal scaling form of the density for @xmath490 is @xmath491 here the constants @xmath492 and @xmath493 are the background densities near the critical points @xmath494 and @xmath495 .
these constants , together with @xmath87 and @xmath88 , are known explicitly in terms of @xmath496 @xcite .
the universal scaling form for the compressibility is @xmath497 where @xmath498 are also known @xcite . in the gaudin - yang model , the above density and compressibility can be cast into the universal scaling forms @xmath499 with dimensionality @xmath500 . here
the scaling functions are @xmath501 and @xmath502 from which one can read off the dynamical critical exponent @xmath503 and correlation length exponent @xmath504 for different phases of the spin states .
such results illustrate the microscopic origin of the quantum criticality of different spin states , i.e. , the singular parts in ( [ scaling - n ] ) and ( [ scaling - kappa ] ) characterize sudden changes of the density of state of either excess fermions or bound pairs .
the tll is maintained below the cross - over temperature @xmath505 which indicates a universal crossover from a relativistic dispersion into a nonrelativistic dispersion @xcite .
the quantum criticality driven by the external field @xmath46 gives rise to the same universality class .
using the lda presented in ( [ n_p ] ) , quantum criticality of the bulk system can be mapped out through finite temperature density profiles in the trapped gas . for small polarization ,
the chemical potential passes the lower critical point @xmath506 from the vacuum into the fully paired phase then passes the upper critical point @xmath470 from the fully paired phase into the fflo - like phase . at finite temperatures , quantum criticality of the fermi gas
can be seen clearly from contour plots of entropy in @xmath507 plane see fig .
[ fig : qc-1](a ) .
the typical @xmath508-shape crossover temperature @xmath505 separates the quantum critical regimes where @xmath509 .
the crossover temperatures are determined by minimums or maximums of the magnetization or by the breakdown of the linear - temperature - dependent entropy @xcite .
[ fig : qc-1 ] ( b ) and ( c ) show that the unpaired density curves for different temperatures intersect at the critical points @xmath510 and @xmath470 , respectively . the phase boundary separating the fully paired phase from the fflo - like phase can be mapped out from the density profiles of unpaired fermions in the trapped gas at finite temperatures .
as the temperature decreases , the compressibility evolves a round peak sitting in the phase of the higher density of state .
it diverges at zero temperature .
similarly , for the high polarization case , the density profiles of unpaired and paired atoms can be used to map out the phase boundaries @xmath469 ( @xmath511 ) and @xmath512 ( @xmath513 ) , respectively @xcite .
this signature can be used to confirm the quantum critical law , as per the recent experimental measurements @xcite . plane .
symbols indicate the crossover temperature @xmath505 separating the quantum critical regimes from vacuum , single component @xmath441 of paired fermions and two - component @xmath443 of fflo - like states .
the intersection of the density curves at different temperatures can map out the critical points ( b ) @xmath510 and ( c ) @xmath470 .
the corresponding compressibility curves ( d ) and ( e ) intersect at the same critical points after a subtraction of the background compressibility . from yin _ et al . _
( 2011 ) . ]
the universal scaling behaviour of the homogenous system can be mapped out through the density profiles of the trapped gas at finite temperatures .
however , inhomogeneity caused by the finite - size scaling effect is evident in the scaling analysis @xcite .
it has been proved that quantum criticality of the bulk system can be revealed from the singular part of a thermodynamic quantity near the trapping centre @xmath514 .
the scaling behaviour exists in the limit of large trapping size .
e.g. , under a scale change @xmath88 , the singular part of the density below the critical dimension @xmath515 can be written @xcite @xmath516 with @xmath517 and @xmath518 .
choosing @xmath519 , the scaling function with finite - size trapping is @xmath520 with @xmath521 .
the scaling behaviour is revealed through plotting @xmath522 against @xmath35 at @xmath514 for different temperatures with a fixed @xmath523 .
this means that the scaling behaviour of the homogeneous system could be extracted from the trapping centre in a small window @xcite .
nevertheless , the finite - size error lies within the current experimental accuracy @xcite .
one can either lower the temperature or increase the interaction strength such that all data curves for the physical properties at different temperatures collapse into a single curve with a proper scaling in the trapped gas .
landau s fermi liquid theory provides a universal description of low energy physics of interacting electron systems in higher dimensions , where interactions only lead to finite renormalizations of physical properties @xcite .
the quasiparticle excitations have a divergent lifetime when the excitation energy goes to zero .
renormalization of individual quasiparticles leads to a similar fermi liquid behaviour , e.g. , a finite density of states and a step - like singularity in momentum distribution at zero temperature .
the deviations from the values of physical properties of noninteracting systems present the interaction effect .
however , the low energy physics of 1d interacting many - body systems does not have such quasiparticle - type excitations . in 1d many - body systems ,
all particles participate in the low energy excitations and form collective motions of the charge and spin densities with different velocities @xcite .
thus the low energy physics only depends on the tll parameter and the velocities of collective charge and spin oscillations . introducing charge and spin boson fields @xmath524
, the low energy physics of the 1d spin-@xmath525 repulsive fermi gas can be described by an effective hamiltonian @xcite @xmath526 the fields @xmath527 and @xmath528 obey the standard bose communication relations @xmath529=\mathrm{i } \delta_{\nu\mu}\delta(x - y)$ ] with @xmath530 .
the parameter @xmath531 is a short - distance cutoff .
the last term in the effective hamiltonian ( [ effective ] ) characterizes the backscattering process , i.e , it corresponds to a @xmath532 scattering .
the 1d interacting fermi gas separates into charge and spin parts @xmath533 the coefficients for different processes are given phenomenologically @xcite .
the coefficient @xmath534 is the energy cost for changing the particle density while @xmath535 determines the energy for creating a nonzero spin polarization .
the compressibility and susceptibility are given by @xmath536 and @xmath537 , respectively .
at fixed point @xmath538 the specific heat is given by a linear temperature dependent relation @xmath539 where @xmath540 . here
@xmath541 is the specific heat coefficient of noninteracting fermions .
the susceptibility is given by @xmath542 where the noninteracting susceptibility reads @xmath543 .
thus the wilson ratio at the fixed point is given by @xmath544 this ratio gives a universal feature of collective motions of particles .
the tll parameter @xmath545 and charge and spin velocities @xmath546 determine universal behaviour of correlation functions @xcite .
on the other hand , in terms of the tba formalism , the physical quantities are described by the set of nonlinear integral equations @xcite @xmath547 with @xmath548 . here
@xmath549 is given in takahashi ( 1999 ) .
the free energy per unit length @xmath550 and the pressure @xmath70 follow as @xmath551 where @xmath61 denotes the particle density . the spin wave bound states give an antiferromagnetic ordering at low temperatures . the population imbalance associated with the fredholm equations ( [ be - r ] ) lead to three distinguishable phases : the spin singlet groundstate with magnetization @xmath158 ( zero magnetic field ) ; a magnetic phase with finite magnetization ( @xmath552 ) and a fully polarized phase with @xmath553 ( @xmath554 ) .
the critical field , at which the density of down - spin atoms is zero , is given by @xcite @xmath555 the phase diagram in the chemical potential - magnetic field plane is shown in fig .
[ fig : phase - r ] . .
@xmath556 denotes partially polarized phase with finite magnetization . @xmath557
denotes the fully polarized phase .
, title="fig : " ] spin charge separation is a hallmark of the tll physics for 1d interacting fermions . for arbitrary repulsive coupling @xmath55 in arbitrary magnetic field @xmath558 , the tba equations ( [ tba - lambda ] )
yield @xcite @xmath559 where @xmath560 and @xmath390 are , respectively , the holon and spinon excitation velocities @xmath561 however , the velocities @xmath560 and @xmath390 can be analytically calculated only for strong and weak interactions .
the numerical solutions of spin and charge velocities have been discussed in the literature @xcite , see fig .
[ fig : velocities ] . in a harmonic trap
, there is an imbalanced mixture of two - component fermions in the trapping center and fully polarized fermions at the edge @xcite .
the exact analytical solution of quasi - one - dimensional spin-@xmath525 fermions with infinite repulsion for an arbitrary confining potential was present in @xcite . in the context of exact solutions , considerable
work has been done to derive low temperature analytic results for ba solvable models .
these include the work on the free energy of spin chains at low temperatures under a small magnetic field @xcite and the calculation of the leading temperature dependent terms in the free energy of the massive heisenberg model by johnson and mccoy ( 1972 ) .
later , filyov _ et al . _
( 1981 ) derived an exact solution to the s - d exchange model expressed as a series in terms of the temperature . following the method proposed @xcite for small external field @xmath562 , lee _ et al .
_ , @xcite solved the tba equations ( [ tba - lambda ] ) by the weiner - hopf method , where the dressed energy potential is simplified as @xmath563 with the cutoff chemical potential @xmath564 . here
@xmath215 is the pressure .
the terms in the function @xmath565 provide insights on spin and charge density fluctuations . with the help of this potential and using sommerfeld s lemma @xcite , the free energy of the system defined by eq .
( [ pressure - repulsive ] ) is given by @xmath566 which gives the universal low temperature form of spin - charge separation theory ( [ free - dress ] ) with the excitation velocities @xmath567 for the external field approaching the saturation field @xmath568 , the charge and spin velocities can be derived from the relations ( [ fermi - v ] ) .
the leading terms in the velocities are then found to be @xmath569 the susceptibility values for different values of the chemical potential are consistent with the field theory prediction @xmath570 with @xmath571 @xcite for strong repulsion . on the other hand , the tba equations ( [ tba - lambda ] ) show that the spin - spin interaction for the system of the polarized fermions with strong coupling @xmath59 can be effectively described by the isotropic spin-1/2 heisenberg chain with a weak antiferromagnetic coupling @xmath572 . for the isotropic spin-1/2 heisenberg chain , the susceptibility at @xmath573 is given by @xmath574 which coincides with the field theory prediction @xmath575 @xcite .
furthermore , for strong repulsion , the pressure of the gas with a weak magnetic field is given by @xcite @xmath576 this result provides the low temperature thermodynamics which extends beyond the range covered by spin - charge separation theory . with the low temperature expansion
, the pressure ( [ pres - rep - l ] ) significantly evolves into the thermodynamics of two free gaussian fields at criticality .
the result for the free energy at low temperature gives a universal signature of tlls where the leading low temperature contributions are solely dependent on the spin and charge velocities .
this opens a way to experimentally explore how the low temperature thermodynamics of a 1d many - body system naturally separates into two free gaussian field theories .
this may possibly be used to test the spin and charge velocities in experiment with an ultracold atomic 1d fermi gas in a harmonic trap @xcite , where the spin and charge of the ultracold atoms refer to two internal atomic hyperfine states and the atomic mass density , respectively , as per the theoretical scheme to explore spin - charge separation waves in fig.[fig : s - c - separation ] .
this phenomenon has already been experimentally observed in electron liquids @xcite .
the experimental advances in trapping and cooling ultracold quantum gases have also led to realizations of degenerate fermi - bose mixtures with various combinations of fermionic and bosonic atoms such as @xmath577li-@xmath578li @xcite , @xmath577li-@xmath579na @xcite , @xmath580k-@xmath581rb @xcite , @xmath577li-@xmath581rb @xcite , @xmath29yb-@xmath582yb @xcite , and @xmath583li-@xmath582yb @xcite and @xmath583li-@xmath584cs @xcite .
this experimental success opens up a further gateway for exploring striking quantum many - body phenomena though tuning interactions between inter- and intra - species of atoms . in light of the experimental realizations of fermi - bose mixtures ,
various theoretical methods have been used to study phases of superfluids and mott insulators , instabilities of collapse and demixing and quantum correlations of the 1d fermi - bose mixtures , such as the mean - field approach @xcite , tll theory @xcite and numerical methods @xcite .
the tll field theory @xcite predicts that the binary mixtures of bosons and spin - polarized fermions with population imbalance present competing ordering
i ) strong attraction between the two species leads to collapse ; ii ) a strong repulsion leads to demixing ; iii ) subtle tuning of the intra - and inter - particle scattering leads to pairing and two - component tlls . on the other hand ,
the 1d fermi - bose mixture with equal masses of bosons and spin - polarized fermions and with the same strength of delta - function interaction between boson - boson , boson - fermion and fermions with different spins was solved a long time ago @xcite .
the exactly solved model provides a benchmark towards understanding various quantum many - body effects in 1d fermi and bose mixtures .
recently , particular theoretical interest has been paid to the groundstate properties @xcite , magnetism @xcite , correlation functions @xcite , thermodynamics and quantum criticality @xcite .
the 1d @xmath42-function interacting mixture of @xmath585 spinless bosons and @xmath586 fermions with spin - up and @xmath587 fermions with spin - down is described by the hamiltonian ( [ ham ] ) with zeeman term @xmath588 .
@xmath46 is an external magnetic field .
the bethe wave function for this model requires symmetry under exchange of spatial and internal spin coordinates between two bosons or fermions with different spin states and antisymmetry for two fermions with the same spin state .
the ba equations for this fermi - bose mixture with an irreducible representation @xmath589 $ ] are @xcite @xmath590 where @xmath591 . in these equations
@xmath592 , with @xmath593 , are the quasimomenta of the particles and @xmath127 , with @xmath594 , are parameters for fermions with spin - down and the bosons .
@xmath595 with @xmath596 are the parameters for the bosons .
lai and yang ( 1971 ) showed by numerically solving the set of coupled integral equations that the energy of the system for the mixture of bosons and polarized fermions is a monotonic decreasing function with respect to the ratio of bosons and the total number of particles in the system .
the groundstate properties of the system have been further studied @xcite .
the magnetic properties of the mixture of bosons and polarized fermions provide further insight into the competing ordering .
application of the external magnetic field to the polarized fermions causes zeeman splitting of the spin - up and spin - down fermions into different energy levels .
the groundstate can only accommodate fermions that are in the lower energy level .
therefore it is expected that when the direction of the magnetic field is along the spin - up ( @xmath597 ) direction , spin - down fermions can no longer populate the groundstate .
the free energy of the strongly coupled mixture is given in the form @xmath598 where the groundstate energy per unit length is given by @xcite @xmath599.\label{e - fb - m}\end{aligned}\ ] ] here @xmath600 is the boson number density .
if the external field exceeds the critical field @xmath601 , where the pressure per unit length is given by @xmath602 $ ] , the system enters a phase of fully - polarized fermions .
the susceptibility in the vicinity of the critical field @xmath603 diverges as @xmath604 this van hove type of singularity is subtly different from the linear field - dependent magnetization in the two - component attractive fermi gas with polarization , where the effective interaction between the bosonic pairs is weakly attractive in the tonks - girardeau limit @xcite .
in contrast , for weak repulsion the groundstate energy presents a mean field effect , i.e. , the energy per unit length is @xmath605.\ ] ] the magnetization @xmath606 and the susceptibility @xmath607 follow from this equation .
these results show that in the weak coupling regime the square - root field - dependent behavior of magnetization emerges for finite external field .
furthermore , using cft , frahm and palacios @xcite have computed the asymptotics of boson and fermion green s functions for a mixtures of bosons and polarized and unpolarized fermions .
the fourier transform of equal time boson green s function has a form @xmath608 .
the response function of fermions follows a form @xmath609 .
the exponents @xmath600 and @xmath610 are determined by finite - size energies and momenta .
the presence of spin population imbalance of fermions gives rise to different critical exponents .
in fact , the exact result indicates that no demixing occurs in the fermi - bose mixture with a repulsive interaction .
a thorough study of the ground state properties , including density distributions and the single particle correlation functions was reported in @xcite .
using exact solution with local density approximation in a harmonic trap , the density profiles of the fermi - bose mixture were predicted . in the weakly interacting regime
, bosons can condense to the trapping centre whereas fermions spread out due to the fermi pressure .
for strong repulsion , the boson density distribution is extended in a wider region and the fermion density shows strong non - monotonous behavior , see fig.[fig : f - b - density ] . . at zero temperature ,
fermions sit at the wings while the trapping centre comprises of a mixture of fermions and bosons . from imambekov and demler ( 2006b ) .
, title="fig : " ] the fully - polarized fermi - bose mixture is described by the hamiltonian @xmath611 where @xmath612 , @xmath613 are boson and fermion field operators , @xmath614 , @xmath615 are the masses , @xmath616 , @xmath617 are chemical potentials of bosons and fermions , and @xmath618 , @xmath619 are boson - boson and boson - fermion interaction strengths , respectively . for bosons and fermions of equal mass and equal interaction strength @xmath620 , a different set of ba equations from ( [ ba - m ] )
were derived @xcite with @xmath621 where @xmath622 and @xmath623 .
here @xmath585 is the number of bosons .
the ba equations ( [ ba - m ] ) with a particular choice of fermion spin polarization are physically equivalent to ( [ ba - m - f ] ) .
this can be seen from the fact that the groundstate energy of the model ( [ h_second_form ] ) is given by the same expression as ( [ e - fb - m ] ) , where the spin - down fermions are gapfull . at finite temperatures ,
the equilibrium states become degenerate .
the thermodynamics of the model are determined from the integral equations @xcite @xmath624 for fixed temperature @xmath1 and chemical potential @xmath625 , @xmath626 , the pressure is given by @xmath627 . in this grand canonical ensemble , it is convenient to use the chemical potential @xmath35 and the chemical bias @xmath628 to discuss the zero temperature phase diagram of the model @xcite , see fig .
[ fig : phase - mixture ] .
the phase boundaries were determined by analysing the dressed energy equations obtained from the tba equations ( [ tba_nonlinear ] ) in the limit @xmath430 .
the critical line @xmath629\ ] ] separates the mixed phase and the pure fermion phase . here
dimensionless units have been used , i.e. , @xmath630 and @xmath631 with @xmath632 . in a harmonic trap , this phase diagram can be presented within the lda , i.e. , @xmath633 and @xmath634 .
it was found @xcite that for both strong and weak interactions bosons and fermions coexist in the central part and fermions sit at the wings . for weak interaction , bosons can condense into the centre while the fermions spread out in the whole trapping space due to fermi pressure .
however , for the strong interaction regime , the fermi density shows strong non - monotonous behaviour @xcite .
the significant feature of correlation functions of the mixture in the tonks - girardeau limit were discussed in detail @xcite . in particular , the fourier transform of the bose - bose correlation function is governed by the tll parameter @xmath635 via @xmath636 for @xmath637 and the fourier transform of the fermi - fermi correlation function has singularities at @xmath638 .
the discontinuity at @xmath639 indicates an interaction effect
. plane .
three distinguished phases result from varying the chemical potential and the chemical potential difference @xmath628 : the pure boson phase for @xmath640 and @xmath641 ; the pure fermion phase below the phase boundary @xmath642 in the region @xmath643 and the mixture of bosons and fermions above the phase boundary @xmath642 in the region @xmath597 . from yin _ et al . _
, title="fig : " ] vs chemical potential from the exact solution , showing quantum criticality driven by chemical potential for @xmath644 .
the crossover temperatures ( white squares and triangles ) separate the vacuum , @xmath442 and @xmath645 from the quantum critical regimes . from yin _ et al . _
, title="fig : " ] the tba equations ( [ tba_nonlinear ] ) have been used to explore scaling behaviour of the thermodynamics in the fermi - bose mixture .
the universal leading order temperature corrections to the free energy @xcite @xmath646 indicate a collective tll signature at low temperatures . in the strongly repulsive regime for @xmath562 @xmath647\nonumber .\ ] ] as already remarked , the tll description is incapable of describing quantum criticality since it does not contain the right fluctuations in the critical regime . for the physical regime ,
i.e. , @xmath648 , or @xmath649 , the pressure is @xmath650 here the functions @xmath203 and @xmath565 are determined by @xmath651 the function @xmath652 can be obtained from ( [ tba_nonlinear ] ) by iteration @xcite . the equation of state ( [ pressure - fm ] ) can be used to explore the critical behaviour of the model .
e.g. , the entropy in fig .
[ fig : entropy - mixture ] shows that the tll is maintained below a crossover temperature at which the linear temperature - dependent entropy breaks down . in fig .
[ fig : entropy - mixture ] , @xmath442 denotes a tll of fully - polarized fermions and @xmath645 denotes a a two - component tll of the mixture of bosons and fermions described by ( [ ll ] ) . as the temperature is tuned over the crossover temperatures the scaling function of thermodynamic properties give rise to the free fermion universality class of criticality that entirely depends on the symmetry excitation spectrum and dimensionality of the system .
the equation of state pressure , determined by ( [ pressure - fm ] ) , contains universal scaling functions which control the thermodynamic properties in the quantum critical regimes .
near the critical point , the thermal dynamical properties can be cast into universal scaling forms , e.g. , ( [ scaling - n ] ) and ( [ scaling - kappa ] ) for a free fermi theory of criticality , i.e. , with @xmath653 , @xmath654 and @xmath655 .
the detailed analysis of quantum criticality of the fermi - bose mixture has been presented in yin _
as for the fermi system , with the help of the exact solutions the critical properties of the bulk system can be mapped out from the density profiles of the trapped fermi - bose mixture gas at finite temperatures .
a pseudo spin-1/2 system of interacting atomic fermions has been experimentally realized by loading atoms within two lowest hyperfine levels , i.e. , states @xmath656 and @xmath657 .
the interaction strength can be controlled by tuning the scattering length through feshbach resonance .
weakly bound molecules exist in the phase where the scattering length is small and positive .
these molecules can form a bec .
the tunability of the scattering length across the feshbach resonance leads to divergent scattering length . as a result
the interactions can be effectively enhanced . in the strong interacting limit
, these molecules may continuously transform into bcs pairs such that the system reaches the bec - bcs crossover @xcite .
universal many - body behaviour is expected in the crossover ( unitarity ) regime @xcite .
it is of a great interest that the third pseudo spin state @xmath658 is added to the two - component fermi gas @xcite .
in contrast to the two - component case , three - component fermions possess new features @xcite . as a consequence , bcs pairing can be favored by anisotropies in three different ways , namely atoms in three low sublevels denoted by @xmath656 , @xmath657 and @xmath658 can form three possible pairs @xmath659 , @xmath660 and @xmath661 @xcite .
degenerate three - component fermions with tunable interparticle scattering lengths @xmath662 , @xmath663 and @xmath664 open up the possibility of novel many - body phenomena .
specifically , strongly attractive three - component atomic fermions can form spin - neutral three - body bound states called ` trions ' .
a degenerate fermi gas of atoms in three different hyperfine states of @xmath665li @xcite has been created at a temperature @xmath666 , where @xmath667 is the fermi temperature .
the spin state mixture of ultracold fermionic atoms has been further used to study subtle physics of three - body recombination , atom - dimer scattering , the atomic efimov trimer , association and disassociation of the atom - dimer collisions _ etc _ @xcite . in particular , the three - component interacting fermions have received considerable interest in the study of the quantum phase transition from a colour superfluid to singlet trions @xcite , see fig .
[ fig : pair - trion ] .
this study also sheds light on colour superconductivity of quark matter in nuclear physics . , whereas the trionic state occurs for @xmath668 . from rapp _ et al . _
, title="fig : " ] loading three - component fermions with contact interaction on a 1d optical lattice , the system , i.e. , the 1d multi - component hubbard model , is no longer ba solvable @xcite . in terms of bosonization approach ,
it has been found @xcite that the low - energy physics of the 1d three - component hubbard model shows the formation of three - atom bound states , i.e. , a quantum phase transition from a colour superfluid to the singlet trion state .
motivated by such exotic phases , the 1d integrable three - component fermi gas with delta - function interactions @xcite has been studied to give a precise understanding of colour pairing and the trionic state @xcite .
the first quantized many - body hamiltonian of the three - component @xmath42-function interacting fermi gas is as defined in eq .
( [ ham ] ) , with an additional zeeman energy term @xmath669 . in this system
, the fermions can occupy three possible hyperfine levels ( @xmath670 , @xmath671 and @xmath672 ) with particle number @xmath673 , @xmath674 and @xmath675 , respectively @xcite .
the zeeman energy levels @xmath676 are determined by the magnetic moments @xmath677 and the magnetic field @xmath678 . by convention ,
particle numbers in each hyperfine states satisfy the relation @xmath679 .
thus the particle numbers of unpaired fermions , pairs , and trions are respectively given by @xmath680 , @xmath681 and @xmath682 for the attractive regime .
the unequally spaced zeeman splitting in the three hyperfine levels can be characterized by two independent parameters @xmath683 and @xmath684 .
here @xmath685 is an average zeeman energy . for the 1d three - component fermi gas ,
the energy eigenspectrum is given in terms of the quasimomenta @xmath118 of the fermions via eq .
( [ ek ] ) , which satisfy the ba equations @xcite @xmath686 here @xmath593 , @xmath687 and @xmath688 .
the parameters @xmath689 are the rapidities for the internal hyperfine spin degrees of freedom . it is assumed that there are @xmath587 fermions in state @xmath658 , @xmath690 fermions in state @xmath657 and @xmath691 fermions in state @xmath656 . for the irreducible representation @xmath692 $ ] , a three - column young tableau encodes the numbers of unpaired fermions , bound pairs and trions given by @xmath693 , @xmath694 and @xmath695 . in the attractive regime , the ba equations ( [ be3 ] ) admit complex string solutions for @xmath696 , i.e. , three - body bound states ( trions ) and two - body bound states ( colour pairing ) @xcite . for arbitrary spin polarization , ( i )
there are @xmath697 spin - neutral trions in the quasimomentum @xmath6 space accompanied by @xmath697 spin bound states in the @xmath698-parameter space and @xmath697 real roots in the @xmath699-parameter space , ( ii ) @xmath700 bcs bound pairs in @xmath6 space accompanied by @xmath700 real roots in @xmath698 space and ( iii ) @xmath701 unpaired fermions in @xmath6 space .
these root patterns are depicted in fig .
[ fig : pattern3 ] .
trions , @xmath702 colour pairs and @xmath703 excess single fermions in the groundstate . for strongly attractive interaction , the unpaired and paired quasimomenta can penetrate into the central region occupied by tightly bound trions . from lee
_ et al . _
( 2011).,title="fig : " ] for weak attraction , we redefine the polarizations @xmath704 with @xmath705 and the energy @xmath706 with dimensionless parameter @xmath707 .
the groundstate energy follows from the ba equations ( [ be3 ] ) as @xcite @xmath708+o(\gamma^2).\label{energy - w - g } \end{aligned}\ ] ] this result was also obtained from the fredholm equations @xcite .
the leading order of the interaction energy gives a two - body mean field interaction energy between the particles with different internal spin states .
the kinetic energy part indicates three free - fermi gases of different species .
for strong attraction ( @xmath431 ) these charge bound states are stable and the system is strongly correlated .
the corresponding binding energies of the charge bound states are given by @xmath709 . explicitly , the bcs pair binding energy is @xmath710 and the binding energy for a trion is @xmath711 . without loss of generality ,
the numbers of trions @xmath697 , pairs @xmath700 and unpaired fermions @xmath701 are assumed to be even .
the explicit root patterns to the ba equations ( [ be3 ] ) for trions , pairs and single fermions are then @xmath712 where @xmath713 , @xmath714 and @xmath715 , see fig .
[ fig : pattern3 ] .
the real parts of the bound states have a hard - core signature , explicitly , @xmath716 where @xmath717 with @xmath718 . here
@xmath719 with the function @xmath720 where @xmath721 and @xmath722 is the step function .
this result points to the interference effects among the molecule states and excess single fermions . from these roots ,
the groundstate energy in the thermodynamic limit is given explicitly by @xcite @xmath723 with @xmath724 . here
@xmath725 and @xmath726 is the number of @xmath727-atom molecule states .
in general the 1d interacting fermi gases with higher spin symmetries have two distinguishing features : ( i ) mean field theory for the weak coupling regime , and ( ii ) strongly correlated molecule states of different sizes in the strongly attractive regime .
the fundamental physics of the model is determined by the set of equations ( [ be3 ] ) which can be transformed to generalised fredholm types of equations in the thermodynamic limit .
the asymptotic expansion solution of the fredholm equations for arbitrary component fermi gas has been thoroughly studied @xcite . and
( b ) weak interaction @xmath728 with @xmath301 .
the figure reveals a trion phase @xmath432 , a pairing phase @xmath678 , an unpaired phase @xmath565 and four different mixtures of these states .
good agreement is found between the analytical critical fields ( black lines ) and the numerical solutions ( white lines ) of the tba equations ( [ tba - f3 ] ) . the pure trion phase @xmath432 is present in the strong coupling regime , whereas the colour pairing phase is favoured in the weak coupling limit . from kuhn and foerster ( 2012 ) . ,
title="fig : " ] and ( b ) weak interaction @xmath728 with @xmath301 .
the figure reveals a trion phase @xmath432 , a pairing phase @xmath678 , an unpaired phase @xmath565 and four different mixtures of these states .
good agreement is found between the analytical critical fields ( black lines ) and the numerical solutions ( white lines ) of the tba equations ( [ tba - f3 ] ) . the pure trion phase @xmath432 is present in the strong coupling regime , whereas the colour pairing phase is favoured in the weak coupling limit . from kuhn and foerster ( 2012 ) .
, title="fig : " ] the groundstate energies ( [ energy - w - g ] ) and ( [ e ] ) provide full phase diagrams in the @xmath729 plane for the weak and strong coupling regimes , see fig .
[ fig : e3 ] . the fields @xmath730 and @xmath731
are determined through the relations @xcite @xmath732 with the constraint condition @xmath733 . for the strong coupling regime in the absence of zeeman splitting ,
i.e. , @xmath734 , trions form a singlet groundstate .
however , the zeeman splitting can lift the @xmath735 degeneracy and drive the system into different phases . for small @xmath730 , a transition from a trionic state into a mixture of trions and pairs
occurs as @xmath731 exceeds the lower critical value @xmath736 .
when @xmath731 is greater than the upper critical value @xmath737 , a pure pairing phase takes place .
trions and bcs pairs coexist when @xmath738 .
these critical fields are derived from the relations ( [ h1-h2 ] ) @xcite @xmath739 the phase transitions from the colour paired phase @xmath678 to the mixed phase @xmath740 of pairs and unpaired fermions induced by increasing @xmath730 are reminiscent of those for the two - component system discussed in sec .
[ fermi - fflo ] .
this mixed phase consisting of the bcs - pairs and unpaired fermions is referred to the fflo phase , see fig .
[ fig : e3 ]
. for small @xmath731 , a phase transition from a trionic into a mixture of trions and unpaired fermions occur as @xmath730 increases .
using the relations ( [ h1-h2 ] ) , the trionic phase with zero polarization forms a singlet groundstate of trions when the field @xmath741 , whereas when @xmath730 is greater than the upper critical value @xmath742 all trions are broken and the state becomes a normal fermi liquid , see fig .
[ fig : e3 ] . here
@xmath743 in addition , for a certain regime of @xmath730 and @xmath731 , there is a phase transition from the trionic state into the mixture of trions , pairs and unpaired fermions , see fig .
[ fig : e3 ] . in the weak coupling regime , the critical fields can be obtained from the relations @xmath744 we see that either a mixture of bcs pairs and unpaired fermions or a mixture of trions and unpaired fermions or a mixture of trions , pairs and unpaired fermions populates the groundstate for certain values of @xmath730 and @xmath731 .
these asymptotic results indeed agree well with the full phase diagram determined from numerical solutions @xcite , see fig .
[ fig : e3 ] .
it is interesting to note that the pure paired phase can be sustained under certain zeeman splittings . in this phase ,
the two lowest levels are almost degenerate for certain tuning of @xmath730 and @xmath731 .
the persistence of this colour pairing phase is relevant for the study of phase transition between the colour bcs - pairing phase and the state of trions . in the thermodynamic limit , the grand partition function @xmath745 is given in terms of the gibbs free energy @xmath746 where the chemical potential @xmath35 , the zeeman energy @xmath747 and the entropy @xmath5 are given in terms of the densities of unpaired fermions , charge bound states , trions and spin - strings , which are all subject to the ba equations ( [ be3 ] ) .
the equilibrium states are determined by minimizing the gibbs free energy , which gives rise to a set of coupled nonlinear integral equations
the tba equations for the dressed energies @xmath748 @xcite . in the thermodynamic limit , the pressure @xmath215 is defined in terms of the gibbs energy by @xmath749 , including three parts , @xmath750 , @xmath751 and @xmath752 , for the pressure of unpaired fermions , pairs and trions , respectively , with @xmath753 here we have set the boltzmann constant @xmath754 .
the quantum phase transitions in the model with zeeman splitting may be analyzed via the dressed energy equations , which are obtained from the tba equations in the limit @xmath430 as @xmath755*{\epsilon^ { ( 2)}}(\lambda)-\left[k_2+k_4\right]*{\epsilon^ { ( 3)}}(\lambda),\nonumber\\ \epsilon^{(2)}(\lambda)&=&2\lambda^2 - 2\mu -\frac{c^2}{2}-h_2-k_1*{\epsilon^ { ( 1)}}(\lambda)\nonumber \\ & & -k_2*{\epsilon^{2}}(\lambda ) -\left[k_1+k_3\right]*{\epsilon^{(3)}}(\lambda),\nonumber\\ \epsilon^ { ( 1)}(k)&=&k^2-\mu -h_1-k_1*{\epsilon^ { ( 2)}}(k ) -k_2*{\epsilon^{(3)}}(k ) .
\label{tba - f3}\end{aligned}\ ] ] here we denote @xmath756 .
the integration boundaries @xmath757 characterize the `` fermi surfaces '' at @xmath758 .
the chemical potential and magnetization are determined by @xmath730 , @xmath731 , @xmath759 and @xmath61 through the relations @xmath760 .
the dressed energy equations ( [ tba - f3 ] ) indicate effective interactions among trions , pairs and single fermions . if we denote effective chemical potentials @xmath761 for trions , @xmath762 for bound pairs and @xmath763 for unpaired fermions , the energy transfer relations among the binding energy , the zeeman energy and the variation of chemical potentials between different fermi seas are given by @xcite @xmath764
these equations determine the full phase diagram and the critical fields triggered by the zeeman splitting @xmath765 and @xmath766 .
indeed , the relations ( [ e - h - relation ] ) give the results ( [ energy - w - g ] ) and ( [ e ] ) in the weak and strong coupling regimes . in the phase diagrams fig . [ fig : mu - h-3 ] , the phase boundaries can be also analytically determined by analyzing the band fillings through the dressed energy equations ( [ tba - f3 ] ) .
phase diagrams for ( a ) equally spaced zeeman splitting and unequally - spaced zeeman splitting ( b ) , ( c ) and ( d ) . from he _
, title="fig : " ] at low temperature regimes , the tba equations provide universal thermodynamics of the attractive three - component fermi gas @xcite . in the grand canonical ensemble ,
the unequal spaced zeeman splitting lead to various phases in the @xmath304 plane , see fig .
[ fig : mu - h-3 ] .
this @xmath304 phase diagram can be determined by the dressed energy equations ( [ tba - f3 ] ) .
the quantum phases at zero temperature persist due to the nature of collective motion ( forming tll phases ) for a certain temperature range .
although there is no quantum phase transition in 1d many - body systems at finite temperatures , quantum criticality leads to a crossover from a relativistic dispersion to a nonrelativistic dispersion between the tll and the quantum critical regime .
the nature of the tll physics is revealed from the universal leading temperature corrections to the free energy in the different phases demonstrated in fig .
[ fig : e3 ] , i.e. , @xmath767 where the sum involves a term for each cluster component in the phase
. e.g. , @xmath768 for phase @xmath769 . for strong attraction ,
@xmath770 with @xmath771 are the velocities for unpaired fermions , pairs and trions , respectively . here the function @xmath772 are given by ( [ a3 ] ) .
on the other hand , the pressure @xmath773 of trions , pairs and excess fermions can be obtained in an analytical manner using the polylog function in the strong attractive regime , with @xmath774 for @xmath775 .
up to a few leading order terms , the functions @xmath776 are @xmath777 where @xmath778 .
the total pressure @xmath779 provides a high precision equation of state through iterations with ( [ bbb ] ) .
the thermodynamics and critical behaviour can be worked out in a straightforward manner in terms of the polylog function .
the equation of state ( [ ppp ] ) covers not only the zero temperature result of the three - component strongly attractive fermi gas but also the tll thermodynamics .
this result opens up further study of quantum criticality with respect to the phase transitions when the parameters drive the system across the phase boundaries of fig .
[ fig : mu - h-3 ] . for the repulsive regime , the thermodynamics of the three - component fermi gas is determined by another set of tba equations @xcite . in this regime ,
spin - charge separation is a hallmark of the 1d three - component fermi gas . in the low - lying excitations , interacting particles `` split '' into spins and charges as the temperature tends to absolute zero . the collective motion of fermions with only spin or charge , called spinons and chargons / holons ( the antiparticle of a chargon ) , which have different velocities .
the three - component fermi gas has @xmath780 symmetry that leads to two sets of spin waves .
it is shown @xcite that the low temperature thermodynamics of such a gas naturally separates into free gaussian field theories for the @xmath781 charge degree of freedom and two spin rapidities .
the free energy gives a universal low temperature tll behaviour , namely @xmath782 the spin and charge velocities can be derived @xcite from the relations @xmath783 and @xmath784 . for three - component fermions , there are two spin velocities @xmath785 and @xmath786 , where @xmath787 for pure zeeman splitting . the central charge for the spin part is @xmath788 and for the charge part @xmath789 .
the reason for the value @xmath788 is because the @xmath735-invariant fermion model has two spin `` fermi seas '' whose dependence on @xmath46 are equal , i.e. , we considered the case where @xmath790 . following the wiener - hopf method developed for the study of the thermodynamics of heisenberg spin chains with @xmath791 and @xmath735 symmetries @xcite , the groundstate energy of the gas in a weak magnetic field
is given by @xmath792 .
explicitly , the spin and charge velocities @xmath793 are derived in the strong repulsive regime .
the spin velocity tends to zero while the charge velocity tends to the fermi velocity as the interaction strength @xmath346 .
the low temperature properties of the equation of state follow from @xmath794 where the potential function @xmath795 .
the thermodynamics obtained from this pressure covers the universal thermodynamics of the tll given by ( [ f3r ] ) .
spinor bose gases with spin - independent short - range interactions have a ferromagnetic groundstate , i.e. , the groundstate is always fully polarized .
in contrast to the spinless bose gas , the spinor bose gases with spin - exchange interactions can display a different groundstate , i.e. , either a ferromagnetic or an antiferromagnetic groundstate solely depending on the spin - exchange interaction @xcite . in this regard ,
the 1d spinor fermi gases with a short - range delta - function interaction and spin - spin exchange interaction are particularly interesting due to the existence of various bcs - like pairing phases of quantum liquids associated with the ba @xcite .
large spin fermionic systems also exhibit many new phases of matter which do not appear in the usual spin-1/2 systems .
in particular , spin-3/2 systems possessing a generic @xmath19 symmetry have attracted much theoretical interest @xcite . in these systems ,
s - wave scattering acquires interaction in total spin singlet and quintet channels .
the two interacting channels present a hidden @xmath19 symmetry without fine tuning .
the spin-3/2 systems exhibit a quintet pairing phase with total spin-2 and quartetting order as a four - fermion counterpart of the cooper pairing .
the hamiltonian describing these @xmath19-invariant systems reads @xcite @xmath796 where @xmath797 is the dimensionality and @xmath35 is the chemical potential .
the operators @xmath798 and @xmath799 denote the spin singlet and quintet pairing operators given by @xmath800 with @xmath801 and @xmath802 . using the dirac matrices , the @xmath19 algebra has been constructed explicitly @xcite . the spin singlet pair and quintet pair have been constructed in terms of the @xmath19 scalar and vector operators .
the @xmath19 generators commute with the hamiltonian ( [ ham-3 - 2 ] ) .
the models restore @xmath21 symmetry when @xmath803 .
the fermi liquid theory of @xmath19 symmetry describes different competing orders in the spin-3/2 continuum and lattice models . the lattice version of the hamiltonian ( [ ham-3 - 2 ] )
can be viewed as the one - band generalized hubbard model exhibiting @xmath19 symmetry at arbitrary filling and @xmath804 symmetry at half - filling @xcite .
the @xmath19 invariance can apply equally well in the one - dimensional continuum model and the lattice model @xcite .
writing the left and right moving currents in terms of the @xmath19 scalar , vector and tensor currents @xcite , the low energy physics can be described by an effective hamiltonian density , which can be treated by bosonization .
the result indicates two spin gap phases and various order parameters , see fig .
[ fig : wu - phase ] .
the tll phase exists in the repulsive region @xmath805 where @xmath806 .
the quartetting phase @xmath807 has two orders the quasi - long range ordered superfluidity ( qros ) @xmath808 and charge density wave ( cdw ) of quartets @xmath809 .
the pairing phase @xmath810 separates into a spin singlet pairing phase @xmath811 and dimenerization of spin peierls order @xmath812 . here
the competition between the quartetting and pairing phases is characterized by the ising duality . in this scenario ,
the low energy physics of 1d arbitrary half - spin fermions of ultracold atoms has been studied by bosonization @xcite .
see also a recent study on competing orders in 1d half - filed fermionic ultracold atoms in an optical lattice @xcite .
two different superfluid orders , a confined bcs pairing phase and a confined molecular superfluid , were found for @xmath813 fermionic ultracold atoms .
symmetry in terms of the singlet and quintet interaction channel parameters @xmath814 and @xmath815 .
various competing orders of singlet and cdw pairing , as well as qros and cdw of quartets are formed . from wu ( 2005 ) .
, title="fig : " ] controzzi and tsvelik ( 2006 ) write that although high symmetries do not occur frequently in nature , they deserve attention since every new symmetry brings with itself a possibility of new physics " indicates a perspective of large spin fermionic atoms .
fortunately the 1d realization of spin-3/2 fermions with @xmath19 symmetry is exactly solved under a suitable condition @xcite .
in particular , jiang _ et al . _
( 2009 ) found that the hamiltonian ( [ ham-3 - 2 ] ) reduces to an integrable many - body hamiltonian @xmath816 on a line @xmath817 . here
the coupling constants are given by @xmath818 and @xmath819 .
the integrable hamiltonian ( [ ham-1d - so5 ] ) possesses @xmath19 symmetry .
but the individual spin components are no longer conserved .
however , @xmath820 , @xmath821 and @xmath822 are three independent conserved quantities .
the integrability is guaranteed by the two - body scattering matrix @xmath823 which satisfies the ybe @xmath824 in the above equation , @xmath825 is the projection operator onto the spin @xmath45 channel . the solution has been derived in terms of the ba @xcite . in the repulsive regime , the low energy physics can be described by the spin - charge separation theory .
the elementary spin excitations , including a spin-3/2 spinon , a neutral spinon and a spin-1/2 spinon , has been studied @xcite , see fig .
[ fig : jiang - excitation ] . here
we see that the charge excitations ( a ) indicate a particle - hole type .
the spin excitations ( b ) and ( c ) involve two real @xmath699 and @xmath35 holes , respectively .
the spin excitations ( d ) give two string-2 hole excitations .
these spin excitations are different from the case of @xmath21 symmetric fermions . for an attractive interaction ,
competing pairing orders lead to quantum phases of pairs and quartets .
however , the ba equations give very complicated root patterns .
the study of the attractive integrable @xmath19 symmetric model still remains an open problem .
-holes in the spin-@xmath699 sector ; ( c ) spin excitations of two real @xmath35-holes in the spin-@xmath35 sector and ( d ) spin excitations of two string-2 @xmath699 holes in the spin-@xmath699 sector . from jiang _ et al . _
, title="fig : " ] although integrable spin systems with higher symmetries have been extensively studied in the literature , exactly solved models of fermionic ultracold atoms with higher spin symmetry are still restricted to @xmath826 @xcite and @xmath31 @xcite . for spin - dependent interaction , the spin exchange interaction between the @xmath827- and @xmath828-th particles can be written as a summation of spin projection operators @xmath829 in the channels with even total spin @xmath830 .
the integrable @xmath31-invariant models of atomic fermions are artificially written as @xmath831 where the interaction potential reads @xmath832cp^0_{ij}+c\sum_{m=2,4,\ldots 2s}p^{m}_{ij}.\nonumber\end{aligned}\ ] ] the ba equations for the model ( [ ham - sp ] ) have been derived @xcite .
in addition , melo and martins ( 2007 ) derived the ba solution of a many - body problem of interacting spin-@xmath833 particles with an arbitrary @xmath781 factorized @xmath5-matrix .
the stability of spinor fermi gases in tight waveguides was discussed in del campo _ et al . _
( 2007 ) .
fermionic alkaline - earth atoms provide unique opportunities to study exotic many - body physics in the context of higher spin symmetry , e.g. , @xmath834 with @xmath835 , where @xmath24 is the nuclear spin @xcite . to avoid any confusion with the number of particles @xmath44 . ]
de salvo _ et al . _
( 2010 ) have created a degenerate gas of ultracold fermionic atoms @xmath39sr(@xmath836 ) in an optical trap .
( 2010 ) have reported the realization of a degenerate fermi mixture of two isotopes of ytterbium atoms @xmath837yb ( i=1/2 ) and @xmath838yb ( i=5/2 ) with @xmath839 symmetry .
more recently they @xcite have successfully realized the @xmath30 symmetric mott - insulator state with the atomic fermi gas of @xmath838yb in a 3d optical lattice .
it was found that loading fermions adiabatically into a higher symmetry mott insulating state can achieve lower temperatures than the @xmath791 symmetry state owing to differences in the entropy carried by isolated hyperfine spins .
these alkaline - earth atoms have a particular filled electron shell structure such that their nuclear spins decouple from the electronic angular momentum @xmath347 in these two states .
this decoupling implies that the nuclear spins give the hyperfine spins , i.e. , @xmath840 . for these atomic systems ,
the @xmath841 hyperfine levels are likely to display @xmath842 symmetry where the s - wave scattering lengths are independent of the nuclear spins .
such fermionic systems with enlarged symmetries are motivated to simulate quantum many - body phenomena @xcite which may shed light on physics of strongly correlated transition - metal oxides , heavy - fermion materials and spin - liquids phases .
in contrast to the weaker quantum spin effect for a composite object with larger spin in solid state , large hyperfine spins can be essential to the quantum magnetic states @xcite because spin fluctuations are accommodated in a large number of hyperfine spin states .
large - hyperfine fermions may also be used to study cooper pairing phenomena within hyperfine spins @xcite and quantum information @xcite .
the hamiltonian for the 1d @xmath44-body @xmath42-function interacting fermion problem @xcite is again as defined in eq .
( [ ham ] ) .
there are now @xmath843 possible hyperfine states latexmath:[$|1\rangle , |2\rangle , \ldots , @xmath834 spin symmetry and @xmath781 charge symmetry . for an irreducible representation
@xmath845 $ ] , the young diagram has @xmath843 columns with the quantum numbers @xmath846 . here
@xmath847 is the number of fermions at the @xmath827-th hyperfine level such that @xmath848 .
the groundstate properties and thermodynamics of 1d @xmath834-invariant fermi gases have been studied by means of the ba solution , e.g. , the three - component fermi gas @xcite , @xmath21-invariant spin-3/2 fermions @xcite , and @xmath843-component fermions @xcite . for the groundstate of the @xmath834-invariant fermi gas ,
the generalized fredholm equations for @xmath55 are given by @xcite @xmath849 where @xmath850 and @xmath851 , @xmath852 is the particle quasimomentum distribution function , whereas @xmath853 with @xmath854 are the distribution functions for the @xmath855 spin rapidities .
the groundstate energy @xmath856 per unit length is given by @xmath857 .
for the balanced case , the integration boundaries @xmath858 with @xmath854 are infinitely large .
thus the distributions are given by @xcite @xmath859 , where @xmath860}{\sinh\left[\frac{1}{2}\kappa |\omega |c \right ] } \label{r - m - o}\ ] ] with @xmath861 .
for strong repulsion , i.e. , @xmath862 , the groundstate energy of the balanced @xmath843-component fermi gas follows as @xcite @xmath863 \label{e - r1}\end{aligned}\ ] ] with the constants @xmath864 $ ] and @xmath865 $ ] . here
@xmath866 and @xmath266 are the riemann zeta functions , @xmath867 denotes the euler psi function and @xmath432 is the euler constant . for @xmath868 , it is seen that @xmath869 .
this indicates an insight into the hyperfine spin effect the groundstate energy ( [ e - r1 ] ) reduces to the energy of the spinless bose gas as @xmath346 .
this result was first noticed by yang and you ( 2011 ) by means of the fredholm equations ( [ fe - r ] ) .
the suppression of the hyperfine spin effect is further manifest in the groundstate energy per unit length for the highly polarized case , which reads @xmath870 where @xmath871 with @xmath872 .
this result shows that spin variation does not play an essential role in the groundstate due to the strong repulsion .
however , for small polarization , i.e. , a small external field lifting the @xmath834 symmetry , the integral boundaries @xmath858 with @xmath854 are very large . in this case ,
logarithmic singularities arise in the zero temperature susceptibility .
this configuration is drastically changed as the interaction is decreased . for the weak coupling limit ,
the mean field result for the groundstate energy per unit length is @xmath873 here @xmath874 with @xmath875 denote the polarizations , and @xmath876 is the number of fermions in the @xmath827th level .
the first part is the kinetic energy of the @xmath843-component fermions whereas the second parts is the interaction energy .
this result is valid for arbitrary spin imbalance in the weakly repulsive and attractive regime .
the higher order corrections have not been obtained . for the balanced case ,
i.e. , @xmath877 , the energy is @xmath878 which is the same as for spinless bosons with a weak repulsion as @xmath868 .
the groundstate properties for the limits @xmath879 and @xmath346 have been discussed by schlottmann ( 1997 ) . in the attractive regime , the complex string solutions for @xmath696 form @xmath45-atom bound states up to length @xmath880 with the binding energy for a bound state @xmath881 .
such bound states were studied by gu and yang ( 1989 ) .
a bound state in quasimomentum space of length @xmath45 takes on the form @xmath882 , where @xmath883 . the number of bound states with length @xmath884 is denoted by @xmath885 .
the real part is @xmath886 .
the unpaired atoms have real quasimomenta @xmath124 .
takahashi ( 1970b ) derived the fredholm equations for the attractive fermi gas with an arbitrary number of components @xmath887 where @xmath888 is the density distribution function of single fermions , @xmath889 is the density distribution function for the @xmath45-atom bound state with @xmath890 .
the total number of fermions is given by @xmath891 .
the integration boundaries @xmath892 , characterizing the fermi points in each fermi sea , are determined by @xmath893 .
the groundstate energy per unit length is given by @xmath894 .
for weak attraction @xmath895 , the two sets of the fredholm equations ( [ fe - r ] ) and ( [ fe - a ] ) for the repulsive and attractive regimes preserve the symmetry @xmath896 indeed , the groundstate energy of the @xmath843-component fermi gas with weak attraction has the same closed form as ( [ e - r5r2 ] ) with the replacement @xmath897 .
this means that the groundstate energy of the @xmath843-component gas with arbitrary polarization continuously connects at @xmath284 . for strong attraction @xmath898 ,
the bound states of different sizes form tightly bound molecules of different sizes . in this regime , all the fermi momenta of the molecules are finite , i.e. , @xmath899 with @xmath900 . here
@xmath901 characterizes the fermi momentum of the single spin - aligned atoms .
therefore , in this regime the strong coupling condition allows one to expand the fredholm equations ( [ fe - a ] ) in powers of @xmath902 . the groundstate energy per unit length of the gas with arbitrary polarization in the strong attractive regime is given by @xmath903 .
the energy @xmath904 of the cluster state of an @xmath905-atom is given by @xcite @xmath906\right\}. \nonumber\end{aligned}\ ] ] the coefficients @xmath907 , @xmath908 and @xmath909 can be found in guan _ et al . _
( 2012 ) .
this closed form for the groundstate energy with arbitrary polarization is very accurate for a finitely strong attraction ( for @xmath910 ) .
this high precision groundstate energy has been given for the two - component attractive fermi gas @xcite .
up to order @xmath911 , the above groundstate energy of 1d @xmath843-component fermions with arbitrary population imbalance can be further simplified to the general form given in ( [ e ] ) . from the result ( [ e - a5 ] )
one can obtain full phase diagrams and magnetism at zero temperature .
in particular , for the three - component fermi gases @xcite , spin-3/2 fermi gas with @xmath735 symmetry @xcite , spin-5/2 , 7/2 and 9/2 attractive fermi gases @xcite , and @xmath834 fermi gases @xcite .
the @xmath834 symmetry requires each hyperfine spin state to be conserved .
therefore , there are @xmath843 chemical potentials associated with each spin state . for convenience in analyzing the quantum phases of the multi - component attractive fermi gas ,
the zeemann energy is chosen as @xmath912 where @xmath913 is an effective external field for the cluster state of size @xmath905-atom . here
@xmath730 denotes the chemical potential for unpaired fermions .
the particle numbers are changed by @xmath914 , with @xmath35 the total chemical potential .
these effective fields result in unequally spacing zeeman splitting via @xmath915 . here
@xmath916 because of the spin singlet state . defining the effective chemical potential of the @xmath45-atom bound state by @xmath917
, the result @xmath918 with @xmath919 can be obtained from the groundstate energy . the energy - field transfer relation between @xmath920 and the effective chemical potentials @xmath921 @xcite @xmath922 determines full phase diagrams of the system in terms of the effective fields @xmath923 and chemical potentials . in the special case of pure zeeman splitting where @xmath924 for all @xmath925 , the system has three distinct magnetic phases for a strong attractive regime . for weak coupling , even for pure zeeman splitting ,
the phase diagram is very sophisticated , see fig .
[ fig : phase-4r ] . some subtle phase diagrams for the spin-3/2 , 5/2 , 7/2 and 9/2 attractive fermi gases
have been studied @xcite .
plane for @xmath926 .
the shaded area corresponds to the vacuum . from schlottmann and zvyagin ( 2012a ) .
, title="fig : " ] furthermore , for the polarized phase , the cluster states form multi - component tlls in the low energy physics .
the low - lying excitations are described by the linear dispersion relations @xmath927 , where the velocities can be calculated by @xmath928 } \label{volocity - v - r}\ ] ] for a system featuring galilean invariance . for a strong attractive interaction , the velocities for unpaired fermions and charge bound state of @xmath727-fermions
are given by @xmath770 with @xmath929 .
the function @xmath930 is given by ( [ a3 ] ) .
these dispersion relations naturally lead to the universal form for finite - size corrections to the groundstate energy @xmath931 @xcite , @xmath932 the central charge @xmath434 for @xmath781 symmetry . here
the universal finite - size corrections ( [ fsc ] ) indicate the tll signature of the many - body physics . in this case
the low energy excitations of the system are described by the cft of the gaussian model . in the next subsection
, we shall discuss the thermodynamics of these cluster states .
schlottmann ( 1993 ) derived the tba equations for @xmath834 fermions with repulsive and attractive interaction .
( 2011b ) derived a different set of tba equations which are more convenient for analysis of thermodynamics and phase transitions for the attractive fermi gas . to understand the thermodynamics of this model ,
it is crucial to separate different physical regimes , i.e. , 1 ) groundstate for @xmath430 ; 2 ) tll phases , @xmath933 and @xmath934 ; 3 ) quantum criticality , @xmath935 and @xmath936 ; 4 ) high temperatures , @xmath937 .
the tba equations involve an infinite number of coupled nonlinear integral equations . at zero temperature
, the phase diagrams can be analytically or numerically obtained through the dressed energy equations which can be derived from the tba equations in the limit @xmath430 @xcite . in the strongly attractive regime ,
the effective ferromagnetic spin - spin coupling constants are given by @xmath938 for @xmath939 .
@xmath940 is the pressure for charge @xmath727-atom bound states . in this sense
, we may simply view the non - neutral charge bound state as a molecule with spin @xmath941 , which could flip its spin to form the spin wave bound states ( spin strings ) due to thermal fluctuations .
however , in the physically interesting regime where @xmath942 , @xmath943 and @xmath944 the breaking of charge bound states and spin wave fluctuations are strongly suppressed .
the spin string contributions to thermal fluctuations in this regime can be asymptotically calculated from the tba i.e. , @xmath945 .
it is obvious that @xmath946 becomes exponentially small as @xmath947 .
thus each dressed energy can be written in a single particle form @xmath948 , where the marginal scattering energies among composites and unpaired fermions as well as spin - wave thermal fluctuations are considered in the chemical potentials @xmath949 . with the help of this simplification ,
the thermodynamics at finite temperatures have been given as @xcite @xmath950 for @xmath951 .
the total pressure of the system is given by @xmath952 . here
@xmath953 is the standard polylog function .
furthermore , the suppression of spin fluctuations leads to a universality class of a multi - component tll in each gapless phase , where the charge bound states of @xmath727-atoms form hard - core composite particles , i.e. , the leading low temperature corrections to the free energy reads @xmath954 the velocity @xmath955 of the @xmath727-atom bound state is given by ( [ volocity - v - r ] ) .
this result is consistent with the finite - size correction result ( [ fsc ] ) . in the above equation @xmath956 .
this result proves the existence of tll phases in 1d gapped systems at low temperatures .
the existence of the tll leads to a crossover from a relativistic dispersion to a nonrelativistic dispersion between different regimes at low temperatures @xcite .
linear zeeman splitting may result in a two - component luttinger liquid in a large portion of zeeman parameter space at low temperatures .
it is well established that the universality classes of critical behaviour of two - dimensional systems are described by a rational conformal field theory ( cft ) , where the hilbert space of states contains a direct sum of irreducible representations of a virasoro algebra . using the cft
one can calculate the critical exponents that characterize power law decay of correlation functions at large distance @xcite .
in general , cft predicts that the two - point correlation function for primary fields with conformal dimension @xmath957 is of the form @xmath958 here @xmath959 is the euclidean time ( @xmath960 ) and @xmath961 is the velocity of light .
the conformal dimensions @xmath962 can be read off from the finite - size corrections of low - lying excitations of 1d systems via @xmath963 where @xmath964 is the conformal dimension and @xmath965 is the conformal spin .
the non - negative integers @xmath966 and @xmath967 label the level of the descendant .
@xmath968 represents the number of particles backscattered .
moreover , it was shown by affleck ( 1986 ) that conformal invariance gives a universal form for the finite temperature effects in the free energy by replacing @xmath969 with @xmath1 in the conformal map @xmath970 . considering a conformal mapping of the complex plane without the origin ( corresponding to @xmath40 ) onto a strip of width @xmath971 in the imaginary time direction , the two point correlation function at finite temperatures reads @xmath972 with @xmath973^{2\delta^{\pm } } $ ] . for such critical phenomena ,
the critical hamiltonian of the gapless sector exhibits not only global scale invariance but also local conformal invariance , i.e. , a local version of the scale invariance
. thus the critical hamiltonian can be approximately described by the conformal hamiltonian which is described by the generators of the underlying virasoro algebra with central charge @xmath432 .
the qism provides a systematic way to calculate the critical exponents for ba integrable models .
( 1986 ) derived explicit critical exponents for the @xmath974 and @xmath975 chains . in this approach
, the dressed charge matrix @xmath976 formalism presents a unified framework to calculate the conformal towers through the finite - size corrections to the eigenspectrum of multi - component ba systems @xcite . the universality class of critical exponents is uniquely determined by the symmetry of the models associated with the quantum @xmath4-matrix .
consequently , the asymptotic behaviour of correlation functions for integrable models , such as impenetrable bose gas @xcite , the supersymmetric @xmath0 model @xcite and the hubbard model @xcite , has been studied in the framework of the qism . in addition , exact results for the form factors have been derived for several cases @xcite .
further developments in the theoretical study of correlation functions have been reported @xcite . as remarked in sec .
i , for a system with partial polarization , the fermi energies of spin - up and spin - down electrons are unlikely to match .
this leads to a non - standard form of pairing known as the fflo state .
the power - law decay of the pair correlation @xmath14 with the spatial oscillations depending solely on the mismatch of the fermi surfaces @xmath15 has been identified numerically .
this spatial oscillation is a typical characteristic of the fflo pairing .
the fflo - like pair correlations and spin correlation for the attractive hubbard model were investigated by several groups via various methods , such as dmrg @xcite , qmc @xcite , mean field theory and other methods @xcite .
the fflo signature is displayed by the on - site pair correlation function @xmath977 in the 1d attractive hubbard model @xcite where @xmath978 is the eigenstate for the groundstate of the system . in fig .
[ fig : pair - correlation ] , for polarization @xmath979 , the on - site pair correlation indicates a power - law decay without changing the sign ( i.e. , spatial oscillation ) . for low polarization @xmath980 ,
the pair correlation function still has a power - law decay but periodically changes its sign . for @xmath981 , the oscillation frequency of the pair correlation
becomes faster due to a large mismatch of the two fermi energies .
the wave vector @xmath205 of oscillations @xmath982 .
feiguin and heidrich - meisner ( 2007 ) showed that the pair momentum distribution function has a peak at the mismatch of the fermi surfaces @xmath983 , i.e. , @xmath984 where the pair correlation @xmath985 , see fig .
[ fig : pair - correlation - p ] .
the correlation exponent @xmath986 depends on both the polarization @xmath70 and interaction strength .
such a fflo pairing wave number was also confirmed by the occurrence of a peak in the pair momentum distribution corresponding to the difference between the fermi momenta of individual species @xcite . in the @xmath987 plane for polarization @xmath988 and @xmath989 .
right panels : the axial density distribution profiles in a harmonic trap .
numerical setting with interacting strength @xmath990 . from tezuka and ueda ( 2008 ) .
, title="fig : " ] .
( b ) power - law decay behaviour in real space and comparison of the dmrg result ( symbols ) to the bosonization ( lines ) .
( c ) the positions of the peaks in the momentum distribution function @xmath991 . from feiguin and heidrich - meisner ( 2007 ) .
, title="fig : " ] on the other hand , the critical behaviour of 1d many - body systems with linear dispersion in the vicinities of their fermi points can be described by conformal field theory .
the critical behaviour of the hubbard model with attractive interaction was investigated by bogoliubov and korepin ( 1988,1989,1990,1992 ) . as demonstrated in previous sections , the low - energy physics of the homogeneous 1d fermi gases with polarization
is described by the tlls of bound pairs and excess unpaired fermions in the charge sector and ferromagnetic spin - spin interactions in the spin sector @xcite .
this paves a way to study asymptotic behaviour of various correlation functions by using cft .
the method used to study correlation functions of the spin-1/2 fermi gas with attractive interaction follows closely the method set out in the literature @xcite .
consequently , the asymptotic correlation functions and fflo signature of the 1d attractive spin-1/2 fermi gas have been analytically studied by the dressed charge formalism @xcite .
this study can be carried out naturally for 1d attractive multi - component fermi gases @xcite . in the gapless phase ,
the bound pairs and excess unpaired fermions form two fermi seas which can be described by a two - component tll , where the spin fluctuations are strongly suppressed at low temperatures .
the conformal dimensions of two - point correlation functions of the 1d attractive spin-1/2 fermi gas can be calculated from the elements of the dressed charge matrix @xmath992 describing the finite - size corrections for the low - lying excitations .
the long distance asymptotics of various correlation functions are then examined through the dressed charge formalism at the @xmath993 .
the finite - size corrections to the groundstate energy were computed from the ba equations , with result @xcite @xmath994 where @xmath435 and @xmath436 are the velocities of unpaired fermions and bound pairs , respectively .
three types of low - lying excitations are considered in the calculations of finite - size corrections : a type 1 excitation is characterized by moving a particle close to the right or left fermi points outside the fermi sea ; a type 2 excitation arises from the change in total number of unpaired fermions or bound pairs ; a type 3 excitation is caused by moving a particle from the left fermi point to the right fermi point and vice versa .
such kinds of excitation are also known as backscattering .
all three types of excitations can be unified in the following form of the finite - size corrections for the energy and total momentum of the system @xcite @xmath995 with the notation @xmath996 here the quantum numbers @xmath997 are characterized by the change in quantum numbers @xcite @xmath998 the dressed charges @xmath999 , @xmath1000 , @xmath1001 and @xmath1002 satisfy a set of dressed charge equations @xcite .
the finite - size spectrum is described by a critical theory based on the product of two virasoro algebras each of which has a central charge @xmath434 .
the finite - size scaling form of the energy ( [ e - finite ] ) determines the critical exponents of two - point correlation functions between primary fields @xmath1003 .
at zero temperature , the two - point correlation functions take the form @xmath1004 the conformal dimensions are given by @xmath1005 here @xmath1006 ( @xmath1007 ) characterize the descendent fields from the primary fields . in this way
the quantum numbers for the low - lying excitations completely determine the nature of the asymptotic behaviour of these correlations .
the exponential oscillating term in the asymptotic behaviour comes from the backscattering process .
various correlation functions , e.g. , the single particle green s function @xmath1008 , charge density correlation function @xmath1009 , spin correlation function @xmath1010 and pair correlation function @xmath1011 , can be derived based on the choice of @xmath1012 which define the quantum numbers ( [ qn - d ] ) @xcite .
we now discuss these correlation functions in detail .
the asymptotic form of the single particle green s function @xmath1013 is given by @xmath1014 where the critical exponents for the strong coupling regime are given in terms of the polarization by @xmath1015 the constants @xmath1016 and @xmath1017 can not be derived from the finite - size corrections for low - lying excitations .
the asymptotic form of the pair correlation function @xmath1018 is given by @xmath1019 where the critical exponents in the strong coupling regime are @xmath1020 it is clearly apparent that the leading order for the long distance asymptotics of the pair correlation function @xmath1011 oscillates with wave number @xmath1021 , where @xmath1022 .
the cooper pair correlation has been further discussed with a connection to cft @xcite .
moreover , the leading order for the charge density correlation function @xmath1023 and the spin correlation function @xmath1010 oscillates twice as fast with wave number @xmath1024 , namely @xmath1025 here @xmath1026 and @xmath1027 are constants and the correlation exponents are given by @xmath1028 the oscillations in @xmath1011 , @xmath1009 and @xmath1010 are caused by the mismatch in fermi surfaces between both species of fermions .
these spatial oscillations give a novel signature of the larkin - ovchinikov pairing phase @xcite .
this finding is consistent with the numerical results from dmrg @xcite and qmc @xcite .
it is remarkable to see that the asymptotics of the spatial oscillation terms in the pair and spin correlations are a consequence of type 3 excitations , i.e. , backscattering for bound pairs and unpaired fermions .
the asymptotic behaviour of the fermi field , cooper pair and charge density wave correlation functions for the 1d attractive hubbard model were computed for the magnetic field @xmath1029 at the half - filled band @xcite .
furthermore , the correlation functions in momentum space can be given by taking the fourier transform of their counterparts in position space , i.e. , the fourier transform of equal - time correlation functions reads @xmath1030 where @xmath1031 is given by @xmath1032^{2s}|k - k_{0}|^{\nu}.\ ] ] here the conformal spin of the operator is @xmath1033 and the exponent @xmath37 is expressed in terms of the conformal dimensions by @xmath1034 .
however , experimental observation of this momentum distribution remains very difficult because the transverse expansion dominates the expansion along the axial direction @xcite . in this regard ,
it is practicable to consider the long - time behaviour of the distribution during the sudden expansion of spin - imbalanced ultracold lattice fermions with an attraction after turning off the longitudinal confining potential .
however , the existence of the fflo signature in the expanding gas is still in question @xcite . for fields above the lower critical field @xmath312 ,
these correlators decrease as power - laws .
thus the pairs lose their dominance , i.e. , three types of ordering coexist superconductivity , charge density waves and spin density waves .
indeed the fflo correlation is more robust in the 1d homogenous attractive fermi gas . in the experimental setting discussed in sec .
vii , the quasi-1d system of ultracold fermionic atoms is formed by loading into a bunch of quasi-1d tubes created by two counter - propagating laser beams . by tuning the intensity of the lasers one could adiabatically tune the tunnelling between two neighbour tubes .
such josephson - like tunnelling is likely to lead to a 3d long - range order of the superconducting phase . a quasi-1d model consisting of attractive hubbard chains with interchain tunnelling @xmath1035 @xcite , where @xmath1035 is much less than the energy gap @xmath295 . for a singlet pair state ,
the charge density wave is suppressed , thus the system shows an anomalous average value , i.e. , @xmath1036 .
this means that a superconductive current exists .
through analysis of the instability of the normal state ( refer to the tll phase ) to the superconductor transition , the critical temperature is estimated to be @xmath1037^{1/(2-\gamma ' ) } $ ] , where the @xmath1038 is the critical pair correlation exponent of the 1d homogeneous attractive hubbard chain @xcite . for @xmath1039 , the gapless excitation spectrum of
the cooper pair leads to power - law behaviour of pair correlation function . in sec .
[ s - c - separation ] we saw that for 1d repulsive spin-1/2 fermions , the low energy physics of the model can be reformulated as two massless bosonic theories for the charge and spin degrees of freedom with dispersions @xmath1040 .
based on this spin - charge separation , the bosonization approach has been used to compute correlation functions of the 1d hubbard model @xcite . due to the fact that the spin and charge excitations are independent of each other
, the description of critical phenomena in the repulsive hubbard model has been made by the conformal field theory approach @xcite .
using the critical theory based on the product of two virasoro algebras with central charge @xmath434 , one can systematically compute asymptotics of correlation functions from the finite - size spectrum of low - lying excitations in terms of the ba solution @xcite . in the same fashion
, the long distance asymptotics of various correlation functions of the spin-1/2 fermi gas have been investigated in the strong repulsive regime @xcite .
the model is gapless and thus critical at zero temperature . at @xmath993 the correlation functions decay as some power of distance governed by the critical exponents .
for @xmath1041 the decay is exponential . in the context of spin - charge separation ,
the general two - point correlation function for primary fields @xmath1042 with conformal dimensions @xmath1043 at @xmath993 and @xmath1041 are given by @xmath1044 and @xmath1045 \sinh^{2\delta_{a}^{-}}\left[\frac{\pi t}{v_a}(x+\mathrm{i } v_{a}t)\right ] } \nonumber\end{aligned}\ ] ] where @xmath1046 are the fermi momenta , @xmath1047 and @xmath1048 is euclidean time .
the conformal dimensions of the fields can be written in terms of the elements of the dressed charge matrix as @xmath1049 here the dressed charge matrix is denoted by @xmath1050 which can be obtained from the dressed charge equations , see @xcite .
@xmath1051 are the fermi boundaries for charge and spin degrees of freedom .
the non - negative integers @xmath1052 , @xmath1006 and the parameter @xmath1053 where @xmath1054 characterize the three types of low - lying excitations . @xmath1052 denotes the change in the number of down - spin fermions .
@xmath1006 characterizes particle - hole excitations where @xmath1055 ( @xmath1056 ) is the number of particles at the right ( left ) fermi level jumps to .
@xmath1053 represents fermions that are backscattered from one fermi point to the other , i.e. , @xmath1057 with the expressions ( [ correlation-0 ] ) and ( [ correlation - t ] ) the general two - point correlation functions for the operator @xmath1058 , namely @xmath1059 , can be written as a linear combination of primary fields with conformal dimensions @xmath1043 and their descendent fields from the finite - size spectra of the model . various correlation functions of operators can be written in terms of the field operators @xmath1060 and @xmath1060 where @xmath1061 .
e.g. , \(i ) one particle green s function : @xmath1062 \(ii ) charge density correlation function : @xmath1063 where @xmath1064 and @xmath1065 .
\(iii ) longitudinal spin - spin correlation function : @xmath1066 where @xmath1067 .
\(iv ) transverse spin - spin correlation function : @xmath1068 where @xmath1069 , and @xmath1070 .
\(v ) pair correlation function : @xmath1071 the critical exponents for each of the above correlation functions are determined by the values of the quantum state with @xmath1072 with @xmath1073 for every case .
the explicit results for these correlation functions for @xmath562 and @xmath316 are given by solving the dressed charge equations @xcite .
in the zero field limit @xmath562 the dressed charge equations were solved by the wiener - hopf method , explicitly , @xmath1074 here @xmath1075 with the critical field value @xmath1076 for strong repulsion . for the approach to the critical field
@xmath1077 the dressed charge equations can be solved by asymptotic expansion @xcite , with result @xmath1078 a few terms of the asymptotic expansion for the dressed charges together with the above values for the low - lying excitations determine the asymptotics of the correlation functions . as an illustration of this approach , in the zero field limit ,
the asymptotics of the density - density correlation function are given by @xmath1079 where the critical exponents are @xmath1080 the presence of the external field does not change the form of the correlator ( [ g - nn ] ) . in the large field limit , i.e. , @xmath316 ,
the exponents are given by @xcite @xmath1081 the constants @xmath1082 with @xmath705 depend on the interaction .
for @xmath1083 , we see that the density - density correlation contains @xmath532 and @xmath1084-oscillations
. however , in the strong coupling limit , the system behaves like non - interacting spinless free fermions , thus the @xmath532 terms should vanish , i.e. , @xmath1085 .
the leading contributions to the asymptotics of the density - density correlation function are from the @xmath1084-oscillation term because this oscillation is a consequence of interactions @xcite .
furthermore , the large distance behaviour of the two - point correlation function determines the singularities of spectral functions near @xmath1086 .
the correlation functions in momentum space can be determined by fourier transforming the asymptotics of the above correlators .
of particular interest are the dynamical response functions such as the spectral function @xmath1087 which can be obtained from the imaginary part of the retarded single particle green s function .
the interacting spectral function often has a non - zero width .
there are singularities in the spectral function @xmath1088 for @xmath1089 @xcite , with @xmath1090 here the exponent @xmath1091 , which can be obtained from fourier transformation , is always greater than zero .
using tll theory , the fourier transforms of the zero - temperature single particle green s function have been computed @xcite .
the fourier transform of the @xmath532 luttinger liquid density correlation function was calculated recently in @xcite . in general , the explicit calculation of the spectral function is much more involved @xcite . in general , @xmath834 wess - zumino - witten ( wzw ) theory can be used to capture the low energy behaviour of a family of critical quantum spin chains @xcite .
the @xmath834 wzw models of level @xmath905 describe integrable higher spin models with critical points characterised by the central charge @xmath1092 and the scaling dimensions of the primary fields @xcite .
the @xmath1093 symmetry interacting fermions in 1d have @xmath843 branches of states characterised by one charge degree of freedom and @xmath855 spin rapidities , see eq .
( [ fe - r ] ) .
the low - lying excitations are described by the linear dispersion relations @xmath1094 with the charge velocity @xmath560 and spin velocities @xmath955 for @xmath1095 , respectively .
the ba result @xcite predicts that the energy has a universal finite - size scaling in the low energy excitation spectrum , i.e. , @xmath1096 .
the low energy spectrum can be interpreted in terms of a product of @xmath843 independent virasoro algebras with the same central charge @xmath434 . for vanishing magnetic field ,
the spin velocities are equal .
thus we have a spin - charge separation of the @xmath434 gaussian field theory ( in the charge sector with @xmath781 symmetry ) and @xmath1097 wzw theory ( in the spin sector with @xmath834 symmetry ) . using the ba solution , frahm and schadschneider ( 1993 ) and kawakami ( 1993 )
have calculated finite - size corrections and the spectrum of the low - lying excitations in the 1d multi - component degenerate hubbard model .
the asymptotic behaviour of various correlation functions has been determined from these low - lying spectrum . in the same fashion
, the asymptotics of correlation functions for the bose - fermi mixtures have been studied @xcite .
the general structure of the critical exponents for the multiple nested ba solvable models has been well understood , see a review by schlottmann ( 1997 ) .
the low - lying excitations above the groundstate energy of the critical fermi gases with @xmath834 symmetry have been obtained @xcite @xmath1098 where @xmath1099 and @xmath1100 are positive integers characterizing the excited states at the branches @xmath1101 . here
@xmath1102 stands for the charge degree of freedom . @xmath1103 and @xmath1104 are vectors , where @xmath1105 in vector @xmath1106 denotes the changes of the total numbers in each branch .
the values of @xmath1107 in vector @xmath1108 are integer or half - odd integer depending on @xmath1109 , i.e. , @xmath1110 with @xmath1111 and @xmath1112 .
the conformal dimensions @xmath1113 of the primary field are given in terms of the dressed charge matrix @xmath1114 @xcite @xmath1115 the dressed charges @xmath1116 can be obtained by solving the dressed charge equations obtained from the dressed energy equations by definition @xcite .
consequently , different dressed charge equations are accordingly derived for the multi - component fermi gases in the repulsive and attractive regimes . the dressed charges for the degenerate @xmath834 hubbard model
was calculated explicitly in the absence of magnetic field @xcite . the single particle green s function and charge density - density correlation
were thus studied . in general , using the cft result , the asymptotics of correlators for primary fields at @xmath993 are given by @xmath1117 } } { \prod_{r=0}^{\kappa-1 } ( x - iv_{r}t)^{2\delta^{+}_{r}}(x+iv_{r}t)^{2\delta^{-}_{r}}}.\label{correlation - n}\ ] ] here @xmath1118 is the fermi momentum of each branch .
the critical exponents of the correlation functions of 1d multi - component fermi gases at zero temperature in external magnetic fields can be calculated in a straightforward way .
some response functions in multi - component luttinger liquids were computed analytically @xcite .
the quantum numbers of low - lying excitations , such as @xmath1119 and @xmath1120 , completely determine the nature of the asymptotic behaviour of the correlation functions .
there are many superfluid orders in multi - component attractive fermi gases , for which the operators for the superfluidity @xmath1121 are written in terms of fermion anniliation operators of length-@xmath727 .
the instability of superfluidity and normal phase can be determined by the correlation function ( [ correlation - n ] ) . in the context of ultracold gases with higher spin symmetries , schlottmann and zvyagin ( 2012a,2012b )
have studied various superfluidity ordering in spin-3/2 and spin-5/2 attractive fermi gases , in which the relevant superfluidity operators of few - body bound states were labeled . however
, the study of colour superfluidity , the fflo - like modulated ordering in 1d interacting fermions with large spins still remains preliminary .
it remains to further investigate magnetism , superfluidity , charge and spin density waves by using cft and the ba solutions .
the universal nature and phenomena of interacting fermions , such as landau s fermi liquid theory , luttinger liquid theory , and quantum criticality , have always attracted great attention from theory and experiment .
tan ( 2008a , 2008b,2008c ) has shown that a few universal relations for two - component interacting fermions involve an extensive quantity called the universal contact @xmath1122 .
the first tan relation is for the tails of the momentum distribution which exhibits a universal @xmath1123 decay as the momentum tends to infinity .
here we denote @xmath1124 .
the constant @xmath1122 measures the probability of two fermions with opposite spin at the same position .
secondly , this contact also reflects the rate of change of the energy or free energy due to a small change in the inverse scattering length @xmath1125 for fixed entropy @xmath833 or temperature @xmath1 @xmath1126 tan s additional relation states that the pressure and the energy density are related by @xmath1127 .
the tan relations were found as a consequence of operator identities following from a wilson operator product expansion of the one - particle density matrix @xcite .
these relations hold more generally as long as the interaction range @xmath1128 is much smaller than any other characteristic length scale like the average interparticle distance and the thermal wavelength @xcite .
prior to tan s results , the derivative of the energy @xmath856 with respect to the inverse scattering length @xmath87 has been experimentally examined from measurements of the photoassociation rate of a trapped gas of @xmath577li atoms @xcite . the static and dynamic structure factor for the trapped gas of @xmath577li atoms
has been studied the by using bragg spectroscopy @xcite . in particular , tan s universal contact and thermodynamic relations
have been confirmed for a trapped gas of @xmath580k atoms @xcite .
recent developments on the tan relations for fermions with large scattering length have been reviewed by braaten ( 2012 ) .
the generalization of tan relations to 1d interacting fermions has been studied by barth and zwerger ( 2011 ) using the operator product expansion method developed by kadanoff ( 1969 ) and wilson ( 1969 ) .
the tan adiabatic theorem has been generalized to the 1d gaudin - yang model for which the 1d analog of the tan adiabatic theorem is given in the form @xmath1129 . tan s
universal contact also exists in the asymptotic behaviour of the tail momentum distribution of the 1d fermi gas . in terms of the fields @xmath1130 and @xmath1131 the momentum distribution given by @xmath1132 . using the operator product expansion method ,
barth and zwerger ( 2011 ) found that the momentum distribution of the 1d two - component fermi gas behaves as @xmath1133 as @xmath1134 .
furthermore , using scaling analysis , they also derived the pressure relation which connects pressure and energy density @xmath1135 via the universal contact @xmath1136 for the 1d fermi gas .
tan s universal contact also arises in the additional relation for the pair distribution function and the related static structure factor . from the short distance expansion of the two particle density matrix barth and zwerger ( 2011 ) found the total pair distribution function at short distance @xmath1137 this short distance singularity gives rise to a @xmath1138 power law tail in the associated static structure factor @xmath1139 , i.e. , @xmath1140 with the dimensionless interaction constant @xmath1141 .
in fact , for the 1d two - component fermi gas , the universal contact is obtained by calculating the change of the interaction energy with respect to interaction strength by the hellman - feynman theorem , i.e. , @xmath1142 .
the local pair correlation @xmath1143 is accessible via the exact ba solution through the relation @xmath1144 . here
@xmath54 and @xmath856 is the groundstate energy per unit length . in a similar way , for the 1d @xmath843-component fermi gas
, there exists a 1d analog of the tan adiabatic theorem where the universal contact is given by the local pair correlations for two fermions with different spin states .
the two - body local pair correlation function is given by the expectation value of the four - operator term in the second quantized hamiltonian , explicitly , @xmath1145 with @xmath1146 @xcite .
@xmath856 is again the groundstate energy per unit length . from the asymptotic expansion result for the groundstate energy of the balanced fermi gas obtained in section [ fermi - n ] , we easily find @xmath1147 as @xmath1148 .
the local pair correlation can be obtained from the groundstate energy ( [ e - r1 ] ) of the balanced gas with strong repulsion , @xmath1149 .
\label{g2-rs}\end{aligned}\ ] ] this local pair correlation reduces to that for the spinless bose gas @xcite as @xmath868 . for strong attractive interaction , the local pair correlation for two fermions with different spin states is given by @xcite @xmath1150+o(1/\gamma^4).\label{g2as}\end{aligned}\ ] ] here @xmath1151 and @xmath1152 .
it is noted that for the whole interaction regime , the local pair correlations for the balanced @xmath843-component fermi gas tend to the limiting value for the spinless bosons as @xmath868 .
[ fig : g2-n ] shows the local pair correlation for two fermions with different spin states in the multi - component fermi gas .
vs @xmath1141 for the balanced multi - component fermi gas for @xmath1153 and for the spinless bose gas .
the solid lines are the numerical solutions obtained from the two sets of fredholm equations .
the dashed lines are the asymptotic limits obtained from the first term in ( [ g2as ] ) . from guan _ et al . _
( 2012 ) . ]
in order to realize 1d systems in experiments , one has to apply strong confinement in two transverse directions and allow free motion along the longitudinal direction .
1d systems have been simulated in various settings from solid electronic materials to ultracold atomic gases .
although some 1d features have been observed in solid electronic materials , it is still challenging to control material parameters and separate imperfect effects caused by defects and coulomb interactions etc .
in contrast , ultracold atomic gases trapped in external potentials have high controllability and clean environment . in this section ,
we briefly review the experimental developments in confining quantum systems of ultracold atoms in 1d .
neutral atoms can be trapped by coupling their permanent or induced dipole moments to electromagnetic field gradients . by using a laser field ( or an inhomogenous magnetic field ) , it is possible to trap neutral atoms via the interaction between the induced electric dipole moment and the laser electric field ( or via the interaction between the permanent magnetic dipole moment and the magnetic field ) . correspondingly , there are two typical techniques for realizing 1d quantum atomic gases : optical lattices and atom chips .
optical lattices are created by superimposing one or more pairs of counter - propagating laser beams @xcite , see fig .
[ fig : ol ] . the laser - atom interaction results in an ac - stark shift dependent on the laser intensity and the detuning from resonance @xmath1154 , in which @xmath1155 is the laser frequency and @xmath1156 is the atomic transition frequency .
therefore , due to the reason that optical lattices have periodic laser intensities , their ac - stark shifts provide periodic potentials for confining atoms . for a negative detuning ( red detuning )
, atoms are attracted to the region of large laser intensities . on the contrary , for a positive detuning ( blue detuning ) , atoms are attracted to the region of small laser intensity .
similar to the electrons in a conductor , even if the lattice well depth exceeds the kinetic energy of the atoms , the atoms confined in optical lattices can move among neighbouring lattice sites via quantum tunnelling . in the case of bose atoms ,
the quantum phase transition between superfluid and mott insulator phases has been experimentally observed @xcite .
the groundstate appears as a superfluid when the system is dominated by the quantum tunnelling among neighbouring sites . whereas the groundstate appears as a mott insulator if the system is dominated by the on - site interaction between atoms . for fermionic atoms , the metal - insulator transition and neel anti - ferromagnetic states etc
have been demonstrated in recent experiments @xcite .
ultracold atomic systems confined in optical lattices have highly controllable parameters such as the inter - site hopping strength , on - site interaction strength and dimensionality .
the inter - site hopping strength can be adjusted by tuning the laser intensity of the optical lattices .
the on - site interaction strength can be tuned from positive infinity to negative infinity by using feshbach resonances . in particular
, the dimensionality can be tuned from 3d to 1d by controlling the spatial configuration of the optical lattices , see fig .
[ fig : ol ] .
the 1d lattice can be created by a pair of counter - propagating laser beams to form of a standing - wave , @xmath1157 the 2d square lattice can be formed by two orthogonal standing waves , @xmath1158.\ ] ] the 3d simple cubic lattice can be produced by three orthogonal standing waves , @xmath1159.\ ] ] by imposing laser beams under different spatial configurations , it is also possible to make more complex lattices such as the triangle , honeycomb and kagome lattices .
it has been demonstrated that an array of 1d systems can be created by superposing a harmonic potential onto a 2d optical lattice .
the potential for such a system is of the form @xmath1160 under the tight - binding condition , each lattice well can be regarded as an independently harmonic potential .
that is , the potential @xmath1161 around the lattice site @xmath1162 can be expressed approximately as @xmath1163 + \frac{1}{2}m \omega_{z}^{2}z^2.\ ] ] if the lattice depth is sufficiently large , we have @xmath1164 and so all atoms will always stay in the lowest transverse vibrational state along the @xmath371 and @xmath1165 directions .
the absence of transverse excitations means that each lattice tube along the @xmath36-axis can be viewed as a quasi-1d system .
experimentally , the longitudinal frequency @xmath1166 is around @xmath1167 hz and the transverse frequency @xmath1168 is around @xmath1169 khz @xcite .
atom chips are a kind of nanofabricated atom - optical circuit used to trap and manipulate neutral atoms .
the basic idea of atom chips is to build atomic traps through a bias magnetic field and a system of electric wires fabricated on chip surfaces .
thus the magnetic potential of a straight wire has a hole along the wire . to close the magnetic potential with endcaps ,
there are two simple schemes : one scheme is achieved by combining a straight wire with an inhomogeneous bias field , the other scheme is achieved by combining a bent wire with a homogeneous bias field , see fig .
[ fig : atomchip ] .
the basic property , practical design and experimental procedure of atom chips have been reviewed in the literature @xcite . here ,
we briefly review how to use atom chips to realize 1d traps with neutral atoms . and the chip wire @xmath1170 .
the longitudinal confinement is generated by the other two chip wires @xmath1171 and @xmath1172 . from the phd thesis of a. van amerongen ( university of amsterdam ) , which can be downloaded from http://staff.science.uva.nl/ walraven
/ walraven / theses.htm . ]
one can obtain a highly elongated trap if the longitudinal confinement is very weak compared to the transverse confinements . as illustrated in fig .
[ fig : atomchip ] , the wire with current @xmath1173 and the homogenous bias field @xmath1174 form a waveguide along the longitudinal direction .
this waveguide is closed by two perpendicular wires with currents @xmath1175 and @xmath1176 .
near the trap center , the magnetic field is approximated by @xmath1177 with @xmath1178 denoting the value of the magnetic field in the trap center .
the corresponding magnetic potential is given as @xmath1179 with the trapping frequencies @xmath1180 @xmath1181 where @xmath1182 is the atomic magnetic moment and @xmath45 is the single atomic mass . at finite temperature ,
a real system enters its 1d regime when both the residual thermal energy ( of order @xmath1183 ) and the chemical potential @xmath35 are far less than the transverse excitation energy @xmath1184 , i.e. , @xmath1185 .
experimentally , the atom chips with a z - shaped wire can successfully generate quasi-1d potential traps of high frequency ratio @xmath1186 up to a few hundred @xcite . typically , the transverse frequency @xmath1187 is in an order of khz and the longitudinal frequency @xmath1188 is about 10 hz .
it has also been demonstrated that a 1d - box trap , which has nearly constant potential at the trap minimum combined with tight harmonic confinement in the transverse directions , can be formed by positioning two wiggles in a long straight wire @xcite .
the 1d quantum gases of @xmath42-function contact interaction @xmath1189 are characterized by the dimensionless parameter @xmath252 @xcite , which is defined as the ratio between the interaction energy @xmath1190 and the kinetic energy @xmath1191 , @xmath1192 . here
, @xmath759 denotes the coupling constant and @xmath1193 is the 1d number density .
the coupling constant @xmath759 is determined by the 3d scattering length @xmath87 and the transverse width of the wave function @xcite , @xmath1194 where the constant @xmath1195 and the transverse width @xmath1196 . if @xmath1197 , the atom - atom scattering acquires a 3d character , the interaction strength @xmath759 is given as @xmath1198 .
theoretically speaking , by varying the interaction strength @xmath759 , the gradual transitions from quasi bec and the tg gas appear in 1d bose systems , the bcs - bec - like crossover takes place in 1d two - component fermi systems .
experimentally , to explore different regimes of 1d quantum atomic gases , one may tune the 3d scattering length @xmath87 via the well - developed techniques of feshbach resonance .
feshbach resonance takes place when the energy of a bound state ( a state belonging to a closed channel ) for the inter - particle potential is close to the kinetic energy of the two colliding particles .
the scattering length of the two colliding ultracold atoms undergo a feshbach resonance if the energy of a molecular state is close to the kinetic energy of the the colliding pair of atoms .
the interaction strength between a pair of ultracold atoms is proportional to the scattering length . a positive ( negative ) scattering length gives repulsive ( attractive ) interaction .
the scattering length @xmath87 near a magnetic feshbach resonance point @xmath1178 is given by @xmath1199 with the background scattering length @xmath1200 , the magnetic field @xmath678 and the resonance width @xmath295 .
obviously , the sign and strength of the scattering length can be tuned by adjusting the magnetic field @xmath678 .
feshbach resonances provide an excellent tool for controlling the ultracold interaction between atoms and have been widely used for exploring different regimes of quantum atomic gases .
the theory of feshbach resonances in ultracold atomic gases has been introduced in @xcite .
the experimental developments in this field can be found in a recent review paper @xcite .
there are two usual probes for exploring the physics of quantum atomic gases .
one probe is the correlation function , which measures correlations between the atomic fields at different positions and times .
the correlation function has been extensively used for exploring many - body coherence and distinguishing order and disorder .
the other probe is the dynamical structure factor , which describes the total probability to populate any excited state by transferring both momentum and energy from an initially equilibrium state .
the dynamical structure factor is a powerful quantity for characterizing the low - energy excitations .
the first - order correlation function ( also called the single - particle correlation function ) is @xmath1201 here @xmath1202 and @xmath1203 denote the field operators for atoms in states @xmath1204 and @xmath1205 .
obviously , the first - order correlation function becomes the density @xmath1206 if @xmath1207 and @xmath1208 .
the density profile can be mapped out by the well - developed techniques of optical imaging , in which information about an atomic gas is encoded onto the absorption , phase shift or polarization information of the probe laser .
one popular optical imaging technique is absorption imaging , which is implemented by comparing imagines taken with and without the atomic gas within the field of view and recording the fractional absorption of the probe laser .
however , absorption imaging shows large uncertainty in the measured column density in probing high density atomic gases .
therefore , absorption imaging is usually taken after the atomic gas is released from the trap and expanded for a certain time .
this imaging method is known as time - of - flight imaging @xcite . in time - of - flight imaging ,
the fractional absorption is proportional to the expectation value of the atomic density operator @xmath1209 . by assuming negligible atomic collision effects in the ballistic expansion during the time of flight ,
the density @xmath1210 is proportional to the initial momentum distribution @xmath1211 with the corresponding wave vector @xmath1212 , i.e. , the time - of - flight imaging gives the initial momentum distribution @xmath1213 however , if the time - of - flight is not long enough and the atomic collision effects can not be ignored during the time of flight , the measured density distribution @xmath1210 will not be proportional to the initial momentum distribution @xmath1211 with @xmath1212 @xcite . recently , different from free expansion via time - of - flight , a controlled expansion via shortcuts to adiabaticity has been proposed @xcite . by analysing the correlation of different time - of - flight images , one may reconstruct the density - density correlation function @xmath1214 .
this technique is known as noise spectroscopy @xcite . in different runs of experiments ,
the time - of - flight images have technical noises and quantum fluctuations at the same time .
each time - of - flight imaging is a measurement of the density operator . to explore the quantum correlation of the density operator
, all technical noises should be reduced below the quantum fluctuations .
therefore , the measurement times @xmath241 of imaging must be sufficiently large so that the statistical error @xmath1215 is small enough . the density - density correlation function is a special case of the second - order correlation function @xmath1216 which is also called the two - particle correlation function .
to reconstruct the full correlation functions , in addition to the measurement of density ( diagonal correlations ) , one has to measure the non - diagonal correlations .
it has been suggested that the non - diagonal correlations @xmath1217 can be measured by atomic interferometry @xcite . the fourier sampling of time - of - flight imagines
may also be used to reconstruct the full one - particle and two - particle correlation functions @xcite . in this detection scheme ,
two consecutive raman pulses at the beginning of the free expansion are used to induce a tunable momentum difference to the correlation terms so that both diagonal and nondiagonal correlations in the momentum space can be detected . the first raman operation is implemented by two travelling - wave beams of different wave vectors @xmath1218 and @xmath1219 .
the corresponding effective raman rabi frequency has a spatial dependent phase with @xmath1220 , where @xmath1221 and @xmath1222 is a constant phase . the second raman operation is implemented by two counter - propagating laser beams of the resonant frequency for the transition between the two involved hyperfine levels .
in contrast to the first raman operation , the effective raman rabi frequency for the second raman operation is a spatial independent constant @xmath1223 .
therefore , through time - of - flight imaging , the real and imaginary parts of the nondiagonal correlation function @xmath1224 can be measured by choosing the relative phase @xmath1225 and @xmath1226 , respectively .
combined with the techniques of noise spectroscopy , through analyzing the correlations of imagines corresponding to @xmath1224 and @xmath1227 , respectively , the two - particle correlation function @xmath1228 can be reconstructed .
if the atomic gases have two spin components @xmath1229 and @xmath1230 , the spin - spatial correlations @xmath1231 can be reconstructed by a combination of fourier sampling with a pair of raman pulses which mixes the two spin components .
therefore , in addition to the single - spin density - density correlation functions @xmath1228 , the opposite - spin density - density correlation functions @xmath1232 can be reconstructed .
based upon the backward - scattering of incident particles ( such as photons , electrons and atoms ) from a target sample , bragg scattering is a powerful tool for determining both the momentum and the energy absorbed by the target sample . in experiments with ultracold atomic gases , unlike photons are diffracted by an atom grating , atoms are diffracted on a grating of coherent photons ( lasers ) .
bragg scattering provides an effective tool of spectroscopy , called bragg spectroscopy , which can be used to detect the dynamical structure factor and explore the intrinsic mechanism of the low - energy excitations .
the dynamical structure factor @xmath1233 describes the total probability to populate an excited state by transferring a momentum @xmath1234 and energy @xmath1235 . at nonzero temperatures ,
the dynamical structure factor is @xcite @xmath1236 where @xmath1237 and @xmath1238 are initial and final states of the many - body system , with energies @xmath1239 and @xmath1240 .
the operator @xmath1241 creates a particle of momentum @xmath1242 .
the function @xmath1243 with @xmath1244 is the partition function . at zero temperature
, the equilibrium system can only occupy its groundstate and the dynamical structure factor is expressed as @xmath1245 in a two - photon bragg transition , the momentum shift @xmath1246 is controlled by wave vectors of the two bragg lasers and the frequency @xmath1247 is determined by the frequency difference .
the techniques of bragg spectroscopy have been successfully applied to measure the dynamical structure factor of bose condensed atoms @xcite and strongly interacting 3d bose @xcite and fermi @xcite atomic gases near feshbach resonance .
the techniques of bragg spectroscopy have been employed to detect the elementary excitations in an array of 1d bose gases @xcite . by varying the detuning of the bragg lasers , it is possible to probe both the spin and density dynamical structure factors for multi - component fermi systems @xcite . over the last decade ,
there have been many breakthrough experiments with 1d quantum atomic gases .
these experiments are implemented by using an ensemble of ultracold bose or fermi alkaline atoms occupying one or multiple hyperfine levels .
the key experiments in 1d quantum atomic gases are listed in table [ tab : exp_groups ] .
.[tab : exp_groups]key experiments in 1d quantum atomic gases . [ cols="<,<,<,<,<",options="header " , ] the 1d lieb - liniger bose gas is a prototypical many - body system featuring rich many - body physics
. this gas features three different regimes in quasi-1d traps : a true condensate regime , a quasi - condensate regime and the tonks - girardeau regime @xcite . in the limit of weak interaction , @xmath1248 ,
i.e. , in the quasi - condensate regime , where bose - einstein condensation may take place and mean - field theory well describes the low - energy physics . in this regime , due to the intrinsic nonlinearity from s - wave scattering between atoms , matter - wave solitons
have been observed in several experiments @xcite . in the limit of strongly repulsive interaction , @xmath1249 , i.e. , in the tonks - girardeau regime
, fermionization of bose atoms occurs . by loading ultracold bose atoms into a 2d optical lattices ,
the effective mass of quasi - particles can be increased by applying an additional periodic potential along the longitudinal direction @xcite and therefore an array of tonks - girardeau gases of quasi - particles can be prepared by increasing the effective mass .
the dimensionless parameter @xmath252 can also be changed without modulating the longitudinal trapping enabling the realization of a set of parallel 1d bose atomic gases in the tonks - girardeau regime @xcite . in the quasi - condensate regime of intermediate values of @xmath252 ,
the experimental measurements suggest that both mean - field theory and fermionization fail to describe this crossover @xcite .
the _ in situ _ measurements of the linear density of a nearly 1d trapped bose gas on an atom chip indicates a good agreement with theoretical prediction from the yang - yang thermodynamics equations @xcite .
breathing modes and dipole modes are two usual collective excitations in trapped quantum atomic gases .
the breathing modes can be excited by time - periodically modulating the longitudinal harmonic potential .
the dipole modes can be excited by suddenly displacing the longitudinal harmonic potential .
the ratio of the frequencies of the lowest breathing modes and the dipole modes @xmath1250 has been measured in an array of 1d gases in 2d optical lattices @xcite .
the experimental data show @xmath1251 for a lieb - liniger gas and @xmath1252 for a thermal gas , which are consistent with the theoretical results @xcite .
the experimental observation of strongly damped dipole oscillations due to quantum fluctuations @xcite has been reported @xcite . the 1d-3d crossover and quantum phase transitions from a 1d superfluid to a mott insulator have been observed in quasi-1d systems of an additional lattice potential along the longitudinal direction @xcite . for weak interaction ,
the system is still a superfluid at finite lattice depth and the transition to the mott insulator is induced by increasing the lattice depth . for strong interaction , an arbitrary perturbation by a lattice potential may induce a sine - gordon quantum phase transition from a superfluid luttinger liquid to a mott insulator .
these quantum phases can be well distinguished by detecting their low - energy excitations via the technique of bragg spectroscopy @xcite .
the techniques of atomic interferometry have been widely used to explore the phase coherence and fluctuations between different 1d gases @xcite .
the second - order correlation functions , which relate to the density fluctuations and spatial correlations , have been probed by _ in situ _ measurements of density fluctuations @xcite and _ ex situ _ measurements of photo - association rates @xcite .
these experiments confirm that the local second - order correlation functions are about @xmath1253 , @xmath1254 and @xmath1255 for an ideal 1d bose gas , a quasi - condensed 1d bose gas and a tonks - girardeau gas , respectively .
in addition to the density correlations , phase correlations have been probed by matter wave interferometry @xcite . moreover ,
the three - body correlation functions of 1d bose gases have been obtained by measuring the three - body recombination rate @xcite and the _ in situ _ measurements of third - order number fluctuations @xcite .
in addition to the measurements of local correlation functions @xcite , the temporal two - body correlation function has been measured by the techniques of scanning electron microscopy @xcite . beyond investigating equilibrium behaviour ,
there have been several experimental studies on non - equilibrium behaviour in 1d bose gases .
the coherence dynamics in both isolated and coupled 1d bose gases have been explored by using an atom chip @xcite .
the absence of thermalization in a 1d bose gas has been confirmed by the time evolution from an out - of - equilibrium state @xcite .
the non - equilibrium dynamics of an impurity in 1d bose gases , such as large density fluctuations , multiple scattering events and interaction dependent quadruple oscillations , have been studied in some recent experiments @xcite .
most recently , the experimental observations of quantum dynamics of interacting bosons @xcite and fast relaxation towards equilibrium of quasi - local densities , currents and coherence in an isolated strongly correlated 1d bose gas @xcite give a precise understanding of quantum dynamics and correlations , including observation of the spin dynamics in 1d two - component bose gases @xcite .
as remarked , quasi-1d trapped systems are created by tight transverse confinement .
a confinement - induced - resonance ( cir ) takes place when the incident channel of two incoming atoms and a transversally excited molecular bound state become degenerate .
it has been demonstrated that a cir takes place if the 3d scattering length approaches the length scale of the transverse trap .
the cir has been used to drive a crossover from the tonks - girardeau gas with a strongly repulsive interaction to a super tonks - girardeau gas with strongly attractive interaction @xcite .
in particular , the stable highly excited gas - like phase known as the super tonks - girardeau gas has been realized in the strongly attractive regime of bosonic cesium atoms . in a 3d
trapped fermi gas , the bound diatomic molecules exist only when the scattering length between the atoms is positive under s - wave interaction , i.e. , when @xmath1256 @xcite . in the limit @xmath1257 , the system is a weakly attractive fermi gas .
thus the groundstate is a bcs pair state for a negative scattering length @xcite .
however , in a 1d two - component trapped fermi gas , the scattering properties of two colliding atoms are altered by the tight transverse confinement .
the existence of a bound state does not rely on the sign of the scattering length . using ratio - frequency spectroscopy in an array of 1d fermi atomic gases trapped within a 2d optical lattice , moritz and coworkers
@xcite reported this that the two - body bound state exists irrespective of the sign of scattering length , see fig .
[ fig : binding - energy ] . in this experiment
, they further demonstrated that the bound states for negative scattering length can only be stabilized by the tight transverse confinement @xcite .
very recently , the particle and hole dynamics of fermionic atoms in amplitude - modulated 1d lattices was reported @xcite . the strong transverse confinement of a waveguide not only gives rise to an effective 1d s - wave interaction , but also alters the p - wave interaction in spin - polarized fermions .
due to the absence of s - wave interaction in the spin - polarized fermi gas , the p - wave interaction becomes dominant under resonant scattering conditions @xcite . in a spin - polarized fermi gas ,
the angular part of asymptotic collision wave functions is either the spherical harmonic @xmath1258 if the scattering state is parallel to the quantization axis or @xmath1259 if the scattering state is perpendicular to the quantization axis . therefore ,
in both the 2d and 3d cases , due to the coexistence of two collision channels and the breakdown of degeneracy between two collision channels @xcite , doublet structures of p - wave feshbach resonance have been observed @xcite .
however , for the 1d spin - polarized system , only one of two collision channels is involved .
thus there is a single peak structure of p - wave feshbach resonance in 1d spin - polarized fermions .
the experimental observation has confirmed such a particular signature @xcite , see fig .
[ fig : resonance - peak ] . by loading the spin - polarized fermi atoms into a deep 3d optical lattice
, one can prepare a band insulator of localized atoms in potential wells , which is viewed as a zero - dimensional ( 0d ) system .
there is no resonance feature in such a 0d system because the p - wave scattering is completely inhibited . when the geometry of the gas is tuned from 3d to 2d , the shift of the p - wave feshbach resonance depends on the depth of the optical lattice .
in contrast , 1d fermi gases show a further shift of resonance and give rise to a broadening of the loss feature . and
@xmath1260 ) give rise to a doublet feature .
( c ) in a 1d fermi gas with the spin aligned orthogonal to the atomic motion , the existence of only the collisional channel of @xmath1260 gives rise to a single peak feature .
( d ) in a 1d fermi gas with the spin aligned orthogonal to the atomic motion , the existence of only the collisional channel of @xmath1261 gives rise to another single peak feature .
( e ) in a 0d fermi gas within a deep 3d optical lattice , there is no resonant peak due to the absence of all collisional channels . from gnter _
( 2005 ) . ]
the 1d two - component fermi gas defined by the gaudin - yang model is an ideal system for exploring novel pairing mechanisms . in particular ,
the formation of a fflo - like state with a nonzero centre - of - mass momentum gives rise to a precise understanding of the coexistence of bcs pairs and polarizations . in 3d
, the fflo state occupies a tiny portion of the phase diagram and it is very difficult to observe in experiments .
in contrast , in the 1d gaudin - yang model @xcite the fflo state becomes much more robust due to the band fillings of two fermi seas related with one another , and it occupies major parts of the phase diagram , as demonstrated in fig .
[ fig : phase - s ] and fig .
[ fig : phase - attraction ] .
see also the experimentally measured phase diagram fig .
[ fig : phase - mu - h - e ] .
the system has spin population imbalance caused by a difference in the number of spin - up and spin - down atoms .
the key features of these @xmath993 phase diagrams have been experimentally confirmed using finite temperature density profiles of trapped fermionic @xmath577li atoms @xcite .
experimental observation reveals that the system has a partially polarized core surrounding by either fully paired or fully polarized wings at low temperatures that is in agreement with theoretical prediction of the zero temperature phase diagram within the lda @xcite .
the quantum phases of pairs , excess fermions as well as the mixture of pairs and excess fermions in the phase diagram fig .
[ fig : phase - mu - h - e ] are also consistent with the analysis of several other groups @xcite .
the finite temperature phase boundaries do not indicate a solid phase transition , detailed analysis can be seen in yin _
( 2011b ) . in a 3d
trapped fermi gas , the fully paired core is surrounded by a shell of excess fermions @xcite . ) at temperature @xmath1262nk .
experimental observation is in reasonable agreement with the theoretical prediction @xcite for the zero temperature phase diagram of the trapped gas .
this phase diagram experimentally verifies the coexistence of pairing and polarization at quantum criticality @xcite .
from liao _ et al . _
( 2010 ) . ]
the key method to map out the phase diagram in fig .
[ fig : phase - mu - h - e ] in the experiment @xcite is to measure the _ in situ _ densities of the two spin species .
the 1d spatial density profiles @xmath1263 can be expressed in terms of chemical potential @xmath1264 and effective external field @xmath1265 . here
@xmath1266 is the chemical potential at the trapping centre and @xmath1267 is the effective magnetic field of the homogeneous system .
they can be obtained from the relations ( [ mu - h ] ) or the equation of state ( [ eos ] ) . within the lda ( [ n_p ] ) ,
theoretical density profiles can be used to fit the experimental data obtained by inverse abel transformation of the radial profiles , as per the method section in liao _
( 2010 ) . by changing the polarization ,
the threshold values of different phases in the density profiles can be read off the phase boundaries of the phase diagram of fig .
[ fig : phase - mu - h - e ] . from the density profiles given in fig .
[ fig:1dfermidensityprofile ] , we see that at low polarization below the critical value , the system has a partially polarized core surrounded by fully paired edges . at the critical polarization , almost the whole region is partially polarized . at high polarization above the critical value , the system has a partially polarized core surrounded by fully polarized edges . .
the black circles represent the majority , the blue diamonds represent the minority , and the red squares show the difference .
( a ) at low @xmath1268 , the partially polarized core is enclosed by the fully paired edge .
( b ) for increasing @xmath1269 , the partially polarized core grows and the fully paired edge shrinks .
( c ) near @xmath1270 , where almost the entire cloud is partially polarized .
( d ) for the polarization exceeding the critical value @xmath1271 , the edge of the cloud becomes fully polarized . from liao _ et al . _
( 2010 ) . ]
systems with a small number of trapped atoms are also experimentally feasible @xcite .
most recently , the 1d fermionization of two distinguishable fermions has been experimentally studied by using two fermionic @xmath577li atoms @xcite .
the energy of the two - particle system in the state @xmath1272 is determined by tuning the trapping potential barrier through which the particles can tunnel out .
the fermionization of two distinguishable fermions was identified by measuring the tunnelling time constants for different values of the 1d interacting strength .
it is particularly interesting that for a magnetic field below the confined induced resonance two interacting fermions form a tonks - girardeau state whereas a super tonks - girardeau gas is created when the magnetic field is above the resonance value , see fig .
[ fig:1dfermionization ] . in this case , the two - particle super tonks - girardeau state is stable against three - body collisional losses since there is no third particle present .
in previous sections we have seen how results for exactly solved models of 1d fermi gases provide valuable insights into a wide range of many - body phenomena including fermi polarons , fulde - ferrel - larkin - ovchinnikov - like pairing , the few - body physics of trions , tomonaga - luttinger liquids , spin - charge separation , universal contact , quantum criticality and universal scaling .
this included discussion of the exotic many - body physics of 1d fermi - bose mixtures and 1d multi - component interacting fermions , in particular with @xmath735 , @xmath19 and @xmath9 symmetries .
we reviewed experimental progress on the realisation of 1d quantum atomic gases and the experimental confirmation of the phase diagram of the gaudin - yang model in a 1d harmonic trap . in the present section ,
we discuss some of the promising developments and provide an outlook for future research .
the key points we have identified are as follows .
\(a ) _ high spin symmetries .
_ very recent experimental exploration of highly symmetric mott insulators @xcite gives insight into the pomeranchuk ( 1950 ) type of cooling due to the fact that the spin degrees in the mott insulating state can hold more entropy than the fermi liquid does . in this mott insulating state , the entropy per site increases as the spin degrees of freedom increase .
it in turn leads to a temperature reduction through transfer of entropy from particle motion to spin degrees of freedom .
this exotic nature is different from large spins in solids where the quantum fluctuations of large spins are weak .
such higher symmetry opens up further study of magnetically ordered phases in ultracold fermionic atoms .
in particular , loading large spin ultracold fermionic atoms into a 1d quantum wire geometry provides exciting opportunities to test multi - component tll theory and competing superfluid ordering in 1d interacting fermions with higher symmetries .
e.g. , @xmath21 , @xmath19 and @xmath1273 symmetries in spin-3/2 fermions .
it is natural to expect that 1d interacting fermions with large spin resulting from higher mathematical symmetries will greatly expand our understanding of many - body physics .
however , comprehensive understanding of these models still poses theoretical and experimental challenges due to the more complicated grand canonical ensembles involved .
\(b ) _ p - wave bcs pairing and synthetic gauge fields . _
the traditional s - wave bcs pairing models have various applications to problems in condensed matter physics @xcite , nuclear physics @xcite and ultrasmall metallic grains @xcite _ etc_. more work along this line has been achieved in the study of a @xmath1274 pairing model @xcite .
in contrast to the s - wave pairing model , the @xmath1274 pairing model has an exotic zero temperature phase diagram in terms of density and attractive coupling strength @xcite .
it shows two superfluid phases confined strong - pairing and deconfined weak - pairing , separated by a third - order confinement - deconfinement quantum phase transition .
moreover , a type of electron pairing model with spin - orbit interactions shows that the pairing order parameter can always have @xmath1274-wave symmetry regardless of the strength of pairing interactions @xcite .
more work is required along this line .
in particular , recent developments in the study of ultracold atoms have opened up new opportunities to simulate synthetic external abelian and non - abelian gauge fields coupled to neutral atoms by controlling atom - light interactions @xcite and the ultracold - atom analog of the mesoscopic conductor @xcite .
an important application of this synthetic gauge field is the realisation of spin - orbit coupling in degenerate fermi gases @xcite .
the fano - feshbach resonances can be used to modify the strength of atomic interactions without changing the characteristic range of the potential @xcite .
raman laser beams coupled to zeeman states of ultracold fermionic atoms can create synthetic magnetic , electric fields and spin - orbit coupling , see a brief review @xcite .
these new developments inspire further study of spin - orbit coupling , p - wave and d - wave pairing by exact solutions of new mathematical models , with clear applications in physics .
\(c ) _ universal wilson ratio . _
the low - energy physics of interacting fermi systems exhibits universal phenomena , such as tll in 1d and fermi liquids in higher dimensions .
it is highly desirable to find an intrinsic connection between these low - energy theories .
the wilson ratio is the ratio of susceptibility to specific heat .
despite these two physical quantities having different low temperature behaviour , the wilson ratio is a constant at the fixed point of interacting fermionic systems in 3d @xcite . for noninteracting electrons , the wilson ratio is unity .
the value of the ratio indicates interaction effects and quantifies spin fluctuations .
in contrast to the phenomenological tll parameters , this ratio gives a measurable physical quantity which manifests universal tll physics the effective fixed point model of the tll universality class .
experimental measurement of the wilson ratio of the tll in a spin-1/2 heisenberg ladder has been recently reported @xcite .
it is naturally to believe that this ratio can be used to quantify the magnetic phases ( the fflo - like phases ) in 1d attractive fermi gases of ultracold atoms with different symmetries .
we remark that at the critical point it is a constant solely dependent on the tll parameter and onset charge velocities of different states .
this ratio is experimentally measurable with ultracold fermi atomic gases trapped in 1d .
\(d ) _ diffraction vs non - diffraction . _
non - diffraction in the many - particle scattering process is a unique feature of integrable systems @xcite . in principle ,
diffractive and non - diffractive scattering can be tuned via controlling inter- and intra - species scattering lengths .
experimental exploration of a violation of this characteristic would yield valuable insight into understanding the non - diffractive form of the bethe ansatz wave function , i.e. , how weak violations of non - diffractive scattering still give rise to the bethe ansatz wave function in the asymptotic region .
1d many - body systems with a finite range potential between atoms are likely to be an ideal simulator for identifying the consequences of diffractive _ vs _ non - diffractive scattering . in this regard , across a narrow resonance , the 1d effective potential is determined by not only the scattering length , but also the effective range which is introduced by a strong energy - dependent scattering amplitude of two colliding atoms @xcite . beyond the fermionization of two distinguishable fermions @xcite
, it is an immediate goal to experimentally probe three distinguishable fermions in a 1d harmonic trap .
a possible test of the fundamental nature of diffraction _ vs _ non - diffraction for systems with three particles may ultimately lead to experimental tests for yang - baxter integrability in quantum many - body systems .
we thank s. chen , a. del campo , a. foerster , f. goehmann , t .- l .
ho , z .- q .
yang , y .- p .
wang and s .- z .
zhang for valuable comments on the manuscript .
we also thank j .- s .
caux for sharing his english translation of gaudin s book .
this work has been supported by the national basic research program of china under grant no .
2012cb821300 and no .
2012cb922101 as well as the australia research council through grants dp1093353 , dp1096713 and dp130102839 . c. l. has also been partially supported by the national natural science foundation of china under grant no .
11075223 and the program for new century excellent talents in university of the ministry of education of china under grant no .
ncet-10 - 0850 .
g acknowledges the institute for advanced study , tsinghua university for kind hospitality when part of this paper was prepared . | this article reviews theoretical and experimental developments for one - dimensional fermi gases .
specifically , the experimentally realized two - component delta - function interacting fermi gas the gaudin - yang model and its generalisations to multi - component fermi systems with larger spin symmetries .
the exact results obtained for bethe ansatz integrable models of this kind enable the study of the nature and microscopic origin of a wide range of quantum many - body phenomena driven by spin population imbalance , dynamical interactions and magnetic fields .
this physics includes bardeen - cooper - schrieffer - like pairing , tomonaga - luttinger liquids , spin - charge separation , fulde - ferrel - larkin - ovchinnikov - like pair correlations , quantum criticality and scaling , polarons and the few - body physics of the trimer state ( trions ) .
the fascinating interplay between exactly solved models and experimental developments in one dimension promises to yield further insight into the exciting and fundamental physics of interacting fermi systems . |
the phenomenon of superradiance has a long history dating back more than 60 years to the seminal paper by dicke @xcite where the effect was predicted theoretically . over the past few decades
the superradiance has been the subject of intense theoretical and experimental studies in a large variety of systems .
these include molecular aggregates @xcite , cold atoms @xcite and bose - einstein condensates @xcite , atomic nuclei @xcite , magnetic nanoclusters @xcite , heterostructures @xcite and many others . the key process underlying the mechanism of superradiance is phase synchronization of initially independent atoms caused by the coupling with a common environment represented by the electromagnetic field . in order for such process to occur
the phase decoherence time of atoms should be longer than the photon travel time in the sample @xcite . for samples
which size is smaller than the wavelength of radiation , this condition requires the density of atoms to be sufficiently high . the system - environment
( atom - field ) coupling also manifests itself as an additional indirect interaction ( a sort of the transverse dipole - dipole interaction ) which may suppress superradiant transitions depending on the spatial distribution of atoms or the sample geometry @xcite . from the other hand ,
when the density of atoms ( or , more generally , emitters ) is high the direct interparticle interaction starts to play an increasingly important part in determining cooperative behavior of the particles .
in particular , this interaction strongly affects the properties of low - dimensional systems .
the mott - insulator quantum phase transition in optically trapped atomic systems @xcite and in solid structures @xcite , generation of many - particle entangled states or many - particle coherent dynamics , as it is in the case of effectively interacting atoms inside a high quality dissipative cavity @xcite , bose - einstein condensate @xcite and in molecular clusters with strong magnetic @xcite or coulomb @xcite correlations are examples .
the direct interparticle interaction introduces additional correlations between emitters .
these correlations considerably influence the cooperative optical properties of atoms .
firstly , the interaction directly affects the superradiance leading to a number of peculiarities such as changing the order of superradiant phase transitions @xcite .
recently , the possibility of superradiant relaxation in strongly correlated systems was studied theoretically @xcite .
the experimental results for magnetic molecules of mn@xmath4-ac type were also reported in ref .
earlier , systems of ferroelectric type with strong interparticle interaction were regarded as promising candidates for an active medium of the heat pumping laser @xcite .
secondly , the direct interparticle interaction can play the role of a phase synchronizing factor that may lead to the cooperative behavior which , though shares many common properties with the superradiance effect , essentially differs from superradiance in nature . the classical example furnishes the spectrum of p luminescence band in cds and zno where the emission intensity is proportional to the second power of the free exciton number ( pumping intensity ) . in this case
the effect is caused by exciton - exciton scattering @xcite .
recently , such effects were observed in the microcrystalline phase of cspbcl@xmath5 thin films @xcite . for interacting atoms , the interaction can drastically change the regime of atom - photon dynamics by inducing ( otherwise , excluded ) multiphoton transitions @xcite .
it was shown in ref .
@xcite that interatomic interaction can give rise to non - zero multiphoton emission observed with single - molecule spectroscopy technique as a two - photon cooperative effect for strongly dipole - dipole coupled molecules . theoretically , this phenomenon was predicted as a large two - atom two - photon resonant effect for two atoms inside a high - quality cavity @xcite .
so , different regimes of radiative decay in correlated atomic systems are mainly governed by the interatomic interaction . by controlling the interaction radiation properties of such systems
can be widely varied ranging from superradiant transitions to the generation of the fock state of light .
in particular , such control is feasible for the atoms in optical lattices ( see , e.g. , @xcite ) . in this work cooperative radiation of interacting atoms coupled to electromagnetic bath
will be of our primary interest .
we are aimed to study different relaxation regimes determined by the intensity of interatomic coupling .
the paper is organized as follows . in sec .
[ sec : model ] we formulate the model of @xmath6 two - level atoms with ising - type interaction and qualitatively discuss various regimes of relaxation by considering realignment of the atomic energy spectrum at different values of the interatomic coupling constant @xmath1 .
there are two limiting cases of weak and strong interaction with @xmath0 and @xmath3 , respectively ( @xmath2 is the atomic frequency ) .
we find that , for weakly interacting atoms , ising interaction would affect dynamical behavior of the system leading to the transition to collective pulse relaxation and enhancement of superradiance . for strong interaction , the regime of multiphoton relaxation is predicted to occur .
it is also shown that dependence of the radiation intensity peak on the number of particles can be anomalous at long - range interatomic interaction .
derivation of the master equation for weakly interacting atoms is presented in sec .
[ sec : master - equation ] .
we show that , for certain atomic configurations , dephasing effects of induced dipole - dipole interaction can be suppressed and dynamics of the atomic system can be described by the simplified master equation . in sec .
[ sec : coop - rad - weakly ] , the effects for weakly interacting atoms briefly discussed in sec .
[ sec : model ] are investigated in detail . by applying the mean field approximation
we obtain the results that agree very well with those calculated by solving the equations for atomic variables numerically .
the regime of multiphoton relaxation that takes place at strong interatomic interaction due to inhibition of one - particle one - photon transitions is described in sec .
[ sec : multiphoton ] . finally , in sec .
[ sec : disc - concl ] , we draw together the results , discuss solitonic mechanism of relaxation and make some concluding remarks .
details on some technical results are relegated to appendixes [ ap - born - markov]- [ ap - fast - var ] .
in order to illustrate a possibility of different relaxational regimes , caused by direct interatomic interaction , we consider a simplest model of a chain of two - level atoms with nearest neighbor ising - type interaction . typically , the models of this type are used to describe molecular systems with coulomb and magnetic interactions such as ferroelectric @xcite and magnetic @xcite clusters , interacting electrons in the tightly - binding approximation @xcite , interacting atoms in optical lattice @xcite , where the extended hubbard model is usually introduced .
the interaction takes into account repulsion of the neighboring atoms so that the system may reveal the correlated many - particle behavior
. when the energies of the neighbors are different tunneling transitions between neighboring sites are inelastic .
in addition to the ising - type direct interaction , the atoms interact with a common electromagnetic field ( the atom - field interaction ) .
these two interactions are combined to maintain correlations in the atomic subsystem thus crucially affecting the regimes of cooperative optical transitions .
the hamiltonian of the model atomic chain with the short - range ising - type interaction can be written in the following form @xmath7 where @xmath8 is the hamiltonian of atomic subsystem , @xmath9 is the hamiltonian of electromagnetic field and @xmath10\ ] ] is the hamiltonian of the atom - field interaction and h.c .
stands for hermitian conjugation ; @xmath2 is the frequency of atomic transition , @xmath11 is the coupling constant of the nearest neighbor interaction , @xmath6 is the number of atoms , @xmath12 is the @xmath13-component of the pseudo - spin operator determining the population of the @xmath14th atom , @xmath15 ( @xmath16 ) is the raising ( lowering ) operator of the pseudo - spin , @xmath17 ( @xmath18 ) is the photon annihilation ( creation ) operator characterized by the wavenumber vector @xmath19 , the frequency @xmath20 and the polarization vector @xmath21 , @xmath22 is the coupling constant of the dipole interaction between the @xmath14th atom and transverse electromagnetic field , @xmath23 is the effective dipole moment of @xmath14th two - level atom with the vector of spatial coordinates @xmath24 , @xmath25 is the vacuum dielectric constant and @xmath26 is the volume of field quantization . the hamiltonian is distinguished from standard models of two - level atom interacting with the electromagnetic field by the presence of the interatomic interaction in @xmath27 [ see eq . ] . in what follows
the dynamics of the collective relaxation of initially inverted atomic subsystem will be of our primary concern .
more specifically , we shall study various dynamical scenarios that occur at different values of the coupling constant @xmath1 .
in the remaining part of this section we first discuss energy spectrum of the atomic subsystem and qualitatively describe these dynamical regimes .
the transition energy of an atom depends on the states of its neighbors as a result of interatomic interaction .
this effect can be described as renormalization of the transition frequency of the @xmath14th atom @xmath28 which can also be readily seen from the mean field expression for the hamiltonian @xmath27 : @xmath29 s^z_i=\sum_i\hbar\tilde\omega_i s^z_i.\ ] ] sufficiently weak interaction , @xmath30 , results in inhomogeneous broadening of the radiation spectrum .
when the coupling constant of interatomic interaction @xmath1 increases and @xmath0 , the spectrum of the atomic subsystem undergoes more pronounced rearrangement . as is shown in fig .
[ spectra ] , the spectrum is no longer equidistant , but its structure is identical to that for non - interacting atoms with @xmath31 . in particular
, the latter implies that the state with all the atoms excited @xmath32 ( spin up represents the excited state of the atom ) gives the level of highest energy and the level of lowest energy corresponds to the case where atoms are all in the ground states @xmath33 .
similar to the case of non - interacting atoms , the dynamics of radiative decay is dominated by one - particle one - photon transitions and can be described using the born - markov approximation
. a cooperative relaxation of the system from an excited state can be realized by different schemes of transitions .
it is a possibility to have a generation in the two mode regime with the frequencies @xmath34 ( see fig .
[ spectra]a ) arising due to renormalization of the atomic frequency @xmath35 that governs the intensity of radiation .
it is well known @xcite that the intensity of atom radiation @xmath36 is proportional to the fourth power of the transition frequency @xmath37 , where @xmath38 is the transverse dipole moment of the atomic transition . for @xmath6
weakly interacting atoms , the intensity @xmath36 is defined by the renormalized frequency and can be represented by a sum of the coherent and the incoherent parts , @xmath39 , where @xmath40 and @xmath41 .
dependence of the renormalized frequency on the states of neighboring atoms introduces additional correlations into the incoherent part of radiation .
such correlations may result in a cooperative radiation of fundamentally different origin than superradiance described by the coherent part @xmath42 .
the effect of superradiance , for its part , can be enhanced by the ising - type interaction which may act as a phase synchronizing factor .
it is also worth noting that , for a long - range interparticle interaction , the renormalized frequency @xmath35 depends on the number of atoms , @xmath43 , leading to anomaly in the @xmath6-dependence of the intensity peak of collective radiation .
figure [ spectra]a shows that , when the atomic subsystem is not completely inverted , an excited state may relax through a series of transitions which frequencies are identically equal to the resonant frequency @xmath2 .
this regime of relaxation resembles the solitonic mechanism where the process of emission is connected with inelastic motion of defects ( bloch walls ) induced by a short - range interaction .
we shall discuss this point at greater length in sec .
[ sec : disc - concl ] . in the case of strong interaction with @xmath3 , arrangement of the energy level
significantly differs from that of non - interacting atoms . now
the state @xmath44 does not give the level of highest energy and @xmath45 , where @xmath46 is the energy of the state with @xmath47 excited atoms .
such realignment indicates that many - atom multiphoton transitions will dominate the process of relaxation .
figure [ spectra]b illustrates that , in the limiting case with @xmath48 , the states @xmath44 and @xmath49 give two lowest lying energy levels .
they are separated from the other excited states by the gap of the width @xmath50 .
relaxation of fully inverted state ( transition from @xmath44 to @xmath49 ) can be achieved through a @xmath6-photon process . at this stage
it should be stressed that an accurate description of the interaction between the system and the field in the presence of strong interatomic correlations requires careful consideration of interplay between many - atom multiphoton dipole and multipole transitions . in a sense
, our system can be regarded as an intermediate case between a collection of two - level atoms independently interacting with the field and a single many - level system .
a good example provides the spin - phonon interaction in magnetic molecules mn@xmath4o@xmath4 of the multipole form @xmath51 which symmetry is consistent with the anisotropy energy @xcite .
similarly , the spin superradiance in atomic nuclei can be dominated by multipole transitions @xcite . for the multipole superradiance
@xcite , the exponent of the @xmath6-dependence of the radiation intensity peak ( @xmath6 is the number of emitters ) can be greater than two , @xmath52 with @xmath53 .
this is the case for the cooperative spin - phonon relaxation in mn@xmath4o@xmath4 where a rough estimate gives @xmath54 .
the standard approach to the dissipative dynamics of atoms relies on the lindblad equation for the reduced density matrix of the atomic subsystem derived by eliminating boson variables from the master equation for the density operator of atom - field system .
this master equation is obtained using the born - markov approximation @xcite that neglects multiphoton transitions and thus is applicable only for weakly interacting atoms at @xmath0 . in the subsequent section
we discuss the derivation procedure for the lindblad equation for the density matrix of weakly interacting atoms .
in this section we derive and discuss the master equation for the reduced atomic density operator @xmath55 for the case of weak interaction at @xmath0 .
we show that the direct interaction renormalizes both the induced dipole - dipole interaction and the damping operator in the lindblad equation . in addition , we find that , under certain conditions , destructive effect of the dipole - dipole interaction on supperradiance is suppressed . as it was pointed out in sec .
[ sec : model ] , for the @xmath14th atom , the energy of transition depends on the state of the neighboring atoms .
alignment of the energy spectrum of the atomic subsystem and , as a result , the character of atom - field interaction are determined by the value of the ratio @xmath59 .
when @xmath3 and @xmath60 , ordering of the energy levels changes and certain transitions appear to be excluded . in eq .
the frequency of the dipole transition is effectively described by the operator @xmath61 entering the hamiltonian of interaction @xmath62 .
as opposed to the case with @xmath63 , the eigenvalues of the operator @xmath64 can be negative at @xmath60 depending on @xmath65 and @xmath14 . in this case , for example
, the energy @xmath66 of the `` monodomain '' state @xmath67 is below the energy @xmath68 of the state where only one atom is in the ground state @xmath69 .
the result is that relaxation of the completely inverted state @xmath67 can not occur as a one - photon process .
it should be noted that , for @xmath60 , the change of sign in the eigenvalues of the operator @xmath64 introduces new resonant terms of the form @xmath70 .
however , as it was discussed in sec .
[ sec : strong - interaction ] , this case is additionally complicated by multipole transitions that need to be taken into consideration .
when @xmath0 and @xmath63 , the eigenvalues of the operator @xmath64 are all positive .
so the ising - type interaction just renormalizes the transition frequencies inducing additional correlations in the atomic subsystem .
dynamics of the atomic relaxation is now governed by one - photon processes and , in the remaining part of this section , we restrict ourselves to this case of weakly interacting atoms with @xmath63 .
we can now use the born approximation to derive the equation for the reduced density matrix of the atomic subsystem @xcite , @xmath71 . for simplicity , we consider the zero - temperature case where only vacuum fluctuations of the electromagnetic field are taken into account .
so , the density operator @xmath56 can be written in the following form @xmath72 we additionally assume that the wavelength of radiation is much longer than the size of the system and use the markov approximation @xcite . omitting the technical details that can be found in appendix [ ap - born - markov ] ,
the resulting equation is given by @xmath73 + \nonumber\\ & + & [ r_j(t),\tilde{\rho}(t)s_i^-(t)\frac{i\mathcal{p}}{\omega+{\omega_0 } \hat{\gamma_i}}]+h.c.\biggr\},\end{aligned}\ ] ] where @xmath74 and @xmath75\frac{\sin{kr_{ij}}}{kr_{ij}}+ [ 1 - 3(\bar{\mathbf{d}}\cdot\bar{\mathbf{r}}_{ij})^2]\left(\frac{\cos{kr_{ij}}}{(kr_{ij})^2}-\frac{\sin{kr_{ij}}}{(kr_{ij})^3}\right)\right\},\ ] ] @xmath76 , @xmath77 is the separation distance between @xmath14th and @xmath47the atoms , we assume @xmath78 , @xmath79 , @xmath80 and @xmath81 for brevity , in eq .
we used operator functions defined on eigenvalues of @xmath82 , so that if @xmath83 then , e.g. , @xmath84 .
it should be stressed once again that eq .
is valid only for the case of weakly interacting atoms , when @xmath0 and the born - markov approximation is applicable .
principal values of the integrals that enter the right hand side of the master equation define the dipole - dipole interatomic interaction ( see appendix [ ap - born - markov ] ) and the bethe - part of the lamb shift , which depend on the direct interparticle interaction .
equation can be further simplified under the condition that the characteristic wavelength of radiation @xmath85 is much longer than the size of the sample .
it is well known @xcite that in this case the dipole - dipole interaction is of quasi - static character and is independent of the frequency of atomic transition . since the direct interaction in eq
just renormalizes the atomic frequencies @xmath86 , the asymptotic form of the dipole - dipole interaction in the limit @xmath87 must be identical to that for the system of non - interacting atoms with @xmath31 . at the same time
the relaxation terms proportional to @xmath88 essentially depend on the interatomic interaction . the first term on the right hand side of eq
. describes renormalization of the damping term by the interatomic interaction . after neglecting the rapidly oscillating terms ( the rotating wave approximation ) in eq . and
going back to the schrdinger representation , we derive the master equation for the case of weakly interacting atoms , @xmath63 , in the final form : @xmath94+i\sum_{i\neq{j}}\omega_{ij}[s_i^+ s_j^-,\rho(t)]+ \nonumber\\ & + & \gamma_0\sum_{i , j}\left([s_j^-,\rho(t)\hat{\gamma}_i^3s_i^+]+ [ s_i^-\hat{\gamma}_i^3\rho(t),s_j^+]\right ) . \label{me1}\end{aligned}\ ] ] eq
. describes dynamics of the ising - like interacting atoms coupled to electromagnetic bath .
the key feature of this equation is that only the damping terms are renormalized by the direct interatomic interaction at @xmath95 .
it is known ( see , e.g. , reviews @xcite ) that the dipole - dipole interaction may destroy the cooperative radiation due to the frequency chirping effect .
this effect is caused by the dynamical shift of energy levels that results in a rapid loss of phase synchronization with the characteristic time of phase decoherence much shorter than the rate of relaxation @xcite .
influence of the dipole - dipole interaction on the dynamics of the atomic subsystem is strongly affected by the spatial configuration of atoms and the angular distribution of dipole moments .
it turns out that , for certain configurations , dephasing induced by the dipole - dipole interaction is inhibited and the cooperative radiation takes place @xcite . in particular
, the latter is the case when atoms are all in identical environment and the system is invariant under permutations of atoms , @xmath96 .
these conditions are met for a two atom system or an atomic ring @xcite . for disordered atomic configuration ( atomic cloud ) ,
the superradiance effect depends on the sample shape .
it can occur for pencil - shape patterns as a result of suppression of the destructive effect of the dipole - dipole interaction @xcite . in our case ,
when the interatomic spacings is much smaller than the wavelength ( @xmath95 ) , the form of the dipole - dipole interaction is the same for both interacting and non - interacting atoms .
symmetry of the renormalized damping terms is also identical to the symmetry of the system of non - interacting atoms with @xmath31 .
so , atomic configurations suppressing the dephasing effect of the dipole - dipole interaction are the same for both cases : @xmath31 and @xmath97 .
for the simplest case of two atomic configuration , this point can be illustrated explicitly .
similar to the case of two non - interacting atoms considered in ref .
@xcite , the master equation for two atoms with @xmath97 is exactly solvable .
the exact solution described in appendix [ ap-2atbm ] shows that relaxation is independent of the dipole - dipole interaction .
the analytical results presented in appendix [ ap-2atbm ] can also be used to demonstrate the effect of additional correlations induced by the interaction . for this purpose
, we consider the expression for the relaxation rate of initially excited atoms @xmath98,\ ] ] where @xmath99 and @xmath100 . in fig .
[ 2atbm ] the relaxation rate is plotted as a function of time at various values of @xmath65 .
it is seen that the interaction noticeably influences the character of relaxation even at small values of @xmath65 .
by contrast to the case of non - interacting atoms with @xmath101 , the system features collective decay with a pronounced emitted pulse arising as a result of additional correlations induced by ising - type interaction . now it is our primary task to examine how the cooperative relaxation is influenced by the interatomic interaction .
so , in what follows we shall restrict our study to the atomic configurations where the frequency chirping effect can be disregarded .
for such configurations , the master equation can be considerably simplified by neglecting the terms describing the dipole - dipole interaction .
from the master equation without the dipole - dipole interaction , @xmath102 , it is straightforward to deduce a system of equations for averages of the atomic variables @xmath103 and @xmath104 .
the result is given by [ bloch ] @xmath105 where @xmath106 is one - half the spontaneous decay rate for an isolated atom .
the relaxation rate of the atomic subsystem @xmath107 is of primary importance in our subsequent analysis .
the parameter @xmath108 is defined in the right hand side of eq . .
by contrast to the case of non - interacting atoms , there is no an intimate connection between the relaxation rate @xmath108 and the total intensity of radiation @xmath109 which is proportional to @xmath110 .
the reason is that the interatomic interaction renormalizes the frequency of the dipole transition @xmath111 , whereas @xmath112 and @xmath113 . the right hand side of eq .
can be conveniently rewritten as a sum of the coherent ( @xmath114 ) and incoherent ( @xmath115 ) parts @xmath116 where we used the identities @xmath117 and @xmath118=0 $ ] . in order to decouple correlations in eq .
, we use the bloch representation for the wave functions of two - level atoms @xmath119 @xcite @xmath120 and obtain the closed system of equations for the averages of atomic variables @xmath121 of the following form : @xmath122 where @xmath123 , @xmath124 and @xmath125 there is a rapidly oscillating term @xmath126 in the system of equations .
in contrast to the case of non - interacting atoms with @xmath127 , this term can not be removed by the phase shift @xmath128 . in appendix [ ap - fast - var ]
we apply the method of multitime scales @xcite to eliminate the rapidly oscillating terms and deduce the following equations for slowly varying atomic amplitudes @xmath129 where @xmath130 denotes averaging over the time @xmath131 of fast motion , @xmath132 where @xmath133 and functions @xmath134 describe correlations due to the direct interatomic interaction .
it is known @xcite that relaxation of the atomic subsystem essentially depends on the number of atoms @xmath6 . for non - interacting atoms ( @xmath31 and @xmath127 ) , the first term on the right hand side of eq
. describes processes of spontaneous incoherent decay .
transition to the regime of superradiant relaxation , where correlations are caused by interaction with the electromagnetic field in the vacuum state , occurs when the term describing the coherent part in eq .
( a sum with @xmath135 ) start to play the dominating role .
such transition can be conveniently described using the order parameter of the form ( see , e.g. @xcite ) : @xmath136 and takes place if the number of atoms @xmath6 exceeds its critical value @xmath137
@xcite .
the presence of the factor @xmath138 in eq .
indicates that the direct interaction introduces additional correlations into both the coherent and the incoherent parts of radiation .
it is the effect of such correlations on the incoherent part is responsible for the regime of pulse radiation @xcite discussed in sec .
[ sec : introduction ] . in fig .
[ cor ] we present the results for @xmath6 dependence of the order parameter @xmath139 computed by solving the system numerically at different values of the interaction parameter @xmath59 .
the curves show that the critical number of atoms and the supperradiance transition are almost insensible to the interatomic interaction .
so , similar to the case of non - interacting atoms , we have two qualitatively different regimes of relaxation : the incoherent regime at @xmath140 and the regime of superradiance at @xmath141 .
let us consider decay of initially inverted system of interacting atoms at subcritical number of the atoms @xmath140 in more details .
dependence of the relaxation rate on time calculated by solving eq . with the initial conditions @xmath142
is shown in fig .
[ n10 ] . referring to fig .
[ n10 ] , it is seen that , in the absence of interaction ( @xmath31 and @xmath127 ) , relaxation occurs as an incoherent spontaneous decay of excited atoms .
this regime is characterized by a monotonic exponential decrease of the relaxation rate . when the interaction parameter @xmath65 increases , spontaneous processes are suppressed and the time dependence of the relaxation rate reveals a non - monotonic behavior with a pronounced peak ( see fig .
[ n10 ] ) .
this is the regime of collective pulse relaxation induced by the direct interatomic interaction .
collective pulse relaxation can be easily described using the mean field approximation for decoupling of correlations . to this end
we assume that , for small number of atoms , the relaxation rate is predominantly determined by the first term ( incoherent part ) describing incoherent mechanism of decay .
so , the equation takes the simplified form : @xmath143 we can now substitute the ansatz @xmath144 assuming that @xmath145 , into eq . and use the relation to derive the equation governing dynamics of atomic subsystem @xmath146 where @xmath147 it can be shown that the mean field solution @xmath148 is a step - like function of time and is stable at @xmath149 .
the curves calculated in the mean field approximation for the relaxation rate @xmath150 are shown in fig .
[ n10 ] as dashed lines .
it can be seen that the results are in excellent agreement with the data of numerical analysis .
it should be noted that the peak intensity is proportional to the number of atoms @xmath6 which is a consequence of a short - range character of interatomic interaction . by contrast , for the regime of superradiance , the intensity is typically proportional to @xmath151 . for long - range interaction ,
the @xmath6 dependence of the peak intensity may change .
we demonstrate this effect for the interaction of the form @xmath152 .
in this case the expression for the operator @xmath64 has to be replaced by the relation @xmath153 that gives an additional @xmath6 dependent factor .
the modified expression for the relaxation rate is given by @xmath154\langle\tilde{s}^z\rangle \nonumber\\ & + & 3\beta^2(n-1)(n-2)\langle\tilde{s}^z\rangle^2-\beta^3(n-1)(n-2)(n-3)\langle\tilde{s}^z\rangle^3\bigl\}.\end{aligned}\ ] ] since the relaxation rate @xmath108 reaches the peak at negligibly small @xmath155 , @xmath156 , from eq .
its intensity can be estimated to be an increasing function of @xmath65 proportional to @xmath151 : @xmath157\propto n^2.\ ] ] so , we have shown that the direct ising - type interaction has a synchronizing effect on the system behavior .
the result is that the regime of incoherent spontaneous decay changes to the regime of collective pulse relaxation characterized by an increase in the radiation time . in the subsequent section
we discuss similar effects for the regime of superradiance . as it was previously discussed in sec .
[ subsec : order - parameter ] , the effect of superradiance dominates the regime of relaxation of the atomic subsystem at supercritical values of the number of atoms , @xmath141 .
figure [ n100 ] presents the time dependence of the relaxation rate for @xmath158 at various values of the interaction parameter @xmath65 . at @xmath101 , the peak with the intensity proportional to @xmath151 corresponds to the superradiance effect in the system of non - interacting atoms . in this system
interatomic correlations are solely due to interaction between atoms and electromagnetic field .
synchronizing effect of the ising - type interaction characterized by the parameter @xmath65 manifests itself in enhancement of superradiance and an increase in the peak intensity .
the effect of enhancement can be analyzed in the mean field approximation by applying the procedure described in the previous section to the coherent part ( the second term ) of the relation .
so , the equation for the average of atomic population @xmath159 is now given by @xmath160 where @xmath161 is given in eq . and
the identity ( the conservation law for the pseudo - spin ) @xmath162 is used .
the peak of the relaxation rate @xmath108 is at @xmath156 and eq .
combined with the relation provide an estimate for the gain factor @xmath163 . as for the case of @xmath140 , the interatomic interaction delays the time of radiation . for long - range interaction ,
following the lines of reasoning presented in sec .
[ supr - spont ] , we derive dependence of the relaxation rate on the number of atoms @xmath6 @xmath164\propto n^3.\ ] ] this result differs from the @xmath6 dependence given at the end of the previous section for the regime of collective pulse relaxation .
it is known that interaction between atoms at interatomic spacings smaller than the wavelength may bring about multiphoton processes in atomic system @xcite . at weak interaction with @xmath0 , probability of multiphoton transitions
is much smaller than that of single - photon transitions .
so , the first order of the perturbative expansion over the atom - field interaction can be used to describe dynamical behavior of the system . at strong interaction with @xmath3 ,
single - photon processes are non - resonant and multi - photon transitions start to play an increasingly important role .
this is the so - called regime of multiphoton relaxation . in order to illustrate how this regime may occur we qualitatively consider a model system of two two - level atoms with the hamiltonian @xmath165 .
its energy spectrum is schematically represented in fig .
[ spectra2 ] .
if the hamiltonian of atom - field interaction has the form of a sum @xmath166 , where @xmath167 ( @xmath168 ) describes photon emission ( absorption ) , then the second order perturbative expression for the probability of two - photon transition @xmath169 is given by latexmath:[\[\label{two - phot - tr } w^{(2)}=\frac{2\pi}{\hbar^2}\left|\frac{\langle\downarrow\downarrow|h_+ { \hbar\omega_0(1-\beta/2)-\hbar\omega}+\frac{\langle\downarrow\downarrow|h_+ for strong interaction with @xmath3 , the energy level of one - particle excited states , @xmath171 and @xmath172 , is higher than the energy of the state @xmath173 ( see fig . [ spectra2]b ) , so that one - photon relaxation of the two - particle excited state is excluded by the energy conservation law . as a consequence ,
relaxation occurs as a two - photon process .
probability of multiphoton transitions for @xmath6 excited atoms in free space is small because its magnitude is determined by @xmath6-th order term of perturbative expansion over the atom - field coupling constant @xmath174 .
so , we arrive at the conclusion that the effect of interatomic interaction can be the formation of long living excited states in the system .
this conclusion , however , is not strictly valid if the density of electromagnetic modes has a singularity near the energy of atomic transition , @xmath175 .
an example is a high quality resonant cavity . in this case ,
dynamical behavior of the atom - field system is characterized by rabi oscillations involving many - particle and multiphoton states @xcite .
we demonstrate this effect for the two atomic system embedded into a high - q single - mode resonant cavity .
the hamiltonian of the model is @xmath176,\ ] ] where @xmath177 is the photon annihilation ( creation ) operator for the cavity mode , @xmath178 is the coupling constant of the interaction between atoms and cavity mode .
the model is exactly solvable and the wave function for the system initially prepared in the state @xmath179 , where @xmath180 is the number of photons , can be written in the explicit form : @xmath181|\uparrow\uparrow , n\rangle+ \nonumber\\ & + & \sqrt{(n+1)(n+2)}\left[\frac{g^2}{d}\left(\frac{e^{idt}}{d - j'}+\frac{e^{-idt}}{d+j'}\right)- \frac{e^{ij't}}{2n+3}\right]|\downarrow\downarrow , n+2\rangle- \nonumber\\ & -&i\frac{g\sqrt{n+1}}{d}\sin dt\left(|\downarrow\uparrow , n+1\rangle+|\uparrow\downarrow , n+1\rangle\right)\biggr\},\end{aligned}\ ] ] where @xmath182 in the absence of the interatomic interaction ( @xmath31 ) , the amplitudes of one- and two - photon processes are of the same order of magnitude .
but , at strong interaction with @xmath183 , the two - particle two - photon amplitudes dominate and the wave function can be approximated as follows @xmath184|\uparrow\uparrow , n\rangle+ \nonumber\\ & + & 2i\frac{\sqrt{(n+1)(n+2)}}{2n+3}\sin\delta t|\downarrow\downarrow , n+2\rangle\biggr\},\end{aligned}\ ] ] where @xmath185 and for @xmath186 @xmath187 this means the build - up of two - atomic two - photon rabi oscillations and can be used to generate a schrdinger cat - like entangled atom - field state . thus , interparticle interaction may give rise to the regime of many - particle multiphoton dynamics . when interaction is strong as compared with the energy of atomic transition , inhibition of one - photon processes in the system
is accompanied by transition from the regime of cooperative pulse radiation ( superradiance ) to the generation of fock state of light . in a cavity ,
multiphoton dynamical effects come into play under the condition @xmath188 , so that the level of intermediate energy is essentially shifted away from one - photon resonance ( see fig .
[ spectra2 ] ) .
we studied effects of direct interatomic interaction on collective processes in atom - photon dynamics using , as an example , a simple model of two - level atoms with ising interaction of ferromagnetic type .
we have found that this interaction influences radiation processes of atomic ensemble acting as an additional synchronizing factor . for weakly interacting atoms at @xmath0 , we have shown that interatomic interaction results in inhibition of incoherent spontaneous decay of atoms and dynamical behavior of the system is governed by the regime of collective pulse relaxation .
this regime , though it bears a resemblance to superradiance , has nothing to do with the effect of phase synchronization induced by fluctuations of the electromagnetic field . for a example , in solid state structures , collective pulse relaxation is caused by inelastic exciton - exciton scattering and is characterized by quadratic dependence of the radiation peak on the number of particles @xcite .
we have also found that interaction induced synchronization enhances superradiance and can be responsible for anomalous dependence of the radiation peak on the number of particles . in the presence of interparticle interaction collective pulse radiation and enhanced superradiance
are both characterized by an increase in the delay time of emission . at the end of sec .
[ sec : weak - interaction ] , we have pointed out that when the excited state is not fully inverted its relaxation can be determined by transitions with frequencies equal to the atomic resonant frequency @xmath2 .
if we consider non - excited atoms of the initial state as defects of the atomic chain , such relaxation scheme can be referred to as the solitonic mechanism of relaxation .
the solitonic mechanism implies that in the course of relaxation atoms undergo transition to the ground state successively one after another .
so , such behavior can be interpreted as a defect motion . more generally ,
since the system of equations governing dynamics of atoms at @xmath0 is similar in structure to the volterra system and the toda lattice @xcite , it might be expected that the system posses soliton - like solutions .
figure [ soliton ] presents the numerical results for the excited state with one initially non - excited atom .
the number of atoms is small , @xmath140 , and the relaxation rate is computed from the mean field equation . referring to fig .
[ soliton ] , it is seen that relaxation of the atomic ring can be described as motion of the defect ( bloch wall ) . when the interaction parameter @xmath65 approaches the limit of strong interaction , the relaxation rate @xmath108 assumes the kink - like form and is determined by defect velocity .
the peak is caused by collective pulse relaxation indicating that the retardation time of emission is too short for the defect to travel through the ring .
however , it should be noted that we have neglected the dipole - dipole interaction by considering permutationally invariant atomic configurations , the approximation that is not valid in the presence of defect in the system .
this interaction may have a destructive effect on the solitonic mechanism of relaxation .
by contrast to the case of weakly interacting atoms where dynamics of the atomic subsystem is governed by one - particle one - photon transitions , at strong interatomic interaction with @xmath48 , these transitions appear to be excluded . in this case
multiphoton transitions will determine relaxation of the excited atomic subsystem .
this is what we called the regime of multiphoton relaxation .
the regime is characterized by transition from generation of superradiant pulse to generation of fock quantum state of light .
we have used a simple model of two atoms in a high - q single mode cavity to show that such transition is accompanied by rabi oscillations involving many - atom multiphoton states . in other words , it means generating many - particle entangled atom - field state @xcite .
interestingly , transition to multiphoton dynamics is analogous to the mott - insulator quantum phase transition in optically trapped atomic systems where fock state characterizing the number of localized atoms is formed @xcite . in our case ,
the energy of atomic transition @xmath175 and the interatomic coupling @xmath1 play the role of kinetic and potential energy , respectively .
so , the transition to generation of the fock state of light takes place when the potential energy @xmath1 becomes greater than @xmath175 . in lattice atomic systems with inelastic tunneling transitions between neighboring wells ,
the mott or peierls transitions and the transition to multiphoton relaxation can be related to each other .
we illustrate such a possibility by the simple example of an one - dimensional periodic chain of potential wells , fig .
[ lattice ] . in the half - filling case , when there are half as many atoms as there are wells and the energy of repulsion between neighboring particles
@xmath1 is much larger then the hopping energy of tunneling between neighboring wells @xmath189 , the ground state for identical wells should be degenerate in energy , which corresponds to the atoms localized in either odd or even wells , see fig .
. such ordering of particles is similar to the wigner crystal or the charge density waves observed in low - dimensional conductors @xcite .
possibility of the density waves mott - insulator phase for atoms embedded into an optical lattice was also discussed in ref .
these ordered states are separated from the excited ones by the gap of width @xmath190 .
local tunneling transitions may lift this degeneracy of the ground state . in this case , at @xmath191 , the ground state can be described , at least for a short chain , as a many - particle entangled state of the schrdinger cat type , i.e. , @xmath192 ( the so - called ghz state @xcite ) , associated with many - particle tunneling oscillations between odd and even wells , ( see , e.g. , ref .
@xcite ) , which is similar to the tunneling creep of a charge density waves @xcite . if the wells are energetically inequivalent , fig .
[ lattice]b , the tunneling many - particle transitions from the excited state @xmath193 are inelastic and can be accompanied by multiphoton processes or the cooperative dynamic regimes discussed above .
but our model is oversimplified and , strictly speaking , can not be applied to inelastic tunneling transitions of atoms between energetically different wells .
the hamiltonian of atoms @xmath27 need to be modified to take into account overlapping of the particle wave function in the neighboring well .
this requires additional terms proportional to @xmath194 .
symmetry of interaction between particles and field @xmath195 also has to be of more complicated form @xmath196 .
another limitation of our model is neglecting multipole transitions that , in the case of strong interaction , may compete with multiphoton transitions @xcite . among important omissions in this paper
are spectral characteristics of the radiation . in our case
there are only two modes with the frequencies @xmath197 .
we also have not discussed superradiant and subradiant dicke states .
these have the standard form because the ising - type interparticle interaction makes the atomic spectrum non - equidistant but does not affect symmetry of these states .
authors thank alexei d. kiselev for careful reading of the manuscript and useful suggestions , and p. m. tomchuk for helpful discussions .
in this section we concentrate on the case of weakly interacting atoms with @xmath0 and describe the dissipative dynamics of the atomic subsystem using the born approximation . to this end
we apply to the master equation the standard technique of elimination of bosonic variables @xcite . for simplicity , we consider the zero temperature case and choose an initial state with no correlations between the atomic subsystem and the vacuum field .
so , the density operator can be taken in the factorized form .
substituting eq . into the equation and taking the trace over the field variables , we obtain @xmath198 + \nonumber\\ & + & [ r_j(t),\tilde{\rho}(t-\tau)s_i^-e^{-i{\omega_0}\hat{\gamma_i } ( t-\tau)}]\bigrb)+h.c.\bigr\},\end{aligned}\ ] ] where the sum over @xmath199 is transformed into the integral form @xcite @xmath200 and the functions @xmath201 , @xmath202 , @xmath203 , @xmath204 are described by eqs . - .
the next step is to employ the markov approximation .
this approximation implies that a reservoir relaxation time is much shorter than a time - scale of the atomic subsystem evolution , so that @xmath205 can be replaced by @xmath206 in the right hand side of eq . and
the upper limit of the integral can be extended to infinity . for @xmath0 ,
the eigenvalues of the operators @xmath64 are all positive . using the relation @xmath207 where the symbol @xmath208 stands for the cauchy principal value of the integral
, we can perform the integral in eq . and derive eq . .
we have used the operator functions defined on the eigenvalues of @xmath64 so that if @xmath209 then , e.g. , @xmath210 .
this result can also be obtained directly by representing the terms @xmath211 that enter eq . as follows @xmath212 where @xmath213 the principal values of the integrals in with @xmath214 can be estimated as : [ d - d - accur ] @xmath215a(k_0\hat\gamma_ir_{ij})+ \nonumber \\ & & + \frac{[1 - 3(\bar{\mathbf{d}}\cdot\bar{\mathbf{r}}_{ij})^2]b(k_0\hat\gamma_ir_{ij})- [ 1-(\bar{\mathbf{d}}\cdot\bar{\mathbf{r}}_{ij})^2]}{(k_0\hat\gamma_ir_{ij})^2}\biggr\ } , \\
\nonumber\\ & & \mathcal{p}\int_0^\infty{d\omega}\omega^3\frac{f_{ij}(\omega)}{\omega-\omega_0\hat{\gamma}_i } = \frac{3\pi}{2}\left(\omega_0\hat\gamma_i\right)^3\biggl\{[1-(\bar{\mathbf{d}}\cdot\bar{\mathbf{r}}_{ij})^2 ] \frac{\cos(k_0\hat\gamma_ir_{ij})}{k_0\hat\gamma_ir_{ij}}- \nonumber\\ & & -[1 - 3(\bar{\mathbf{d}}\cdot\bar{\mathbf{r}}_{ij})^2]\left[\frac{\sin(k_0\hat\gamma_ir_{ij})}{(k_0\hat\gamma_ir_{ij})^2 } + \frac{\cos(k_0\hat\gamma_ir_{ij})}{(k_0\hat\gamma_ir_{ij})^3}\right]\biggr\ } -\mathcal{p}\int_0^\infty{d\omega}\omega^3\frac{f_{ij}(\omega)}{\omega+\omega_0\hat{\gamma}_i},\end{aligned}\ ] ] where @xmath216 , @xmath217 and @xmath218 and @xmath219 @xcite .
in the limit @xmath220 , eq . assumes the asymptotical form coincident with the standard expression for the dipole - dipole interaction that does not depend on the coupling constant of ising interaction @xmath1 .
for the case of two non - interacting atoms with @xmath31 , exact solution of the master equation was previously obtained in ref .
it was shown that the decay of initially exited atoms demonstrates the superradiant regime and the dipole - dipole interaction does not influence the cooperative behavior of atoms . in this section
we show that the master equation for two ising - like interacting atoms , @xmath221 , can be solved along similar lines .
the hamiltonian of two interacting atoms is given by @xmath222 in the basis of atomic states @xmath173 , @xmath171 , @xmath172 , @xmath223 the density operator @xmath55 can be written the matrix form @xmath224 and eq .
reduces to the system @xmath225 where @xmath226 , @xmath227 , @xmath228 , @xmath229 , @xmath99 , @xmath100 and @xmath230 is the constant of the dipole - dipole interaction . for the initial conditions @xmath231 , @xmath232 describing two initially excited atoms , it is easy to write down the solution of eq .
@xmath233 and also @xmath234 .
it is seen that similar to the case of non - interacting atoms the solution and the decay rate of the atomic subsystem @xmath235 are both independent of the constant of the dipole - dipole interaction @xmath236 .
in order to eliminate from the rapidly oscillating terms @xmath126 we apply the method of multitime scales @xcite representing the atomic variables @xmath237 as functions of two time scales @xmath238 , where @xmath239 is a characteristic time of fast motion .
we use the following power series expansion of @xmath240 and @xmath241 over the small parameter @xmath242 in order to eliminate the imaginary term in the right hand side of eq .
we choose @xmath251 in the form @xmath252 where @xmath253 is the zero order term of the expansion for @xmath254 given by , to yield the equation @xmath255 equation is the second equation of the system .
the complex conjugate of eq . gives the equation for @xmath260 .
along the same line averaging the equation for @xmath241 from eq . provides the last equation of eq . .
f. meinardi , m. cerminara , a. sassella , r. bonifacio , and r. tubino , phys .
lett . * 91 * , 247401 ( 2003 ) .
s. h. lim , t. g. bjorklund , f. c. spano , and c. j. bardeen , phys .
lett . * 92 * , 107402 ( 2004 ) .
m. lippitz , c. g. hubner , t. christ , h. eichner , p. bordat , a. herrmann , k. mullen , and t. basche , phys .
* 92 * , 103001 ( 2004 ) .
e. m. h. p. van dijk , j. hernando , j. j. garcia - lopez , m. crego - calama , d. n. reinhoudt , l. kuipers , m. f. garcia - parajo , and n. f. van hulst , phys .
lett . * 94 * , 078302 ( 2005 ) .
y. yoshikawa , y. torii , and t. kuga , phys .
lett . * 94 * , 083602 ( 2005 ) .
t. wang and s. f. yelin , phys .
a * 72 * , 043804 ( 2005 ) . s. inouye , a. p. chikkatur , d. m. stamper - kurn , j. stenger , d. e. pritchard , and w. ketterle , science * 285 * , 571 ( 1999 ) d. schneble , y. torii , m. boyd , e. streed , d. e. pritchard , and w. ketterle , science * 300 * , 475 ( 2003 ) .
d. schneble , g. k. campbell , e. streed , m. boyd , d. e. pritchard , and w. ketterle , phys .
a * 69 * , 041601(r ) ( 2004 ) .
g. c. baldwin and j. c. solem , rev .
* 69 * , 1085 ( 1997 ) .
v. i. yukalov and e. p. yukalova , phys .
nukl . * 35 * , 348 ( 2004 ) .
j. tejada , e. m. chudnovsky , j. m. hernandez , and r. amigo , appl .
* 84 * , 2373 ( 2004 ) .
v. k. henner and i. v. kaganov , phys .
b * 68 * , 144420 ( 2003 ) . e. m. chudnovsky and d. a. garanin , phys . rev
. lett . * 89 * , 157201 ( 2002 ) .
t. brandes , phys .
408 * , 315 ( 2005 ) . c. bardot , m. schwab , m. bayer , s. fafard , z. wasilewski , and p. hawrylak , phys
b * 72 * , 035314 ( 2005 ) .
b. laikhtman and l. d. shvartsman , phys .
b * 72 * , 245333 ( 2005 )
. m. gross and s. haroche , phys . rep . *
93 * , 301 ( 1982 ) .
menshikov , phys .
usp . * 42 * , 107 ( 1999 ) .
o. morsch and m. oberthaler , rev .
phys . * 78 * , 179 ( 2006 ) .
g. mazzarella , s. m. giampaolo , and f. illuminati , phys .
a * 73 * , 013625 ( 2006 ) .
n. f. mott , _ metal - insulator transition _
( taylor&francis , london , 1974 ) .
v. j. emery , phys .
b * 14 * , 2989 ( 1976 ) .
m. imada , a. fujimori , and y. tokura , rev .
phys . * 70 * , 1039 ( 1998 ) .
a. gabris and g. s. agarwal , phys .
a * 71 * , 052316 ( 2005 ) .
r. g. unanyan and m. fleischhauer , phys .
* 90 * , 133601 ( 2003 ) .
t. miyakawa , c. p. search , and p. meystre , phys .
a * 70 * , 053622 ( 2004 ) . m. n. leuenberger and d. loss , nature * 410 * , 789 ( 2001 ) .
e. m. chudnovsky and j. tejada , _ macroscopic quantum tunneling of the magnetic moments _ ( cambridge university press , new york , 1998 ) . v. v. krasnoholovets , p. m. tomchuk , and s. p. lukyanets , adv .
phys . * 125 * , 351 ( 2003 ) .
p. m. tomchuk and s. p. lukyanets , j. mol .
struct . * 513 * , 35 ( 1999 ) . c. f. lee and n. f. johnson , phys .
93 * , 083001 ( 2004 ) .
n. n. bogolubov , jr .
, v. n. plechko , and a. s. shumovsky , fiz .
elem . chastits . at .
yadra * 14 * , 1443 ( 1983 ) [ sov .
* 14 * , 607 ( 1983 ) ] . c. hettich , c. schmitt , j. zitzmann , s. kuhn , i. gerhardt , and v. sandoghdar , science * 298 * , 385 ( 2002 ) .
p. k. pathak and g. s. agarwal , phys .
a * 70 * , 043807 ( 2004 ) .
m. greiner , o. mandel , t. w. hansch , and e. bloch , nature * 419 * , 51 ( 2002 ) .
r. blinc and b. zeks , _ soft modes in ferroelectrics and antiferromagnetics _
( north - holland , amsterdam , 1974 ) . | we study effects of direct interatomic interaction on cooperative processes in atom - photon dynamics . using a model of two - level atoms with ising - type interaction as an example ,
it is demonstrated that interparticle interaction combined with atom - field coupling can introduce additional interatomic correlations acting as a phase synchronizing factor .
for the case of weakly interacting atoms with @xmath0 , where @xmath1 is the interparticle coupling constant and @xmath2 is the atomic frequency , dynamical regimes of cooperative relaxation of atoms are analyzed in the born - markov approximation both numerically and using the mean field approximation .
we show that interparticle correlations induced by the direct interaction result in inhibition of incoherent spontaneous decay leading to the regime of collective pulse relaxation which differs from superradiance in nature . for superradiant transition , the synchronizing effect of interatomic interaction is found to manifest itself in enhancement of superradiance .
when the interaction is strong and @xmath3 , one - particle one - photon transitions are excluded and transition to the regime of multiphoton relaxation occurs . using a simple model of two atoms in a high - q single mode cavity
we show that such transition is accompanied by rabi oscillations involving many - atom multiphoton states .
dephasing effect of dipole - dipole interaction and solitonic mechanism of relaxation are discussed . |
one of the challenging problems in pulsar astrophysics is to find a consistent and observationally verified explanation for the high peculiar velocities of pulsars . these velocities can be as high as 1000 km / s and have a mean value of 450 km / s , much greater than the random velocities of ordinary stars ( harrison et al .
@xcite ; lyne & lorimer @xcite ) .
several mechanisms have been put forward in the past to explain the origin of the large proper motions .
since it is believed that pulsars are born during the first stages of some type ii or core - collapsing supernovae , an asymmetric core collapse or explosion could give the pulsars the observed velocities ( shklovskii @xcite ; woosley @xcite ; woosley & weaver @xcite ; janka & mller @xcite ; burrows & hayes @xcite ) .
the evolution of close binary systems could also be responsible for the large pulsar velocities ( gott et al .
@xcite ) . alternatively ,
the emission of electromagnetic radiation during the first months after the supernova explosion , stemming from an off - centered rotating magnetic dipole in the newborn pulsar , could give the pulsar a substantial kick ( harrison & tademaru ) .
another approach is based on the assumption that most of the energy and momentum released during a type ii supernova explosion ( @xmath0 erg ) are carried off by the relativistic neutrinos , as was observationally confirmed by the detection of a neutrino burst associated with sn1987a ( hirata et al .
@xcite ; bionta et al .
therefore , an asymmetric neutrino emission , caused for example by convection ( woosley @xcite ; woosley & weaver @xcite ; janka & mller @xcite ) or strong magnetic fields ( chuga@xmath1 @xcite ; dorofeev et al .
@xcite ; vilenkin @xcite ; horowitz & piekarewicz @xcite ) , may play an important role when trying to understand the origin of pulsar velocities .
not all these mechanisms , however , seem to be able to produce the required high velocities and many of them suffer from the fact that they need magnetic fields of the order of @xmath2 g , pulsar periods of @xmath3 ms or that they are not sustained by other observed pulsar properties ( duncan & thompson @xcite ) . in a recent paper kusenko & segr ( @xcite ) , hereafter ks , have proposed a new mechanism of asymmetric neutrino emission based on resonant neutrino oscillations .
the three types of neutrinos , @xmath4 , @xmath5 and @xmath6 , are abundantly produced in the core of a collapsing star which later on may become a pulsar .
the matter density is so high in the core that the neutrinos do not escape but get trapped .
they undergo a diffusion process until they reach higher radii , where the density has decreased and they can freely stream away . the emission surface , the so - called neutrino sphere , is not the same for the three types of neutrinos . since electron neutrinos can interact via both charged and neutral currents they interact more strongly in the protoneutron star than muon and tau neutrinos
hence , the electron neutrino sphere is at a larger radius than the muon and tau neutrino spheres .
the authors in ks showed that under these conditions neutrinos @xmath6 can resonantly turn into @xmath4 , by means of the adiabatic msw effect ( smirnov @xcite ) , in the region between the tauonic and the electronic neutrino spheres and @xmath6 have been discussed with @xmath7 and small mixing . ] .
the emerging electron neutrino , however , will be absorbed by the medium and therefore the resonant surface becomes the effective surface of emission of the @xmath6 .
neutrinos propagating in media with a longitudinal magnetic field @xmath8 have different electromagnetic properties than in the vacuum case .
they acquire an effective electromagnetic vertex which is induced by weak interactions with the charged particles in the background and generates a contribution @xmath9 to the effective self - energy of the neutrino , @xmath10 being the neutrino momentum ( esposito & capone @xcite ; dolivo & nieves @xcite ; elmfors et al .
the induced vertex modifies the flavor transformation whilst preserving chirality and , as a result , the location at which the resonance occurs is affected , leading to the spherical symmetry of the effective emission surface being replaced by a dipolar asymmetry .
the condition for resonant oscillations to take place is accordingly given by @xmath11 where @xmath12 is the neutrino vacuum mixing angle , @xmath13 the fermi constant , @xmath14 the charge density of the degenerate electron gas in which the neutrino propagates and @xmath15 the radial coordinate .
neutrinos emitted from the two magnetic poles of the resonant surface then have slightly different temperatures because the two poles are at slightly different radii .
the outcome is an asymmetric emission of momentum carried by neutrinos which gives the neutron star a kick in the direction of the magnetic field and thus leads to a recoil velocity in agreement with observational data .
quantitatively the kick is described by the asymmetry in the third component of momentum and estimated by @xmath16 since the total momentum carried away by neutrinos emitted by the protoneutron star is times the momentum of the proper motion of the pulsar , an asymmetry of 1% would give a kick in agreement with observation .
assuming an average energy for the tau neutrinos leaving the protoneutron star of @xmath17 mev , which corresponds to @xmath18 mev , the authors in ks obtain the desired asymmetry of 1% for values of @xmath19 g. as an advantage over other neutrino emission mechanisms the one discussed here works for smaller magnitudes of the magnetic field and does not demand for any constraints on the pulsar period .
if the resonant neutrino conversion turned out to be the origin of pulsar velocities , one could use pulsar observations to obtain information on neutrino masses .
the implications for particle physics models would be of great importance . a possible weak point of the model
, however , lies in the fact that the mass of the tau neutrino is required to be @xmath20 ev in order to have a resonant conversion in the protoneutron star between the electron and the tau neutrino spheres .
although such a mass is not excluded by laboratory experiments , it is difficult to accommodate it in the standard big bang cosmological model .
indeed , the present age of the universe can be used to set an upper bound on the mass of any light , stable neutrino species ( gerstein & zeldovich @xcite ; cowsik & mcclelland @xcite ) : @xmath21 ev , where @xmath22 is the hubble constant in units of 100 km s@xmath23/mpc .
nevertheless , this problematic point might be overruled by unstable tau neutrinos or by a better understanding of the evolution of the universe .
another comment can be made in the context of the supernova physics . from treatments of neutrino transport
it was found that the average energy for tau neutrinos leaving the supernova can be as high as 27 mev ( janka @xcite ) .
such a value , however , would demand for a higher magnetic field in eq .
( [ kick ] ) in order to still explain the observed pulsar velocities thus weakening one of the advantages of the proposed model . on the other hand ,
it is interesting to note that neutrino oscillations between @xmath4 and a tau neutrino with a mass of @xmath24 mev could , for a large range of mixing angles , help to explode supernovae while leaving r - process nucleosynthesis unscathed ( raffelt @xcite ) . a better understanding of supernova explosions and data from future galactic supernovae could be used to test whether the asymmetric resonant neutrino emission model is actually realistic . since none of the above mentioned points represent a serious danger to the discussed mechanism , but its implications for astrophysics and physics beyond the standard model of the fundamental interactions would be significant , it is worthwhile to examine further related observational consequences with the aim of arriving at independent constraints . in addition , the derived constraints can be used to test the viability of some of the other proposed models for the origin of pulsar velocities . in this paper
we derive , using the mechanism introduced in ks , a new expression relating the transverse velocity , the magnetic field and geometric parameters of fast spinning pulsars .
we analyze observational data on pulsars to see whether this relation is fulfilled .
our work , however , seems to indicate that pulsars do not satisfy this relation , and therefore present data do not support the neutrino oscillation mechanism .
according to the authors of ks the asymmetric neutrino emission from a magnetically distorted neutrino sphere gives a net momentum to a pulsar along the magnetic axis .
it is believed that , in general , the spin and magnetic axes of pulsars are not aligned , and thus the magnetic field rotates around the spin axis with the period @xmath25 of the pulsar . from this
it can be seen that a net velocity along the magnetic axis is only obtained if the pulsar period is much larger than the characteristic time scale @xmath26 on which the neutrino flux from the cooling protoneutron star remains constant .
for such a case , the length of the velocity vector @xmath27 and magnetic field vector @xmath8 in the core of the pulsar should be proportional : @xmath28 in a different paper kusenko & segr ( @xcite ) apply the model to pulsars which are slow spinning at birth , @xmath29 s and @xmath30 s , and argue that this @xmath31 correlation should become increasingly more significant as the period of rotation approaches a few seconds .
they claim that observational data show such an increase in the correlation for the examined periods , which would support the investigated model . in the present paper we consider fast spinning pulsars , @xmath32 s , instead .
the characteristic time @xmath26 is of the order @xmath33 s ( suzuki @xcite ) .
therefore , a reliable test of the mentioned correlation between @xmath34 and @xmath35 for slow spinning pulsars should only be possible for periods @xmath36 s which are not observed .
we believe that pulsars with @xmath37 s form a better set to test the model .
a period @xmath37 s is sufficiently smaller than the time interval in which neutrinos are emitted and therefore it can be safely assumed that the magnetic field and thus the net momentum vector rotate around the spin axis several times during the neutrino emission .
the momentum component on the plane orthogonal to the spin axis averages out and only the component parallel to the spin axis survives .
thus , for fast spinning pulsars , the proper motion should be in the direction along the pulsar spin axis . as an important consequence of this alignment of the velocity and spin axis for fast spinning pulsars
one can easily prove that the relation @xmath38 must apply where @xmath39 is the projected pulsar velocity in the sky , @xmath40 the angle between the magnetic and spin axes , and @xmath41 is the angle between the magnetic axis and the line of sight ( impact parameter ) .
we will see that observational data allow a test of this relation . pulsar proper motions are determined by radio interferometry ( harrison et al .
@xcite ) and by interstellar scintillation observations ( cordes @xcite ) .
in order to determine the emission geometry parameters @xmath40 and @xmath41 one has to rely on some emission model for pulsars .
the so - called magnetic - pole model assumes that the electromagnetic radiation originates in the vicinity of a magnetic pole .
the radiation is emitted in a narrow cone - shaped beam centered around the magnetic axis .
the nonvanishing angle @xmath40 between the spin axis and the magnetic axis produces the characteristic pulses every time the beam sweeps the line of sight ( lyne & graham - smith ) .
two different groups ( lyne & manchester ; rankin @xcite , @xcite , @xcite ) have determined @xmath40 and @xmath41 for more than one hundred pulsars . the method used by both groups is based on the apparent increase of pulse width as @xmath42 .
the classification scheme of pulsars , however , differs between the two groups and , for calculating the angles , they find different empirical laws relating the pulse width and the pulsar period : while lyne & manchester ( ) use width @xmath43 , rankin ( @xcite , @xcite , @xcite ) employs width @xmath44 .
the results of both groups show , in general , a remarkable agreement for angles @xmath45 but they can widely differ for larger angles ( miller & hamilton ) . our aim has been to find out whether the relation given by eq .
( [ vbangles ] ) is corroborated by observational data .
bearing in mind the lack of complete agreement between the different measurements of the emission angles , we have studied three different sets of data : ( a ) angles from lyne & manchester ( ) , ( b ) angles from rankin ( @xcite ) and ( c ) the average of the angles from sets ( a ) and ( b ) for those pulsars which are common in both of these sets and which have similar @xmath40 values , namely @xmath46 .
furthermore , it has to be taken into account that eq .
( [ vbangles ] ) , true at the pulsar birth , may be spoiled by the temporal evolution of pulsar parameters ( angles and magnetic field ) .
it is believed that the magnetic field of a pulsar decays with a characteristic time @xmath47 years ( harrison et al .
there is also observational evidence supporting that the magnetic and spin axes become aligned when the pulsar ages ( lyne & manchester ) . to avoid these problems associated with time evolution we have selected from the data the pulsars younger than @xmath48 years and , to be even safer , younger than @xmath49 years ( lorimer et al .
@xcite ) thus creating two subsets for each of the sets ( a ) , ( b ) and ( c ) .
we disregard older pulsars since their present properties may be rather different from those at birth .
we do not consider pulsars in binary systems either since for them eq .
( [ vbangles ] ) could be affected by the evolution of close binary systems .
the selected pulsars for the three cases ( a ) , ( b ) and ( c ) are listed in tables [ tab1 ] and [ tab2 ] .
lll psr b0136 + 57@xmath50 & psr b0329 + 54 & psr b0450@xmath5118@xmath50 + psr b0458 + 46@xmath50 & psr b0559@xmath5105 & psr b0656 + 14@xmath50 + psr b0736@xmath5140 & psr b0823 + 26 & psr b1508 + 55@xmath50 + psr b1642@xmath5103 & psr b1706@xmath5116@xmath50 & psr b1929 + 10 + psr b1933 + 16@xmath50 & psr b1946 + 35@xmath50 & psr b2020 + 28@xmath50 + [ tab1 ] lll psr b0136 + 57@xmath50 & psr b0329 + 54 & psr b0355 + 54@xmath50 + psr b0450@xmath5118@xmath50 & psr b0450 + 55@xmath50 & psr b0458 + 46@xmath50 + psr b0736@xmath5140 & psr b0740@xmath5128@xmath50 & psr b0823 + 26 + psr b0919 + 06@xmath50 & psr b1449@xmath5164@xmath50 & psr b1508 + 55@xmath50 + psr b1642@xmath5103 & psr b1706@xmath5116@xmath50 & psr b1818@xmath5104@xmath50 + psr b1822@xmath5109@xmath50 & psr b1842 + 14 & psr b1933 + 16@xmath50 + psr b1946 + 35@xmath50 & psr b2020 + 28@xmath50 & psr b2021 + 51@xmath50 + psr b2217 + 47 & psr b2224 + 65@xmath50 & + [ tab2 ]
we plot @xmath52 $ ] versus @xmath53 for young , fast spinning pulsars in figs .
[ figu1 ] , [ figu2 ] and [ figu3 ] , for the cases ( a ) , ( b ) and ( c ) , respectively .
the transverse velocity @xmath39 is calculated from the measured angular velocity @xmath54 using the pulsar distance @xmath55 derived from its dispersion measure .
the pulsar velocities have been corrected for galactic rotation by removing a flat rotation curve of rotational velocity 225 km / s , with the sun at a distance of 8.5 kpc from the galactic center .
they have also been corrected for the peculiar solar motion of the sun which was assumed to to have a velocity of 15.6 km / s in the direction @xmath56 and @xmath57 ( murray @xcite ) .
the parameters @xmath55 , @xmath34 , @xmath25 and the pulsar ages as well as the parameters needed to compute @xmath54 have been taken from taylor et al .
( @xcite , @xcite ) , an updated catalog containing 706 pulsars .
the scatter of points in the plots we depict suggests that there is hardly any correlation . to evaluate rigorously their significance ,
we use the spearman rank - order correlation coefficient @xmath58 and its probability @xmath59 .
the coefficient @xmath58 ranges from @xmath60 ( perfect correlation ) to @xmath61 ( perfect anticorrelation ) .
further information is given by @xmath59 , the probability that @xmath58 for two uncorrelated data sets would be larger than the spearman coefficient found .
hence , to have a significant correlation between two sets , one must obtain @xmath62 and @xmath63 .
the calculated @xmath58 and @xmath59 corresponding to our three cases are shown in table [ tab3 ] .
we calculate the two numbers using pulsars younger than @xmath49 yr and pulsars younger than @xmath48 yr .
we are unable to find any correlation of the sort given by eq .
( [ vbangles ] ) ( except possibly for the case ( b ) with pulsars younger than @xmath49 yr ) .
ccccc case & age[@xmath48 yr]@xmath64 & number & @xmath58 & @xmath65 + ( a ) & 1 & 15 & @xmath66 & 0.83 + ( a ) & 0.3 & 9 & 0.25 & 0.52 + ( b ) & 1 & 23 & 0.31 & 0.15 + ( b ) & 0.3 & 17 & 0.68 & @xmath67 + ( c ) & 1 & 10 & @xmath68 & 0.78 + ( c ) & 0.3 & 7 & @xmath69 & 0.88 + [ tab3 ] it is interesting to point out the existence of fast spinning pulsars like psr b0833 - 45 which have large peculiar velocities whilst the magnetic and spin axes are nearly perpendicular .
clearly , the mechanism presented in ks can not account for the proper motion of these pulsars .
we have also examined the case of young , fast spinning pulsars with small angles @xmath70 . in the framework of the studied mechanism
those pulsars should have small transverse velocities @xmath39 ( since @xmath41 is small for most pulsars ) .
for the mean transverse velocity of the pulsars fulfilling the selection criteria we obtain 344 km / s .
this has to be compared with the mean average transverse velocity for pulsars without any restrictions on @xmath40 or the pulsar period .
such mean values were computed to be @xmath71 km / s for 29 pulsars which are younger than 3 myr , and @xmath72 km / s for 99 pulsars without the age constraint ( lyne & lorimer @xcite ) .
the fact that there is no significant difference between the mean transverse velocity we obtain and the latter ones above does not support the model examined here .
there is yet another important consequence of the alignment of the spin axis and the velocity vector for fast spinning pulsars , which has been explored some time ago ( tademaru @xcite ; anderson & lyne @xcite ) in the context of the asymmetric emission of electromagnetic radiation mentioned in the introduction of this paper .
if these two vectors are aligned , the difference between the projected pulsar velocity angle in the sky @xmath73 and the position angle of the linear polarization vector in the center of a pulse ( corrected for faraday rotation ) @xmath74 should be either @xmath75 or @xmath76 depending on the emission mechanism and the inclination of the spin and magnetic axes to the line of sight . while tademaru ( @xcite ) found that the histogram of @xmath77 showed two peaks at @xmath75 and @xmath76 , a subsequent study performed by anderson & lyne ( @xcite ) concluded that such peaks were not present .
the latter result was argued to be in disagreement with the mechanism introduced by harrison & tademaru ( ) . it would not support the resonant neutrino emission model discussed here either .
given the significant increase in the amount and quality of available pulsar data over the last years it would be of interest to carry out the @xmath78 test again .
this could not only help to resolve the contradicting conclusions of the previous studies but also yield an additional test of the neutrino emission model if only fast spinning pulsars are selected from the updated and enlarged sets of observational data .
moreover , both tests have different sources of errors . while the @xmath78 test does not depend on the angles
@xmath40 and @xmath41 the derivation of which is model dependent , it relies on the measurement of the rotation measure rm of a pulsar .
the contribution from faraday rotation , rm@xmath79 , has to be subtracted from the measured polarization angle to obtain the intrinsic polarization angle @xmath74 .
our test relies on the measurement of @xmath40 and @xmath41 but does not need the value of rm for each pulsar which can be a significant source of errors . in this sense ,
both tests constitute complementary ways of checking whether the model given in ks and related models are realistic or not .
we have looked into possible ways of testing the new mechanism suggested by ks to explain the large proper motion of pulsars .
it was argued that for fast spinning pulsars this mechanism forces the vector velocity of a pulsar to be aligned with its spin axis . making use of this alignment
we have found eq .
( [ vbangles ] ) , which is a relation between the magnetic field of a pulsar , its proper motion , the angle @xmath40 between the magnetic and spin axes and the impact parameter @xmath41 .
this relation should be true as long as pulsar velocities are produced by asymmetric emission from a magnetically distorted resonant neutrino - sphere .
we have studied whether pulsar data actually show this correlation . to carry out this study
we have had to face the fact that there is some controversy about the true values of @xmath40 and @xmath41 in the sense that the angles from two different groups show discrepancies for @xmath80 .
therefore , we have in a first step considered separately the data from each group . in addition , we have also selected the pulsars studied by both groups for which there is a reasonable agreement between the angles measured by either group .
we believe that the results we derive in the latter case are the most reliable . in this case , observational data do not show any significant correlation and therefore , the mechanism introduced in ks is unlikely to explain the proper motion of pulsars .
we have indicated that the temporal evolution of pulsars may spoil any initial correlation . to diminish this effect
we have only considered young pulsars .
another possible caveat is the fact that eq .
( [ vbangles ] ) involves the magnetic field in the core of the pulsar whilst what is measured is the surface magnetic field .
it is , however , plausible and accepted to assume that both fields are proportional ( manchester & taylor @xcite ; ruderman @xcite ) .
looking at table [ tab3 ] one can see that the number of pulsars with available data which satisfy all the requirements regarding period , age and availability of emission angles as well as proper motions is small in all the cases considered .
although our results cast doubt on the resonant neutrino emission as the mechanism responsible for the proper motion of pulsars , clearly one needs more data , and clarify the uncertainties related to the true emission angles , to rule out definitely or support this mechanism . as a final remark we would like to point out that our conclusions also apply to other mechanisms proposed to explain the large velocities of pulsars
this is the case for fast spinning pulsars in all models where the momentum kick is collinear to the magnetic axis and the resulting velocity proportional to @xmath34 . under these conditions velocities along the spin axis
are predicted .
hence , the work by chuga@xmath1 ( @xcite ) and dorofeev et al .
( @xcite ) is affected .
they study how the polarization of electrons and positrons by a high magnetic field in a newborn neutron star will generate an anisotropic neutrino emission .
also the mechanism suggested by vilenkin ( @xcite ) is disfavored for a uniform magnetic field , as is the mechanism in horowitz & piekarewicz ( @xcite ) .
pulsar velocities from resonant spin - flavor precession of neutrinos are studied by akhmedov et al .
( @xcite ) . again
, this mechanism is put under pressure by our work as well as the model suggested by kusenko & segr ( @xcite ) where a resonant conversion to sterile neutrinos induced by neutral currents is considered .
we would like to thank dr .
p. podsiadlowski and dr .
s. sarkar for helpful discussions as well as prof .
manchester , prof .
rankin and dr .
k. xilouri for useful information .
m.b . gratefully acknowledges financial support from the fellowship hsp ii / aufe of the german academic exchange service ( daad ) .
r.t . would like to thank dr .
s. sarkar for his invitation to oxford university and the european theoretical astroparticle network for financial support under the eec contract no .
chrx - ct93 - 0120 ( direction generale 12 coma ) . | it has been recently suggested that magnetically affected neutrino oscillations inside a cooling protoneutron star , created in a supernova explosion , could explain the large proper motion of pulsars .
we investigate whether this hypothesis is in agreement with the observed properties of pulsars and find that present data disfavor the suggested mechanism .
the relevance of our results for other models proposed to understand the origin of pulsar velocities is also discussed . |
nowadays , it is very clear how special relativity effects influence on measured data .
the first celebrated example of this fact was the atmospheric muons decay explanation as a time dilation effect .
this is the rossi - hall experiment @xcite . considering the mrzke - wheeler synchronization @xcite as the natural generalization to accelerated observers of einstein synchronization in special relativity ,
we wonder whether mrzke - wheeler effects influence on measured data in nature .
this question is also motivated by the fact that recently the twin paradox was completely solved in ( 1 + 1)-spacetime by means of these effects @xcite and it is natural to ask for empirical confirmation .
of course these effects comprehend the well known special relativistic ones for inertial observers as well as the new ones .
these new effects can be seen as corrections of the special relativistic ones due to the acceleration of the involved observer .
+ a small deviation towards the sun from the predicted pioneer acceleration : @xmath0 for pioneer 10 and @xmath1 for pioneer 11 , was reported for the first time in @xcite .
the analysis of the pioneer data from 1987 to 1998 for pioneer 10 and 1987 to 1990 for pioneer 11 made in @xcite improves the anomaly value and it was reported to be @xmath2 .
this is known as the pioneer anomaly .
+ considering that mrzke - wheeler tiny effects are difficult to measure , we careful looked for some observational object for which the searched effect could be appreciable .
this search led us to the pioneer 10 .
in fact , through a simple analytic formula for the mrzke - wheeler map exact calculation developed in this letter , computing the acceleration difference between the mrzke - wheeler and frenet - serret coordinates for the earth s translation around the sun , we see that this mrzke - wheeler long range effect is between @xmath3 and @xmath4 of the pioneer anomaly value .
unfortunately , due to statistical errors in the measured anomaly , it is not possible to confirm the influence of the mrzke - wheeler acceleration effect on the measured pioneer data .
moreover , a recently numerical thermal model based on a finite element method @xcite has shown a discrepancy of @xmath5 of the actual measured anomaly and due to the mentioned statistical errors , it was concluded there that the pioneer anomaly has been finally explained within experimental error of @xmath6 of the anomaly value : + _ ... to determine if the remaining @xmath5 represents a statistically significant acceleration anomaly not accounted for by conventional forces , we analyzed the various error sources that contribute to the uncertainties in the acceleration estimates using radio - metric doppler and thermal models ... we therefore conclude that at the present level of our knowledge of the pioneer 10 spacecraft and its trajectory , no statistically significant acceleration anomaly exists . _
+ although it is tempting to think that the @xmath5 discrepancy found in @xcite is due to a long range mrzke - wheeler acceleration effect , it can not be confirmed .
we hope that the ideas presented here could encourage other research teams in the search for other observational objects that could finally answer the question posed in this letter . +
consider the @xmath7-spacetime @xmath8 spanned by the vectors @xmath9 with the lorentz metric : @xmath10 respect to the basis @xmath11 .
an observer is a smooth curve @xmath12 naturally parameterized with timelike derivative at every instant ; i.e. @xmath13 .
we will say a vector is spatial if it is a linear combination of @xmath14 .
a spatial vector @xmath15 is unitary if @xmath16 .
+ consider a timelike vector @xmath17 in @xmath8 ; i.e. @xmath18 .
we define the scaled lorentz transformation @xmath19 : @xmath20 where @xmath21 is the orthocronous lorentz boost transformation sending @xmath22 to the unitary vector @xmath23 ; i.e. the original and transformed coordinates are in standard configuration ( @xmath24 , @xmath25 and @xmath26 are colinear with @xmath27 , @xmath28 and @xmath29 respectively where the prime denote the spatial transformed coordinates and the others denote the original spatial coordinates ) . the scaled lorentz transformation has the following properties : @xmath30 @xmath31 + a smooth map @xmath32 is a mrzke - wheeler map of the observer @xmath33 if it verifies : @xmath34 for every real @xmath35 , positive real @xmath36 and unitary spatial vector @xmath15 ( see figure [ mw_coord ] ) .
this map @xcite ,
@xcite , @xcite is clearly an extension of the einstein synchronization convention for non accelerated observers ; i.e. it is the natural generalization of a lorentz transformation in the case of accelerated observers .
+ [ mwformula ] consider an observer @xmath12 .
then , @xmath37 is a mrzke - wheeler map of the observer @xmath33 such that @xmath15 is a unitary spatial vector . _
proof : _ recall that for every @xmath17 such that @xmath18 we have that @xmath38 . this way , @xmath39 because @xmath40 . from the formula
it is clear that @xmath41 is smooth .
@xmath42 + the last mrzke - wheeler map formula was written for the first time in @xcite for @xmath43-spacetime where it was shown , in this particular case , that it is actually a conformal map .
moreover , the twin paradox is solved in @xmath43-spacetime . in the general case treated here ,
the mrzke - wheeler map is no longer conformal .
+ as an example , consider the uniformly accelerated observer in @xmath44-spacetime along the @xmath45 axis : @xmath46 where @xmath35 is its natural parameter and @xmath47 such that @xmath17 is the observer acceleration .
its mrzke - wheeler map is : @xmath48\sigma_{0 } \\ & & + r\cosh \left(\frac{s}{r}\right)\left[\cosh \left(\frac{s}{r}\right ) + \sinh \left(\frac{s}{r}\right)\frac{x}{r}\right]\sigma_{1 } \\ & & + \frac{r}{r}\sinh \left(\frac{r}{r}\right)\left [ y\ \sigma_{2 } + z\ \sigma_{3 } \right ] \\\end{aligned}\ ] ] where @xmath49 . in this example , it is interesting that besides @xmath41 restricted to the @xmath50 plane is a conformal map ( as it was expected from @xcite ) , it is also also conformal restricted to the @xmath51 plane .
the pioneer 10/11 data is measured from earth s dsn antennas ( deep space network ) and we wonder whether this data is affected by earth s translation around the sun .
we comment about earth s rotation at the end of the section .
we model the earth s translation as the uniformly rotating observer @xmath52\ ] ] where @xmath35 is its natural parameter , @xmath53 is its lorentz contraction factor and @xmath54 .
its mrzke - wheeler map is : @xmath55\sigma_{0 } \\ & & + \left [ r\cos\left(\frac{\omega}{ck}r\right ) + x\ \sqrt{\frac{1}{k^{2 } } - \left(\frac{r}{r}\sin\left(\frac{\omega}{ck}r\right)\right)^{2 } } \right]\vec{a}(s ) \\ & & + \frac{1}{k}\ y\ \vec{b}(s ) \\ & & + z\ \sqrt{\frac{1}{k^{2 } } - \left(\frac{r}{r}\sin\left(\frac{\omega}{ck}r\right)\right)^{2}}\ \sigma_{3 } \\\end{aligned}\ ] ] where @xmath49 .
we have chosen the framing @xmath56 corresponding to the @xmath57 coordinates such that @xmath58 is the frenet - serret framing of the observer ( see figure [ frenetserret ] ) : @xmath59 @xmath60 this expression was also obtained in @xcite in the particular case @xmath61 .
it is interesting to notice the oscillatory term of the above map . in order to compare the spatial mrzke - wheeler coordinates with the frenet - serret
coordinates we consider the difference @xmath62 . because @xmath63 and @xmath64 , we have that @xmath65 and restricted to the region @xmath66 we have the approximation : + @xmath67\ \vec{a}(s ) \\ & & + \ y\
\vec{b}(s)+ z\ \sigma_{3 } \\\end{aligned}\ ] ] because the @xmath22 component is zero , we have the following transformation between the spatial mrzke - wheeler coordinates and the frenet - serret coordinates : @xmath68 where @xmath69 . because the pioneer s velocity and acceleration are very small respect to the natural scale of the problem @xmath70 , differentiating the above expression we have : @xmath71 where @xmath36 is the distance from the sun and @xmath17 is the acceleration . because the recorded pioneer data ( at least for pioneer 10 ) corresponds to the region between @xmath72 and @xmath73 , we can consider that @xmath74 where @xmath75 is the pioneer s distance from the sun and we have the following approximation : @xmath76 where @xmath77 is the pioneer s speed and @xmath78 is the angle between its radius vector from the sun and its velocity vector . computing the acceleration difference @xmath79 between the mrzke - wheeler and frenet - serret
coordinates at the pioneer s maximal speed @xmath80 , we have the result : @xmath81 and we see that it is between @xmath3 and @xmath82 of the measured pioneer anomaly @xmath83 . + the calculated difference @xmath79 points towards the @xmath29 edge when @xmath84 and in the opposite direction when @xmath85
. this would contradict the claim that the anomaly always points towards the sun made in the data analysis @xcite and @xcite .
however , in the data analysis made in @xcite , it is claimed that it can not be confirmed whether the anomaly is sunwards , contrary to the earlier claim .
+ finally , we would like to comment about a possible numerical analysis on the influence of earth s rotation on the measured data . in order to do so , we define the following framing dependent non abelian product of observers : @xmath86 this product is the generalization of the special relativistic velocities addition and has the following property : @xmath87 this way
, the observer @xmath88 is the one we should consider and its mrzke - wheeler map is just the composition of the previously exactly calculated map of the uniformly rotating observer .
unfortunately , the map gets really involved and the analysis must be done numerically .
an analysis of the parameters involved in the rotation analysis , shows that the magnitude order of the long range mrzke - wheeler acceleration effect coincides with the one of the pioneer anomaly and should also be considered .
although after strongly numerical evidence it is tempting to think that the @xmath5 discrepancy of the anomaly value found in @xcite is due to a long range mrzke - wheeler acceleration effect described in this letter , due to statistical errors in the measured anomaly it can not be neither confirmed nor neglected .
we hope that the ideas presented here could encourage other research teams in the search for other observational objects that could finally answer whether mrzke - wheeler effects influence on measured data in nature .
mrzke m , wheeler j , _ gravitation as geometry - i : the geometry of spacetime and the geometrodynamical standard meter _ gravitation and relativity , h .- y .
chiu and w. f. hoffmann , eds .
, w. a. benjamin , new york - amsterdam ( 1964 ) 40 . | we wonder whether mrzke - wheeler effects influence on measured data in nature . through a formula developed in this letter for the calculation of the mrzke - wheeler map of a general accelerated observer , we study the influence of the mrzke - wheeler acceleration effect on the nasa s pioneer anomaly and found that it is about a fifth of the anomaly value . due to statistical errors in the measured anomaly ,
it is not possible to neither confirm nor neglect the influence of the mrzke - wheeler acceleration effect on the measured pioneer data .
we hope that the ideas presented here could encourage other research teams in the search for other observational objects that could finally answer the question posed in this letter . |
symmetry occurs in many constraint satisfaction problems .
for example , in scheduling a round robin sports tournament , we may be able to interchange all the matches taking place in two stadia .
similarly , we may be able to interchange two teams throughout the tournament . as a second example , when colouring a graph ( or equivalently when timetabling exams ) , the colours are interchangeable
. we can swap red with blue throughout .
if we have a proper colouring , any permutation of the colours is itself a proper colouring .
problems may have many symmetries at once .
in fact , the symmetries of a problem form a group . their action is to map solutions ( a schedule , a proper colouring , etc . ) onto solutions .
symmetry is problematic when solving constraint satisfaction problems as we may waste much time visiting symmetric solutions .
in addition , we may visit many ( failing ) search states that are symmetric to those that we have already visited .
one simple but effective mechanism to deal with symmetry is to add constraints which eliminate symmetric solutions @xcite .
unfortunately eliminating all symmetry is np - hard in general @xcite .
however , recent results in parameterized complexity give us a good understanding of the source of that complexity . in this survey paper ,
i summarize results in this area . for more background ,
see @xcite .
to illustrate the ideas , we consider a simple problem from musical composition
. the all interval series problem ( prob007 in csplib.org @xcite ) asks for a permutation of the numbers 0 to @xmath0 so that neighbouring differences form a permutation of 1 to @xmath0 . for @xmath1
, the problem corresponds to arranging the half - notes of a scale so that all musical intervals ( minor second to major seventh ) are covered .
this is a simple example of a graceful graph problem in which the graph is a path .
we can model this as a constraint satisfaction problem in @xmath2 variables with @xmath3 iff the @xmath4th number in the series is @xmath5 .
one solution for @xmath6 is : @xmath7 the differences form the series : @xmath8 .
the all interval series problem has a number of different symmetries .
first , we can reverse any solution and generate a new ( but symmetric ) solution : @xmath9 second , the all interval series problem has a value symmetry as we can invert values .
if we subtract all values in ( 1 ) from @xmath10 , we generate a second ( but symmetric ) solution : @xmath11 third , we can do both and generate a third ( but symmetric ) solution : @xmath12 to eliminate such symmetric solutions from the search space , we can post additional constraints which eliminate all but one solution in each symmetry class . to eliminate the reversal of a solution , we can simply post the constraint : @xmath13 this eliminates solution ( 2 ) as it is a reversal of ( 1 ) . to eliminate the value symmetry which subtracts all values from @xmath10
, we can post : @xmath14 this eliminates solutions ( 2 ) and ( 3 ) . finally , eliminating the third symmetry where we both reverse the solution and subtract it from @xmath10 is more difficult .
we can , for instance , post : @xmath15 & \leq_{\rm lex } & [ 10-x_{11 } , \ldots , 10-x_{1}]\end{aligned}\ ] ] note that of the four symmetric solutions given earlier , only ( 4 ) with @xmath16 , @xmath17 and @xmath18 satisfies all three sets of symmetry breaking constraints : ( 5 ) , ( 6 ) and ( 7 ) .
the other three solutions are eliminated .
we will need some formal notation to present some of the more technical results . a _ constraint satisfaction problem _ ( csp ) consists of a set of variables , each with a finite domain of values , and a set of constraints @xcite .
each _ constraint _ is specified by the allowed combinations of values for some subset of variables .
for example , @xmath19 is a binary constraint which ensures @xmath20 and @xmath21 do not take the same values .
global constraint _ is one in which the number of variables is not fixed .
for instance , the global constraint @xmath22,n)$ ] ensures that @xmath2 variables , @xmath23 to @xmath24 , take @xmath25 different values @xcite .
that is , @xmath26 .
constraint solvers typically use backtracking search to explore the space of partial assignments .
after each assignment , constraint propagation algorithms prune the search space by enforcing local consistency properties like domain or bound consistency .
a constraint is _ domain consistent _
( _ dc _ ) iff when a variable is assigned any of the values in its domain , there exist compatible values in the domains of all the other variables of the constraint .
such values are called a _
support_. a csp is domain consistent iff every constraint is domain consistent .
recently , bessiere _
_ have shown that a number of common global constraints are intractable to propagate @xcite .
for instance , enforcing domain consistency on the constraint is np - hard @xcite .
parameterized complexity can provide a more fine - grained view of such results , identifying more precisely what makes a global constraint ( in)tractable .
we will say that a problem is _ fixed - parameter tractable _
( _ fpt _ ) if it can be solved in @xmath27 time where @xmath28 is _ any _ computable function , @xmath29 is some parameter , @xmath30 is a constant , and @xmath2 is the size of the input .
for example , vertex cover ( `` given a graph with @xmath2 vertices , is there a subset of vertices of size @xmath29 or less that cover each edge in the graph '' ) is np - hard in general , but fixed - parameter tractable with respect to @xmath29 since it can be solved in @xmath31 time @xcite .
hence , provided @xmath29 is small , vertex cover can be solved effectively .
as we have argued , symmetry is a common feature of many real - world problems that dramatically increases the size of the search space if it is not factored out .
symmetry can be defined as a bijection on assignments that preserves solutions .
the set of symmetries form a group under the action of composition .
we focus on two special types of symmetry .
a _ value symmetry _ is a bijective mapping , @xmath32 of the values such that if @xmath33 is a solution then @xmath34 is also .
for example , in our all interval series example , there is a value symmetry @xmath32 that maps the value @xmath4 onto @xmath35 .
a _ variable symmetry _
, on the other hand , is a bijective mapping , @xmath36 of the indices of variables such that if @xmath33 is a solution then @xmath37 is also .
for example , in our all interval series example , there is a variable symmetry @xmath36 that maps the index @xmath4 onto @xmath38 .
this swaps the variable @xmath39 with @xmath40 .
a simple and effective mechanism to deal with symmetry is to add constraints to eliminate symmetric solutions @xcite .
the basic idea is very simple .
we pick an ordering on the variables , and then post symmetry breaking constraints to ensure that the final solution is lexicographically less than any symmetric re - ordering of the variables .
that is , we select the `` lex leader '' assignment . for example , to break the variable symmetry @xmath36 , we post the constraint : @xmath41 & \leq_{\rm lex } & [ x_{\theta(1 ) } , \ldots , x_{\theta(n)}]\end{aligned}\ ] ] efficient inference methods exist for propagating such constraints @xcite .
the symmetry breaking constraints in our all interval series example can all be derived from such lex leader constraints . in theory
, the lex leader method solves the problem of symmetries , eliminating all symmetric solutions and pruning many symmetric states .
unfortunately , the set of symmetries might be exponentially large ( for example , in a graph @xmath29-colouring , there are @xmath42 symmetries ) .
there may therefore be too many symmetry breaking constraints to post .
in addition , decomposing symmetry breaking into many lex leader constraints typically hinders propagation .
we focus on three special but commonly occurring cases where symmetry breaking is more tractable and propagation can be more powerful : value symmetry , interchangeable values , and row and column symmetry . in each case , we identify islands of tractability but show that the quick - sands of intractability remain close to hand .
value symmetries are a commonly occurring symmetry that are more tractable to break @xcite . for instance , puget has proved that a linear number of symmetry breaking constraints will eliminate any number of value symmetries in polynomial time @xcite .
given a set of value symmetries @xmath43 , we can eliminate all value symmetry by posting the global constraint @xmath44)$ ] @xcite .
this is a conjunction of lex leader constraints , ensuring that , for each @xmath45 : @xmath46 unfortunately , enforcing domain consistency on this global constraint is np - hard .
however , this complexity depends on the number of symmetries . breaking all value symmetry is fixed - parameter tractable in the number of symmetries .
enforcing domain consistency on @xmath44)$ ] is np - hard in general but fixed - parameter tractable in @xmath47 .
* proof : * np - hardness is proved by theorem 1 in @xcite , and fixed - parameter tractability by theorem 7 in @xcite . one situation where we may have only a small number of symmetries
is when we focus on just the _ generators _ of the symmetry group @xcite .
this is attractive as the size of the generator set is logarithmic in the size of the group , many algorithms in computational group theory work on generators , and breaking just the generator symmetries has been shown to be effective on many benchmarks @xcite . in general , breaking just the generators may leave some symmetry .
however , on certain symmetry groups ( like that for interchangeable values considered in the next section ) , all symmetry is eliminated ( theorem 3 in @xcite ) .
by exploiting special properties of the value symmetry group , we can identify even more tractable cases .
a common type of value symmetry with such properties is that due to interchangeable values .
we can break all such symmetry using the idea of value _ precedence _ @xcite .
in particular , we can post the global symmetry breaking constraint @xmath48)$ ] .
this ensures that for all @xmath49 : @xmath50 that is , the first time we use @xmath5 is before the first time we use @xmath29 for all @xmath49 .
for example , consider the assignment : @xmath51 this satisfies value precedence as 1 first occurs before 2 , 2 first occurs before 3 , etc .
now consider the symmetric assignment in which we swap @xmath52 with @xmath53 : @xmath54 this does not satisfy value precedence as @xmath53 first occurs before @xmath52 .
a constraint eliminates all symmetry due to interchangeable values . in @xcite
, we give a linear time propagator for enforcing domain consistency on the constraint . in @xcite
, we argue that @xmath55 can be derived from the lex leader method ( but offers more propagation by being a global constraint ) .
another way to ensure value precedence is to map onto dual variables , @xmath56 which record the first index using each value @xmath5 @xcite .
this transforms value symmetry into variable symmetry on the @xmath56 .
we can then eliminate this variable symmetry with some ordering constraints : @xmath57 in fact , puget proves that we can eliminate _ all _ value symmetry ( and not just that due to value interchangeability ) with a linear number of such ordering constraints .
unfortunately , this decomposition into ordering constraints hinders propagation even for the tractable case of interchangeable values ( theorem 5 in @xcite ) . indeed , even with _
just _ two value symmetries , mapping into variable symmetry can hinder propagation .
this is supported by the experiments in @xcite where we see faster and more effective symmetry breaking with the global constraint .
this global constraint thus appears to be a promising method to eliminate the symmetry due to interchangeable values .
a generalization of the symmetry due to interchangeable values is when values partition into sets , and values within each set ( but not between sets ) are interchangeable .
the idea of value precedence can be generalized to this case @xcite .
the global @xmath58 constraint ensures that values in each interchangeable set occur in order .
more precisely , if the values are divided into @xmath59 equivalence classes , and the @xmath5th equivalence class contains the values @xmath60 to @xmath61 then @xmath58 ensures @xmath62 for all @xmath63 and @xmath64 . enforcing domain consistency on @xmath58
is np - hard in general but fixed - parameter tractable in @xmath65 @xcite .
another common type of symmetry where we can exploit special properties of the symmetry group is row and column symmetry @xcite .
many problems can be modelled by a matrix model involving a matrix of decision variables @xcite
. often the rows and columns of such matrices are fully or partially interchangeable @xcite .
for example , the equidistant frequency permutation array ( efpa ) problem is a challenging combinatorial problem in coding theory .
the aim is to find a set of @xmath66 code words , each of length @xmath67 such that each word contains @xmath68 copies of the symbols 1 to @xmath69 , and each pair of code words is at a hamming distance of @xmath70 apart .
for example , for @xmath71 , @xmath72 , @xmath73 , @xmath74 , one solution is : @xmath75 this problem has applications in communications , and is closely related to other combinatorial problems like finding orthogonal latin squares .
_ @xcite consider a simple matrix model for this problem with a @xmath66 by @xmath67 array of variables , each with domains @xmath76 to @xmath69 .
this model has row and column symmetry since we can permute the rows and the columns of any solution .
although breaking all row and column symmetry is intractable in general , it is fixed - parameter tractable in the number of columns ( or rows ) . with a @xmath2 by @xmath77 matrix , checking lex leader constraints that break all row and column symmetry is np - hard in general but fixed - parameter tractable in @xmath78 .
* proof : * np - hardness is proved by theorem 3.2 in @xcite , and fixed - parameter tractability by theorem 1 in @xcite . note that the above result only talks about _ checking _ a constraint which breaks all row and column symmetry .
that is , we only consider the computational cost of deciding if a complete assignment satisfies the constraint .
propagation of such a global constraint is computationally more difficult . just row or column symmetry on their own are tractable to break . to eliminate all row symmetry we can post lexicographical ordering constraints on the rows .
similarly , to eliminate all column symmetry we can post lexicographical ordering constraints on the columns .
when we have both row and column symmetry , we can post a @xmath79 constraint that lexicographically orders both the rows and columns @xcite . this does not eliminate all symmetry since it may not break symmetries which permute both rows and columns .
nevertheless , it is more tractable to propagate and is often highly effective in practice .
note that @xmath79 can be derived from a _ strict _ subset of the constraints . unfortunately propagating such a constraint
completely is already np - hard . with a @xmath2 by @xmath77 matrix , enforcing domain consistency on @xmath79 is np - hard in general .
* proof : * see threorem 3 in @xcite .
there are two special cases of matrix models where row and column symmetry is more tractable to break .
the first case is with an all - different matrix , a matrix model in which every value is different .
if an all - different matrix has row and column symmetry then the lex - leader method ensures that the top left entry is the smallest value , and the first row and column are ordered @xcite .
domain consistency can be enforced on such a global constraint in polynomial time @xcite .
the second more tractable case is with a matrix model of a function .
in such a model , all entries are 0/1 and each row sum is 1 .
if a matrix model of a function has row and column symmetry then the lex - leader method ensures the rows and columns are lexicographically ordered , the row sums are 1 , and the sums of the columns are in decreasing order @xcite .
domain consistency can also be enforced on such a global constraint in polynomial time @xcite .
the study of computational complexity in constraint programming has tended to focus on the structure of the constraint graph ( e.g. especially measures like tree width @xcite ) or on the semantics of the constraints ( e.g. @xcite ) .
however , these lines of research are mostly concerned with constraint satisfaction problems as a whole , and do not say much about individual ( global ) constraints . for global constraints of bounded arity , asymptotic analysis has been used to characterize the complexity of propagation both in general and for constraints with a particular semantics . for example
, the generic domain consistency algorithm of @xcite has an @xmath80 time complexity on constraints of arity @xmath2 and domains of size @xmath70 , whilst the domain consistency algorithm of @xcite for the @xmath2-ary constraint has @xmath81 time complexity .
_ showed that many global constraints like are also intractable to propagate @xcite .
more recently , samer and szeider have studied the parameterized complexity of the constraint @xcite .
szeider has also studied the complexity of symmetry in a propositional resolution calculus @xcite .
see chapter 10 in @xcite for more about symmetry of propositional systems .
we have argued that parameterized complexity is a useful tool with which to study symmetry breaking . in particular , we have shown that whilst it is intractable to break all symmetry completely , there are special types of symmetry like value symmetry and row and column symmetry which are more tractable to break . in these case ,
fixed - parameter tractability comes from natural parameters like the number of generators which tend to be small . in future
, we hope that insights provided by such analysis will inform the design of new search methods . for example
, we might build a propagator that propagates completely when the parameter is small , but only partially when it is large . in the longer term , we hope that other aspects of parameterized complexity like kernels will find application in the domain of symmetry breaking .
crawford , j. , ginsberg , m. , luks , g. , roy , a. : symmetry breaking predicates for search problems . in : proceedings of the 5th international conference on knowledge representation and reasoning , ( kr 96 ) .
( 1996 ) 148159 walsh , t. : symmetry within and between solutions . in zhang , b.t . ,
orgun , m.a .
pricai 2010 : trends in artificial intelligence , 11th pacific rim international conference on artificial intelligence , proceedings .
volume 6230 of lecture notes in computer science . , springer ( 2010 ) 1113 katsirelos , g. , walsh , t. : symmetries of symmetry breaking constraints . in coelho ,
h. , studer , r. , wooldridge , m. , eds . :
ecai 2010 - 19th european conference on artificial intelligence , proceedings .
volume 215 of frontiers in artificial intelligence and applications .
, ios press ( 2010 ) 861866 gent , i. , walsh , t. : : a benchmark library for constraints .
technical report , technical report apes-09 - 1999 ( 1999 ) a shorter version appears in the proceedings of the 5th international conference on principles and practices of constraint programming ( cp-99 ) .
pachet , f. , roy , p. : automatic generation of music programs . in jaffar , j. ,
ed . : proceedings of fifth international conference on principles and practice of constraint programming ( cp99 ) , springer ( 1999 ) 331345 bessiere , c. , hebrard , e. , hnich , b. , walsh , t. : the complexity of global constraints . in : proceedings of the 19th national conference on ai , association for advancement of artificial intelligence ( 2004 ) bessire , c. , hebrard , e. , hnich , b. , walsh , t. : the tractability of global constraints . in : 10th international conference on principles and practices of constraint programming ( cp-2004 ) .
volume 3258 of lecture notes in computer science . , springer ( 2004 ) 716720 bessire , c. , hebrard , e. , hnich , b. , kiziltan , z. , walsh , t. : filtering algorithms for the nvalue constraint . in : integration of ai and or techniques in constraint programming for combinatorial optimization problems , 2nd international conference ( cpaior-2005 ) .
( 2005 ) downey , r.g . , fellows , m.r . ,
stege , u. : parameterized complexity : a framework for systematically confronting computational intractability . in : contemporary trends in discrete mathematics : from dimacs and dimatia to the future .
volume 49 of dimacs series in discrete mathematics and theoretical computer science .
( 1999 ) 4999 walsh , t. : general symmetry breaking constraints . in benhamou , f. , ed .
: principles and practice of constraint programming - cp 2006 , 12th international conference , cp 2006 , nantes , france , september 25 - 29 , 2006 , proceedings .
volume 4204 of lecture notes in computer science . , springer ( 2006 ) katsirelos , g. , narodytska , n. , walsh , t. : combining symmetry breaking and global constraints . in oddi ,
a. , fages , f. , rossi , f. , eds . : recent advances in constraints , 13th annual ercim international workshop on constraint solving and constraint logic programming , csclp 2008 .
volume 5655 of lecture notes in computer science . , springer ( 2008 ) 8498 law , y.c . , lee , j. , walsh , t. , yip , j. : breaking symmetry of interchangeable variables and values . in : 13th international conference on principles and practices of constraint programming ( cp-2007 ) , springer - verlag ( 2007 ) hnich , b. , kiziltan , z. , walsh , t. : combining symmetry breaking with other constraints : lexicographic ordering with sums . in : proceedings of the 8th international symposium on the artificial intelligence and mathematics .
( 2004 ) frisch , a. , hnich , b. , kiziltan , z. , miguel , i. , walsh , t. : global constraints for lexicographic orderings . in : 8th international conference on principles and practices of constraint programming ( cp-2002 ) , springer ( 2002 ) puget , j.f .
: breaking all value symmetries in surjection problems . in van beek ,
: proceedings of 11th international conference on principles and practice of constraint programming ( cp2005 ) , springer ( 2005 ) bessire , c. , hebrard , e. , hnich , b. , kiziltan , z. , quimper , c.g . ,
walsh , t. : the parameterized complexity of global constraints . in fox , d. , gomes ,
c. , eds . : proceedings of the 23rd national conference on ai , association for advancement of artificial intelligence ( 2008 ) 235240 law , y. , lee , j. : global constraints for integer and set value precedence . in : proceedings of 10th international conference on principles and practice of constraint programming ( cp2004 ) , springer ( 2004 ) 362376 walsh , t. : symmetry breaking using value precedence . in brewka , g. , coradeschi , s. , perini , a. , traverso , p. , eds .
: ecai 2006 , 17th european conference on artificial intelligence , august 29 - september 1 , 2006 , riva del garda , italy , including prestigious applications of intelligent systems ( pais 2006 ) , proceedings .
volume 141 of frontiers in artificial intelligence and applications .
, ios press ( 2006 ) 168172 flener , p. , frisch , a. , hnich , b. , kiziltan , z. , miguel , i. , pearson , j. , walsh , t. : breaking row and column symmetry in matrix models . in : 8th international conference on principles and practices of constraint programming ( cp-2002 ) , springer ( 2002 ) flener , p. , frisch ,
a. , hnich , b. , kiziltan , z. , miguel , i. , walsh , t. : matrix modelling .
technical report apes-36 - 2001 , apes group ( 2001 ) presented at formul01 workshop on modelling and problem formulation , cp2001 post - conference workshop .
flener , p. , frisch , a. , hnich , b. , kiziltan , z. , miguel , i. , walsh , t. : matrix modelling : exploiting common patterns in constraint programming . in : proceedings of the international workshop on reformulating constraint satisfaction problems .
( 2002 ) held alongside cp-2002 .
huczynska , s. , mckay , p. , miguel , i. , nightingale , p. : modelling equidistant frequency permutation arrays : an application of constraints to mathematics . in gent , i. , ed .
: principles and practice of constraint programming - cp 2009 , 15th international conference , cp 2009 , lisbon , portugal , september 20 - 24 , 2009 , proceedings .
volume 5732 of lecture notes in computer science . , springer ( 2009 ) 5064 katsirelos , g. , narodytska , n. , walsh , t. : on the complexity and completeness of static constraints for breaking row and column symmetry . in cohen ,
d. , ed . : principles and practice of constraint programming - cp 2010 , 16th international conference , proceedings .
lecture notes in computer science , springer ( 2010 ) flener , p. , frisch , a. , hnich , b. , kiziltan , z. , miguel , i. , pearson , j. , walsh , t. : symmetry in matrix models .
technical report apes-30 - 2001 , apes group ( 2001 ) presented at symcon01 ( symmetry in constraints ) , cp2001 post - conference workshop .
bessiere , c. , rgin , j. : arc consistency for general constraint networks : preliminary results . in : proceedings of the 15th international conference on ai , international joint conference on artificial intelligence ( 1997 ) 398404 | symmetry is a common feature of many combinatorial problems .
unfortunately eliminating all symmetry from a problem is often computationally intractable .
this paper argues that recent parameterized complexity results provide insight into that intractability and help identify special cases in which symmetry can be dealt with more tractably . |
two distinct glasses have been predicted and identified in hard spheres with short range attractions : a repulsion driven glass at high density , and an attraction driven one , formed at low temperatures ( or high attraction strength ) @xcite . whereas the former one is caused by the steric hindrance of the particle cores and the so - called _ cage effect _ , the latter forms due to the bonding between particles .
this system is realized experimentally by a colloid - polymer mixture , where the effect of the polymers is to induce an effective attraction between the colloids @xcite .
both glasses have been indeed identified , although the attractive one , which at low concentrations is termed ` gel ' , often competes with ( and inhibits ) liquid - gas phase separation @xcite . dynamical heterogeneities ( dh )
have been found in the proximity of repulsion driven glass transitions by computer simulations , i.e. in lennard - jones mixtures @xcite , or hard spheres @xcite . in these cases , while the system is structurally liquid - like ( homogeneous ) , a population of particles of increased mobility is observed .
as the glass transition is approached from the fluid side , the heterogeneities become more pronounced , but decrease again deeper in the glass @xcite . the role of these dynamical heterogeneities in the glass transition
is as yet unclear ; whereas mode coupling theory focusses on averaged quantities and neglects them @xcite , the so - called facilitated dynamics theories give dh the central role for their description of the glass transition @xcite . in recent works
, it has been shown that dh can be found also in attractive glasses , by studying the distribution of particle displacements in the system @xcite .
in fluid states close to the transition two populations of particles were found , separated by a minimum in the displacement distribution .
a similar feature has been found also in repulsive glasses , which could imply a common origin @xcite .
however , the low density of the attractive glass , as low as @xmath0 , causes structural heterogeneities as well ; the system forms a percolating cluster of high density material , leaving voids with no particles .
a correlation between structural and dynamical heterogeneities is thus possible , showing that ` fast ' particles are in the surface of the cluster , whereas the ` slow ' ones are mostly trapped in the inner parts of it @xcite . in this work
, we study the dh inside the non - ergodic region , for two different states , and compare them with those of the equilibrium systems .
only one population of particles can be identified from the distribution of particle displacements , and the distribution is narrower for the state with stronger attractions . moreover , as the systems age , they become more and more homogeneous , from the point of view of the dynamics .
both results indicate that the strongest dh are obtained in the fluid side of the phase diagram , close to the glass transition . as a side remark
, it must be noted that the structural heterogeneities mentioned above persist in the out - of - equilibrium systems , and thus are not the sole origin of the dh in attractive glasses .
we have performed computer simulations of a system composed of @xmath1 soft core ( @xmath2 ) particles with attractive interactions given by the asakura - oosawa ( ao ) potential @xcite .
it models a mixture of colloids with non - adsorbing polymers , and the range of attraction is set by the polymer size . in order to guarantee full access to the whole parameter space , phase separations
have been inhibited .
crystallization is avoided by polydispersity ( flat distribution , @xmath3 width ) , and liquid - gas demixing by a repulsive barrier extending to two mean diameters .
further details of the interaction potential can be found in previous works @xcite .
length is measured in units of the average radius , @xmath4 , and time in units of @xmath5 , where the thermal velocity , @xmath6 , was set to @xmath7 .
equations of motion were integrated using the velocity - verlet algorithm , in the canonical ensemble ( constant ntv ) , to mimic the colloidal dynamics , with a time step equal to @xmath8 .
every @xmath9 time steps , the velocity of the particles was re - scaled to assure constant temperature .
the range of the attraction is set to @xmath10 .
density is reported as volume fraction , @xmath11 , with @xmath12 the number density , and the attraction strength is measured in units of the polymer volume fraction @xmath13 ( at contact the ao attraction strength is @xmath14 ) .
the attractive glass transition for this system has been studied previously @xcite .
an mct analysis of the results ( diffusion coefficient , time scale and viscosity ) yields a transition point at @xmath15 for the colloid density @xmath16 . for the study of aging here ,
the systems were equilibrated without attraction ( @xmath17 ) at @xmath16 , and then instantaneously _ quenched _ to the desired @xmath13 at zero time , @xmath18 .
two attraction strengths have been studied , @xmath19 and @xmath20 , lying beyond the nonergodicity transition . in both cases , @xmath21 independent simulations have been performed , and the evolution of the system has been followed as a function of the time elapsed since the quench , called waiting time , @xmath22 .
correlation functions thus depend on two times : @xmath23 and @xmath24 .
, @xmath25 , @xmath26 , @xmath27 , ... , and for different polymer fractions , @xmath13 , as labeled .
note that as @xmath13 increases , two populations of particles with different mobilities appear in the system .
the glass transition is located at @xmath15 , estimated from mct analysis ( power law fittings ) @xcite.,width=491 ] in fluid states close to the attractive glass , increasing dh have been found , the stronger the attraction @xcite .
two populations of particles are observed to appear as the attraction is increased , one of mobile particles and another one of quasi - immobile particles ( see fig . [ fig1 ] ) @xcite .
( the minimum in the distribution of squared displacements allows for an unambiguous identification of almost every particle . )
these two populations are structurally well differentiated : the particles in the ` skin ' of the gel , with a small number of bonds , are mobile , and the particles in the inner parts of the gel , the ` skeleton ' , are quasi - immobile .
the populations were observed to be stable for long times , although particles can change from one to the other .
it is thus an equilibrium feature of the system , partly induced by the structural heterogeneity . and
@xmath20 , as labeled , for different waiting times ; from left to right , @xmath28 , @xmath29 , @xmath30 , @xmath31 , @xmath32 , @xmath33 .
the dashed line shows the msd for hard spheres at this density.,width=491 ] at a long time , @xmath34 .
( qualitatively similar snapshots are obtained for @xmath20.),width=415 ] all of the states presented in fig .
[ fig1 ] are , nevertheless , fluid states .
the structural properties do not depend on time , nor the dynamical ones on the initial time , and the correlation functions averaged throughout the whole system decay to zero . by further increasing the attraction strength , however , the system falls out of equilibrium and shows aging . in fig .
[ msd ] , the mean squared displacement ( msd ) is presented for the two quenches studied here , @xmath19 and @xmath20 ; different lines are the msd for different waiting times , @xmath22 . as @xmath22 increases
, bonds are formed between particles , as shown in fig .
[ nneigh ] for both states , which hinders the motion of the particles , causing the dynamical arrest . in the msd
, a plateau develops at short distances , signaling the localization length , clearer in the case of the higher @xmath35 , where the bonds are stronger and the _ cage _ ( network of bonds ) of the particles is tighter , as shown by the shorter localization length ( implying higher non - ergodicity parameters in the density correlation function ) .
the mean number of bonds per particle is , however , lower .
also , the aging is more dramatic and at long waiting times the msd hardly reaches the range of the attraction , contrary to previous works where the plateau in msd , or the density correlation function , was not observed @xcite . here
we will concentrate on the dh , and a full analysis of the aging will be presented elsewhere @xcite .
and @xmath20 and three waiting times , as labeled .
the different lines represent the distribution at @xmath36 , @xmath25 , @xmath26 , @xmath27 , @xmath37 , @xmath38 , @xmath39 , @xmath40 and @xmath41 , from left to right , respectively.,width=415 ] , as a function of its mean number of neighbours during this time , for @xmath19 and @xmath42 .
( only ten simulations are considered , i.e. @xmath43 particles ) .
note the vertical bands at integer numbers of neighbours ( especially for @xmath44 ) , due to particles that have not changed their neighbours.,width=491 ] a snapshot of the system with @xmath19 is presented in fig .
the system forms an intricate structure , with voids and tunnels , similar to the fluid states presented above .
therefore , the skeleton and skin " picture presented above , with two different populations of particles , could still be applicable . in fig .
[ distributions ] we present the distribution of squared displacements at different times for both quenches .
strikingly , the distribution is monomodal in all cases , and the peak evolves to larger displacements as time proceeds .
the strong dh observed in the fluid states , have , therefore , disappeared in the glass side .
moreover , the deeper quench , @xmath20 , shows narrower distributions at all waiting times than the quench at @xmath19 , indicating that the dynamics is more homogeneous the deeper the state is in the glass , in agreement with findings for lennard - jones ( repulsive ) glasses @xcite . at @xmath19 ,
however , a tail in the distribution to long distances can be observed , caused by some particles that can travel long distances ; the number of which decreases with @xmath22 .
this feature is reminiscent of the population of fast particles observed in the fluid ( fig .
[ fig1 ] ) .
the origin of this population is studied in fig .
[ nneigh - msd ] , where the squared displacement of every particle for a given time is correlated with the mean number of neighbours of the particle during this time .
the plot shows that indeed the fastest particles in the system have less neighbours on average , whereas the particles with many neighbours , move very little .
therefore , the simple picture of skeleton and skin " , can still be applied for the attractive glass close to the transition .
the number of these fast particles is , nevertheless , decreasing with waiting time , as observed by comparing similar times @xmath45 for different waiting times @xmath22 in fig .
[ distributions ] .
accordingly , the mean number of bonds per particle increases ( see fig . [ nneigh ] ) , implying a compaction in the system , and thus having fewer particles in the skin , and even those , more tightly trapped .
however , we can not state whether the population of fast particles will reach a steady state or if it will vanish eventually at very long waiting times . at @xmath20 , on the other hand , the tail of fast particles is absent in the distribution of squared displacements , although similar structural heterogeneities are observed .
we have shown that the strong dh found in fluid states close to the attractive glass transition in colloids with short range attractions decrease again deep into the non - ergodic region .
the distribution of squared displacements is monomodal and no particles with increased mobility are observed ( @xmath20 ) .
however , close to the glass transition , in the glass side , aging is slower and some dh can still be detected : a tail in the distribution of squared displacements indicates fast particles , that can be identified with particles in the outer parts of the particle network .
this feature is reminiscent of the population of fast particles found in the fluid states close to the transition .
however , the stability of this population of fast particles in the glass can not be established , but should not be present in a truly arrested glassy state .
the results presented here agree with those of repulsive glasses , contrary to other comparions between attractive and repulsive glasses @xcite .
a.m.p . acknowledges financial support by the dgcyt ( project mat2003 - 03051-co3 - 01 ) .
this work was funded in part ( m.e.c . ) by epsrc gr / s10377 .
a.m.p . and m.f .
were partially funded by ai - daad project no .
ha2004 - 0022 .
pham , a.m. puertas , j. bergenholtz , s.u .
egelhaaf , a. moussaid , p.n .
pusey , a.b .
schofield , m.e .
cates , m. fuchs , w.c.k .
poon , science , * 296 * , 104 ( 2002 ) .
f. sciortino , nature materials , * 1 * 145 ( 2002 ) . c. n. likos , 2001 _ phys .
rep . _ * 348 * 267 ; s. asakura and f. oosawa , j. chem . phys . *
22 * 1255 ( 1954 ) .
pham , s.u .
egelhaaf , p.n .
poon , phys .
e * 69 * 011503 ( 2004 ) .
w. kob , c. donati , s.j .
plimpton , p.h .
poole , s.c .
glotzer ; phys .
e * 79 * 2827 ( 1997 ) ; c. donati , s.c .
glotzer , p.h .
poole ; phys .
lett . * 82 * 5064 ( 1999 ) .
e. flenner , g. szamel , phys .
e * 72 * 011205 ( 2005 ) .
e.r . weeks , j.c .
crocker , a.c .
levitt , a. schofield , d.a .
weitz ; science * 287 * , 627 ( 200 ) .
b. doliwa , a. heuer , phys .
lett * 80 * 4915 ( 1998 ) ; phys .
e * 61 * , 6898 ( 2000 ) .
k. vollmayr - lee , w. kob , k. binder , a. zippelius , j. chem .
phys * 116 * 5158 ( 2002 ) k. vollmayr - lee , a. zippelius , cond - matt/0507107 .
w. gtze in _ liquids , freezing and glass transition _ ed .
hansen , d. levesque and j. zinn - justin , amsterdam , north - holland ( 1991 ) .
l. berthier , j.p .
garrahan , phys .
e * 68 * 041201 ( 2003 ) ; j.p .
garrahan , d. chandler , phys .
lett . * 89 * 035704 ( 2002 ) .
a.m. puertas , m. fuchs , m.e .
cates , j. chem . phys . * 121 * 2813 ( 2004 ) a.m. puertas , m. fuchs , m.e .
cates , j. phys .
b * 109 * 6666 ( 2005 ) ; phys .
e * 67 * , 031406 ( 2003 ) .
reichman , e. rabani , p.l .
geissler , j. phys .
b * 109 * 14654 ( 2005 ) .
a.m. puertas , e. zaccarelli , f. sciortino , j. phys .
* 17 * l271 ( 2005 ) .
e. zaccarelli , g. foffi , f. sciortino , and p. tartaglia , phys . rev .
lett . * 91 * 108301 ( 2003 ) a.m. puertas , m. fuchs , m.e .
cates . in preparation | the dynamical heterogeneities ( dh ) in non - ergodic states of an attractive colloidal glass are studied , as a function of the waiting time . whereas the fluid states close to vitrification showed strong dh , the distribution of squared displacements of the glassy states studied here only present a tail of particles with increased mobility for the lower attraction strength at short waiting times .
these particles are in the surface of the percolating cluster that comprises all of the particles , reminiscent of the fastest particles in the fluid .
the quench deeper into the attractive glass is dynamically more homogeneous , in agreement with repulsive glasses ( i.e. lennard - jones glass ) . ,
, 64.70.pf , 82.70.dd , 61.20.lc |
lm , sl and ac acknowledge the italian miur for financial support .
pu wishes to thank a.j . dean and j.b .
stephen for useful scientific discussions . | using recent polarimetric observations of the crab nebula in the hard x - ray band by integral , we show that the absence of vacuum birefringence effects constrains @xmath0 lorentz violation in qed to the level @xmath1 at @xmath2 cl , tightening by more than three orders of magnitude previous constraints .
we show that planned x - ray polarimeters have the potential to probe @xmath3 by detecting polarization in active galaxies at red - shift @xmath4 .
experimental constraints on the parameters quantifying lorentz invariance violation ( lv ) are of fundamental importance .
because the lowest order corrections predicted in the photon dispersion relation imply the vacuum is birefringent , observations of polarized photons from distant astronomical sources provide very promising tests . in this letter
we exploit the recently discovered linear polarization of hard x - rays from the crab nebula ( cn ) @xcite .
these observations show a remarkably high degree of linear polarization ( @xmath5 ) and very close alignment of the polarization vector with both the optical polarization vector and the projection on the sky of the spin axis of the central neutron star . the high degree of polarization together with the lack of detectable rotation of the polarization vector of these @xmath6kev photons whilst propagating over the intervening @xmath7 cm enables us to tighten existing constraints by three orders of magnitude .
recent years have witnessed a growing interest in the possible high energy violations of local lorentz invariance as well as a flourishing of observational tests .
indeed , specific hints of lv arose from various approaches to quantum gravity @xcite .
however , most tests require a well established theoretical framework to calculate reaction rates and describe the particle dynamics . here , we work within the framework of effective field theory with non - renormalizable , mass dimension 5 lv operators ( see @xcite and references therein ) restricted to qed , for which the most general dispersion relations for photons and electrons are @xmath8 where ( [ eq : disp_rel_phot ] ) refers to photons and ( [ eq : disp_rel_ferm ] ) to fermions . @xcite . ]
we assume @xmath9 to be comparable to the planck mass @xmath10 .
the constants @xmath11 and @xmath12 indicate the strength of the lv .
the @xmath13 signs denote right and left circular polarization in ( [ eq : disp_rel_phot ] ) , and positive and negative helicity states of the fermion in ( [ eq : disp_rel_ferm ] ) .
equation ( [ eq : disp_rel_phot ] ) implies that the direction of polarization rotates during propagation due to the different velocities of the right- and left - handed circular polarizations , @xmath14 .
this effect is known as vacuum birefringence ( vb ) .
although it may seem hopeless to search directly for effects suppressed by the planck energy scale , even tiny corrections can be magnified to measurable ones when dealing with high energies , long distances of signal propagation or peculiar reactions ( see , e.g. , @xcite ) .
recently @xmath15 have been constrained to have a magnitude less than @xmath16 at 95% confidence level ( cl ) by a detailed analysis of the synchrotron component of the cn broadband spectrum @xcite , while the constraint @xmath17 has been obtained by @xcite considering the absence of vb effects during the propagation of optical / uv polarized light from gamma - ray bursts ( grb ) , were claimed in @xcite from grb 021206 observations @xcite ; however the result was later contested @xcite . ] .
there are also preliminary indications , based on an analysis of the photon fraction in ultra - high - energy cosmic rays , that these coefficients might be less than @xmath18 , though nothing conclusive can be claimed yet @xcite . in this work
we tighten the current constraints on @xmath19 suppressed lv by about three orders of magnitude for photons , by considering the limits on vb effects implied by the recently detected @xcite polarized hard x - rays from the cn .
firstly , we set such constraints following the arguments by @xcite , an approach robust against systematic uncertainties related to astrophysical modeling .
we then infer tighter limits that exploit and rely on modeling of the crab nebula and pulsar . finally , we consider the constraints which future x - ray polarization measurements of extragalactic objects , e.g. active galactic nuclei ( agn ) will allow .
this is of particular interest in the light of current experimental efforts to build x - ray polarimeters @xcite . during propagation over a distance @xmath20 ,
the ( cosmological ) distance is given by @xmath21 , which includes a @xmath22 factor in the integrand to take into account the red - shift acting on the photon energies .
as usual , @xmath23 is the present value of the hubble parameter , while @xmath24 and @xmath25 represent the density fractions of cosmological constant and matter in the universe , respectively . ] , the polarization vector of a linearly polarized plane wave with momentum @xmath26 rotates through an angle @xcite , @xmath27 observations of polarized light from a distant source can constrain @xmath28 in two ways , depending on the amount of available information on both the observational and the theoretical ( i.e. source modeling ) side : 1 . since detectors have a finite energy bandwidth , eq .
( [ eq : theta ] ) is never probed in real situations .
rather , if some net amount of polarization is measured in the band @xmath29 , an order - of - magnitude constraint arises from the fact that if the angle of polarization rotation ( [ eq : theta ] ) were to differ by more than @xmath30 over this band , the detected polarization would fluctuate sufficiently for the net signal polarization to be suppressed @xcite . from ( [ eq : theta ] ) , this constraint is @xmath31 this just requires that any intrinsic polarization ( at source ) is not completely washed out during signal propagation .
it thus relies on the mere detection of a polarized signal , without considering the observed polarization degree .
a more refined limit can be obtained by calculating the maximum observable polarization degree , given the maximum intrinsic value @xcite : @xmath32 where @xmath33 is the maximum intrinsic degree of polarization , @xmath34 is defined in eq .
( [ eq : theta ] ) and the average is weighted over the source spectrum and instrumental efficiency , represented by the normalized weight function @xmath35 @xcite .
conservatively , one can set @xmath36 , but a lower value can sometimes be justified on the basis of source modeling .
using , one can then cast a constraint by requiring @xmath37 to exceed the observed value .
2 . suppose that polarized light measured in a certain energy band has a position angle @xmath38 with respect to a fixed direction .
at fixed energy , the polarization vector rotates by the angle ( [ eq : theta ] ) ; if the position angle is measured by averaging over a certain energy range , the final net rotation @xmath39 is given by the superposition of the polarization vectors of all the photons in that range : @xmath40 where @xmath34 is given by ( [ eq : theta ] ) . if the position angle at emission @xmath41 in the same energy band is known from a model of the emitting source , a constraint can be set by imposing @xmath42 although this limit is tighter than that obtained from the previous methods , it clearly hinges on assumptions about the nature of the source , which may introduce significant uncertainties . in the case of the crab nebula ,
a @xmath43% degree of linear polarization in the @xmath44 band has recently been measured by the integral mission @xcite .
this measurement uses all photons within the spi instrument energy band .
however the convolution of the instrumental sensitivity to polarization with the detected number counts as a function of energy , @xmath35 , is maximized and approximately constant within a narrower energy band ( 150 to 300 kev ) and falls steeply outside this range @xcite .
for this reason we shall , conservatively , assume that most polarized photons are concentrated in this band .
given @xmath45 , @xmath46 and @xmath47 , eq .
( [ eq : decrease_pol ] ) leads to the order - of - magnitude estimate @xmath48 .
a more accurate limit follows from ( [ eq : pol ] ) . in the case of the cn
there is a robust understanding that photons in the range of interest are produced via the synchrotron proces , for which the maximum degree of intrinsic linear polarization is about @xmath49 ( see e.g. @xcite ) .
figure [ fig : casea ] illustrates the dependence of @xmath50 on @xmath11 for the distance of the cn and for @xmath51 . the requirement @xmath52 ( taking account of a @xmath2 offset from the best fit value @xmath53 ) leads to the constraint ( at 99% cl ) @xmath54 on @xmath11 for the distance of the cn and photons in the 150300 kev range , for a constant @xmath35 . ]
it is interesting to notice that x - ray polarization measurements of the cn already available in 1978 @xcite , set a constraint @xmath55 , only one order of magnitude less stringent than that reported in @xcite .
constraint ( [ eq : constraint - degree ] ) can be tightened by exploiting the current astrophysical understanding of the source .
the cn is a cloud of relativistic particles and fields powered by a rapidly rotating , strongly magnetized neutron star . both the _ hubble space telescope _ and the _ chandra _ x - ray satellite have imaged the system , revealing a jet and torus that clearly identify the neutron star rotation axis @xcite . the projection of this axis on the sky lies at a position angle of @xmath56 ( measured from north in anti - clockwise ) .
the neutron star itself emits pulsed radiation at its rotation frequency of 30 hz . in the optical band
these pulses are superimposed on a fainter steady component with a linear polarization degree of 30% and direction precisely aligned with that of the rotation axis @xcite . the direction of polarization measured by integral - spi in the @xmath57-rays is @xmath58 ( @xmath59 error ) from the north , thus also closely aligned with the jet direction and remarkably consistent with the optical observations .
this compelling ( theoretical and observational ) evidence allows us to use eq .
( [ eq : constraint - caseb ] ) . conservatively assuming @xmath60 ( i.e. @xmath2 from @xmath41 , 99% cl ) , this translates into the limit @xmath61 and @xmath62 for a @xmath63 deviation ( 95% cl ) .
figure [ fig : caseb ] shows @xmath64 as function of @xmath11 .
the left hand panel reports the global dependence ( the spikes correspond to rotations by @xmath65 ) , while the right hand panel focuses on the interesting range of values . ) rules out the possibility that the polarization angle is close to the expected one after rotating by some multiple of @xmath66 ( the polarization angle is defined on the interval @xmath67 $ ] ) . ]
the constraints presented in ( [ eq : constraint - degree ] ) and ( [ eq : constraint - serious - crab ] ) are remarkably strong .
although based on a cumulative effect , they are achieved using a local ( galactic ) object .
the reason lies , on the one hand , in the quadratic dependence of @xmath34 on the photon energy , in constrast with the linear gain given by distance ( see e.g. eq .
( [ eq : theta ] ) ) . on the other hand ,
the robust theoretical understanding of the cn has enabled us to strengthen the constraints significantly .
further improvements on lv constraints via birefringenge are expected thanks to the forthcoming high - energy polarimeters , such as xeus @xcite , pogolite @xcite , polar - x @xcite and gamma ray imager @xcite which will provide an unprecedented sensitivity , sufficient to detect polarized light at a few % levels also in extragalactic sources .
the lv limits will be optimized by balancing between source distance and observational energy range depending on the detector sensitivity .
this is illustrated in fig .
[ fig : plot ] , where the strength of the possible constraints ( cast with the first , most general method described above ) is plotted versus the distance of sources ( in red - shift @xmath68 ) and for different energy bands ( medium x- and @xmath57-rays ) .
remarkably , constraints of order @xmath69 could be placed if some polarized distant sources ( @xmath70 ) will be observed by such instruments at 1 mev . |
in recent years , higher - order operators have become the object of intense study in the search for possible effects of lorentz invariance violation @xcite .
these planck - mass suppressed higher - order operators allows to describe new physics beyond those obtainable from renormalizable operators , that is , operators with mass dimension four or less @xcite .
for example , the higher - order effective theory may involve additional degrees of freedom associated to ultra - high energies which do not converge perturbatively to the normal ones when taking the limit of the dimensionless parameters in the effective terms to zero .
lee and wick studied these exotic modes in the context of negative metric theories @xcite and in spite of the ghost states that appear , they showed that unitarity can be preserved by demanding all stable particles to be positive norm states @xcite . here
we check perturbative unitarity in a qed consisting of higher - order myers and pospelov photons @xcite and standard fermions .
the myers - pospelov lagrangian density for photons is given by @xmath0 where @xmath1 is a four - vector defining a preferred reference frame , @xmath2 is the planck mass and @xmath3 is a dimensionless parameter
. we can always select a real basis of four - vectors @xmath4 to be orthonormal and to satisfy the properties described in ref . . in analogy with the left and right handed polarizations of usual electrodynamics
we can switch to a basis of complex four - vectors @xmath5 and define the orthogonal projectors @xmath6 as @xmath7 where @xmath8 . to derive the dispersion relation
we can expand the gauge field in term of this complex basis and replace in the equations of motion to arrive at @xmath9 in agreement with the work in ref . .
here we check perturbative unitarity in the process of electron - positron scattering @xmath10 . for this
we use the optical theorem which relates the imaginary part of the forward scattering amplitude @xmath11 with the total cross section as @xmath12 where the sum runs over all intermediate physical states . considering the qed extension model
the amplitudes that contribute to the @xmath13-matrix are the direct amplitude @xmath14 and the exchange amplitude @xmath15 where @xmath16 , @xmath17 and @xmath18 , @xmath19 and where @xmath20 is the usual fermionic normalization constant .
let us start with the left hand side of the unitarity condition ( [ unitarity ] ) .
a similar calculation has been given in the minimal sector of the standard - model extension , see ref . .
to simplify we will consider the lightlike case where we have a ghost state with frequencies @xmath21 and two photons with frequencies @xmath22 , see ref . , and the propagator @xmath23 where and we have included the @xmath24 prescription .
we are interested in the imaginary part of the forward scattering amplitude , therefore let us set @xmath25 and @xmath26 .
moreover , we can see that the direct process does not contribute since the virtual photon can never be on shell for non - zero external momenta , hence @xmath27=0 $ ] .
let us find the contribution of the exchange process and substitute the propagator ( [ propagator ] ) in ( [ amplitud2 ] ) @xmath28 because only the poles can contribute to the imaginary part and due to energy conservation encoded in @xmath29 , we have that only the positive poles of the virtual photon have a chance to contribute .
we can discard the ghost contribution since its energy @xmath30 lies beyond the region of validity of the effective theory .
that is , the external fermions will always satisfy the condition @xmath31 .
hence , we have @xmath32\nonumber \\&= & - e^2 \int d k^0 \int \frac{d^3 \vec k}{(2\pi)^3}\delta^4(p_1+p_2-k ) v^{\mu } v^ { * \nu } \sum_{\lambda}\frac { p^{\lambda}_{\mu \nu } \delta ( k_0-\omega_1^{\lambda } ) } { 2g\lambda ( k_0-\omega_0^{\lambda } ) ( k_0- \omega_2^{\lambda } ) } , \nonumber \\ & = & e^2\int \frac{d^3 k}{(2\pi)^3 } \delta^4(p_1+p_2-k ) v^{\mu } v^ { * \nu } \sum_{\lambda } \frac { \varepsilon_{\mu}^{\lambda } \varepsilon_{\nu}^{*\lambda } } { 2g\lambda ( \omega_1^{\lambda}-\omega_0^{\lambda } ) ( \omega_1^{\lambda}-\omega_2^{\lambda } ) } , \nonumber \\ & = & \int \frac{d^3 k}{(2\pi)^3 } \delta^2(p_1+p_2-k ) \sum_{\lambda } \left| \mathcal m_{\lambda } \right|^2,\end{aligned}\ ] ] where we have used the notation @xmath33 for the physical process @xmath34 and we have introduced the normalization constant @xmath35 . finally we have @xmath36= \int \frac{d^3 k}{(2\pi)^3 } \delta^2(p_1+p_2-k ) \left| \mathcal m_{\rm phys } \right|^2,\end{aligned}\ ] ] and therefore the unitarity condition is satisfied in this scattering process .
with an explicit calculation we have verified that the unitarity condition in the process of electron - positron scattering at tree level order is satisfied .
a next step is to verify unitarity to order @xmath37 that will require to analyze more diagrams .
some of them contain loops where the ghosts can appear off - shell , thus , introducing an extra difficulty .
checking the unitarity condition to these order will give us a robust support in order to make physical predictions in the theory .
this work was supported by the direccin de investigacin de la universidad del bo - bo grant 123809 3/r and fapei .
myers and m. pospelov , phys .
lett . * 90 * ( 2003 ) 211601 ; p.a .
bolokhov and m. pospelov , phys .
rev . * d 77 * , 025022 ( 2008 ) .
kostelecky and m. mewes , phys .
d * 80 * , 015020 ( 2009 ) ; v.a .
kostelecky and m. mewes , phys .
d * 85 * , 096005 ( 2012 ) .
t. mariz , phys .
d * 83 * , 045018 ( 2011 ) ; t. mariz , j.r .
nascimento and a.y .
petrov , phys .
d * 85 * , 125003 ( 2012 ) ; r. casana , m.m .
ferreira , e.o .
silva , e. passos and f.e.p.d .
santos , phys .
d * 87 * , 047701 ( 2013 ) ; c.m .
reyes , l.f .
urrutia , j.d .
vergara , phys . rev . *
d78 * , 125011 ( 2008 ) .
d. colladay and v.a .
kostelecky , phys .
d * 55 * , 6760 ( 1997 ) ; d. colladay and v.a .
kostelecky , phys .
d * 58 * , 116002 ( 1998 ) .
kostelecky and n. russell , rev .
phys . * 83 * , 11 ( 2011 ) .
lee and g.c . wick , nucl .
b * 9 * , 209 ( 1969 ) ; t.d .
lee , g.c . wick , phys .
* d2 * , 1033 - 1048 ( 1970 ) .
cutkosky , p.v .
landshoff , d.i .
olive , j.c .
polkinghorne , nucl .
* b12 * , 281 - 300 ( 1969 ) .
reyes , phys .
d * 87 * , 125028 ( 2013 ) .
reyes , phys .
d * 82 * , 125036 ( 2010 ) .
m. schreck , phys .
d * 86 * , 065038 ( 2012 ) ; f.r .
klinkhamer and m. schreck , nucl .
b * 848 * , 90 ( 2011 ) . | the unitarity of a lorentz - invariance violating qed model with higher - order myers and pospelov photons coupled to standard fermions is studied .
as expected , we find ghost states associated to the higher - order terms that may lead to the loss of unitarity .
an explicit calculation to check perturbative unitarity in the process of electron - positron scattering is performed and it is found to be possible to be preserved . |
the accreted stellar halo ( ash ) of a galaxy represents a record of the accretion history of the galaxy itself .
its assembly is determined by a large number of free parameters , including the structural properties of each accreted satellite ( virial mass , concentration , stellar content , morphology ) , the orbital properties of each accretion event ( energy and angular momentum at infall ) , the structural properties of the host itself during accretion .
this implies a significant degree of stochasticity , as shown by the observed halo - to - halo scatter ( e.g. , van dokkum et al .
2014 ) and by the dichotomy between the ` broken ' and sharply declining density profile of the stellar halo of the milky way ( e.g. , deason et al .
2013 ) and the more extended halo of andromeda , whose density profile is well described by a single power - law ( e.g. , gilbert et al .
2012 , ibata et al .
in amorisco ( 2015 ) i have isolated the main ingredients that shape the contribution of each accreted satellite to the ash .
i adopted a simplified approach and assumed that the contributing dwarfs are dark matter dominated ( as expected for the case of an @xmath1 host ) and ignored the gravitational influence of the host s disk .
combined with the halo mass - concentration relation ( e.g. , gao et al .
2008 , ludlow et al .
2014 ) , this reduces the structural properties of each minor merger to two dimensionless parameters .
two additional parameters characterise the orbital properties of the satellite at accretion ( e.g. , benson 2005 , jiang et al .
the locus of this parameter space that is relevant to a @xmath0cdm cosmology is explored with a library of minor merger n - body simulations , in which stars are assigned to the most bound 5@xmath2 of the satellite s particles , using a standard particle tagging technique ( e.g. , bullock & johnston 2005 , cooper et al .
this study shows that dynamical friction is a major player in shaping stellar deposition , allowing only the most massive ( and/or concentrated ) accreted satellites to deposit their stars in the innermost regions of the host .
orbital radialisation by dynamical friction causes the stellar populations deposited by such most massive accretion events ( virial satellite - to - host mass ratio at accretion @xmath3 ) to bear little memory of the details of the orbital properties of the progenitor at infall .
galaxies sharing a final virial mass of @xmath4 . panels
b ) and c ) : an example for the procedure of assembly of the accreted stellar halo using the combination of a given accretion history and of a library of minor merger simulations ; each contribution to the halo is color - coded by accretion redshift .
panel c ) : the spherically averaged density profiles of the 150 accreted stellar haloes corresponding to the accretion histories of panel a ) , color - coding refers to the local mass - weighed virial satellite - to - host mass .
panel e ) : the correlation between the total accreted stellar mass in the halo and the number of main accretion events ( see text for details ) . ]
i use the library just described to assemble 500 ashs , for galaxies that share a virial mass of @xmath5 .
i use monte carlo generated accretion histories ( fakhouri et al .
2010 ) , 150 of which are displayed in _ panel a ) _ of fig . 1 , color - coded by the total accreted stellar mass .
satellite stellar masses are assigned based on a redshift - independent abundance matching relation ( garrison - kimmel et al .
2014 , 0.3 dex scatter ) .
_ panels b ) _ and _ c ) _ exemplify the assembly of an individual ash : each accreted and disrupted satellite ( color - coded by its accretion redshift ) is associated to the spherically averaged density profile of the stars it deposits in the halo .
these are retrieved from the library using the relevant time - interval between accretion and @xmath6 , and are re - scaled to physical quantities according to the dimensional properties of the merger at hand .
_ panel d ) _ displays the spherically averaged density profiles of 150 ashs built in this manner . at each radius , the halo - to - halo scatter approaches 2 dex , and increases at @xmath7 , together with an increasing amount of not fully phase - mixed substructure from recent accretion events .
although with a significant scatter , on average , ashs share a logarithmic density slope @xmath8 within 20 kpc , and become steeper with radius , as shown by the dashed guiding lines .
the details of this steepening are highly variable : some profiles have marked and sharp breaks , others ` roll ' gently towards steeper and steeper slopes , other remain comparatively shallower .
the radii where such transitions take place are equally variable . the color - coding in _
panel d ) _ indicates the local mass - weighted satellite - to - host virial mass ratio . on average , the innermost regions of the ash
are contributed by satellites that have larger virial mass ratios at accretion .
this gradient has been observed in cosmological hydrodynamical simulations ( rodriguez - gomez et al .
2016 ) and i conclude is a direct consequence of dynamical friction ( e.g. , amorisco 2015 ) .
color - coding in _
panel d ) _ reveals that the local mean virial mass ratio also correlates positively with the local density .
_ panel e ) _ confirms this link by showing a scatter plot of the total accreted stellar mass of the ash against the ` number of main accretion events ' @xmath9 ( i.e. the ratio between the total accreted stellar mass of the ash and the mean stellar mass of the contributing satellites , mass - weighted by stellar mass itself ) . the most massive ashs result from the accretion of just one / two particularly massive satellites , which dominate the ex - situ mass .
although this technique represents a highly simplified approach , it allows for an efficient exploration of the significant stochasticity of ashs .
physical ingredients that are neglected here ( the host s stellar disk , the morphologies of the accreted dwarfs , any post - accretion star formation etc . ) will result in even increased variability .
detailed analysis of a large sample of ashs concentrating on the systematic correlations that connect density profile and accretion history is the subject of a forthcoming work ( amorisco 2016 ) . | i use a library of controlled minor merger n - body simulations , a particle tagging technique and monte carlo generated @xmath0cdm accretion histories to study the highly stochastic process of stellar deposition onto the accreted stellar halos ( ashs ) of @xmath1 galaxies .
i explore the main physical mechanisms that drive the connection between the accretion history and the density profile of the ash .
i find that : i ) through dynamical friction , more massive satellites are more effective at delivering their stars deeper into the host ; ii ) as a consequence , ashs feature a negative gradient between radius and the local mass - weighed virial satellite - to - host mass ratio ; iii ) in @xmath1 galaxies , most ashs feature a density profile that steepens towards sharper logarithmic slopes at increasing radii , though with significant halo - to - halo scatter ; iv ) the ashs with the largest total ex - situ mass are such because of the chance accretion of a small number of massive satellites ( rather than of a large number of low - mass ones ) . |
galaxy clusters provide crucial information to our understanding of the large - scale cosmic structure and to constrain cosmological models .
they populate the high - mass tail of the mass function of virialized galaxy systems ; their abundance and redshift distribution depend on the average density of the universe and the normalization of the power spectrum of the initial density perturbations ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
clusters are a hostile environment to galaxies and are thus also a unique tool to investigate the connection between environment and galaxy properties ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
separating the galaxies that do actually belong to the cluster from the interlopers -the galaxies that happen to lie in the field of view but are not dynamically linked to the cluster- is crucial to derive accurate estimates of the cluster properties , including its mass @xcite , or the color and star formation gradients of its galaxy population @xcite .
interloper rejection techniques are numerous and their sophistication has progressively increased over the years , thanks to the increased quality and richness of the observational data : over the last decade , the handful of clusters with tens of measured redshifts within @xmath10 of the cluster center has increased by at least a factor of 10 ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
early observations of galaxy clusters do not usually extend into the outer regions of the system .
early interloper rejection techniques identify galaxy members solely on the basis of their redshift separation from the cluster center .
the gravitational potential well can however become substantially shallower at increasing radius and the combination of velocity and radial distance is now an essential ingredient for the identification of galaxy members in samples that extend to the cluster virial radius and beyond .
the caustic technique @xcite identifies the escape velocity profile of galaxy clusters from their center to radii as large as @xmath5 , where @xmath8 is the radius of the sphere whose average density is 200 times the critical density of the universe .
the technique was thus applied to estimate the gravitational potential well and the mass profiles of galaxy clusters to radii that extend to the cluster infall region ( see reviews in @xcite and @xcite ) .
where the cluster is in the appropriate redshift range for weak lensing mass estimation and a comparison is thus possible , caustic and lensing masses agree within 30% at the virial radius ( diaferio et al 2005 , geller et al 2013 ) , whereas at smaller and larger radii the two mass estimates show a systematic offset of at most 50% and 20% respectively @xcite .
because the technique measures the escape velocity profile , a byproduct of the caustic procedure is the identification of interlopers . compared to other interloper rejection algorithms the caustic technique has two major advantages : ( 1 ) it does not require the system to be in dynamical equilibrium and ( 2 ) it does not rely on the derivation of the cluster mass profile to remove interlopers .
these advantages enable the technique to identify interlopers both in the central and outer regions of clusters , where other techniques can not be applied .
the caustic technique assumes spherical symmetry , an assumption that is common to most methods .
in addition , when used as a mass estimator method , the caustic technique returns correct mass estimates if clusters form by hierarchical clustering and thus they have the internal kinematical and dynamical properties , including the shape of the velocity anisotropy profile , that clusters generally have in these models . the caustic technique as an interloper rejection algorithm , or some simplified versions of it , was applied to real clusters to investigate the dependence of galaxy properties on environment ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and to provide robust estimates of the cluster velocity dispersion and mass ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
thanks to the approximate self - similarity of self - gravitating systems , the technique can also be applied to reject stellar interlopers in galaxies : @xcite used the caustic method results to estimate the velocity dispersion profile of the stars in the milky way halo , and @xcite demonstrated that a proper stellar interloper rejection alleviates the tension between the internal velocity dispersion profiles of the milky way dwarf satellites and the expectations of modified newtonian dynamics . @xcite also probed the dark matter distribution in the outer regions of disk galaxies by identifying their satellites with the caustic technique . despite this extensive application ,
the caustic technique has never been exhaustively explored as a method to identify interlopers . here , we provide a thorough analysis of its performance and of its random and systematic errors . in section [ sec : review ] we briefly describe the caustic technique , whereas in section [ sec : cat ] we present the mock cluster catalogs . in section [ sec : idmem ] we discuss the technique performance .
we finally investigate the impact of our interloper rejection on the cluster mass estimates in section [ sec : massestid ] .
we compare the performance of our method with other rejection techniques in section [ sec : disid ] .
conclusions are presented in section [ sec : conid ] .
in hierarchical clustering , clusters of galaxies form by the aggregation of smaller systems .
the accretion is not purely radial ( e.g. , * ? ? ?
* ) , because galaxies within the falling clumps have velocities with a substantial non - radial component .
therefore , the galaxy velocities are set by the local gravitational potential more than by the radial infall expected in the spherical collapse model @xcite .
when observed in the redshift diagram -the plane of the line - of - sight velocity @xmath11 of the galaxies in the cluster rest frame versus their projected distance @xmath12 from the cluster center- the cluster members populate a region with a trumpet shape approximately symmetric along the @xmath12 axis @xcite .
the caustics define the boundaries of this region whose amplitude @xmath13 decreases with increasing @xmath12 .
@xcite demonstrate that @xmath13 is a combination of the profile of the escape velocity from the cluster and the profile of the velocity anisotropy parameter @xmath14 , where @xmath15 , @xmath16 , and @xmath17 are the longitudinal , azimuthal and radial components of the velocity @xmath18 of a galaxy , respectively , and the brackets indicate an average over the velocities of the galaxies in the volume @xmath19 centered on position @xmath20 . in a spherically symmetric system ,
the average square of the velocity of the system members at radius @xmath12 is @xmath21 where @xmath22 is the component of the line - of - sight velocity and @xmath23 for the escape velocity at radius @xmath12 , we have @xmath24 , where @xmath25 is the gravitational potential .
if the amplitude @xmath13 measures the average component along the line of sight of the escape velocity at radius @xmath12 , namely @xmath26 , we obtain the relation @xmath27 this equation shows the dynamical information contained in the observable caustic amplitude @xmath13 .
being a combination of the gravitational potential profile and the function @xmath28 , @xmath13 can provide the estimate of both the escape velocity profile from the cluster and the mass profile of the cluster @xcite .
we emphasize that the entire argument outlined above holds regardless of the stability of the system . to measure @xmath13 we need to locate the caustics in the redshift diagram .
the technique consists of three major steps : ( 1 ) the construction of a binary tree based on the projected galaxy pairwise energy ; ( 2 ) the determination of a threshold to cut the binary tree ; ( 3 ) the identification of the cluster center to obtain the redshift diagram and determine the galaxy number density on this diagram . .
the square and the triangle show the nodes @xmath29 and @xmath30 respectively .
the curve between @xmath29 and @xmath30 is the @xmath31 plateau .
the dashed line shows the line - of - sight velocity dispersion of the particles within the sphere of radius @xmath5 . ] at the first step , all the galaxies are arranged in a binary tree according to their pairwise binding energy @xmath32 where @xmath33 is the pair projected separation , @xmath34 is the line - of - sight velocity difference and @xmath35 m@xmath36 are the two galaxy masses assumed to be constant .
the binary tree is built as follows : ( i ) initially each galaxy is a group @xmath37 ; ( ii ) the binding energy @xmath38 , where @xmath39 is the binding energy between the galaxy @xmath40 and the galaxy @xmath41 , is associated with each group pair @xmath42 ; ( iii ) the two groups with the smallest binding energy @xmath43 are replaced with a single group @xmath44 and the total number of groups is decreased by one ; ( iv ) the procedure is repeated from step ( ii ) until only one group is left .
figure [ fig : dendrogram ] shows the binary tree of a random sample of 200 particles extracted from a simulated halo selected from the @xmath2-body simulation described in the next section , whereas figure [ fig : sky ] shows the celestial coordinates of the same particles with the same color code as in figure [ fig : dendrogram ] .
the second step of the caustic technique procedure is the threshold choice .
the tree arranges the galaxies in potentially distinct groups ; however , to get effectively distinct groups and to specifically define the set of candidate members , we need to cut the tree at some level .
this level sets the node from which the candidate members hang .
all these candidate members do not necessarily coincide with the optimal members that are determined by the caustic location .
below we will extensively illustrate the reason for this distinction between candidate and optimal members . in order to choose the threshold to cut the binary tree , we identify the main branch as the branch that emerges from the root and contains the nodes from which , at each level , the largest number of galaxies ( or leaves ) hangs .
the leaves hanging from each node @xmath45 of the main branch provide a velocity dispersion @xmath46 .
when walking along the main branch from the root to the leaves , @xmath46 rapidly decreases due to the progressive loss of galaxies that are most likely not associated with the cluster ( figure [ fig : plateau ] ) ; then @xmath47 reaches a `` @xmath31 plateau '' at some node @xmath29 .
most of the galaxies hanging from this node are members ; in fact , the system is nearly isothermal and the removal of the less bound galaxies does not affect the value of @xmath46 .
at some point of the walk along the main branch , the loss of the most bound galaxies , whose binding energy is very small , causes @xmath46 to drop again .
this second rapid drop identifies the nodes @xmath30 which sets the limit of the @xmath31 plateau .
the first node @xmath29 closest to the root is the appropriate level for the identification of the system and we define the galaxies hanging from it the candidate members of the cluster .
they determine the center of the system , its radius , and its line - of - sight velocity dispersion .
these quantities are used to build the redshift diagram .
the third step of the procedure is the location of the caustics in the redshift diagram .
the caustics are the curves satisfying the equation @xmath48 . here
@xmath49 is the galaxy number density in the redshift diagram , namely the plane @xmath50 , and @xmath51 is the root of the equation @xmath52 the function @xmath53 is the mean caustic amplitude within @xmath54 , @xmath55 , @xmath56 is the velocity dispersion of the candidate members , @xmath54 is their mean projected separation from the center , and @xmath57 is a smoothing parameter ( see @xcite for details ) .
figure [ fig : rediag ] shows the result of this procedure on the redshift diagram .
we stress that the procedure to locate the caustics is independent of any assumption on the dynamical equilibrium of the system , of the shape of @xmath28 and of the gravitational potential profile @xmath25 ; this procedure actually measures the combination of @xmath28 and @xmath25 expressed by the caustic amplitude @xmath13 ( equation [ eq : rig - pot ] ) .
we use the synthetic galaxy clusters described in @xcite selected from the @xmath2-body simulation of @xcite .
the simulation models a cubic volume of 192 @xmath58 mpc on a side of a flat @xmath3cdm model , with matter density @xmath59 , hubble parameter @xmath60 , normalization of the power spectrum @xmath61 and baryon density @xmath62 .
the simulation contains @xmath63 dark matter particles with mass @xmath64 m@xmath36 and , initially , @xmath63 gas particles with mass @xmath65 m@xmath36 .
the simulation was run with gadget-2 @xcite .
further details of the simulations and the dark matter halo identification are given in @xcite . in the following ,
we limit our analysis to the gravitational dynamics of the dark matter distribution .
in fact , both @xmath2-body simulations ( e.g. , * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and observations ( e.g. , * ? ? ?
* ) indicate that any velocity bias between galaxies and dark matter is negligible .
we consider the 100 dark matter halos with mass @xmath66 at redshift @xmath67 .
we locate each halo at @xmath68 and redshift @xmath69 km s@xmath70 .
we simulate the compilation of the redshift catalog of a galaxy cluster by projecting each halo along 10 random lines of sight . for each of these lines of sight ,
we choose two additional directions orthogonal to the first one and to each other .
we end up with @xmath71 mock redshift catalogs .
each catalog contains a random sample of @xmath72 particles distributed within a rectangular parallelepiped centered on the cluster with a squared field of view of @xmath4 mpc on a side and 192 @xmath58 mpc deep . with this number of particles in the field of view
, we obtain a distribution of the number of particles within the sphere of radius @xmath8 that has median @xmath73 and percentile range [ 10% , 90% ] equal to @xmath74 $ ] ; the median number of particles within @xmath5 is 185 and the percentile range [ 10% , 90% ] is @xmath75 $ ] .
these numbers are comparable to the sample sizes of recent large galaxy redshift surveys of clusters and their surroundings , such as cirs @xcite and hecs @xcite .
the binary tree algorithm applied to the individual mock catalogs gives a center of the cluster and a velocity dispersion of the candidate members .
the center and velocity dispersion determined with the binary tree are close to the correct quantities in most cases @xcite .
specifically , in 2678 mock catalogs ( 89% of the cases ) the algorithm locates the center on the expected cluster ; in the remaining 11% of the cases , the field of view is particularly crowded with numerous groups and clusters , and the cluster of interest might not be the most massive cluster in the field . in these cases
, the algorithm identifies the center of a different cluster . in a similar situation happening with catalogs of real clusters
we will relocate the center on the cluster of interest . here , we simply remove these problematic catalogs . among the 2678 correctly identified clusters ,
the estimated velocity dispersion within @xmath5 is within 5 ( 30)% of the real one in 50 ( 95)% of the systems ; the center deviations are smaller than @xmath76 on the sky and 250 km s@xmath70 along the line of sight in 90% of the clusters .
the largest discrepancies between the correct center and the center found by the algorithm occur in systems with evident substructures that produce multiple peaks of the particle number density distribution .
when happening with catalogs of real clusters , these cases can yield off - centered redshift diagrams .
this problem can be removed by relocating the center on the most luminous galaxy of the cluster or on the peak of the x - ray emission . in our mock catalogs
, we do not keep these systems , but further remove those catalogs where the center found by the algorithm has an offset greater than @xmath77 on the sky or greater than @xmath78 km s@xmath70 along the line of sight .
thus , the final number of mock catalogs reduces from 2678 to 2420 .
and @xmath8 of each individual cluster , respectively .
the number density distribution includes all the particles from the entire sample of 2420 mock catalogs . ]
a galaxy is a cluster member if its binding energy is negative , namely if its velocity @xmath11 is lower than the velocity @xmath79 required to escape the cluster when the galaxy is at distance @xmath12 from the cluster center : @xmath80 .
figure [ fig : enerdist ] shows the number density distribution of the dark matter particles in our mock cluster catalogs in the plane of binding energy versus the three - dimensional ( 3d ) clustrocentric distance .
the plot includes the entire sample of 2420 cluster catalogs .
figure [ fig : enerdist ] shows that a substantial fraction of bound particles have clustrocentric distance much larger than @xmath8 @xcite .
therefore , in principle , we might use a different and simpler criterion to define a cluster member : a galaxy whose clustrocentric distance is smaller than , for example , @xmath5 .
this criterion is actually more restrictive than the criterion based on the binding energy , as figure [ fig : enerdist ] suggests .
nevertheless , we include this criterion in the following analysis , for the sake of comparison .
hereafter , we call 3d members these two sets of members defined on the basis of the 3d information
. we expect that the caustic technique identifies the 3d members as the galaxies within the caustics in the redshift diagram . in the following
we will also consider a criterion based on the binary tree .
as described in section [ sec : review ] , the caustic technique arranges the galaxies in a binary tree according to their pairwise projected binding energy ; by cutting the tree at the @xmath31 plateau we define a set of candidate members . in the following analysis , we show that the choice of this name is appropriate , because the interloper contamination of this set of candidate members is larger than the contamination of the set of members determined by the caustic location . in conclusion
, we consider two definitions of 3d members : ( a ) galaxies with negative binding energy ; ( b ) galaxies within @xmath5 ; and two possible criteria for their identification : ( 1 ) galaxies within the caustics in the redshift diagram ; ( 2 ) galaxies on the main branch of the binary tree cut at the @xmath31 plateau .
hereafter , we refer to the members identified with methods ( 1 ) or ( 2 ) as 2d members
. in sections [ sec : causid ] and [ sec : treeid ] we focus on the performance of the first and second criteria , respectively .
we compute two relevant quantities : the completeness @xmath81 , which is the fraction of 3d members that are also identified as 2d members , and the contamination @xmath82 , which is the ratio between the number of particles taken as 2d members that are actually interlopers and the total number of 2d members .
figure [ fig : rediagbound ] shows the redshift diagram of a cluster from our sample , with the bound galaxies defined as the 3d members , shown as blue dots
. as expected , most of the 3d members are within the caustics .
to illustrate how the method performs on average in this case , we compute the completeness and contamination profiles of each cluster . at each radius , we consider the median of the set of profiles and their dispersion .
the upper left panel of figure [ fig : evel - fg ] shows the median differential profile of the completeness @xmath81 and the regions containing 50% , 68% and 90% of the profiles . only at small radii
the caustic algorithm removes a few per cent of the 3d members , because the caustic amplitude is slightly underestimated , as can be seen in the example of figure [ fig : rediagbound ] .
the caustic criterion thus provides a completeness close to 0.95 at radii smaller than @xmath83 , and increases to 1.0 at larger radii . the median differential contamination ( figure [ fig : evel - fg ] , bottom left panel ) is larger than 0.1 at radii larger than @xmath84 , but
the cumulative contamination ( figure [ fig : evel - fg ] , bottom right panel ) remains below 0.09 at @xmath5 .
table [ table : comp ] gives the corresponding 68% levels for the cumulative profiles of @xmath81 and @xmath82 .
figure [ fig : rediag3r200 ] reproduces the same redshift diagram of figure [ fig : rediagbound ] with the red dots the particles within @xmath5 from the cluster center defined as the 3d members . in this case , most 3d members are within the caustics , but , at large radii , many particles within the caustics are not 3d members .
in fact , in a sample of particles extracted from a spherical halo whose number density profile decreases with radius , the number of particles within a given 3d radius @xmath12 observed in projection can fall to zero beyond a projected radius @xmath85 ; obviously , @xmath86 decreases with the size of the particle sample . for a @xcite number density profile and for a sample of @xmath87 particles within @xmath5 ( the average number of particles per cluster sample ) , we find that the number of particles with a 3d distance smaller than @xmath5 is zero at projected distances larger than @xmath88 .
the differential contamination @xmath82 for the entire sample remains smaller than 0.1 within @xmath8 , on average , but increases dramatically at larger radii and reaches @xmath89 at @xmath90 .
it follows that the corresponding cumulative profile of the fraction of galaxies identified as members when they are actually interlopers reaches 0.27 at @xmath5 ( table [ table : comp ] ) . on the other hand ,
the profiles of the completeness we obtain with this definition of 3d members are very similar to figure [ fig : evel - fg ] : the method yields a large ( @xmath91 ) and stable completeness to radii as large as @xmath5 ( table [ table : comp ] ) .
we now evaluate the completeness @xmath81 and the contamination @xmath82 of the set of members identified with the binary tree . with the bound galaxies as the 3d members , the completeness @xmath92 is constant over the entire range of @xmath12 ( table [ table : comp ] ) .
the median differential contamination @xmath82 reaches a maximum value of @xmath93 at @xmath94 .
the cumulative profile thus reaches the value 0.13 ( table [ table : comp ] ) at @xmath95 .
the differential profile of @xmath82 shows that , if the bound galaxies are taken as 3d members , using the binary tree procedure introduces interlopers at @xmath96 ; this result is somewhat worse than the caustic location performance shown in the previous section , because the caustic location , on average , provides samples without contamination up to @xmath97 ( figure [ fig : evel - fg ] ) .
overall , however , when we adopt the bound particles as 3d member , both the caustic location and the binary tree give high levels of completeness ( @xmath98 ) and low levels of contamination ( @xmath99 ) within @xmath5 .
in the case of the galaxies within @xmath5 as 3d members ( definition ( b ) in section [ sec : def ] ) , the completeness @xmath81 has a constant median value @xmath100 ( table [ table : comp ] ) . clearly , applying the binary tree procedure to determine the members of a cluster guarantees an extremely high completeness of the sample . on the other hand ,
the median differential contamination @xmath82 is smaller than 0.1 at @xmath101 , and increases at larger radii up to @xmath102 .
this high contamination at large radii translates into a cumulative @xmath82 of 0.31 at @xmath5 ( table [ table : comp ] ) .
the reason for this large contamination at large radii derives from the decreasing of the number density profile , as discussed in section [ sec : causid ] . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ]
in this section , we analyze the effect of our interloper removal methods on the estimation of the mass , because interlopers have a non - negligible impact on the mass estimation , especially at large radii , where interlopers can cause an overestimate of the mass as large as a factor of three @xcite .
we consider the three standard methods described in @xcite : the virial , the average and the median mass estimators .
all estimators assume that the galaxies have equal mass and the system is in a steady state .
the virial mass estimator is @xmath103 where @xmath2 is the number of galaxies with measured redshifts , @xmath104 is the line - of - sight velocity and @xmath105 is the projected separation between galaxy @xmath106 and galaxy @xmath107 .
we do not add the surface pressure term in this analysis , because the correction that it introduces is expected to be smaller than @xmath108% @xcite .
the median mass estimator is supposed to be less sensitive to interlopers , because if interlopers in velocity populate the tails of the distribution of the quantity @xmath109 , which is estimated for each of the @xmath110 pairs , the median of this quantity is a more robust estimate than the mean .
the mass is thus @xmath111 where the coefficient @xmath112 is calibrated with @xmath2-body simulations @xcite .
if we take the mean , rather than the median , we obtain the average mass estimator @xmath113 where @xmath114 is again calibrated with @xmath2-body simulations @xcite .
as mentioned above , one expects that the virial and average mass estimators are more sensitive to interlopers than the median mass estimator .
we apply these three estimators to our samples after removing the interlopers with our two different procedures : ( 1 ) the caustic location and ( 2 ) the binary tree .
the top panel of figure [ fig : massr200 ] shows the results of applying the mass estimators to individual clusters within @xmath8 , @xmath84 and @xmath5 . on average , the mass estimate is unbiased when the member galaxies are identified with the caustic location , whereas it is biased high by at least 20% when the member galaxies are extracted from the main group of the binary tree .
this result confirms our expectation that the caustic location removes interlopers more efficiently than the binary tree procedure .
as expected , the median mass estimator is the less sensitive method to the presence of interlopers : in fact , it yields the values of @xmath115 closest to one when the interlopers are removed with the less efficient binary tree procedure . for comparison
, we also show the mass estimated with the caustic technique applied to the full sample of particles , because , in principle , the technique is not affected by the presence of interlopers .
with all the estimators , the spread increases with radius . for the caustic technique ,
this increase derives from the smaller number of galaxies available for locating the caustics . for the other estimators the system is required to be in virial equilibrium , that does not necessarily hold at radii larger than @xmath8 ; therefore , at these radii , the three standard estimators are more likely to return an incorrect mass .
in addition , part of these spreads derives from the assumption of spherical symmetry .
the bottom panels of figure [ fig : massr200 ] show the mass estimates of the stacked clusters with @xmath116 and @xmath117 .
the spreads for the case @xmath116 are slightly smaller than in the case of individual clusters ( upper panel of figure [ fig : massr200 ] ) , despite the fact that , on average , individual clusters have 185 galaxies within @xmath5 , a factor of 3.7 larger than the @xmath116 stacked cluster ( bottom left panel of figure [ fig : massr200 ] ) . in the stacked clusters , the assumption of spherical symmetry is basically correct
, therefore the spread only derives from the sample size .
in fact , the spreads further reduce by a factor of roughly @xmath118% in the case of the @xmath117 stacked cluster ( bottom right panel of figure [ fig : massr200 ] ) .
we finally note that in the case @xmath116 , the mass estimate is biased low by 20% . in this case , in fact , the number of galaxies within @xmath8 is only @xmath119 , on average , and the velocity field is too poorly sampled to return a correct mass .
proper estimates of the mass of galaxy clusters and of the properties of their galaxy population depend on the accurate separation between the cluster members and those galaxies that appear projected in the cluster field of view but are not dynamically linked to the cluster . numerous methods to identify and remove interlopers in galaxy clusters
have been suggested in the literature .
the algorithms are based either on the line - of - sight velocity separation of the galaxy from the cluster center alone or on both the velocity and the projected separations . the former class of algorithms is suited for galaxy samples that only survey the central regions of the clusters .
these algorithms include the @xmath120 clipping method @xcite , which assumes that the velocity distribution is close to gaussian , the gap method @xcite , and the adaptive kernel method @xcite
. however , not all interlopers have large velocity separations from the cluster , as figure [ fig : rediagbound ] illustrates .
these interlopers are difficult to identify and can generate a rather counterintuitive systematic error : they can cause a slight underestimate , rather than an overestimate , of the cluster velocity dispersion @xcite . rather than iterating over the velocity dispersion ,
like the @xmath120 clipping method does , one can iterate over the virial mass : at each step , one removes the galaxy that causes the largest mass variation @xcite .
the projected and virial mass estimators @xcite are sensitive to the presence of interlopers in different ways ; comparing their mass values can also be used to identify interlopers @xcite .
iterating over the mass provides more robust results than iterating over the velocity dispersion , because interlopers can affect more the estimate of the size of the cluster , which enters the mass estimate , than the velocity dispersion @xcite .
when galaxy catalogs survey large cluster regions , the methods described above can be extended and applied to galaxy subsamples separated into bins of projected distances to the cluster center @xcite .
thus , the velocity distribution assumed to be gaussian at each radius can have different widths at different radii @xcite , or the velocity dispersion in the @xmath120 clipping method can be derived at different radii by solving the jeans equation for a steady state system and isotropic galaxy orbits @xcite .
a step forward an interloper rejection algorithm based on a dynamical approach derives from the following consideration : from an extensive galaxy sample we can actually extract information on the dependence of the escape velocity on the clustrocentric distance @xcite .
based on this idea , most algorithms first estimate the mass profile by assuming dynamical equilibrium , and then , from these mass profiles , derive the escape velocity as a function of the projected distance to the cluster center . the final solution
thus must be obtained by iteratively removing the identified interlopers until the mass profile converges . here , we have shown the performance of the caustic technique , used as an interloper rejection algorithm .
the technique only uses the number density distribution of galaxies in the redshift diagram to estimate directly the escape velocity profile from the system . unlike the methods mentioned above
, the caustic technique relies neither on the assumption of dynamical equilibrium nor on the estimate of the mass profile .
therefore , the technique does not require any iteration ; in addition , the estimate of the cluster mass profile is a further step that is unnecessary for identifying the interlopers and it is a step that we have not taken here . in the presence of extensive surveys ,
interloper rejection algorithms based on the estimate of the mass or on the escape velocity usually perform better than algorithms solely based on the velocity distribution ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
@xcite perform an extensive comparison of a number of different algorithms .
they conclude that the method by @xcite is the most effective at removing interlopers , producing samples with average contaminations @xmath82 in the range @xmath121 within one projected virial radius .
these values are in perfect agreement with our median @xmath122 ( table [ table : comp ] ) . however , there are two noticeable differences between our analysis and theirs : the dynamical state of the clusters and the sample extension .
the caustic technique is independent of the dynamical state of the cluster , and , in fact , we only adopt the cluster mass as the criterion to build our sample of @xmath0 simulated clusters . on the contrary , in their sample of @xmath108 simulated clusters ,
@xcite pay particular attention to only include relaxed systems that have no sign of ongoing mergers , because the @xcite method requires dynamical equilibrium to be effective .
merging clusters require more sophisticated approaches ( see , e.g. , @xcite and references therein ) , like the caustic technique .
in addition , the dynamical equilibrium assumption clearly limits the analysis to projected radii smaller than the virial radius , where the equilibrium is expected to hold , whereas the caustic technique enables the identification of interlopers to much larger radii .
we find that , with the caustic technique , the median contamination increases to @xmath123 and @xmath124 at @xmath84 and @xmath5 , respectively , with a median completeness @xmath81 that remains larger than @xmath91 .
no other methods that remove interlopers in these regions are currently available
. when we use the caustic technique to identify interlopers , the virial mass estimator returns a mass overestimated by 10% within @xmath8 , similarly to the results of @xcite , who removed interlopers with a combination of the gap procedure @xcite and the method of den hartog and katgert @xcite from mock clusters with more than 60 members .
this bias is not present when we use the median and average mass estimators . at radii larger than @xmath8 , where only the caustic technique can be used to remove interlopers , all mass estimates are unbiased .
the caustic technique identifies the escape velocity profiles of galaxy clusters to radii as large as @xmath5 ; we can thus estimate the cluster mass in regions where the cluster is not necessarily in dynamical equilibrium .
the performance of the caustic method as a mass estimator has been tested on both @xmath2-body simulations @xcite and real clusters @xcite , regardless of their dynamical state : when we compare the caustic mass with the gravitational lensing mass in a combined sample of 22 clusters , the two estimates generally agree @xcite . here , we have investigated an additional use of the caustic technique : the interloper rejection method . in this case
, the technique relies only on the location of the caustics on the redshift diagram and makes no use of the mass profile of the cluster .
we have tested the ability of the method to identify the cluster galaxy members by using @xmath0 galaxy clusters with mass @xmath1 extracted from a cosmological @xmath2-body simulation of a @xmath3cdm universe . unlike the case of the mass estimate , where we compare the caustic technique with gravitational lensing , we can not test the interloper rejection method on real clusters . however , the caustic technique is based on the hypothesis that clusters form by hierarchical clustering ; wide observational evidence , based on x - ray and optical data , including gravitational lensing studies ( e.g. * ? ? ?
* ; * ? ? ?
* ) , suggest that this hypothesis is well founded . therefore we expect that @xmath2-body simulated clusters are a reasonable representation of real clusters and that the results of our analysis can be safely applied to real clusters .
our mock catalogs contain 1000 galaxies in the field of view of @xmath4 mpc on a side at the cluster location .
the true 3d members , defined as the gravitationally bound galaxies , are compared to the galaxies identified as members with the caustic technique .
we find a completeness of @xmath6 within @xmath5 , whereas the contamination increases from @xmath7 at @xmath8 to @xmath9 at @xmath5 .
the lack of spherical symmetry in clusters of galaxies causes most of the spread of the completeness and the contamination profiles .
in fact , when applying the technique to samples built from a spherically symmetric stacked cluster the spreads decrease by at least a factor of two . no other technique for the identification of the members of a galaxy cluster
provides such large completeness and small contamination at these large radii .
the mass estimated with the virial theorem within @xmath5 , after removing interlopers in the case of individual clusters , is unbiased and is within 30% of the actual mass . the use of the spherically symmetric stacked cluster decreases the spread to less than 10% . for the sake of clarity ,
we remind the systematic error that our interloper rejection method can introduce : the membership identification is based on identifying the caustic amplitude with the escape velocity _ tout - court _ , whereas the caustic amplitude , which we measure independently of the knowledge of @xmath28 , of the mass profile and of the gravitational potential profile , actually is the escape velocity corrected by the factor @xmath125 ( equation [ eq : rig - pot ] ) . the fact that we neglect this correction factor when we identify the caustic amplitude with the escape velocity can propagate in an incorrect separation of the cluster members from the interlopers .
our excellent results show that , despite this simplification , the caustic method can satisfactorily separate the members from the interlopers . the increasing amount of data in clusters of galaxies @xcite requires adequate tool for extracting the information they contain and properly comparing them with the output of the galaxy formation modeling that is increasingly sophisticated @xcite .
the caustic technique can provide accurate estimates of the dark matter distribution in the outer regions of galaxy clusters and information on the dynamical connection between galaxies and clusters .
the first piece of information is relevant for our understanding of the formation of cosmic structure and can even constrain the properties of dark matter @xcite and the theory of gravity @xcite .
determining the membership of galaxies in the outskirts of clusters is unique to the caustic method . applying the algorithm to a large sample of clusters can provide the first accurate measure of how the gradients of properties of the cluster galaxy population , such as color and star formation rate , merge into the field .
in addition , it might provide the first determination of galaxy membership in the filaments surrounding clusters that represent the preferred path of mass accretion @xcite ; this piece of information can thus enlighten the connection between the formation of galaxies and the large - scale structure . in future work
, we will investigate the reliability of the caustic method in performing these measurements and assess the impact that these measures can have on the models of the formation of the cosmic structure .
we thank giuseppe murante , stefano borgani and the other members of our @xcite collaboration for the use of our @xmath2-body simulations for this project .
we thank margaret geller for enlightening discussion and suggestions and an anonymous referee whose comments prompted us to improve the presentation of the caustic technique and of our results .
the simulations were carried out at the cineca supercomputing centre in bologna ( italy ) , with cpu time assigned through a inaf - cineca key project .
als acknowledges a fellowship by the prin inaf09 project `` towards an italian network for computational cosmology ''
. partial support from the infn grant pd51 and the prin - miur-2008 grant ` 2008nr3ebk_003 ` `` matter - antimatter asymmetry , dark matter and dark energy in the lhc era '' is also gratefully acknowledged .
this research has made use of nasa s astrophysics data system . | the caustic technique measures the mass of galaxy clusters in both their virial and infall regions and , as a byproduct , yields the list of cluster galaxy members .
here we use @xmath0 galaxy clusters with mass @xmath1 extracted from a cosmological @xmath2-body simulation of a @xmath3cdm universe to test the ability of the caustic technique to identify the cluster galaxy members .
we identify the true three - dimensional members as the gravitationally bound galaxies .
the caustic technique uses the caustic location in the redshift diagram to separate the cluster members from the interlopers .
we apply the technique to mock catalogs containing 1000 galaxies in the field of view of @xmath4 mpc on a side at the cluster location . on average
, this sample size roughly corresponds to 180 real galaxy members within @xmath5 , similar to recent redshift surveys of cluster regions .
the caustic technique yields a completeness , the fraction of identified true members , @xmath6 within @xmath5 .
the contamination , the fraction of interlopers in the observed catalog of members , increases from @xmath7 at @xmath8 to @xmath9 at @xmath5 .
no other technique for the identification of the members of a galaxy cluster provides such large completeness and small contamination at these large radii .
the caustic technique assumes spherical symmetry and the asphericity of the cluster is responsible for most of the spread of the completeness and the contamination . by applying the technique to an approximately spherical system obtained by stacking the individual clusters , the spreads decrease by at least a factor of two .
we finally estimate the cluster mass within @xmath5 after removing the interlopers : for individual clusters , the mass estimated with the virial theorem is unbiased and within 30% of the actual mass ; this spread decreases to less than 10% for the spherically symmetric stacked cluster .
cosmology : miscellaneous dark matter galaxies : clusters : general gravitation large - scale structure of universe methods : data analysis techniques : miscellaneous |
the discovery of giant magnetoteresistance in 1998 by the groups of fert and grnberg led to new reading heads for hard disks @xcite . moreover for the first time , a device based on magnetic phenomena replaced a conventional electronics device based on the movement of the electrons charge and thus opened the way to the field of spintronics or magnetoelectronics .
the aim is to replace conventional electronics with new devices where magnetism plays a central role leading to smaller energy consumption .
several architectures have been proposed @xcite but only in 2009 dash and collaborators managed to inject spin - polarized current from a metallic electrode into si , which is a key issue in current research in this field . showing that spintronic devices can be incorporated into conventional electronics @xcite . in order to maximize the efficiency of spintronic devices ,
the injected current should have as high spin - polarization as possible @xcite .
to this respect half - metallic compounds have attracted a lot of interest ( for a review see reference @xcite ) .
these alloys are ferromagnets where the majority spin channel is metallic while the minority - spin band structure is that of a semiconductor leading to 100%spin - polarization of the electrons at the fermi level and thus to possibly 100% spin - polarized current into a semiconductor when half metals are employed as the metallic electrode .
the term half - metal was initially used by de groot et al in the case of the nimnsb heusler alloy @xcite .
ab - initio ( also known as first - principles ) calculations have been widely used to explain the properties of these alloys and to predict new half - metallic compounds .
an interesting case is the transition - metal pnictides like cras and mnas .
akinaga and collaborators found in 2000 that when a cras thin film is grown on top of a zinc - blende semiconductor like gaas , the metallic film adopts the lattice of the substrate and it crystallizes in a meta - stable half - metallic zinc - blende phase @xcite structure .
later cras was successfully synthesized in the zinc - blence structure in the form of multilayers with gaas @xcite and other successful experiments include the growth of zinc - blende mnas in the form of dots @xcite and crsb in the form of films @xcite .
experiments agree with predictions of ab - initio calculations performed by several groups @xcite . in the case of the half - metallic ferromagnets like cras or crse
, the gap in the minority - spin band arises from the hybridization between the @xmath0-states of the @xmath1 atom and the triple - degenerated @xmath2 states of the transition - metal and as a result the total spin - moment , @xmath3 , follows the slater - pauling ( sp ) behavior being equal in @xmath4 to @xmath5 where @xmath6 the total number of valence electrons in the unit cell @xcite .
recently theoretical works have appeared attacking also some crucial aspects of these alloys like the exchange bias in ferro-/antiferromagnetic interfaces @xcite , the stability of the zinc - blende structure @xcite , the dynamical correlations @xcite , the interfaces with semiconductors @xcite , the exchange interaction @xcite , the emergence of half - metallic ferrimagnetism @xcite and the temperature effects @xcite . an extended overview on the properties of these alloys can be found in reference @xcite . of the lattice constant .
note that in the case of the cras / cdse we have two non - equivalent interfaces : ( i ) when the sequence of the atoms is ... -cr - as - cd- ... denoted as cras / cdse-1 and ( ii ) when the sequence is ... -cr - se - cd- ... denoted as cras / cdse-2 .
finally we should note that we have assumed the lattice constant of the two semiconductors ( 0.606 nm ) .
[ fig1 ] ] [ cols="<,^,^,^,^,^ " , ] [ table6 ]
finally , in the last section we will present our results concerning the case of in , for the cras / inas interface , and cd , for both cras / cdse interfaces , impurities at various sites .
all three interfaces show similar behavior and thus in figure [ fig9 ] we present the dos for all possible in impurities for the cras / inas multilayer .
we should note that with respect to the conservation of the half - metallicity this is the most interesting case since for the other two cras / cdse interfaces the half - metallic character is conserved for all cases under study . in table
[ table6 ] we have gathered the atom - resolved spin moments for all cases under study and as it can be easily deduced from the table the variation of the spin moments for the same position of the in(cd ) impurity is similar for all three interfaces and thus we will restrict our discussion to the cras / inas case . .
[ fig9 ] ] we expect that the most frequent case to occur would be the in impurity at the cr site since such an impurity does not disrupt the zinc - blende structure . in atoms have only two valence electrons occupying the deep - energy - lying @xmath7-states and thus for the energy window which we examine the @xmath0-states , which we observe , have their origin at the nearest as neighbors whose @xmath0-states penetrate in the in sites ( cd has only one valence @xmath7-electron ) .
thus the in impurity acts similarly to a void , although it does not lead to such large reorganization of the charge of the neighboring atoms , leading to slightly larger spin moment of the neighboring atoms with respect to the perfect interfaces as shown in table [ table6 ] . due to the small weight of the in @xmath0-states
we have multiplied the corresponding dos with a factor 5 or 10 in figure [ fig9 ] to make it visible . with respect to the case of void impurity at the cr site , here
the shift of the bands of the nearest - neighboring as atoms is smaller keeping the half - metallic character of the interface although the gap is considerably shrinking .
when the in impurity is located at the void1 site , the disturbance of the lattice is smaller with respect to the case just presented , although both cr and void1 sites have the same nearest - neighbors and as shown in figure [ fig9 ] the width of the gap remains unchanged . due to
the negligible weight of the in @xmath0 states also the occurrence of in impurities at void2 and void3 sites leads to a slight variation of the spin moments but the half - metallicity is preserved and the gap retains a large width . to conclude we should discuss also the case of in impurities at as sites .
as atoms at the interface have two cr atoms as nearest neighbors and the hybridization between the as @xmath0- and cr @xmath2-orbitals is strong .
the substitution of an as atom by an in one leads to reduced hybridization for the cr orbitals and cr atoms at the interface regain the charge participating at the bonds with the missing as atom .
this extra charge is accommodated at the cr spin - up states leading to larger spin magnetic moments of the cr atoms at the interface which now are about 3.59 @xmath4 instead of 3.25 @xmath4 in the case of the perfect cras / inas interface presented in table [ table1 ] .
the in impurity atom and its nearest - neighboring in atoms have states within the gap , as shown in figure [ fig9 ] but if we take into account that we have multiplied the in dos by 10 their real weight at the fermi level is negligible with respect to the cr majority - spin dos .
thus we can safely consider that the half - metallicity is conserved although as shown by the cr dos , the gap in the minority - spin band seriously shrinks and the fermi level is near the right edge of the gap .
we have studied using the korringa - kohn - rostoker method the appearance of single impurities at interfaces between the half - metallic ferromagnet cras and the binary inas and cdse semiconductors . in the case of bulk cras studied in reference
@xcite we had shown that most impurities affect the half - metallic character of cras inducing states either at the edges of the gap or in the middle of the gap .
but multilayers are very thin as experiments show @xcite and thus we can not use the impurity calculations for the bulk to discuss the case of interfaces .
we have studied al possible single impurities at the interfaces .
contrary to the bulk systems almost all defects were not affecting the half - metallic character of the perfect interfaces .
the only exception were void impurities at cr or in(cd ) sites .
the missing cr or in(cd ) atom leads to a reorganization of the charge of the surrounding atoms and as a result the @xmath0 bands of the nearest neighboring as(se ) atom shift to higher energies crossing the fermi level and leading to loss of the half - metallicity .
this is the opposite behavior that the one exhibited by the void impurities at cr site in the bulk cras alloy where the dos remained almost unchanged @xcite .
the different behavior of the void impurities should be attributed to the lower dimensionality of the interfaces with respect to the bulk .
but void impurities are schottky - type and we expect them to exhibit high formation energies and thus their number should be small in the experimental systems .
thus contrary to the bulk cras and eventually thick films showing a bulk - like behavior , impurities are expected to affect only marginally the half - metallic character of the interfaces in the case of thin multilayers and the latter ones are promising for spintronic devices .
zhao and a. zunger , phys . rev .
b 71 ( 2005 ) 132403 ; m.s .
miao and w.r.l .
lambrecht , phys .
b 72 ( 2005 ) 064409 ; l.j . shi b.g .
liu , j. phys . : condens .
matter 17 ( 2005 ) 1209 ; b. g. liu , phys .
b 67 ( 2003 ) 172411 . c.y .
fong , m.c .
qian , j.e .
pask , l.h .
yang , s. dag , appl .
84 ( 2004 ) 239 ; m. moradi and z. soltani , j. appl . phys . 105 ( 2009 ) 023701 ; f. ahmadian , m.r .
abolhassani , m. ghoranneviss , m. elahi , physica b : condens .
matter 404 ( 2009 ) 3684 .
e. aioglu , i. galanakis , l.m .
sandratskii , p. bruno , j. phys . : condens .
matter 17 ( 2005 ) 3915 .
i. galanakis , k. zdogan , e. aioglu , b. akta , phys .
b 74 ( 2006 ) 140408(r ) ; k. zdogan , i. galanakis , b. akta , e. aioglu , j. magn .
320 ( 2008 ) 197 .
m. leai , ph .
mavropoulos , j. enkovaara , g. bihlmayer , s. blgel , phys .
97 ( 2006 ) 026404 ; l. chioncel , m.i .
katsnelson , g.a . de wijs , r.a .
de groot , a.i .
lichtenstein , phys .
b 71 ( 2005 ) 085111 ; r. skomski and p. a. dowben , europhys . lett .
58 ( 2002 ) 544 . | we present an extended study of single impurity atoms at the interface between the half - metallic ferromagnetic zinc - blende cras compound and the zinc - blende binary inas and cdse semiconductors in the form of very thin multilayers .
contrary to the case of impurities in the perfect bulk cras studied in [ i. galanakis and s.g .
pouliasis , j. magn .
magn . mat . 321 ( 2009 )
1084 ] defects at the interfaces do not alter in general the half - metallic character of the perfect systems .
the only exception are void impurities at cr or in(cd ) sites which lead , due to the lower - dimensionality of the interfaces with respect to the bulk cras , to a shift of the @xmath0 bands of the nearest neighboring as(se ) atom to higher energies and thus to the loss of the half - metallicity .
but void impurities are schottky - type and should exhibit high formation energies and thus we expect the interfaces in the case of thin multilayers to exhibit a robust half - metallic character . electronic structure , half - metals , cras 75.47.np , 75.50.cc , 75.30.et |
the iter device is a tokamak designed to study controlled thermonuclear fusion . roughly speaking ,
it is a toroidal vessel containing a magnetised plasma where fusion reactions occur .
the plasma is kept out of the vessel walls by a magnetic field which lines have a specific helicoidal geometry
. however , turbulence develops in the plasma and leads to thermal transport which decreases the confinement efficiency and thus needs a careful study .
plasma is constituted of ions and electrons , which motion is induced by the magnetic field .
the characteristic mean free path is high , even compared with the vessel size , therefore a kinetic description of particles is required , see _ dimits _ @xcite .
then the full 6d vlasov - poisson model should be used for both ions and electrons to properly describe the plasma evolution . however , the plasma flow in presence of a strong magnetic field has characteristics that allow some physical assumptions to reduce the model .
first , the larmor radius , i.e. the radius of the cyclotronic motion of particles around magnetic field lines , can be considered as small compared with the tokamak size and the gyration frequency very fast compared to the plasma frequency .
thus this motion can be averaged ( gyro - average ) becoming the so - called guiding center motion . as a consequence
, 6d vlasov - poisson model is reduced to a 5d gyrokinetic model by averaging equations in such a way the 6d toroidal coordinate system @xmath0 becomes a 5d coordinate system @xmath1 , with @xmath2 the parallel and @xmath3 the perpendicular to the field lines components of the particles velocity , @xmath4 the angular velocity around the field lines and @xmath5 the magnetic momentum depending on the velocity norm @xmath6 , on the magnetic field magnitude @xmath7 and on the particles mass @xmath8 which is an adiabatic invariant . moreover ,
the magnetic field is assumed to be steady and the mass of electrons @xmath9 is very small compared to the mass of ions @xmath10 .
thus the cyclotron frequency @xmath11 is much faster for electrons than for ions @xmath12 .
therefore the electrons are assumed to be at equilibrium , i.e. the effect of the electrons cyclotronic motion is neglected and their distribution is then supposed to be constant in time .
the 5d gyrokinetic model then reduces to a vlasov like equation for ions guiding center motion : @xmath13 where @xmath14 is the ion distribution function for a given adiabatic invariant @xmath15 with @xmath16 , velocities @xmath17 and @xmath18 define the guiding center trajectories .
if @xmath19 , then the model is termed as conservative and is equivalent to a vlasov equation in its advective form : @xmath20 this equation for ions is coupled with a quasi - neutrality equation for the electric potential @xmath21 on real particles position , with @xmath22 ( with @xmath23 the larmor radius ) : @xmath24 where @xmath25 is an equilibrium electronic density , @xmath26 the electronic temperature , @xmath27 the electronic charge , @xmath28 the boltzmann constant for electrons and @xmath29 the cyclotronic frequency for ions .
+ these equations are of a simple form , but they have to be solved very efficiently because of the 5d space and the large characteristic time scales considered .
this work is then a contribution in this direction , following _ grandgirard et al _ who develops the gysela 5d code that solves this @xmath30 gyrokinetic model , see @xcite and @xcite .
looking at the model , one notices that the adiabatic invariant @xmath15 acts as a parameter .
therefore for each @xmath15 we have to solve a 4d advection equation as accurately as possible but also taking special care on mass and energy conservation , especially in this context of large characteristic time scales . the maximum principle that exists at the continuous level for the vlasov equation
should also be carefully studied at discrete level : @xmath31 with @xmath32 the value at @xmath33 in cell @xmath34 at time @xmath35 .
there is no physical dissipation process in the gyrokinetic model that might dissipate over / undershoots created by the scheme and the loss of this bounding extrema of the solution at @xmath36 may even eventually crash a simulation .
those studies will be achieved in this paper on a relevant reduced model , the 4d drift - kinetic model described in section [ use ] , which has the same structure than equations .
the geometrical assumptions of this model for ion plasma turbulence are a cylindrical geometry with coordinates @xmath37 and a constant magnetic field @xmath38 , where @xmath39 is the unit vector in @xmath40 direction .
this 4d model is conservative and will be discretized using a conservative semi - lagrangian scheme , the parabolic spline method scheme ( psm , see _ zerroukat et al _ @xcite and @xcite ) .
it is a fourth order scheme which is equivalent for linear advections to the backward semi - lagrangian scheme ( bsl ) currently used in the gysela code ( see _ grandgirard et al _ @xcite ) and introduced by _ cheng - knorr _ @xcite and _ sonnendrcker et al _ @xcite ) .
this conservative psm scheme based on the conservative form of the vlasov equation will be described in section [ use ] and properly allows a directional splitting .
+ in this paper , the bsl and psm schemes will be detailed with an emphasis on their similarities and differences .
we will see that one difference is about the maximum principle .
the bsl scheme satisfies it only with a condition on the distribution function reconstruction and the conservative psm scheme does not satisfy it without an extra condition on the volumes conservation in the phase space .
the last condition is equivalent to try to impose that the velocity field is divergence free at the discrete level .
a scheme is given to satisfy this constraint in the form of an equivalent finite volume scheme .
moreover , we have designed a slope limiting procedure , slope limited splines ( sls ) , to get closer to a maximum principle for the discrete solution , by at least diminish the spurious oscillations appearing when strong gradients exist in the distribution function profile .
+ the outline of this paper is the following : in section [ section_bslvspsm ] will be recalled some important properties of vlasov equations at the continuous level .
then bsl and psm schemes will be described and compared , according to properties of the discrete solutions . in section [ stability ] , a numerical method will be given to improve the respect of the maximum principle by vlasov discrete solutions when using the psm scheme and particularly to keep constant the volume in the phase space .
in section [ use ] , practical aspects of the psm scheme use will be described in the context of the 4d drift - kinetic model and at last we will comment on numerical results .
let us consider an advection equation of a positive scalar function @xmath41 with an arbitrary divergence free velocity field : @xmath42 with position @xmath43 and @xmath44 the advection velocity field . + the solutions satisfy the maximum principle : @xmath45 for any initial time @xmath46 .
+ since @xmath47 , we can also use an equivalent conservative formulation of the vlasov equation : @xmath48 for more details , see _ sonnendrcker _ lecture notes @xcite .
one obvious property of this conservation law ( reynolds transport theorem ) is to conserve the mass in a lagrangian volume @xmath49 , by integrating the distribution function on each lagrangian volume element @xmath50 : @xmath51 let us introduce the convective derivative @xmath52 , thus becomes : @xmath53 considering a lagrangian motion of an infinitely small volume @xmath49 , we have @xmath54 , thus we obtain : @xmath55 obviously , a divergence free flow @xmath56 conserves a lagrangian volume in its motion .
let us consider a vlasov equation in its non conservative form : @xmath57 with @xmath41 a scalar function , position @xmath43 and @xmath44 the advection field .
the bsl scheme , see _ sonnendrcker et al _ @xcite , is based on the invariance property of function @xmath58 along characteristic curves to obtain values @xmath59 at time @xmath60 from the values @xmath61 at @xmath35 : @xmath62 with @xmath63 the eulerian coordinates and the characteristic curves @xmath64 defined as @xmath65 with the initial position @xmath66 at @xmath35 .
let us locate the discrete function values @xmath67 at mesh nodes @xmath68 .
we solve the following nonlinear system which is a second order approximation of @xmath69 : @xmath70 with @xmath71 .
the function @xmath72 is a reconstruction of the solution @xmath73 according known values at nodes @xmath74
using cubic splines basis functions on the domain to obtain the value at @xmath75 , which is not a mesh node in general . [ [ bsl_prop ] ] properties of the bsl scheme + + + + + + + + + + + + + + + + + + + + + + + + + + + + this scheme is formally fourth order in space .
it is second order in time using for instance a leap - frog , predictor - corrector or runge - kutta time integration .
mass is not conserved by this scheme , because it has no conservative form . however , an approximated maximum principle is satisfied .
let us consider @xmath76 the cubic spline interpolation of the distribution function @xmath77 at time @xmath35 , we have for any @xmath63 : @xmath78 it then naturally appears a `` discrete '' maximum principle : @xmath79 comparing with the property , we have here @xmath80 and @xmath81 because the cubic spline reconstruction does not satisfy a maximum principle . if we have a manner to enforce this property to this reconstruction , a maximum principle is granted for the bsl scheme .
no directional splitting is allowed since the bsl scheme is based on the non conservative form of the vlasov equation , see @xcite .
let us consider a vlasov equation in its conservative form : @xmath82 with @xmath41 a scalar function , position @xmath43 and @xmath44 the advection field .
notice that with the hypothesis @xmath83 , conservative form and non - conservative form of the vlasov equation are equivalent .
the psm scheme , see _ zerroukat et al _ @xcite and @xcite , is based on the mass conservation property of function @xmath58 in a lagrangian volume @xmath84 to obtain the value @xmath59 at time @xmath60 : @xmath85 with the characteristic curves @xmath64 defined as @xmath86 and @xmath87 with @xmath35 the initial time , and the volume @xmath88 such that @xmath89 defined by the lagrangian motion with the field @xmath90 .
the important point is that this conservative formalism properly allows a directional splitting without loosing the mass conservation .
indeed , equation may be solved with @xmath91 successive 1d advections still of conservative form : @xmath92.\ ] ] we then approximate a 1d equation for each direction @xmath93 using the conservation property . omitting subscript @xmath93 ,
the psm scheme writes in 1d as follows : @xmath94 with @xmath95 settled as the 1d mesh nodes and @xmath96 the associated foot of the characteristic curve , @xmath97 $ ] and @xmath98 $ ] .
+ let us define the unknowns of the scheme as the average of @xmath58 in cell @xmath34 @xmath99 and the primitive function @xmath100 with the uniform space step @xmath101 and @xmath102 an arbitrary reference point of the domain and for instance the first node of the grid @xmath103 .
therefore , one has to solve a nonlinear system , which is similar to the bsl one , to obtain a discrete solution of equation that writes : @xmath104 with the time step @xmath71 and the uniform space step @xmath105 .
+ the computation of the reconstructed primitive function @xmath106 is based on values at mesh nodes @xmath107 : @xmath108 then this set of values is interpolated by cubic splines functions to obtain an approximated value @xmath109 of the primitive function @xmath110 at any point @xmath40 of the domain : @xmath111 [ [ properties - of - the - psm - scheme ] ] properties of the psm scheme + + + + + + + + + + + + + + + + + + + + + + + + + + + + this scheme is formally fourth order in space and strictly equivalent to the bsl scheme for constant linear advection , see @xcite .
it is second order in time using for instance a leap - frog , predictor - corrector or runge - kutta time integration scheme .
mass is exactly conserved by this scheme for each 1d step @xmath93 of the directional splitting .
however , no maximum principle does exist for each step @xmath93 even for the exact solution : in general @xmath112 , even if the velocity field is divergence free @xmath113 .
let us consider the scheme in @xmath91 dimensions of space : @xmath114 with @xmath41 a scalar function , position @xmath43 and @xmath44 the advection field .
let us consider a cell @xmath34 , where the solution is described at time @xmath60 by its average in cell @xmath34 , the psm scheme then writes : @xmath115 we thus obtain the following relation : @xmath116 with the average of the distribution function in the lagrangian volume at time @xmath35 : @xmath117 here clearly appears two conditions , both difficult to satisfy especially in the context of a directional splitting , to have a maximum principle defined as follows : @xmath118 1 .
maximum principle on the distribution function in @xmath119 : @xmath120 2 .
conservation of volumes in the phase space at the discrete level : @xmath121 the first condition is difficult to ensure in general , because a maximum principle should be satisfied for any average of the distribution function on an arbitrary volume @xmath122 . moreover , in the context of a directional splitting , it is impossible to satisfy a maximum principle for a 1d step @xmath93 , because it does not exist at the continuous level since in general @xmath123 .
therefore it is probably impossible to recover a maximum principle of the reconstruction after all steps of the directional splitting .
+ the second condition is true at the continuous level while @xmath113 , since we have @xmath124 , see equation in section [ subsec : basics ] . as well as for the first condition , in the context of a directional splitting , it is difficult to ensure a constant volume evolution @xmath125 after all steps of the directional splitting , where compressions or expansions of the lagrangian volume occur successively .
+ as a consequence , we will propose a form of the conservative psm scheme that does not use a directional splitting .
however we will not write the semi - lagrangian form of the psm scheme in @xmath91 dimensions of space because it is costly in computational time , because of the reconstruction step , and it is difficult to handle with arbitrary coordinate systems .
the solution we choose is to use an equivalent finite volume form of the psm scheme described in section [ psm_fv ] , which is locally 1d at each face of the mesh .
it is therefore possible to design 1d numerical limiters to try to better satisfy the maximum principle condition .
moreover , we will show that this form allows an exact conservation of the volumes in the phase space . the maximum principle and therefore the robustness of this scheme
will thus be considerably improved .
enforcing the first condition on the maximum principle of the distribution function reconstruction can be really costly in computational time . instead of trying to correct the cubic spline reconstruction
, we will reduce the spurious oscillations , generated by high order schemes when strong gradients appear in the distribution function profile , by using a classical _ van leer _ like slope limiting procedure , see for instance _ leveque _ @xcite .
we propose here to measure the gradients in the flow and to add diffusion where they are detected .
the diffusion is added by mixing the high order psm flux with a first order upwind flux .
the evaluation of the gradient is given by the classical function @xmath126 and we estimate the diffusion needed with a function @xmath127 $ ] based on a minmod like limiter function ( see fig .
[ fig : gamma_theta ] ) .
the resulting limiter we propose here is called sls ( slope limited splines ) , see @xcite for details : @xmath128 where @xmath129 we define @xmath130 as the classical slope ratio of the distribution which depends on the direction of the displacement : @xmath131 however , the classical limiter minmod , where @xmath132 , set @xmath133 to 0 when @xmath134 .
that means that the scheme turns to order 1 when an extrema exists , i.e. the slope ratio @xmath134 .
these extrema are thus quickly diffused and that leads to loose the benefits of a high order method . for sls , the choice is to let the high - order scheme deal with the extrema and only add diffusion when strong gradients occurs , i.e. the slope ratio @xmath135 .
we also introduce a constant k in relation to control the maximum slope allowed without adding diffusion , i.e. mixing with the upwind scheme , see figure [ fig : gamma_theta ] : [ cols="^ " , ] let us set a velocity field @xmath136 such that : @xmath137 let us write the conservative advection equation in polar coordinates : @xmath138 notice that the geometric jacobian @xmath139 for polar coordinates .
+ the psm scheme without directional splitting in the finite volume form reads here : @xmath140 with @xmath141 and @xmath142 positioned at cell faces center and with cell @xmath143 of volume @xmath144 and faces areas @xmath145 and @xmath146 .
the cell averaged values of @xmath147 used in the scheme are : @xmath148 using the integral form of the fluxes : @xmath149 let us introduce the volumes swept by each cell face in its normal motion in accordance with the _ green _ formula and the way of computation of feet of characteristic curves normal to cell faces , i.e. without taking into account the tangential motion at the cell faces or the curvature of the mesh , see fig .
[ fig : polvf ] : @xmath150 @xmath151 therefore we obtain : @xmath152 with @xmath153 we here recover a discrete mass conservation formulation . to obtain @xmath154 , and thus preserve a constant function ,
it yields : @xmath155 using @xmath156 definitions , we thus obtain a discrete divergence formulation in polar coordinates @xmath157 to be nullified : @xmath158 with the following first order definition of the characteristic curves feet computation : @xmath159 as a conclusion , we have presented a general methodology for any coordinate system to compute the associated discrete divergence free condition , by using the approximation of cell edges by straight lines and by only considering the normal to cell faces motion of the volume as it has to be when invoking the _ green _ formula in this finite volume framework .
the discrete divergence formulation is independent of the time integration method .
it is a discrete consistent relation for the advection field of the form @xmath157
. it should be satisfied to get the conservation condition on volumes @xmath160 in the phase space , which is necessary to obtain a maximum principle for the psm scheme or actually for any finite volume scheme .
this condition is also necessary when using the semi - lagrangian psm scheme with directional splitting as described in section [ section_psm ] as well as when using the finite volume form described in section [ psm_fv ] . in fig .
[ nodiv ] , we compare the results of a 4d drift - kinetic benchmark ( see section [ drift ] for details ) obtained with the semi - lagrangian psm scheme [ section_psm ] with an advection field computed : first in such a way the discrete divergence free condition is satisfied and second with an advection field computed by cubic spline interpolation without satisfying this condition . [ !
htbp ] result at time @xmath161 with ( left ) the advection field computed in a way ( see section [ drift ] ) that satisfy the discrete divergence condition and with ( right ) the advection field computed with cubic splines , which do not satisfy this condition .
respecting condition for the advection field not only leads to a better respect of the maximum principle , it is actually necessary to ensure the stability of the scheme .
the result in figure [ nodiv ] diverges from realistic physics .
, title="fig:",height=170 ] result at time @xmath161 with ( left ) the advection field computed in a way ( see section [ drift ] ) that satisfy the discrete divergence condition and with ( right ) the advection field computed with cubic splines , which do not satisfy this condition .
respecting condition for the advection field not only leads to a better respect of the maximum principle , it is actually necessary to ensure the stability of the scheme .
the result in figure [ nodiv ] diverges from realistic physics .
, title="fig:",height=170 ]
this work follows those of _
grandgirard et al _ in the gysela code , see @xcite and @xcite .
the geometrical assumptions of this model for ion plasma turbulence are a cylindrical geometry with @xmath162 coordinates @xmath37 and a constant magnetic field @xmath38 , where @xmath39 is the unit vector in @xmath40 direction .
the model is the 4d drift - kinetic equations described in_grandgirard et al _ @xcite : @xmath163 with @xmath164 and @xmath165 with @xmath166 the electric potential .
+ the 4d vlasov equation governing this system , where the ion distribution function is @xmath167 , is the following : @xmath168 this equation is coupled with a quasi - neutrality equation for the electric potential @xmath169 that reads : @xmath170 with @xmath171 and constant in time physical parameters @xmath25 , @xmath172 , @xmath26 and @xmath27 .
let us notice that the 4d velocity field @xmath173 @xmath174 is divergence free : @xmath175 because of variable independence @xmath176 and @xmath177 and we have @xmath164 , with @xmath165 and @xmath38 , thus @xmath178 and @xmath179 therefore , one can write an equivalent conservative equation to the preceding vlasov equation : @xmath180 we have obtained a discrete form of the velocity field divergence to nullify ( [ div_disc_pol ] ) , as a necessary condition to obtain a numerical solution with a maximum principle .
we saw in ( [ diva ] ) that @xmath181 is satisfied equivalently if @xmath182 ( [ div_rtheta ] ) is satisfied and this is still true at the discrete level ( independence of variables ) .
therefore , the velocity field should nullify the discrete polar divergence ( [ div_disc_pol ] ) : @xmath183 with @xmath184 using definitions given in ( [ def_vgc ] ) .
let us define the electric potential at the nodes of the mesh @xmath185 , whatever the way it is computed .
let us set the following natural finite difference approximation for the velocity field : @xmath186 with this approximated velocity field , the approximation of @xmath182 ( [ divrt_disc ] ) is satisfied .
the proof is easy , we just have to put the velocity field in ( [ divrt_disc ] ) to see that all terms annulate each others .
notice that the electric potential @xmath166 should be computed at nodes @xmath187 of the mesh to obtain velocities at the center of cell faces @xmath188 and @xmath189 .
it is well adapted to the psm schemes , where the displacement should be calculated at cell faces . in this section
, we will compare the numerical methods on a 4d drift - kinetic benchmark , following the paper _ grangirard et al _ @xcite .
the model is described in section [ drift4d ] .
we will compute the growth of a 4d unstable turbulent mode .
the benchmark consists of exciting the plasma mode @xmath190 , with @xmath8 the poloidal mode ( @xmath126 ) and @xmath191 the toroidal mode ( @xmath40 ) .
the initial distribution function is the sum of an equilibrium and a perturbation distribution function @xmath192 .
the equilibrium distribution function has the following form : @xmath193 and the perturbation @xmath194 @xmath195 with @xmath196 and @xmath197 two exponential functions and @xmath198 with @xmath199 the length of the domain in @xmath40 direction , @xmath10 , @xmath200 , @xmath201 physical constant profiles , see @xcite for details .
we have set here @xmath202 and @xmath203 . at the beginning of the time step ,
the distribution function @xmath204 is known at time @xmath35 , with @xmath205 .
the time step @xmath206 is computed at each step with the cfl like condition : @xmath207 with the coefficients @xmath208 , because the flow is highly non - linear in @xmath209 planes thus characteristics should not cross each others during one time step , and @xmath210 because it is linear advection in direction @xmath40 and @xmath211 so characteristics can not cross each others and then we allow a maximum displacement of @xmath212 cells . actually excluding the linear phase , the most restrictive directions for the time step are @xmath213 and @xmath126 , in such a way
this last value ( 8) has a minor importance compare to the leading parameters @xmath208 .
+ the operator splitting between the quasi - neutral equation and the vlasov transport equation is made second order using a predictor - corrector scheme in time : 1 .
time step @xmath214 computation .
quasi - neutral equation solving at @xmath35 using the distribution function @xmath61 ( actually the density ) to obtain the electric potential @xmath215 at time @xmath216 .
the advection field @xmath217 is computed with @xmath215 according to equation and using formula .
4d vlasov equation solving at @xmath35 with time step @xmath218 to obtain the distribution function @xmath219 at time @xmath216 using the advection field @xmath217 .
quasi - neutral equation solving at @xmath216 using the distribution function @xmath220 ( actually the density ) to obtain the electric potential @xmath221 at time @xmath216 .
the advection field @xmath222 is computed with @xmath221 according to equation and using formula .
4d vlasov equation solving at @xmath35 with time step @xmath214 to obtain the distribution function @xmath223 at time @xmath60 using the advection field @xmath222 . in the two following paragraphs
, we describe the schemes for the 4d vlasov equation solving of the algorithm with @xmath224 in the prediction step and @xmath225 in the correction step .
[ [ algosl ] ] 4d semi - lagrangian psm sheme with directional splitting + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + * psm 1d advection of @xmath204 in direction @xmath211 with velocity @xmath226 and time step @xmath227 to obtain @xmath228 .
* psm 1d advection of @xmath228 in direction @xmath40 with velocity @xmath229 and time step @xmath227 to obtain @xmath230 .
* psm 1d advection of @xmath230 in direction @xmath126 with velocity @xmath231 and time step @xmath227 to obtain @xmath232 .
* psm 1d advection of @xmath232 in direction @xmath213 with velocity @xmath233 and time step @xmath234 to obtain @xmath235 .
* psm 1d advection of @xmath236 in direction @xmath126 with velocity @xmath231 and time step @xmath227 to obtain @xmath237 .
* psm 1d advection of @xmath238 in direction @xmath40 with velocity @xmath229 and time step @xmath227 to obtain @xmath239 . * psm 1d advection of @xmath240 in direction @xmath211 with velocity @xmath226 and time step @xmath227 to obtain @xmath241 .
each psm 1d advection is achieved using the standard 1d semi - lagrangian psm scheme as described in section [ section_psm ] .
the directional splitting is second order by using a strang like decomposition .
since we use here a directional splitting and a second order scheme in time for the computation of the characteristic curves , the volumes are not strictly conserved in the phase space , because the scheme does not satisfy the discrete divergence free condition .
[ [ algofv ] ] 4d finite volume form of the psm scheme : + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + * psm 1d advection of @xmath204 in direction @xmath211 with velocity @xmath226 and time step @xmath227 to obtain @xmath228 . * psm 1d advection of @xmath228 in direction @xmath40 with velocity @xmath229 and time step @xmath227 to obtain @xmath230 . *
psm 2d advection of @xmath230 in each plane @xmath209 with velocities @xmath242 and time step @xmath234 to obtain @xmath243 .
* psm 1d advection of @xmath244 in direction @xmath40 with velocity @xmath229 and time step @xmath227 to obtain @xmath239 . *
psm 1d advection of @xmath240 in direction @xmath211 with velocity @xmath226 and time step @xmath227 to obtain @xmath241 .
each psm 1d advection is achieved using the standard 1d semi - lagrangian scheme as described in section [ section_psm ] .
the psm 2d advection in @xmath209 is achieved with the finite volume form as described in section [ psm_fv ] . since we use here the scheme [ psm_fv ]
, the volumes are strictly conserved in the phase space , because the scheme does satisfy the discrete divergence free condition . even if we use the semi - lagrangian psm 1d advection in @xmath40 and @xmath211 directions , the property is kept because the velocity is constant in these directions .
the mesh is @xmath245 cells in @xmath246 directions .
boundary conditions are periodic for directions @xmath126 and @xmath40 and _ neumann _ ( @xmath247 ) in @xmath213 and @xmath211 .
we first ran the reference test case with the non - conservative backward semi - lagrangian ( bsl ) scheme in section [ sec : bslscheme ] which is currently used in the gysela code .
then we ran four test cases to show the influence of each numerical treatment : the standard conservative semi - lagrangian psm scheme in section [ algosl ] with 1d directional splitting ( psm directional splitting 1d ) and the same with the sls limiter ( sls directional splitting 1d ) , the unsplit finite volume form of the psm scheme in section [ algofv ] ( psm finite volume ) and the same with the sls limiter ( sls finite volume ) .
+ the computed 4d distribution functions are pictured in fig .
[ sol1800 ] at time @xmath248 and in fig .
[ sol4400 ] at time @xmath249 .
we only present 2d slices @xmath250 of the distribution function at @xmath251 and for a given value of @xmath252 . in these figures
@xmath64 stands for @xmath213 direction and @xmath253 for the @xmath126 direction . at initial time
@xmath254 , the minimum and maximum values of the distribution function in this slice are @xmath255 and these values should be the same at any time of the computation if the maximum principle would be respected . in fig .
[ sol1800 ] , we show pictures of each scheme result at time @xmath248 , which corresponds approximately to the beginning of the non - linear turbulent phase saturation .
small structures are appearing and interact with each others .
all results are still close qualitatively . however , we already see oscillations in the solution obtained with psm ds ( directional splitting ) , where the minimum and maximum values @xmath256 are already quite different than the one at initial time .
the psm finite volume ( psm fv ) form and the sls ds better keep these extrema , but only sls finite volume keep the extrema unchanged until time @xmath248 with a really similar behaviour of the solution .
* psm finite volume * + simulation with 128x256x32x16 cells psm directional splitting 1d ( up - left ) psm finite volume ( up - right ) sls directional splitting 1d ( down - left ) sls finite volume ( down - right ) time = 1800.,title="fig : " ] simulation with 128x256x32x16 cells psm directional splitting 1d ( up - left ) psm finite volume ( up - right ) sls directional splitting 1d ( down - left ) sls finite volume ( down - right ) time = 1800.,title="fig : " ] + * sls directional splitting * * sls finite volume * + simulation with 128x256x32x16 cells psm directional splitting 1d ( up - left ) psm finite volume ( up - right ) sls directional splitting 1d ( down - left ) sls finite volume ( down - right ) time = 1800.,title="fig : " ] simulation with 128x256x32x16 cells psm directional splitting 1d ( up - left ) psm finite volume ( up - right ) sls directional splitting 1d ( down - left ) sls finite volume ( down - right ) time = 1800.,title="fig : " ] + in fig .
[ sol4400 ] , we show pictures of each scheme result at time @xmath249 when turbulence is well developed .
we see that the standard psm scheme creates a lot of unphysical oscillations ( structures are reaching the boundaries in @xmath213 ) and may crash the computation .
the psm fv form and the sls ds better keep the turbulence structures , but still oscillations are created .
the sls finite volume keep the extrema of the solution reasonably well @xmath257 ( the sls limiter does not provide a maximum principle ) and the solution is smooth .
we may say that the added diffusion with the limiter helps the scheme to diffuse subgrid structures without creating oscillations .
the divergence free property of the finite volume scheme is important to cure to solution from instabilities that can be seen at @xmath213 values close the average value of @xmath213 ( vertical line at the middle of pictures in fig . [ sol4400 ] ) in the sls directional splitting solution compare to the sls finite volume solution .
* psm finite volume * + simulation with 128x256x32x16 cells psm directional splitting 1d ( up - left ) psm finite volume ( up - right ) sls directional splitting 1d ( down - left ) sls finite volume ( down - right ) time = 4400 , with all color tables set to the minimum and maximum value at initial time.,title="fig : " ] simulation with 128x256x32x16 cells psm directional splitting 1d ( up - left ) psm finite volume ( up - right ) sls directional splitting 1d ( down - left ) sls finite volume ( down - right ) time = 4400 , with all color tables set to the minimum and maximum value at initial time.,title="fig : " ] + * sls directional splitting * * sls finite volume * + simulation with 128x256x32x16 cells psm directional splitting 1d ( up - left ) psm finite volume ( up - right ) sls directional splitting 1d ( down - left ) sls finite volume ( down - right ) time = 4400 , with all color tables set to the minimum and maximum value at initial time.,title="fig : " ] simulation with 128x256x32x16 cells psm directional splitting 1d ( up - left ) psm finite volume ( up - right ) sls directional splitting 1d ( down - left ) sls finite volume ( down - right ) time = 4400 , with all color tables set to the minimum and maximum value at initial time.,title="fig : " ] in fig .
[ solbsl ] , we see in the reference bsl solution at time @xmath248 spurious oscillations produced during the reconstruction step of the distribution function , which is the only possibility to break the maximum principle for the bsl scheme : here extrema are @xmath258 instead of values at initial time @xmath255 . at time @xmath249 , we see spurious oscillations as well , but the maximum principle is better satisfied than with the standard psm ds scheme in fig .
[ sol4400 ] , because no conservation of volumes in the phase space has to be satisfied , as it is explained in section [ sec : bslscheme ] .
* bsl time=4400 * + reference bsl simulation with 128x256x32x16 cells bsl at time=1800 with the real color table values ( left ) bsl at time=4400 with the color table set to the minimum and maximum value at initial time ( left).,title="fig : " ] reference bsl simulation with 128x256x32x16 cells bsl at time=1800 with the real color table values ( left ) bsl at time=4400 with the color table set to the minimum and maximum value at initial time ( left).,title="fig : " ] +
the psm scheme has been successfully integrated in the gysela code and has been tested on 4d drift - kinetic test cases .
we had first experimentally stated and afterward explained in this paper that the psm scheme can be unstable without taking care of a velocity field divergence free condition .
the numerical results show that the study of the volume evolution in the phase space is fruitful .
notice that this conservative scheme properly allows a directional splitting , in the semi - lagrangian or in the finite volume form , what is not the case with the bsl scheme .
the slope limited splines ( sls ) limiter is efficient to cut off spurious oscillations of the standard psm scheme by adding diffusion that helps eventually the scheme to manage small structures below the cell size .
of course , the psm scheme should be further validated as well as its integration in the gysela code using the gyrokinetic 5d model in toroidal geometry . in particular , the curvature of the mesh couple several directions by the geometrical _ jacobian _ which makes the divergence free condition more complex , as well as the writing of the quasi - neutral solver and the gyroaverage operator .
v. grandgirard , y. sarazin , p. angelino , a. bottino , n. crouseilles , g. darmet , g. dif - pradalier , x. garbet , ph .
ghendrih , s. jolliet , g. latu , e. sonnendrcker , l. villard , global full - f gyrokinetic simulations of plasma turbulence , plasma phys .
volume 49b , pp . 173182 ( december 2007 ) .
v. grandgirard , m. brunetti , p. bertrand , n. besse , x. garbet , p. ghendrih , g. manfredi , y. sarazin , o. sauter , e. sonnendrcker , j. vaclavik , l. villard , a drift - kinetic semi - lagrangian 4d code for ion turbulence simulation , j. comput .
physics , vol .
395 - 423 ( 2006 ) .
j.guterl , j .-
braeunig , n. crouseilles , v. grandgirard , g. latu , m. mehrenberger , e. sonnendrcker , test of some numerical limiters for the psm scheme for 4d drift - kinetic simulations , inria report 7467 ( november 2010 ) .
f. huot , a. ghizzo , p. bertrand , e. sonnendrcker , o. coulaud , instability of the time splitting scheme for the one - dimensional and relativistic vlasov - maxwell system , j. comput .
185 , issue 2 , pp . 512531
( 2003 ) .
m. zerroukat , n. wood , a. staniforth , application of the parabolic spline method ( psm ) to a multi - dimensional conservative semi - lagrangian transport scheme ( slice ) , j. comput .
phys , vol .
225 n.1 , pp .
935 - 948 ( 2007 ) . | the purpose of this work is simulation of magnetised plasmas in the iter project framework . in this context , kinetic vlasov - poisson like models
are used to simulate core turbulence in the tokamak in a toroidal geometry .
this leads to heavy simulations because a 6d dimensional problem has to be solved , even if reduced to a 5d in so called gyrokinetic models .
accurate schemes , parallel algorithms need to be designed to bear these simulations .
this paper describes the numerical studies to improve robustness of the conservative psm scheme in the context of its development in the gysela code . in this paper , we only consider the 4d drift - kinetic model which is the backbone of the 5d gyrokinetic models and relevant to build a robust and accurate numerical method . numerical simulation , conservative scheme , maximum principle , plasma turbulence 65m08 , 76m12 , 76n99 |
in 1968 , zenger @xcite proved the following result that is known as zenger s lemma .
[ zenger_finite ] let @xmath0 be a norm on @xmath1 , let @xmath2 for all @xmath3 , and let @xmath4 .
then there exists a vector @xmath5 with @xmath6 and @xmath7 such that the functional @xmath8 on @xmath1 defined by @xmath9 has norm one .
zenger applied theorem [ zenger_finite ] in the proof of the fact that the convex hull of the point spectrum of a linear operator is contained in its numerical range ; see also ( * ? ? ?
* section 19 ) .
some years ago , theorem [ zenger_finite ] was also applied in the theory of invariant subspaces ; see @xcite or @xcite .
the aim of this note is to extend the theorem from the space @xmath1 to the classical sequence space @xmath10 .
we omit the proof of the following known lemma , since it can be easily proved by an application of the uniform boundedness principle .
we refer the reader to @xcite or @xcite for details concerning absolute weak topologies .
[ equivalent_topologies ] for a net @xmath11 in @xmath10 and a vector @xmath12 , the following assertions are equivalent : ( a ) the net @xmath11 converges to @xmath13 in the absolute weak topology @xmath14 , i.e. , for each @xmath15 , @xmath16 \(b ) the net @xmath11 converges to @xmath13 in the weak * topology @xmath17 , i.e. , for each @xmath15 ,
@xmath18 \(c ) for each @xmath19 , @xmath20 , and the net @xmath11 is norm bounded , i.e. , @xmath21 .
we will mention several times the order ideal @xmath22 generated by a vector @xmath23 . for this smallest order ideal containing @xmath24 it holds that @xmath25 we refer the reader to @xcite or @xcite for the theory of riesz spaces .
let @xmath0 be a norm on the classical sequence space @xmath10 that is equivalent to the original norm @xmath26 , i.e. , there are numbers @xmath27 such that @xmath28 by @xmath29 we denote the predual norm to the norm @xmath0 , i.e. , the norm of a vector @xmath30 is defined by @xmath31 for @xmath32 , let @xmath33 be the natural projection on @xmath10 defined by @xmath34 since @xmath35 , the operator norm @xmath36 of @xmath33 with respect to the norm @xmath0 is at least @xmath37 .
some properties of these projections are assumed in the following extension of theorem [ zenger_finite ] from the space @xmath1 to the sequence space @xmath10 .
[ zenger ] let @xmath0 be a norm on @xmath10 that is equivalent to the norm @xmath26 , i.e. , ( [ equivalent_norms ] ) holds .
suppose also that @xmath38 let @xmath39 be a sequence of strictly positive numbers such that @xmath40 .
then there exist @xmath41 and @xmath42 such that @xmath43 clearly , we can assume that @xmath44 in ( [ equivalent_norms ] ) . by the banach - alaoglu theorem
, the unit ball @xmath45 is compact in the weak * topology @xmath46 . since the unit ball @xmath47
is contained in @xmath48 , it is also compact in @xmath46 if we show that it is closed in @xmath46 . to end this , pick an arbitrary net @xmath11 in @xmath49 converging to a vector @xmath12 . by lemma [ equivalent_topologies ] , @xmath20 for each @xmath19 , and so @xmath50 for every @xmath32 .
since the norm @xmath0 is equivalent to the norm @xmath26 , we obtain @xmath51 for every @xmath32 .
now , the inequality @xmath52 implies that @xmath53 , and so @xmath54 that is , @xmath55 as desired .
define a function @xmath56 by @xmath57 since @xmath58 for each @xmath55 , @xmath59 for each @xmath60 , and so the series above converges in @xmath61 $ ] .
we claim that the function @xmath62 is upper semicontinuous in the topology @xmath46 .
pick any @xmath63 and any net @xmath11 in @xmath49 converging to @xmath55 .
we must show that @xmath64 if @xmath65 for all @xmath66 .
for each @xmath67 we define @xmath68 by @xmath69 @xmath70 .
observe that @xmath71 $ ] , since @xmath72 . by jensen s inequality , we have , for every @xmath67 and @xmath66 , @xmath73 and so @xmath74 since @xmath75 for all @xmath60 and since @xmath76
, we have @xmath77 , and so lemma [ equivalent_topologies ] implies that @xmath78 consequently , we obtain from ( [ eq1 ] ) that @xmath79 now , by the monotone convergence theorem , @xmath80 and @xmath81 and so we obtain from ( [ eq2 ] ) that @xmath64 as claimed .
since every upper semicontinuous function attains its maximum on a compact set , there exists a vector @xmath82 such that @xmath83 for all @xmath55 . since @xmath84 for @xmath85
, we have @xmath86 . also , @xmath6 and @xmath87 for all @xmath60 .
now , fix @xmath55 such that , for some @xmath88 , @xmath89 for all @xmath19 , that is , @xmath13 belongs to the order ideal @xmath22 generated by @xmath24 . since @xmath90 for all @xmath67
, it holds that @xmath91 which rewrites to the inequality @xmath92
@xmath93 @xmath94 as @xmath95 ( @xmath96 ) .
now we will use the known fact that , for any @xmath97 , the sequence @xmath98 increases to @xmath99 , while the sequence @xmath100 decreases to @xmath99 . since @xmath101
the sums of positive terms @xmath102 increase with @xmath103 to a finite limit @xmath104 by the monotone convergence theorem , while the sums @xmath105 @xmath106 decrease with @xmath103 ( provided @xmath107 ) to the sum @xmath108 by the dominated convergence theorem .
therefore , we conclude from ( [ negative ] ) that @xmath109 and so @xmath110 since we can replace @xmath13 by @xmath111 ( @xmath112 ) , it must hold that @xmath113 that is , @xmath114 where @xmath115 @xmath70 .
given @xmath32 , define @xmath116 by @xmath117 for @xmath118 and @xmath119 otherwise .
then we have @xmath120 . inserting this @xmath13 in the inequality ( [ norm1 ] )
yields @xmath121 . since this holds for any @xmath32
, we conclude that @xmath42 and @xmath122 .
it remains to show that the inequality ( [ norm1 ] ) holds for any @xmath55 .
we may assume that @xmath123 , and so @xmath124 for all @xmath32 large enough . since for such an @xmath125 the vector @xmath126 belongs to @xmath127 , we have @xmath128 and so @xmath129
this shows that @xmath130 . on the other hand , since @xmath131 and @xmath132 , we have @xmath133 .
this completes the proof .
theorem [ zenger ] raises some questions .
two of them are answered by the following examples .
the first example shows that in theorem [ zenger ] the assumptions ( [ p_n ] ) can not be removed . the norm @xmath134 is equivalent to the original norm @xmath26 on @xmath10 , as @xmath135 .
for every @xmath12 and every @xmath32 , we have @xmath136 , and so @xmath137 and @xmath138 . if @xmath139 then @xmath140 , and thus @xmath141 .
however , for @xmath142 we have @xmath143 and @xmath144 , so that the right - hand assumption in ( [ p_n ] ) is not satisfied
. let @xmath145 be a sequence of strictly positive numbers such that @xmath146 . to prove that the conclusion of theorem [ zenger ] does not hold , pick any @xmath41 such that @xmath132 and @xmath87 for all @xmath147 .
since @xmath148 , we have @xmath149 for all @xmath147 and @xmath150 for some positive integer @xmath151 .
therefore , it follows from @xmath152 that there is @xmath32 such that @xmath153 define the vector @xmath154 by @xmath115 for @xmath19 , and the vector @xmath12 by @xmath155 for @xmath118 and @xmath119 otherwise
. then @xmath156 and @xmath157 this implies that @xmath158 , and so the conclusion of theorem [ zenger ] is not valid .
examining the proof of theorem [ zenger ] one may ask whether the vector @xmath159 is necessarily the order unit of @xmath10 , i.e. , the order ideal @xmath22 generated by @xmath24 is equal to @xmath10 .
the following example shows that this is not the case .
set @xmath160 , and define the norm on @xmath10 by @xmath161 since @xmath162 this norm is equivalent to the original norm on @xmath10 .
clearly , @xmath163 and @xmath164 .
since @xmath165 for every @xmath12 and every @xmath32 , we have @xmath166 and @xmath167 and so @xmath168 .
therefore , the norm @xmath0 satisfies all assumptions of theorem [ zenger ] .
define the vectors @xmath145 and @xmath169 by @xmath170 then @xmath171 , so that @xmath172 . in order to show that @xmath173 , choose any @xmath12 with @xmath174 . denoting @xmath175 $ ]
, we have , for all @xmath176 , @xmath177 and so @xmath178 it follows that @xmath179 @xmath180 this shows that @xmath181 , as desired .
in the well - known arrow - debreu model from the mathematical economy ( see @xcite or @xcite ) , various commodities are exchanged , produced and consumed . the classical model deals with finitely many commodities , but already debreu proposed to study commodity spaces with an infinite number of commodities . let us suppose that there are countably many commodities , and that the commodity space is a subset @xmath182 of the real vector space @xmath10 .
each vector @xmath183 represents a _ commodity bundle _ , that is , the number @xmath184 is the amount of the @xmath60-th commodity .
inputs for production are negatively signed , outputs are positively signed .
if @xmath185 is the price for one unit of the @xmath60-th commodity , then we can introduce the _ price vector _ @xmath186 , and define the _ value _ of a commodity bundle @xmath13 at prices @xmath187 by @xmath188 .
the price @xmath185 is negative in the case when the @xmath60-th commodity is noxious , and so the consumer pays to get rid of it .
hence , each price vector defines a linear functional on @xmath10 , if we define the price space as the real vector space @xmath189 . in this theory
every consumer has individual taste or _
preference _ that is a binary relation @xmath190 defined on the commodity space @xmath182 which is reflexive , transitive and total ( see @xcite or @xcite ) . such a preference is usually represented by a _ utility function _
@xmath191 in the following way : @xmath192 an important example of utility functions is a _ cobb - douglas utility function _
defined on the positive cone @xmath193 of @xmath194 by @xmath195 where @xmath2 for all @xmath60 and @xmath4 . since composing a utility function by a strictly increasing function
does not change the preference relation , we can replace the cobb - douglas utility function @xmath196 by the utility function @xmath197 hence , in our commodity space @xmath198 it is natural to take the utility function @xmath199 where @xmath2 for all @xmath60 and @xmath146 .
this is exactly the function defined in the proof of theorem [ zenger ] .
how can we interpret the conclusions of theorem [ zenger ] ?
we consider the preference relation represented by the utility function @xmath62 that is defined on the unit ball @xmath47 , where @xmath0 is an equivalent norm on @xmath10 having the properties ( [ p_n ] ) .
theorem [ zenger ] gives the commodity bundle @xmath82 which maximizes the utility function @xmath62 on @xmath49 , so that @xmath24 is the most desirable bundle in @xmath49 . at the same time
, the result determines the price vector @xmath200 with the property that the value @xmath201 of @xmath55 at prices @xmath8 is maximal for the bundle @xmath202 . c. d. aliprantis , o. burkinshaw , _ locally solid riesz spaces with applications to economics _ , second edition , mathematical surveys and monographs 105 , american mathematical society , providence , ri , 2003 . | we prove an infinite - dimensional generalization of zenger s lemma that was used in the proof of the fact that the convex hull of the point spectrum of a linear operator is contained in its numerical range .
two relevant examples are given , and possible application in the arrow - debreu model is also discussed .
6.6 mm norms , sequence spaces , arrow - debreu model |
" . ] here we summarize which permutations yield equivalent and nonequivalent metastable structures . (
i ) the cyclic variants of any permutations give equivalent structures because of the periodicity of the crystal .
( ii ) pairs of permutations exchanged by @xmath68 ( e.g. , @xmath69 " and @xmath70 " ) yield pairs of structures exchanged by the mirror plane @xmath71 ( plane perpendicular to the units , see fig .
s[fig : permutation - op ] ) .
( iii ) pairs of permutations exchanged by inversion ( e.g. , @xmath69 " and @xmath72 " ) yields structures interchanged by global inversion ( @xmath73 ) .
all the permutations related by these three operations yield equivalent structures in that they give equal values of formation enthalpy .
note that the mirror reflection with respect to plane @xmath74 ( plane parallel to the units , see fig .
s[fig : permutation - op ] ) is represented by the combination of the former operations .
we carried out the first - principles structure optimization for the structures summarized in table [ tab : struct ] , which resulted in the values of the formation enthalpy depicted in fig .
4 ( triangle ) .
note that this list exhausts all the possible structures for @xmath75 , in that any unlisted permutations can be related to either of the listed ones with the operations summarized in fig .
s[fig : permutation - op ] .
the first - principles structure optimization has also been done for other compounds for reference .
the squares in fig .
4 correspond to the body - centered monoclinic ( without spatial modulation ) and @xmath20-po sulfur ( ref . ) , @xmath76- and @xmath77-hs ( ref . ) , @xmath78-h@xmath67s@xmath16 ( ref . ) , @xmath17- and @xmath15-h@xmath14s ( ref . ) , @xmath8- and @xmath9-h@xmath16s ( ref . ) , and @xmath79-h ( ref . ) . here
we summarize the reaction(s ) that can occur in the compressed h@xmath14s and intermediate phases . as discussed in the main text ,
the pressure - induced stabilization of the hs and magnli " phases should stimulate a reaction @xmath18 ( @xmath19)@xmath55@xmath20+hs [ fig .
s[fig : transforms - summary](a ) ] .
there is also another possible reaction forming h@xmath16s slabs @xmath18 ( @xmath19)+h@xmath1@xmath552@xmath20 [ fig .
s[fig : transforms - summary](b ) ] .
the values of formation enthalpy suggests that this reaction is also stimulated upon compression [ figs . 4 and s[fig : transforms - summary](c ) ] .
although the latter reaction requires excess hydrogen atoms , it may also have been relevant in the previous experiments : observation of phases of elemental sulfur @xcite indicates h@xmath14 or any other h - rich phases as their decomposition counterparts .
( @xmath19)@xmath55@xmath20+hs and ( b ) reaction @xmath18 ( @xmath19)+h@xmath1@xmath552@xmath20 , where small and large balls represent hydrogen and sulfur atoms , respectively .
( c ) schematic picture of the possible transformation paths ( see also fig . 5 ) . ] | we theoretically give an infinite number of metastable crystal structures for the superconducting sulfur hydride h@xmath0s under pressure . previously predicted crystalline phases of h@xmath1s and h@xmath2s
have been thought to have important roles for the experimentally observed low and high @xmath3 , respectively .
the newly found structures are long - period modulated crystals where slab - like h@xmath1s and h@xmath2s regions intergrow in a microscopic scale . the extremely small formation enthalpy for the h@xmath1s h@xmath2s boundary indicated with the first - principles calculations suggests possible alloying of these phases through formation of local h@xmath2s regions .
the modulated structures and gradual alloying transformations between them explain peculiar pressure dependence of @xmath3 in sulfur hydride observed experimentally , as well as could they prevail in the experimental samples under various compression schemes .
sulfur hydride has recently been found to become superconductor at extremely high pressure around 200 k @xcite .
this is the first achievement of superconducting transition temperature ( @xmath3 ) exceeding the nitrogen boiling point among the conventional phonon mediated superconductors @xcite , as well as it has broken the long - standing record of 160 k in mercury - cuprate @xcite .
a remarkable feature observed in this superconducting phenomenon is pressure and annealing - scheme dependences of @xmath3 @xcite .
( i ) when the pressure ( @xmath4@xmath5100 gpa ) is applied to the h@xmath1s sample at room temperatures and afterwards cooled down , the observed @xmath3s amount to over 150 k , with 203k being the maximum value ( open circle in fig .
[ fig : crystals ] ) .
( ii ) by pressurizing at temperature around 100 k , on the other hand , the observed @xmath3 remains low and next rapidly increases ( open square in fig .
[ fig : crystals ] ) .
although this behavior suggests a variety of structural phases and their peculiar properties , efficient experimental observations have been obstructed by the extremely high pressure . instead , first - principles calculations have provided insights for the superconducting phases .
it is now established that some of the observed values of @xmath3 with the low- and high-@xmath6 schemes are well reproduced @xcite with crystal structures predicted for compositions of h@xmath1s @xcite , as well as h@xmath7s@xmath1 @xcite and h@xmath2s @xcite ( see fig .
[ fig : crystals ] ) .
however , consistent understanding on the observed behavior still remains unprecedented .
in particular , no clear explanation on the rapid increase of @xmath3 with the low-@xmath6 scheme has been given , despite accumulated first - principles proposals of candidate structures with various composition of h@xmath0s @xcite .
( open circle , square @xcite and star with error bar @xcite ) compared with the first - principles calculations ( solid symbols ) .
the latter data are taken from ref .
( pentagon ) , ref .
( rhombus ) , ref .
( inverted triangle ) , ref .
( circle and square ) , ref .
( double circle ) , ref .
( triangle ) , ref .
( hexagon ) and ref .
( circle with x - mark ) .
the data points in the shaded areas have been obtained from the corresponding predicted crystal structures @xcite , respectively .
note that @xmath8-h@xmath2s is a trigonally distorted variant of @xmath9-h@xmath2s . as discussed later
, we attribute the pressure dependence of @xmath3 ( open square ) to the magnli " phases . the values from ref .
( refs . )
includes the plasmon effect ( phonon anharmonic effect ) . the vertical bars connecting
the rhombi indicate the @xmath3 variation with the empirical coulomb parameter @xmath10 @xcite . ]
the current theoretical attempts have focused on the possible understanding with a minimal number of distinct structural phases .
rather , we provide a different view : not only distinct phases but also their mixture have vital roles .
specifically , we find an infinite number of metastable crystal structures having compositions h@xmath0s@xmath11 with 2/3@xmath12@xmath13@xmath123/4 , which have not been reported in the first - principles structure - search studies .
they can be understood as long - period modulated crystals formed by stacking the h@xmath1s and h@xmath2s slab - like structures , which are reminiscent of the magnli phases in transition - metal oxides @xcite .
all these structures are thermodynamically as stable as the complete separation to h@xmath1s and h@xmath2s phases .
this suggests that the microscopic _ intergrowth _ of the h@xmath1s and h@xmath2s regions requires little activation enthalpy and therefore occurs ubiquitously in the experimental situation , forming , as it were , slab - alloy phase .
we also show that the experimentally observed @xmath3-pressure curve is reproduced from these alloy - like phases .
-h@xmath14s and ( b ) ( 1 0 0 ) view of @xmath15-h@xmath1s , accompanied by their schematic pictures .
the unit cells are depicted in thin lines . in the schematic pictures ,
the bond depicted by panel ( d ) is indicated by thick dashed lines .
( c ) the fundamental unit structure , interlaced square nets " and its schematic picture . in the latter ,
sulfur and hydrogen atoms are located at the corners and the centers of the edges , respectively .
( d ) bonding structure in panels ( a ) and ( b ) , where sulfur and hydrogen atoms are depicted in yellow and red , respectively . ]
we begin with a discussion of the hidden similarity between theoretically predicted relevant structures for h@xmath14s and h@xmath16s . @xmath17-h@xmath1s and @xmath15-h@xmath1s ( fig . [
fig : crystals ] ) are known to give low @xmath3 values @xcite , which agree relatively well with the experimental values with the low-@xmath6 annealing for the low - pressure regime .
@xmath9-h@xmath2s ( fig .
[ fig : crystals ] ) has been thought to give the @xmath3 with the high-@xmath6 annealing cases @xcite , and its presence is recently confirmed with the x - ray diffraction measurements @xcite .
seen from the ( 1 1 0 ) direction [ ( 1 0 0 ) direction ] in the setup of ref . , one can notice that @xmath17- and @xmath15-h@xmath14s can be decomposed into the common unit structures [ figs .
[ fig : relations](a)(b ) ] : two h@xmath14s square nets interlaced with each other [ fig .
[ fig : relations](c ) ] .
the units of this shape are bound with each other so that sulfur atoms adopt local fcc - like stacking [ fig .
[ fig : relations](d ) ] @xcite .
the @xmath17 and @xmath15 structures can now be distinguished by the configuration of the inter - unit bonding . the uniform ( alternating ) bonding orientation yields the @xmath17 ( @xmath15 ) structure .
an important thing is that the @xmath9-h@xmath16s structure can also be formed with the present unit by binding them so that the vertical edges are shared [ fig .
[ fig : unit - bond](c ) ] . " and ( e ) its relaxed counterpart ( see the main text ) .
the @xmath9-h@xmath16s regions are shaded in red . ] the unit - based perspective for h@xmath14s and h@xmath16s is indeed helpful for us to construct a group of metastable structures . in a wide variety of materials such as metal oxides ,
multiple crystalline phases with slightly different stoichiometries emerge depending on the bonding between the unit complexes @xcite .
one of such examples is the magnli phases in molybdenum- @xcite , tungsten- @xcite , titanium- @xcite , and vanadium - oxides @xcite , where two - dimensional defects ( or _ crystallographic shear _
@xcite ) of edge- or face - sharing bonding between metal - oxide complexes are periodically formed . on the analogy to them ,
here we consider possible crystalline phases formed by the structural h@xmath14s unit .
we first define the three types of the inter - unit bonding ( fig [ fig : unit - bond ] ( a)(c ) ) : bonding @xmath18 and @xmath19 correspond to the two bonding orientations seen in @xmath15-h@xmath14s , respectively , whereas bonding @xmath20 corresponds to the bond sharing the edge . with this definition , by arranging the units side by side and next assigning any type of bonding for every neighboring pair of units , we can generate a crystal structure formed by the units . in this sense , every circular permutation composed of desired numbers of @xmath18 , @xmath19 and @xmath20 yields a corresponding long - period modulated crystal structure .
any structures generated in this way have intermediate composition h@xmath21s@xmath11 ( @xmath22 ) , whose unit cells are composed of h@xmath23s@xmath24 with @xmath25 and @xmath26 being the period of the permutation and the number of @xmath20 , respectively .
the
@xmath27 " structure is exemplified in fig .
[ fig : unit - bond](d ) , whose unit - cell formula is h@xmath28s@xmath29 . as the number of @xmath20 increases , h@xmath16s - like structural regions grow [ fig .
[ fig : unit - bond ] ( d)(e ) ] .
note that @xmath17-h@xmath14s , @xmath15-h@xmath14s , and @xmath9-h@xmath16s are generated by permutations
@xmath18 " ( also @xmath19 " ; see supplemental materials @xcite ) ,
@xmath30 " , and @xmath20 " , respectively .
denotes the number of atoms in the unit cell .
the structures corresponding to the points are summarized in supplemental materials @xcite .
the inset panels focus the region between @xmath31 ( h@xmath14s ) and @xmath32 ( h@xmath16s ) , where the formation enthalpies are measured with respect to the values for the phase separation of h@xmath14s and h@xmath16s .
bold lines indicate @xmath3350 kelvin . ] using the structures thus generated from the permutations of period 2 , 3 , and 4 as inputs , we carried out the first - principles structure optimization for various pressures .
our calculations were done using the first - principles code package quantum espresso @xcite with the generalized - gradient approximation for the exchange - correlation potential @xcite .
the unit - cell compositions of the resulting structures are h@xmath34s@xmath16 , h@xmath35s@xmath36 , h@xmath37s@xmath7,h@xmath38s@xmath29 , h@xmath39s@xmath7 , h@xmath28s@xmath29 , h@xmath40s@xmath41 , and h@xmath42s@xmath43 , respectively .
detailed conditions of the calculations are summarized in supplemental materials @xcite .
we have found that the formation enthalpy @xmath44(h@xmath0s@xmath11 ) for all the resulting structures satisfy @xmath45 ( fig .
[ fig : enthalpy ] ) .
we have also confirmed that all these structures retain the bonding characteristics in their initial structures [ fig .
[ fig : unit - bond](e ) ] .
namely , every different permutation gives different optimum metastable structures where well - defined bcc - h@xmath16s slab regions emerge .
although we do not examine the @xmath25@xmath46@xmath47 cases here , an infinite number of metastable structures are expected to be obtained in this way . remarkably , the values of formation enthalpy relative to the complete decomposition into h@xmath14s and h@xmath16s are wholly within 50 k per atom [ insets of fig . [
fig : enthalpy](a)(d ) ] .
this dependence is not an indication of trivial phase separation into h@xmath14s and h@xmath16s because the two regions _ intergrow _ in a microscopic scale in the respective structures ( see fig .
[ fig : unit - bond ] ( e ) ) .
a more appropriate interpretation is that the formation enthalpy for the h@xmath14s
h@xmath16s boundary is extremely small in these structures .
let us argue the relevance of the newly found intermediate phases in the experimental situation .
we here have to bear in mind that the intermediate structures are _ within _ the convex hull of the formation enthalpy in the present pressure regime , apart from the narrow range around 110 gpa ( fig .
[ fig : enthalpy ] and ref . ) .
this means that in the compressed h@xmath14s system , these phases do not grow as the decomposition residues of the emergence of h@xmath16s phase , if annealed with enough time and heat .
however , the decomposition paths otherwise depend on the enthalpy barrier from the pristine h@xmath14s structures .
the transition between two metastable structures with slightly different @xmath13 occurs through transformations of a small fraction of @xmath48-type inter - unit bonds to @xmath20 .
this sporadic character of the bonding transformation and the small formation enthalpy for the h@xmath14s
h@xmath16s boundary revealed above suggest that such transition requires little activation enthalpy .
the intermediate crystalline phases should thus be observable if h@xmath0s systems exhibit either of the @xmath17- or @xmath15-h@xmath1s region and next further compressed at low temperature .
the possible emergence of the intermediate magnli - type phases draws a consistent explanation on the puzzling pressure dependence of @xmath3 observed experimentally . in the present magnli " phases , @xmath3 is expected to increase as the h@xmath2s regions grow . indeed , by selecting plausible phases and calculating their superconducting @xmath3s from the first principles @xcite , we successfully reproduced the experimentally observed @xmath3 behavior ( fig.[fig : crystals ] ) : 41 k ( 130gpa ) , 80 k ( 180gpa ) , 107 k ( 190gpa ) and 121 k ( 200gpa ) . here , the increase of @xmath3 is due to enhancement of electron - phonon coupling ( see supplemental materials @xcite ) .
the succession of the magnli " phases with increasing fraction of the h@xmath2s region is hence the probable origin of the pressure dependence of @xmath3 observed with the low-@xmath6 compression scheme , filling the unsolved gap between the theory and experiments .
although we have selected the @xmath49 " , @xmath50 " , , @xmath51 " , and
@xmath18(@xmath20@xmath529 ) " phases for 130 , 180 , 190 and 200 gpa , respectively , we do not conclude these specific phases were emerging in the experimental situation in ref . but
were more various magnli phases .
for example , there has been a recent first - principles calculation that h@xmath7s@xmath1 , which is actually equivalent to the @xmath50 " phase , also yields a good agreement for @xmath53@xmath12@xmath54 gpa ( circle with x - mark in fig .
[ fig : crystals ] ; ref . ) .
( @xmath19 ) @xmath55 @xmath20+hs " .
the shaded area in the left plot represents the multiple intermediate structures , whereas that in the right indicate the h@xmath2s regions .
] s taken from ref .
, where the horizontal bars indicate regions where we can find similarity to simulated patterns for some magnli phases [ see panel ( c ) ] .
( bottom ) the simulated patterns for the reference structures .
( b ) representative simulated patterns exhibiting multiple diffraction peaks .
( c ) patterns obtained from a selected group of structures which are ordered so that the fraction of @xmath20-bond gradually varies . ]
we can also find a plausible reaction path consistent with the experiment .
the calculated formation enthalpies for hs and the magnli " phases decrease with compression compared with those for h@xmath1s [ see fig .
[ fig : enthalpy ] ( a)(c ) ] .
this pressure - induced phenomenon would stimulate a local reactions forming the h@xmath2s slab : @xmath18 ( @xmath19)@xmath55@xmath20+hs ( fig .
[ fig : transform - scheme ] ; see also supplemental materials @xcite ) .
this supports the present scenario ; the increase of the h@xmath2s region by compression .
the multiplicity of the metastable phases deduces an interesting speculation : arbitrary values of @xmath3 between those of the low- and high-@xmath3 phases in ref .
are observable .
the local h@xmath2s - slab formation can also be stimulated by annealing . by carefully controlling the annealing conditions
, one would observe various values and their temporal evolution even at fixed pressures .
here we note a possible characteristic effect in the present h@xmath0s magnli phases for future studies . in the titanium - oxide magnli phase
, it is known that the local electron - lattice coupling is enhanced by the two - dimensional defects to form bipolarons ( or charge order ) @xcite .
hydrogen atoms are in principle subject to this instability because of their multivalent character .
although we have not found traces of such polaronic phases ( see supplemental materials @xcite ) , in view of the strong electron - phonon coupling , proximity effects of such phases may affect the transport and superconducting properties @xcite . to facilitate experimental exploration of the h@xmath21s magnli phases , we computed theoretical x - ray diffraction patterns at 150gpa using rietan package ( ref . , see supplemental materials@xcite ) .
we focus on two distinct groups of structures .
when the fraction of the @xmath20 bond is small and both @xmath18 and @xmath19 bonds are present , multiple minor peaks are generally seen around the diffraction angle of the dominant peak in @xmath8-h@xmath2s [ ( 110 ) peak ; @xmath56@xmath57@xmath58@xmath59 in fig .
[ fig : xray](a ) ] , as exemplified in fig .
[ fig : xray](b ) .
this behavior is due to subtle distortion of the sulfur lattice and apparently difficult to understand with a simple combination of the patterns for the pristine @xmath8-h@xmath2s , @xmath17-h@xmath1s and @xmath15-h@xmath1s phases .
experimental noise " may actually be contributed to by the diffraction peaks from those phases .
we also show the simulated patterns when the @xmath20 bonds are gradually introduced in either @xmath17- or @xmath15-h@xmath1s phase in fig .
[ fig : xray](c ) .
we observe gradual evolutions between the end points .
remarkably , the intensity of the dominant peak is well retained while peripheral features such as sub - peaks , peak shift and tail are found . in the previous diffraction experiments @xcite , presence of the h@xmath2s phase has been indicated from the ( 110 ) peak , though it was accompanied by subtle sub - peaks depending on the experimental protocol .
the magnli phases of this group may provide a unified understanding on such structures . to demonstrate this , we plotted an experimental pattern quoted from ref . in fig .
[ fig : xray](a ) . the sub - peaks and tails around
@xmath60@xmath61@xmath62@xmath63@xmath64 well resembles to , for example , those for @xmath65 and @xmath66 , as indicated in panels ( a ) and ( c ) .
we additionally assert that not only crystalline magnli phases but also their nanoscale stripes can emerge in an ordered and random fashion , similarly to the cases of the microsyntactic intergrowth @xcite observed in silicon carbide @xcite and metallic oxides including the magnli materials @xcite .
such structures will appear in reality an alloy - like phase formed by the slabs of the metallic h@xmath14s and h@xmath16s .
this alloy phase is a compound analog of the classic superconducting alloys formed by elemental metals @xcite , but quite different from them .
the ingredients of the former alloy h@xmath1s and h@xmath2s slabs can develop in the common h@xmath1s crystalline phases and therefore , even in the pristine sample , their alloying occurs in the self - contained manner through the structural and stoichiometric transformations .
this alloying obviously smears the experimental diffraction pattern , which could appear to be amorphous - like behavior . in summary
, we have provided an interpretation based on the common structural unit for the known crystal structures of the low- and high-@xmath3 h@xmath14s and h@xmath16s and thereby found metastable magnli - type structures formed by their layered microsyntactic intergrowth .
the experimentally observed pressure dependence of @xmath3 is reasonably explained with pressure - induced gradual transformations through such alloy phases .
the present finding gives a new insight into the high-@xmath3 superconducting phenomena in sulfur hydride that the microscopic mixture of the phases can be ubiquitous .
_ -acknowledgment .
_ this work was supported by mext element strategy initiative to form core research center in japan and jsps kakenhi grant numbers 15k20940 ( r. ak ) and 15h03696 ( r. ar ) from japan society for the promotion of science ( jsps ) .
we thank yinwei li and mari einaga for sharing structure data for h@xmath67s@xmath16 in ref . and
x - ray diffraction data in ref . , respectively .
we also thank mikhail eremets , yoshihiro iwasa , hui - hai zhao , and yuta tanaka for enlightening comments .
the crystal structures were depicted using vesta @xcite .
the superconducting properties were calculated at the supercomputer center at the institute for solid state physics in the university of tokyo .
999 a. p. drozdov , m. i. eremets , and i. a. troyan , v. ksenofontov , and s. i. shylin , nature ( london ) * 525 * , 73 ( 2015 ) .
a. p. drozdov , m. i. eremets , and i. a. troyan , arxiv:1412.0460 .
j. bardeen , l. n. cooper , and j. r. schrieffer , phys . rev . * 108 * , 1175 ( 1957 ) . c. w. chu , l. gao , f. chen , z. j. huang , r. l. meng , and y. y. xue , nature ( london ) * 365 * , 323 ( 1993 ) . l. gao , y. y. xue , f. chen , q. xiong , r. l. meng , d. ramirez , c. w. chu , j. h. eggert , and h. k. mao , phys . rev .
b * 50 * , 4260(r ) ( 1994 ) . m. monteverde , c. acha , m. nez - regueiro , d. a. pavlov , k. a. lokshin , s. n. putilin , and e. v. antipov , epl ( europhys . lett . ) * 72 * , 458(2005 ) . n. takeshita , a. yamamoto , a. iyo , and h. eisaki , j. phys .
. jpn . * 82 * , 023711 ( 2013 ) .
m. einaga , m. sakata , t. ishikawa , k. shimizu , m. i. eremets , a. p. drozdov , i. a. troyan , n. hirao , and y. ohishi , arxiv:1509.03156 ; advance online publication in nat .
phys .
y. li , j. hao , h. liu , y. li , and y. ma , j. chem .
phys . * 140 * , 174712 ( 2014 ) .
d. duan , y. liu , f. tian , d. li , x. huang , z. zhao , h. yu , b. liu , w. tian , and t. cui , sci .
reports * 4 * , 6968 ( 2014 ) .
n. bernstein , c. s. hellberg , m. d. johannes , i. i. mazin , and m. j. mehl , phys .
rev .
b * 91 * , 060511(r ) ( 2015 ) . j. a. flores - livas , a. sanna , and e. k. u. gross , eur
. phys .
j. b , * 89 * , 1(2016 ) .
r. akashi , m. kawamura , s. tsuneyuki , y. nomura , and r. arita , phys .
rev .
b * 91 * , 224513 ( 2015 ) . i. errea , m. calandra , c. j. pickard , j. nelson , r. j. needs , y. li , h. liu , y. zhang , y. ma , f. mauri , phys .
rev .
lett . * 114 * , 157004 ( 2015 ) .
i. errea , m. calandra , c. j. pickard , j. r. nelson , r. j. needs , y. li , h. liu , y. zhang , y. ma , and f. mauri , nature ( london ) * 532 * , 81 ( 2016 ) .
t. ishikawa , a. nakanishi , k. shimizu , h. katayama - yoshida , t. oda , and n. suzuki , sci .
reports * 6 * , 23160 ( 2016 ) .
w. sano , t. koretsune , t. tadano , r. akashi , and r. arita , phys .
rev .
b * 93 * , 094525 ( 2016 ) .
y. li , l. wang , h. liu , y. zhang , j. hao , c. j. pickard , j. r. nelson , r. j. needs , w. li , y. huang , i. errea , m. calandra , f. mauri , and y. ma , phys .
rev .
b * 93 * , 020103(r ) ( 2016 ) .
d. duan , x. huang , f. tian , d. li , h. yu , y. liu , y. ma , b. liu , and t. cui , phys .
rev .
b * 91 * , 180502(r ) ( 2015 ) .
a. magnli , acta cryst .
* 6 * , 495 ( 1953 ) .
d. a. papaconstantopoulos , b. m. klein , m. j. mehl , and w. e. pickett , phys .
rev .
b * 91 * , 184511 ( 2015 ) .
y. quan and w. e. pickett , phys .
rev .
b * 93 * , 104526 ( 2016 ) . e. j. nicol and j. p. carbotte , phys .
rev .
b * 91 * , 220507 ( 2015 ) . c. heil and l. boeri , phys . rev .
b * 92 * , 060508(r ) ( 2015 ) . m. komelj and h. krakauer , phys . rev .
b * 92 * , 205125 ( 2015 ) .
a. bianconi and t. jarlborg , nov .
supercond .
mater * 1 * , 37 ( 2015 ) .
a. bianconi and t. jarlborg , epl * 112 * , 37001 ( 2015 ) .
t. jarlborg and a. bianconi , sci .
reports * 6 * , 24816 ( 2016 ) .
l. ortenzi , e. cappelluti , and l. pietronero , arxiv:1511.04304 .
a. p. durajski , r. szczniak , and l. pietronero , ann .
phys .
( berlin ) * 528 * , 358 ( 2016 ) .
l. p. gorkov and v. z. krezin , sci .
reports * 6 * , 25608 ( 2016 ) .
a. f. goncharov , s. s. lobanov , i. kruglov , x .-
m .
zhao , x .-
j .
chen , a. r. oganov , z. konpkov , and v. b. prakapenka , phys .
rev .
b * 93 * , 174105 ( 2016 ) .
note that the theoretically predicted @xmath8 distortion is so small as to be difficult to measure in the experiment .
we can also regard that the dominant inter - unit bonding is the s - s one @xcite instead s - h , though this difference does not change the conclusion of the later discussions .
l. bragg and g. f. claringbull , _ crystal structure of minerals _ ( g. bell and sons , ltd . ,
london , 1965 ) .
l. r. eyring and l. t. tai , in _
treatise on solid state chemistry : volume 3 crystalline and noncrystalline solids _ , edited by n. b. hannay ( plenum press , new york , 1976 ) , chapter 3 .
l. a. bursill , proc .
roy .
soc .
ser .
a * 311 * , 1505 ( 1969 ) .
d. wang , d. su , and r. schlgl , cryst .
res .
technol . * 38 * , 153 ( 2003 ) .
m. sundberg and r. j. d. tilley , phys .
stat .
sol .
a * 22 * , 677 ( 1974 ) ; r. pickering and r. j. d. tilley , j. solid state chem . * 16 * , 247 ( 1976 ) .
s. andersson , b. colln , u. kuylenstierna , and a. magnli , acta chem . scand . * 11 * , 1641 ( 1957 ) ; s. andersson , b. colln , g. kruuse , u. kuylenstierna , a. magnli , h. pestmalis , and s. sbrink , acta chem
. scand . * 11 * , 1653 ( 1957 ) .
o. terasaki and d. watanabe , jpn
. j. appl
. phys . * 10 * , 292 ( 1971 ) .
l. a. bursill and b. g. hyde , prog .
solid state chem .
* 7 * , 177 ( 197 ) s. harada , k. tanaka , and h. inui , j. appl
. phys . * 108 * , 083703 ( 2010 ) .
s. andersson and l. jahnberg , ark .
kemi * 21 * , 413 ( 1963 ) .
y. hirotsu , y. tsunashima , s. nagakura , h. kuwamoto , and h. sato , j. solid state chem .
* 43 * , 33 ( 1982 ) p. c. canfield , j. d. thompson , and g. gruner , phys .
rev .
b * 41 * , 4850(r ) ( 1990 ) .
u. schwingenschlgl and v. eyert , ann .
phys .
( berlin ) * 13 * , 475 ( 2004 ) .
a. d. wadsley , in _ fasciculus extraordinarius alfred werner , 1866 - 1919 : alfred werner commemoration volume _
( verlag helvetica chimica acta , basel , 1967 ) .
see supplemental materials ( url ) , which include refs .
o. degtyareva , e. gregoryanz , m. somayazulu , h. k. mao , and r. j. hemley , phys .
rev .
b * 71 * , 214104 ( 2005 ) . c. j. pickard and r. j. needs , nat . phys . * 3 * , 473(2007 ) .
p. giannozzi _
et al _ , j. phys . :
condens .
matter * 21 * , 395502 ( 2009 ) ; http://www.quantum - espresso.org/. j. p. perdew , k. burke , and m. ernzerhof , phys .
rev .
lett . * 77 * , 3865 ( 1996 ) .
n. troullier and j. l. martins , phys .
rev .
b * 43 * , 1993 ( 1991 ) ; http://www.abinit.org/downloads/psp-links/psp-links/gga_fhi . f. izumi and k. momma , solid state phenom . * 130 * , 15 ( 2007 ) .
m. lders , m. a. l. marques , n. n. lathiotakis , a. floris , g. profeta , l. fast , a. continenza , s. massidda , and e. k. u. gross , phys . rev .
b * 72 * , 024545 ( 2005 ) . m. a. l. marques , m. lders , n. n. lathiotakis , g. profeta , a. floris , l. fast , a. continenza , e. k. u. gross , and s. massidda , phys . rev .
b * 72 * , 024546 ( 2005 ) .
r. akashi and r. arita , phys .
rev .
b * 88 * , 014514 ( 2013 ) .
y. takada , j. phys .
. jpn . * 45 * , 786 ( 1978 ) .
r. akashi and r. arita , phys .
rev .
lett . * 111 * , 057006 ( 2013 ) .
r. akashi and r. arita , j. phys .
. jpn . * 83 * , 061016 ( 2014 ) .
s. baroni , s. de gironcoli , a. dal corso , and p. giannozzi , rev .
mod .
phys . * 73 * , 515 ( 2001 ) .
d. pines , _ elementary excitations in solids _
( benjamin , new york , 1963 ) .
r. akashi , k. nakamura , r. arita , and m. imada , phys .
rev .
b * 86 * , 054513 ( 2012 ) .
m. methfessel and a. t. paxton , phys .
rev .
b * 40 * , 3616 ( 1989 ) .
m. kawamura , y. gohda , and s. tsuneyuki , phys .
rev .
b * 89 * , 094515 ( 2014 ) .
j. rath and a. j. freeman , phys .
rev .
b * 11 * , 2109 ( 1975 ) .
p. e. blchl , o. jepsen , and o. k. andersen , phys .
rev .
b * 49 * , 16223 ( 1994 ) . p. b. allen and r. c. dynes , phys .
rev .
b * 12 * , 905 ( 1975 ) .
g. henkelman , a. arnaldsson , and h. jnsson , comput .
mater .
sci . * 36 * , 254 ( 2006 ) ; e. sanville , s. d. kenny , r. smith , and g. henkelman , j. comp .
chem .
* 28 * , 899 ( 2007 ) ; w. tang , e. sanville , and g. henkelman , j. phys . : condens .
matter * 21 * , 084204 ( 2009 ) .
s. lakkis , c. schlenker , b. k. chakraverty , r. buder , and m. marezio , phys .
rev .
b * 14 * , 1429 ( 1976 ) .
a. b. migdal , zh .
eksp .
. fiz . * 34 * , 1438 ( 1958 ) [ sov .
phys .
jetp * 7 * , 996 ( 1958 ) ] ; g. m. eliashberg , zh .
eksp .
. fiz . * 38 * , 966 ( 1960 ) [ sov .
phys .
jetp * 11 * , 696 ( 1960 ) ] ; d. j. scalapino , in _ superconductivity _ edited by r. d. parks , ( marcel dekker , new york , 1969 ) volume 1 ; j. r. schrieffer,_theory of superconductivity ; revised printing _ , ( westview press , colorado , 1971 ) .
we can also consider another possible reaction forming local h@xmath16s slabs and stimulated by compression : @xmath18 ( @xmath19)+h@xmath1@xmath552@xmath20 @xcite .
b.k .
chakraverty , j. phys .
( paris ) * 40 * , l99 ( 1979 ) ; * 42 * , 1351 ( 1981 ) .
a. s. alexandrov , zh .
fiz .
khimi , * 57 * , 273 ( 1983 ) a. s. alexandrov and j. ranninger , phys .
rev .
b * 24 * , 1164 ( 1981 ) ; a. s. alexandrov , j. ranninger , and s. robaszkiewicz , phys . rev .
b * 33 * , 4526 ( 1986 )
. a. s. alexandrov and n. f. mott , rep .
. phys . * 57 * , 1197 ( 1994 ) ; j. t. devreese and a. s. alexandrov , rep .
. phys . * 72 * , 066501 ( 2009 ) .
h. sato and s. shiozaki , mat .
res .
bull . , * 9 * , 679 ( 1974 ) .
d. j. m. bevan , b. hudson , and p. t. moseley , mat .
res .
bull . , * 9 * , 1073 ( 1974 ) .
y. hirotsu and h. sato , j. solid state chem .
* 26 * , 1 ( 1978 ) .
y. hirotsu , s. p. faile , and h. sato , mat .
res .
bull .
, * 13 * , 895 ( 1978 ) .
y. hirotsu and h. sato , mat .
res .
bull .
, * 15 * , 41 ( 1980 ) .
b. t. matthias , t. h. geballe , and v. b. compton , rev .
mod .
phys . * 35 * , 1 ( 1963 ) . k. momma and f. izumi , j. appl
. crystallogr . * 44 * , 1272 ( 2011 ) . |
of the particles seen so far in collider experiments , the top quark is no doubt the most expensive , and the most glamorous .
it is therefore the center of attention at the tevatron and the lhc , until of course a new star , higgs , comes along .
expressed less colloquially , the top quark is considered an interesting study object because it has many quantum numbers and thus couples to almost all other particles , through various ( chiral , vector , scalar ) structures , all of which bear scrutiny for deviations .
precise scrutiny is feasible because the large top mass implies , first , that it couples strongly to whatever breaks the electroweak symmetry , and second , the resulting large width minimizes obscuring hadronization effects and allows preservation of spin information .
top is also a troublemaker for the standard model , contributing significantly to the quadratic divergences of the higgs self energy , but is at the same time an life raft for beyond the standard model ( bsm ) theories such as the mssm ( raising the upper limit on the light higgs in that theory ) . with
the tevatron having made the first precious thousands top quarks , leading to its discovery and ftests of some of its properties , the lhc is a genuine top quark factory and will allow us to study the top quark in great detail . here
i review some of the interesting aspects of top quark physics .
i will necessarily be short on length and details , and i refer to other excellent recent reviews @xcite for more .
let us recall the various interactions of the top quark field @xmath0 in the standard model lagrangian .
the interaction with gluons is a vectorlike coupling involving an @xmath1 generator in the fundamental representation @xmath2_{ij } t_j(x ) g_\mu^a(x)\,,\ ] ] where @xmath3 label color .
the interaction with photons is simply vectorlike and proportional to the top quarks fractional charge @xmath4 its charged weak interaction is chiral and flavor - changing @xmath5 while it neutral weak interaction is flavor - conserving and parity violating @xmath6 the interaction of the top with the higgs boson of the yukawa type @xmath7 with a coupling constant that is directly related to its mass @xmath8 . beyond these , effective interactions such as for flavor - changing neutral currents , occur due to loop corrections , and are therefore very small .
all of these couplings could be modified in structure and strength by virtual effects due to new interactions associated with physics beyond the standard model .
this is particularly relevant for the top quark if only because it evidently couples strongly to the electroweak symmetry breaking sector ( the yukawa coupling @xmath9 in eq .
( [ eq:4 ] ) is very close to 1 in strength ) .
it is therefore important to test these structures in detail .
such studies can be guided by educated guesses about possible alternatives to the standard model , and so let us briefly review the role of the top quark or its partners in some bsm models .
if it were nt for the top quark corrections to the lightest higgs boson mass , the minimal supersymmetric standard model ( mssm ) would predict it to be lighter than the @xmath10 boson , and would thus already have been ruled out .
the maximum viable mass for this boson is thus about @xmath11 gev .
top plays an even more central role , in that its dominant contribution to the running of a higgs potential parameter down from the gut scale in fact leads to a negative eigenvalue for the higgs mass matrix , thereby explaining electroweak symmetry breaking .
moreover , in the supersymmetry searches at the lhc , both regarding discovery and subsequent unraveling , top would play a key part , as many heavy supersymmetric partners have top among its decay products .
the key idea @xcite is to construct a model in which the higgs boson emerges as a pseudo - goldstone boson , and is therefore naturally light .
this is analogous to explaining the lightness of the pion by its nature as a pseudo - goldstone boson for spontaneously broken chiral symmetry .
a number of such little higgs models have in fact now been constructed , see @xcite for a review . in these models
the top quark plays a key role , simply because its contribution to the higgs mass corrections are the most dominant , and must be cancelled to an appropiate extent , i.e. up to a certain scale .
in fact , ambitions here do not run as high as the gut scale ; the goal is to solve at least the `` little hierarchy '' problem , keeping the higgs mass natural up to a scale a factor 10 above what would still be natural for the standard model , up to about 10 tev .
perhaps the most extensively explored model is the littlest higgs model with @xmath12-parity @xcite . here
, heavier mirror copies of the four standard model electroweak gauge fields appear , which cancel the higgs mass contributions of the latter .
likewise , t - odd and t - even fermionic top - partners are introduced whose higgs mass contribution cancel , to one - loop , that of the top .
the heavy top partners , that also decay to tops , could be visible by lhc experiments . in extra - dimensional scenarios @xcite
top plays less of a central role , as the mechanism for curing the hierarchy problem is not based on particles but spacetime .
nevertheless , the kaluza - klein excited states of gluons might be best visible as resonances in top quark pair production channels , as these are more easily identified .
top has also played a key role in setting up models in which , in analogy with bcs superconductively , the higgs is effectively a fermion ( top ) boundstate , formed by new `` topcolor '' gauge interaction @xcite that views the 3rd generation as special .
evidence for such a mechanism , which would also yield charged and neutral top - pions , and possibly new heavy gauge bosons , could arise in @xmath13 invariant mass distributions . by their nature , these new particles would couple strong to top quarks , and lead to a varied phenomenology involving top final states .
a comprehensive review is ref .
the top quark property that is most readily employed in top physics is its mass .
the tevatron experiments have set the standard to a level that will be hard to pass by the lhc by measuring it to less than 1% accuracy ( @xmath14 gev ) . together with an accurately measured @xmath15 boson mass it severely constrains the mass range of a possible higgs boson both in the standard model and in the mssm .
therefore its precise measurement is of considerable importance , and so also its careful definition .
a natural definition is based on the location of the pole of the full top quark propagator , the pole mass . however , because the top quark , being colored , can never propagate out to infinite times - a requirement for the definition of a particle mass in scattering - such a pole only exists in perturbation theory , and its location is intrinsically ambiguous by @xmath16 @xcite .
a theoretically more precise definition is the @xmath17 mass @xmath18 whose relation to the pole mass is known to sufficiently high order . for @xmath19 one often takes the implicit value found when intersecting the @xmath18 curve with the @xmath20 axis , yielding @xmath21 . for the top quark ,
this value is about 10 gev smaller than the pole mass , and thus the question often arises what mass the tevatron and lhc experiments measure .
experimentally , the top quark mass is reconstructed by collecting jets and leptons .
soft particles arising from both within and outside these jets may enter them , and thus affect the reconstructed mass . moreover
, various experimental methods used ( e.g. track quality cuts ) , and monte carlo based corrections , do not have a clean perturbation theory description .
therefore the question is difficult to answer , but the pole mass should be the closest to the true answer .
the top quark inclusive cross section at hadron colliders has been a constant of theoretical attention over many years , with steady progress toward its more accurate determination .
let us look at some recent developments , for which we need a brief discussion on threshold resummation .
when the top quark pair is produced near threshold , logarithms whose argument represents the distance to threshold in the perturbative series become numerically large . it is important to note here that the definition of the threshold depends on the observable .
thus , for the inclusive cross section threshold is given by @xmath22 .
for the transverse momentum distribution we have @xmath23 , and for the doubly differential distribution in @xmath24 and rapidity we have @xmath25 .
the perturbative series for any of these ( differential ) cross sections can be expressed as @xmath26 plus non - logarithmic terms . here
@xmath27 represents any of the threshold conditions , suitably normalized , for the observables enumerated by @xmath28 .
note that it is of course allowed to use e.g. @xmath29 for the inclusive cross section , by first analyzing @xmath30 and then integrating over @xmath24 .
for any complete fixed order calculation this will give the same answer , but if one only selects the logarithmic terms because the exact answer is unknown , numerical differences will occur which can be seen as an theoretical uncertainty @xcite . the logarithms result from phase space regions where the extra gluons emitted are soft and/or collinear to their massless , on - shell emitter .
resummation concerns itself with carrying out the sum in eq .
( [ eq:7 ] ) . to this end
it is often convenient , in order to account for momentum or energy conservation conditions , to take moments with respect to @xmath27 @xmath31 the resummed result then takes the generic form @xmath32 after resummation , the inverse transform to eq .
( [ eq:8 ] ) should be taken . including up to the function @xmath33 in the exponent amounts to n@xmath34ll resummation
. not surprisingly , for higher @xmath35 these functions are progressively more difficult to determine .
key benefits of resummation are ( i ) gaining all - order control of the terms that make using fixed - order perturbation theory unreliable , thereby restoring predictive power , and ( ii ) reduction of scale uncertainty . for drell - yan or higgs production via gluon fusion
, the threshold is @xmath36 , and the resummed partonic cross section has the form @xcite @xmath37\,.\ ] ] the function @xmath38 is the cusp anomalous dimension and controls soft and collinear radiation , while @xmath39 includes contributions from soft , wide - angle emissions .
both are known to 3rd order .
the resummed top quark pair production cross sections with threshold @xmath40 is similar in form , but here @xmath41 is a matrix in color space .
most of the latest theoretical estimates for the inclusive pair production cross sections take threshold resummation to a certain accuracy into account .
furthermore , uncertainties due to scale variables ( lessened by the resummation ) , and the pdf s are included .
it should be said however that , while the top is always pair - produced fairly close to threshold at the tevatron , at the lhc this is not necessarily true , and hence the impact of threshold - resummation is smaller .
thus , cacciari et al . in @xcite
update their earlier @xcite estimates for the inclusive pair production cross section for a range of masses , for number of pdf sets , and for both nlo and nlo+nll cross sections .
they find that at the lhc the scale uncertainties are significantly larger than the pdf uncertainties , and that it is important to vary the renormalization @xmath42 and factorization scale @xmath43 independently .
nadolsky et al in ref.@xcite , in the context of using the cteq6.6 set , examine the potential of the @xmath13 cross section as a gluon density probe , and its role in normalizing certain classes of lhc cross sections .
kidonakis and vogt @xcite derive nnlo estimates for the inclusive cross section from the nll resummmation for the double differential cross section , and include uncertainties due to pdf , scale , and kinematics choice . moch and uwer @xcite also update the nlo - nll results of the results in ref .
@xcite and then extend the resummation accuracy to nnll , using threshold @xmath40 .
the required 3-loop @xmath38 @xcite and the 2-loop @xmath44 which they were able to constitute from the drell - yan and higgs equivalent functions @xcite , and results and insights from refs .
@xcite . from this result
they construct a second order estimate for the @xmath13 inclusive cross section of the form @xmath45 with @xmath46 all known , including the coulomb corrections .
furthermore , all scale - dependent logartithms , constructed using renormalization group methods , were included .
a much reduced scale uncertainty was found , albeit for @xmath43 and @xmath42 held equal .
results such as those above serve to whet the appetite for an exact nnlo calculation for the inclusive cross section , and also on this front much progress is being made .
the real calculations are now available , thanks in large measure to having the nlo @xmath47 available @xcite .
the 2-loop virtual corrections are have been obtained in the limit @xmath48 from both factorization - based calculations @xcite , and direct calculation using semi - automatized mellin - barnes techniques @xcite .
more recently @xcite , the first full 2-loop virtual results have been derived .
the exact nnlo calculation seems to be no longer infinitely far on the horizon .
single tops are produced by the weak interaction , in processes that are customarily categorized ( fig .
[ fig : singletopprocesses ] ) using born kinematics .
-channel ( 1 ) , @xmath49-channel ( 2 ) processes , and the @xmath50 associated ( 3 ) production channel . ] a particularly interesting aspect of single - top production is the prospect of directly measuring @xmath51 and testing the chiral structure of the associated vertex : top produced singly in this way is highly polarized , and offers a chance to study its spin .
furthermore , the dominant @xmath49 channel at the lhc will , when confronting measurements with a 5-flavor nlo calculation , allow extraction of the @xmath52-quark density .
this will be useful in predicting other production processes at the lhc .
the single top production characteristics are sensitive to new physics , depending on the channel .
thus , the @xmath53-channel will be sensitive to e.g. @xmath54 resonances , the @xmath49-channel to fcnc s .
experimentally , this process turns out to be very difficult to extract from backgrounds , and so far only ( strong ) evidence has been found by the d0 @xcite and cdf @xcite collaborations , with 95% cl lower limits on @xmath51 of 0.68 and 0.66 , resp .
the measured cross sections agree within errors with the nlo calculations @xcite .
the inclusive cross sections at the tevatron are rather small , 0.9 ( @xmath53 ) and 2 ( @xmath49 ) pb , with the @xmath50 channel negligible . at the lhc the numbers are ,
approximately , 10 , 246 and 60 pb , respectively .
clearly at the lhc , the @xmath49-channel will be dominant .
besides interesting in its own right , this process is a background to putative new physics processes , such as higgs production in association with a @xmath15 boson .
while the inclusive cross section has received much theoretical and experimental attention , the interest in distributions in certain variables is increasing , given the increased tevatron data set , and the lhc start .
let us review some recent developments .
the charge asymmetry is the difference in production rate for top and anti - top at fixed angle or rapidity .
while electroweak production via a @xmath10-boson could produce a ( very small ) asymmetry at lo , qcd itself does produce it at @xmath55 through a term proportional to the su(3 ) @xmath56 symbol @xcite cdf and d0 have recently performed first measurements , albeit with still large uncertainties of this asymmetry @xcite .
thus , the impact of even higher orders becomes interesting which at this stage can only be assessed from approximate , resummation based calculations @xcite , and was studied in ref .
the asymmetry was found to be stable with respect to including such higher order corrections , and to be much less sensitive to scale variations . at the lhc , where the @xmath57 channel dominates , the asymmetry is naturally small , but may be enhanced at large invariant mass , where the @xmath58 channel regains strength .
the charge asymmetry is present at leading order in @xmath59 production .
however , nlo corrections @xcite appear to wash out the asymmetry for this reaction .
another important distribution for both the tevatron and the lhc is in the invariant mass @xmath60 .
the standard model shape has relatively small uncertainty but is sensitive to the top mass , and may thus assist in determining it .
shape deviations from the qcd predictions in this distribution ( peaks , peak - dip structures ) are telltales of new physics , such as resonances with various spin , parity and color quantum numbers .
a study employing the flexibility of madgraph in a bottom - up approach was performed in ref .
@xcite , in which only the most generic aspects of new models are used .
[ fig : mttbar ] @xcite contains the invariant mass distribution for a number of @xmath53-channel resonances , showing indeed marked dependence on the quantum numbers of the resonance .
part of the attractiveness of the top quark as a study object is its power to self - analyze its spin , through its purely left - handed sm weak decay .
this is both a useful aid in signal - background separations , and itself a property worthy of detailed scrutiny , as certain new physics models could introduce right - handed parts .
the correlation between top spin and directional emission probability for its decay products is expressed through @xmath61 where @xmath62 , with 1 indication 100% correlation .
the dominant decay mode @xmath63 at lowest order , we have @xmath64 . qcd corrections to these values
are small .
the charged lepton direction ( or the down - type quark in a hadronic decay of the intermediate @xmath15 ) is indeed 100% correlated with the top quark spin .
this is amusingly more than for its parent @xmath15 boson , a consequence of interference of two amplitudes with different @xmath15 polarizations . in single - top quark production , which occurs via the charged weak interaction ,
the top is produced left - handed , so a correlation should be a clear feature of the production process and a discriminant from the background . in fig .
[ fig : stspincorr ] this correlation as computed with mc@nlo @xcite is shown .
-channel single - top production at the tevatron , a clear correlation of the lepton flight direction with the recoiling light quark jet .
the correlation disappears when spin - correlations are turned off in mc@nlo @xcite . ] in top quark pair production a correlation of an individual quark with a fixed direction is almost absent , however there is a clear correlation between the top and anti - top spins .
the size of the correlation depends on the choice of reference axes @xmath65 @xcite . at the tevatron
the beam direction @xmath66 is good choice , at the lhc the helicity axes @xmath67 should give near - maximal correlation @xmath68 indeed , the correlation coefficient @xmath69 depends in fact on the correlation axis .
thus , at lo in qcd , the values for @xmath70 at the tevatron ( lhc ) is @xmath71 ( @xmath72 ) .
nlo corrections modify these number somewhat @xcite .
bsm models that influence the pair production mechanism ( e.g. new resonances ) could noticeably influence these correlations .
many interesting top producing reactions produce other particles in association .
these reactions allow new tests of the top sm interactions , such as its coupling to the photon , @xmath10 or higgs boson
. among the most interesting is @xmath73 , which , if a good sample can be isolated , would allow a direct determination of the top yukawa coupling ( the sm value is very close to 1 ) . while it may take some time to gather sufficient data that allow the higgs to be cleanly identified and reconstructed ( via the @xmath74 decay mode ) and backgrounds
may be large .
a nlo calculation has been carried out @xcite using a variety of methods , the @xmath75 kinematics with different masses of the final state particles making the calculations challenging .
study of associated production with an electroweak boson could reveal anomalous couplings with the top , different from those in section [ sec : top - quark - standard ] .
robust theoretical tools exist @xcite which will allow fairly accurate determinations of these couplings using lhc data .
production of @xmath13 with a jet is another interesting reaction .
correlations of the top quarks with the extra jet in the final state in this process can yield a more detailed understanding the top quark reactions .
as such it is a also an important background to bsm processes .
a nlo calculation was completed recently @xcite .
its role in the nnlo inclusive cross section calculation and the determination of the charge asymmetry have been mentioned already .
it is interesting to note that in order to prevent errors in the face of severe technical complexities , the calculation was purposely carried out , within one collaboration , in two , fully independent calculational chains .
perhaps the most widely useful progress in describing top quark processes at hadron colliders is in the realm of monte carlo .
efforts in recent years have led to descriptions beyond @xmath76 processes in lo qcd ( with subsequent decay and parton showering ) in general purpose monte carlo programs .
these fall short when extra hard jets are present besides the top quarks , nor are they intrinsically normalized as their only scale dependence in the coupling , with no compensating terms in the matrix element . much ingenuity and labor has been brought to bear to remedy these deficiences .
let us review some of this .
higher - multiplicity matrix element monte carlo s now reach @xmath13 plus up to six jets , and use a variety of methods .
alpgen ( @xmath77 jets ) does not use feynman diagrams but recursion relations to compute the matrix element .
comphep ( @xmath78 jets ) uses squared amplitudes .
madgraph / madevent ( @xmath79 jets ) uses complex helicity amplitudes .
however , while matrix element monte carlo s improve the description of radiative hard emission emission events , they should if possible not sacrifice the power of the parton showers to account for collinear and soft radiation .
matching procedures have been defined to this end .
ckkw @xcite uses @xmath80 clustering to separate phase space into two regions in each of which one of the descriptions should hold . to match properly , the matrix elements are reweighted by sudakov form factors and @xmath81 factors at the scales correspond to the nodal branchings . on the ps side , the showers are vetoed to ensure that only emissions below the matching scale are included .
mlm @xcite also reweights the matrix elements , then showers them , but discards events where the shower generates emission harder than the matching scale .
both procedures have been implemented in a number of matrix - element event generators , and extensively compared @xcite .
other very important progress has been in made in matching nlo to parton shower - based monte carlo ( mc@nlo @xcite and powheg @xcite ) .
matching is essentially an issue of avoiding double counting in the one - emission contribution , which can either come from nlo or from the ps , and in the virtual parts , between the virtual nlo part and the sudakov form factors .
mc@nlo matches , in practice , to herwig angular - ordered showers .
the benefit of this matching is clearly visible in fig .
[ fig : pttpair ] @xcite . @xcite . ]
a small percentage of the events mc@nlo generates have a negative weight , reflecting virtual contributions and subtractions present in nlo and matching .
powheg insists on having positive weights , and exponentiates the complete first order real matrix element to that end .
both these frameworks are growing in the list of processes , and realism ( e.g. spin correlations @xcite ) .
agreement is generally very good , see fig .
[ fig : psnlotoppt ] , also with ps - matched matrix - element generators @xcite , although interesting differences exist .
such differences reflect genuine ambiguities .
an interesting issue arises in the @xmath50 mode of single top production .
some diagrams occurring at nlo contain an intermediate anti - top that can become resonant .
these diagrams can be interpreted as lo @xmath13 `` doubly resonant '' production , with subsequent @xmath82 decay , see fig .
[ fig : wtlo ] . production . ]
it thus becomes an issue to what extent the @xmath50 and @xmath13 can be properly defined as individual processes .
several definitions of the @xmath50 channel have been given in the literature , each with the aim of recovering a well - behaved expansion in @xmath81 .
the problem of interference in fact affects any computation that considers contributions beyond the leading order , i.e. at least @xmath83 .
the cross section at this order has been previously presented in refs .
@xcite , where only tree - level graphs were considered , and in refs .
@xcite , where one - loop contributions were included as well . in ref .
@xcite the issue of interference was addressed extensively in the context of event generation , in particular the mc@nlo framework .
two different procedures for subtracting the doubly - resonant contributions and recovering a perturbatively well - behaved @xmath50 cross section were defined . in `` diagram removal ( dr ) '' the graphs in fig .
[ fig : wtlo ] were eliminated from the calculation , while in `` diagram subtraction ( ds ) '' the doubly resonant contribution was removed via a counterterm . while the former method is strictly speaking not gauge - invariant , it was shown , first , that gauge variations are very small , and second that the answer is very close to the gauge invariant second procedure .
the ds procedure leads to the following expression for the cross section @xmath84 where @xmath85 labels the initial state channel in which the doubly - resonant contribution occurs : @xmath86 .
@xmath87 is the square of the non - resonant diagrams , @xmath88 their interference with the square of graphs of fig .
[ fig : wtlo ] @xmath41 .
the subtraction term @xmath89 requires careful construction .
@xmath90 it was shown that , with suitable cuts , the interference terms are small , as shown in fig .
[ fig : ptdrds ] . from eq .
( [ eq:5 ] ) it is straightforward to see that the difference of dr and ds is essentially the interference term . a particularly suitable cut is a puttting a maximum on the @xmath24 of the second hardest @xmath52-flavored hadron , a generalization of a proposal made in ref .
@xcite .
spectra in dr and ds scheme for lepton from top decay @xcite . ]
thus defined , the @xmath50 and @xmath13 cross sections can be separatedly considered to nlo .
their separation at lhc does remain difficult however .
top quark physics is at present at a pivotal point , between the tevatron and the lhc . as an enterprise it must progress from promising to performing .
the standard model behavior of the top quark has withstood the first tests at the tevatron , and must in the next few years face a barrage of highly detailed and varied tests by the lhc experiments .
top s attractiveness as a study object has by no means diminished . on the contrary ,
new observables are being enlisted to this end .
the characteristics of production and decay , in association with other particles , are very revealing .
top does not hide its spin , and awareness of the importance of studying angular distributions of its decays has grown .
the theoretical tools for top physics studies are good , and keep improving with remarkable pace . with these , and the turn - on of the lhc at the moment of writing , it is a very safe bet indeed to state that the top quark will remain special for years to come .
i would like to thank the organizers for a wonderful and stimulating conference .
i am grateful for clarifying discussions with stefano frixione , carlo oleari and chris white .
this work has been supported by the foundation for fundamental research of matter ( fom ) , and by the national organization for scientific research ( nwo ) w. bernreuther , j. phys . * g35 * , 083001 ( 2008 ) , [ 0805.1333 ] . t. han , 0804.3178 .
a. quadt , eur .
j. * c48 * , 835 ( 2006 ) .
d. chakraborty , j. konigsberg and d. l. rainwater , ann . rev
. nucl . part .
sci . * 53 * , 301 ( 2003 ) , [ hep - ph/0303092 ] .
m. beneke _ et al .
_ , hep - ph/0003033 . n. arkani - hamed , a. g. cohen and h. georgi , phys . lett . *
b513 * , 232 ( 2001 ) , [ hep - ph/0105239 ] .
n. arkani - hamed , a. g. cohen , t. gregoire and j. g. wacker , jhep * 08 * , 020 ( 2002 ) , [ hep - ph/0202089 ] . m. schmaltz and d. tucker - smith , ann .
nucl . part .
* 55 * , 229 ( 2005 ) , [ hep - ph/0502182 ] .
cheng and i. low , jhep * 09 * , 051 ( 2003 ) , [ hep - ph/0308199 ] .
cheng and i. low , jhep * 08 * , 061 ( 2004 ) , [ hep - ph/0405243 ] . n. arkani - hamed , s. dimopoulos and g. r. dvali , phys . lett . *
b429 * , 263 ( 1998 ) , [ hep - ph/9803315 ] .
l. randall and r. sundrum , phys .
lett . * 83 * , 3370 ( 1999 ) , [ hep - ph/9905221 ] . c. t. hill , phys
* b266 * , 419 ( 1991 ) . c. t. hill and s. j. parke , phys
. rev . * d49 * , 4454 ( 1994 ) , [ hep - ph/9312324 ] . c. t. hill and e. h. simmons , phys
. rept . * 381 * , 235 ( 2003 ) , [ hep - ph/0203079 ] .
m. beneke and v. m. braun , nucl . phys .
* b426 * , 301 ( 1994 ) , [ hep - ph/9402364 ] . i. i. bigi , m. a. shifman , n. g. uraltsev and a. i. vainshtein , phys . rev .
* d50 * , 2234 ( 1994 ) , [ hep - ph/9402360 ] .
m. c. smith and s. s. willenbrock , phys .
* 79 * , 3825 ( 1997 ) , [ hep - ph/9612329 ]
. n. kidonakis , e. laenen , s. moch and r. vogt , phys . rev . * d64 * , 114001 ( 2001 ) , [ hep - ph/0105041 ] . g. sterman , nucl .
b281 * , 310 ( 1987 ) .
s. catani and l. trentadue , nucl . phys . * b327 * , 323 ( 1989 ) .
s. catani , d. de florian , m. grazzini and p. nason , jhep * 0307 * , 028 ( 2003 ) , [ hep - ph/0306211 ] . s. forte and g. ridolfi , nucl .
b650 * , 229 ( 2003 ) , [ hep - ph/0209154 ] .
t. o. eynck , e. laenen and l. magnea , jhep * 06 * , 057 ( 2003 ) , [ hep - ph/0305179 ] .
e. laenen and l. magnea , phys
. lett . * b632 * , 270 ( 2006 ) , [ hep - ph/0508284 ] . t. becher , m. neubert and g. xu , jhep * 07 * , 030 ( 2008 ) , [ 0710.0680 ] .
v. ravindran , nucl . phys . *
b752 * , 173 ( 2006 ) , [ hep - ph/0603041 ] .
a. idilbi , x .- d .
ji and f. yuan , nucl .
b753 * , 42 ( 2006 ) , [ hep - ph/0605068 ] .
m. cacciari , s. frixione , m. m. mangano , p. nason and g. ridolfi , 0804.2800 .
m. cacciari , s. frixione , m. l. mangano , p. nason and g. ridolfi , jhep * 04 * , 068 ( 2004 ) , [ hep - ph/0303085 ] .
m. nadolsky _
et al . _ ,
phys . rev . * d78
* , 013004 ( 2008 ) , [ 0802.0007 ] . n. kidonakis and r. vogt , 0805.3844 . s. moch and p. uwer , phys . rev . * d78 * , 034003 ( 2008 ) , [ 0804.1476 ] .
s. moch , j. a. m. vermaseren and a. vogt , nucl .
* b688 * , 101 ( 2004 ) , [ hep - ph/0403192 ] .
a. vogt , s. moch and j. a. m. vermaseren , nucl .
* b691 * , 129 ( 2004 ) , [ hep - ph/0404111 ] .
a. vogt , phys .
* b497 * , 228 ( 2001 ) , [ hep - ph/0010146 ] .
h. contopanagos , e. laenen and g. sterman , nucl . phys . * b484 * , 303 ( 1997 ) , [ hep - ph/9604313 ] .
a. mitov and s. moch , jhep * 05 * , 001 ( 2007 ) , [ hep - ph/0612149 ] .
w. bernreuther _ et al .
_ , nucl . phys . *
b706 * , 245 ( 2005 ) , [ hep - ph/0406046 ] .
s. mert aybat , l. j. dixon and g. sterman , phys . rev .
* d74 * , 074004 ( 2006 ) , [ hep - ph/0607309 ] . s. dittmaier , p. uwer and s. weinzierl , phys .
lett . * 98 * , 262002 ( 2007 ) , [ hep - ph/0703120 ] . m. czakon , a. mitov and s. moch , nucl .
* b798 * , 210 ( 2008 ) , [ 0707.4139 ] . m. czakon , phys . lett . * b664 * , 307 ( 2008 ) , [ 0803.1400 ] .
d0 , v. m. abazov _
et al . _ ,
lett . * 98 * , 181802 ( 2007 ) , [ hep - ex/0612052 ] .
cdf , t. aaltonen _ et al .
_ , 0809.2581
. b. w. harris , e. laenen , l. phaf , z. sullivan and s. weinzierl , phys .
rev . * d66 * , 054024 ( 2002 ) , [ hep - ph/0207055 ] .
cao , r. schwienhorst and c. p. yuan , phys .
* d71 * , 054023 ( 2005 ) , [ hep - ph/0409040 ] .
cao , r. schwienhorst , j. a. benitez , r. brock and c. p. yuan , hep - ph/0504230 .
j. campbell , r. k. ellis and f. tramontano , phys . rev .
* d70 * , 094012 ( 2004 ) , [ hep - ph/0408158 ] . j. campbell and f. tramontano , nucl .
b726 * , 109 ( 2005 ) , [ hep - ph/0506289 ] .
p. nason , s. dawson and r. k. ellis , nucl . phys .
* b327 * , 49 ( 1989 ) . w. beenakker , w. l. van neerven , r. meng , g. a. schuler and j. smith , nucl . phys .
* b351 * , 507 ( 1991 )
. j. h. kuhn and g. rodrigo , phys . rev . * d59 * , 054017 ( 1999 ) , [ hep - ph/9807420 ] .
d0 , v. m. abazov _
lett . * 100 * , 142002 ( 2008 ) , [ 0712.0851 ] .
cdf , t. aaltonen _ et al .
_ , 0806.2472
. l. g. almeida , g. sterman and w. vogelsang , phys . rev . *
d78 * , 014008 ( 2008 ) , [ 0805.1885 ] .
r. frederix and f. maltoni , 0712.2355 .
s. frixione , e. laenen , p. motylinski and b. r. webber , jhep * 04 * , 081 ( 2007 ) , [ hep - ph/0702198 ] . w. bernreuther , a. brandenburg and z. g. si , phys .
b483 * , 99 ( 2000 ) , [ hep - ph/0004184 ] .
w. bernreuther , a. brandenburg , z. g. si and p. uwer , phys . lett . *
b509 * , 53 ( 2001 ) , [ hep - ph/0104096 ] .
g. mahlon and s. parke , phys . lett .
* b411 * , 173 ( 1997 ) , [ hep - ph/9706304 ] . w. bernreuther , a. brandenburg , z. g. si and p. uwer , nucl . phys . *
b690 * , 81 ( 2004 ) , [ hep - ph/0403035 ] .
w. beenakker _
et al . _ ,
b653 * , 151 ( 2003 ) , [ hep - ph/0211352 ] .
s. dawson , l. h. orr , l. reina and d. wackeroth , phys .
* d67 * , 071503 ( 2003 ) , [ hep - ph/0211438 ] .
u. baur , a. juste , d. rainwater and l. h. orr , phys .
d73 * , 034016 ( 2006 ) , [ hep - ph/0512262 ] .
u. baur , a. juste , l. h. orr and d. rainwater , phys .
rev . * d71 * , 054013 ( 2005 ) , [ hep - ph/0412021 ] .
a. lazopoulos , t. mcelmurry , k. melnikov and f. petriello , 0804.2220 .
s. catani , f. krauss , r. kuhn and b. r. webber , jhep * 11 * , 063 ( 2001 ) , [ hep - ph/0109231 ] . m. l. mangano , ( 2004 ) , available at ` h`ttp://cern.ch/ mlm / talks / lund - alpgen.pdf .
j. alwall _ et al .
j. * c53 * , 473 ( 2008 ) , [ 0706.2569 ] .
s. frixione and b. r. webber , jhep * 06 * , 029 ( 2002 ) , [ hep - ph/0204244 ] .
p. nason , jhep * 11 * , 040 ( 2004 ) , [ hep - ph/0409146 ] . s. frixione , p. nason and b. r. webber , jhep * 08 * , 007 ( 2003 ) , [ hep - ph/0305252 ]
. m. l. mangano , m. moretti , f. piccinini and m. treccani , jhep * 01 * , 013 ( 2007 ) , [ hep - ph/0611129 ] .
s. frixione , p. nason and g. ridolfi , jhep * 09 * , 126 ( 2007 ) , [ 0707.3088 ] .
t. m. p. tait , phys . rev . * d61 * , 034001 ( 2000 ) , [ hep - ph/9909352 ] .
a. s. belyaev , e. e. boos and l. v. dudko , phys .
* d59 * , 075001 ( 1999 ) , [ hep - ph/9806332 ] . b. p. kersevan and i. hinchliffe , jhep * 09 * , 033 ( 2006 ) , [ hep - ph/0603068 ] . s. zhu , phys .
b524 * , 283 ( 2002 ) .
cao , 0801.1539 .
s. frixione , e. laenen , p. motylinski , b. r. webber and c. d. white , jhep * 07 * , 029 ( 2008 ) , [ 0805.3067 ] . | i review how the top quark is embedded in the standard model and some its proposed extensions , and how it manifests itself in various hadron collider signals . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.