article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the measurement and generation of photons is one of most challenging tasks in nowadays quantum optics .
since the photon can be considered as the carrier of information , it plays a fundamental role in quantum information and quantum communications @xcite .
the generation of single photons is typically described by conditional measurements of quantum correlated light fields with single photon detectors @xcite .
the sophisticated task is to find a proper device that can detect at least in a good approximation single photons . among other approaches ,
e.g. , @xcite , an avalanche photo diode in geiger mode is an experimentally accessible device , being close to a single photon counter @xcite .
however , this detector not only produces a click signal in the case of a single incident photon , but also when multiple photons have been absorbed . a possible way to partially overcome this ambiguity is given by a joint measurement of a signal with multiple click detectors .
one implementation is given by so - called multiplexing detection schemes , see , e.g. , @xcite , and another one employs detector arrays , see , e.g. , @xcite . in both scenarios ,
the incident light is split equally into @xmath0 modes , and each mode can be measured with one of those on - off diodes .
the probability @xmath1 for a total number of @xmath2 clicks is described through the click counting distribution @xcite , @xmath3 with @xmath4 denoting the number of clicks , @xmath0 being the number of on - off detectors , @xmath5 the quantum efficiency , and @xmath6 denotes the normal ordering prescription .
based on the variance of these statistics , it is possible to identify in theory and experiment nonclassical , i.e. sub - binomial , light @xcite .
a source of correlated photon pairs together with the considered class of detectors can be used , for example , to identify spatial correlations @xcite or to perform a detector calibration @xcite . moreover ,
these states allow one to predict the presence of a photon in one mode , if a photon is detected in the other one .
for the first time , this so - called heralded generation of single photons has been experimentally realized in ref .
recent developments led to a further enhancement of this kind of single - photon source @xcite .
other protocols to manipulate quantum states in theory and experiment in the single photon regime are known as photon addition and photon subtraction @xcite .
these engineering protocols can be also used to add or subtract several photons @xcite or to probe experimentally fundamental commutation relations @xcite .
another prominent example for controlled state manipulation is noiseless amplification @xcite .
all these engineering processes are able to enhance quantum properties , such as entanglement , for applications in quantum information science , see , e.g. , @xcite . in the present contribution
, we study the quantum state engineering for conditional measurements with on - off detector systems
. we will consider three schemes .
first , the conditional measurement in one mode of a quantum correlated bipartite states .
second , a multi - photon subtraction protocol , and , third , a multiple photon addition process . in all scenarios , we consider imperfect detectors for a realistic description of the underlying physical situation . as an example , we study a noiseless amplification scenario and apply it to the conversion of coherent light into different types of nonclassical states .
the paper is structured as follows . in sec .
[ sec : heralding ] , we consider the heralded quantum state manipulation by measurements with on - off detector systems . in secs .
[ sec : subtraction ] and [ sec : addition ] , we describe the multi - photon subtraction and addition processes , respectively , applying the same detection process . a combination of addition and subtraction yields the noiseless amplification procedure , which is applied in sec .
[ sec : na ] to the engineering of nonclassical states .
we conclude in sec .
[ sec : conclusions ] .
let us start with the heralding scheme , see fig .
[ fig : heraldingscheme ] . a quantum correlated light field may be generated by a parametric process in order to certify a photon - photon correlation between two modes : @xmath7 and @xmath8 . in this scenario ,
one part of the radiation field ( mode @xmath8 ) is measured with a click detector system .
in such a detector the incident light beam is split into @xmath0 output states with equal intensities .
each of the resulting beams is measured with a single avalanche photo diode .
the joint number of clicks @xmath2 yields the click counting statistics @xmath1 in eq . , for @xmath4 .
the remaining part of the field ( mode @xmath7 ) is further processed , solely if @xmath2 clicks have been measured in mode @xmath8 .
click event of the detector .
, width=158 ] before we start , let us briefly review some properties of a click counter as a photon number resolving device . the click counting statistics in eq
is the quantum analog to the binomial statistics . whereas the true photon statistics of a quantum state @xmath9 is given by @xmath10 for @xmath11
resembling a poissonian form @xcite .
in previous works , we have shown that the binomial form requires a reformulation of nonclassicality conditions @xcite .
moments of the click counting statistics can be used to identify quantum correlations even between multiple detector systems . moreover , an extension to general photon absorption processes led to the description of on - off diodes working in a non - linear interaction regime , e.g. , two - photon absorption .
similarly , it allows the description of noise models , e.g. , dark counts .
it is worth mentioning that the click counting statistics approaches the photoelectric counting theory in the limit of an infinite number of avalanche diodes , @xmath12 .
since we aim to describe protocols for quantum state engineering , it is important to describe the click counting statistics in terms of a positive operator valued measure ( povm ) . in general
, the povm element , with @xmath13 for @xmath14 , can be written as @xmath15 hence , all povm elements can be written as a linear combination of the operator @xmath16 for different values of @xmath17 . using the @xmath18-symbol , cf .
appendix [ app : dsym ] , we get @xmath19 in the following , we are going to use this representation for conditional measurements .
a general , bipartite input state @xmath20 is given in fock basis expansion as @xmath21 using the povm element @xmath22 for @xmath2 clicks , we find @xmath23\right)\\ = & \sum_{p , q=0}^\infty\left[\sum_{m = k}^\infty \mathcal d^{1-\eta,\eta}_{k , m}\rho_{p , q , m , m}\right]|p\rangle\langle q| .
\label{eq : heraldingout } \end{aligned}\ ] ] this result already represents the output state triggered to @xmath2 clicks of the scheme in fig . [
fig : heraldingscheme ] for @xmath0 diodes , including the quantum efficiency @xmath5 . as an example
let us consider a two - mode squeezed - vacuum state undergoing a full phase diffusion , @xmath24 with @xmath25 .
interestingly , although this state is considered to be classically correlated with respect to a number of notions of quantumness , it has been shown to be two - mode quantum correlated @xcite . in fig .
[ fig : heralding ] , we show the photon statistics @xmath26 of the heralded output states in eq . , depending on the number of clicks @xmath2 of the detector , @xmath27 the aim to generate a @xmath2-photon state by a post - selection is properly approximated even with a finite quantum efficiency , @xmath28 .
note that the photon distribution is normalized to one for each number of clicks @xmath2 .
the contribution of higher photon numbers is not negligible for larger values of @xmath2 , however it can be reduced by increasing the number of diodes . in the limit , @xmath12 and @xmath29 ,
the output state would be a perfect @xmath2-photon state , cf .
appendix [ app : erre ] . is shown for the photon - photon correlated input in eq .
( @xmath30 ) triggered to @xmath2 clicks .
for a small number of clicks , @xmath31 , and a high quantum efficiency , @xmath32 , the heralding yields a good estimate to a @xmath2-photon state . , width=317 ]
the idea of a single - photon subtraction is the implementation of the operation : @xmath33}. \end{aligned}\ ] ] such an operation is a classical one , cf .
, e.g. , @xcite .
applying this operation @xmath2 times would yield a @xmath2-photon subtraction .
we aim to formulate this operation within a single detection process with detector systems of multiple on - off diodes . for this reason ,
we study the scenario in fig . [
fig : subtractionscheme ] .
a light beam enters one input of a beam splitter
vacuum is supposed at the other input port .
the transmissivity and reflectivity of the beam splitter are @xmath34 and @xmath35 , respectively , with @xmath36 .
note that , without loss of generality , @xmath34 and @xmath35 can be chosen to be positive real numbers . an on - off detector system is applied to measure one of the outputs ( mode @xmath8 ) .
we post - select those output states in mode @xmath7 , which corresponds to @xmath2 clicks of the detector .
clicks of the detector will be further processed .
, width=158 ] let us start with a coherent input field @xmath37 .
the beam splitter transforms the input field as @xmath38 we perform a partial trace with @xmath39 , cf .
eq . , in the second mode
. hence , we get a resulting state as @xmath40\right ) = { \rm e}^{-\lambda r^2|\alpha|^2}|t\alpha\rangle\langle t\alpha| \\=&\sum_{m=0}^\infty \frac{1}{m!}\left(-\lambda\frac{r^2}{t^2}\right)^m \hat a{}^m|t\alpha\rangle\langle t\alpha|\hat a^\dagger{}^m .
\label{eq : subtractioncoherent } \end{aligned}\ ] ] let us note that the scaling @xmath34 of the coherent amplitude @xmath41 can be understood as a loss process .
a general input state @xmath20 may be given in the glauber - sudarshan representation @xcite as @xmath42 then a loss operation @xmath43 can be written as @xmath44 with the scaled @xmath45 function : @xmath46 using the decomposition in eq . together with eq . , we get for the conditional measurement of @xmath2 clicks , @xmath22 , the output state : @xmath47\frac{r^2}{t^2}\right)^m\right ] \hat a{}^m\lambda^{\rm ( loss)}_{t}(\hat\rho_{\rm in})\hat a^\dagger{}^m \\=&\sum_{m = k}^\infty \frac{1}{m!}\mathcal d_{k , m}^{-\eta',\eta'}\hat a{}^m\lambda^{\rm ( loss)}_{t}(\hat\rho_{\rm in})\hat a^\dagger{}^m , \label{eq : subtractionkrauss } \end{aligned}\ ] ] with an effective efficiency of @xmath48 where we used the property of the @xmath18-symbol that it vanishes , @xmath49 , for @xmath50 , cf . appendix [ app : dsym ] .
the process in eq . , @xmath51 , describes the @xmath2 click conditioned measurement protocol in fig .
[ fig : subtractionscheme ] .
the output state , @xmath52 , is normalized to the probability @xmath53 to have @xmath2 clicks .
let us note that the process in eq .
not only subtracts @xmath2 photons .
it also includes higher numbers of subtractions .
moreover , the output state undergoes a loss process @xmath54 .
this feature is a consequence of the process itself and not a result of the imperfect detector , since it is independent of the quantum efficiency @xmath5 . in fig .
[ fig : subtractionexample ] , we study this quantum process for a thermal input field , @xmath55 with a mean photon number @xmath56 . applying the loss process
@xmath54 is equivalent to a scaling of the mean photon number : @xmath57 .
the output @xmath45 function triggered to a @xmath2 click event is @xmath58 cf .
appendix [ app : dths ] , where we additionally give analytical expressions for displaced thermal states .
the plots of @xmath59 in fig .
[ fig : subtractionexample ] are additionally normalized to one , @xmath60 .
the number of individual click diodes , @xmath61 , corresponds to a @xmath62 array detector or to a multiplexing detector with a depth @xmath63 , @xmath64 , with a quantum efficiency of @xmath65 .
since the considered process is a classical one , the gaussian input state is deformed into another classical ( non - negative ) output distribution .
function of a @xmath2-click subtraction process for @xmath61 on - off diodes with a quantum efficiency @xmath66 .
the transmission is @xmath67 , the mean thermal photon number of the input state is @xmath68 .
, width=317 ]
so far , we adapted the photon - subtraction protocol for measurements with on - off detector systems .
another frequently studied operation is a single - photon addition : @xmath69}. \end{aligned}\ ] ] the photon addition is known to be a nonclassical process and , hence , can be used to generate nonclassical output states from classical inputs , cf .
, e.g. , @xcite .
now , let us study this operation for click counting devices . in fig .
[ fig : additionscheme ] , a scenario for multi - photon addition is outlined .
an incoming signal is combined with an externally pumped parametric process , which generates two output beams . in particular , blocking the signal , a pump photon will generate two output photons propagating in different directions . if one of the photons is measured with an on - off detector system in this case , the corresponding twin photon is propagating in the other spatial mode . in the general case of fig . [
fig : additionscheme ] , the signal field undergoes a parametric amplification process and we post - select the output states to @xmath2 clicks of the detector .
clicks of the detector .
, width=158 ] first , we consider a coherent input field @xmath37 and the operator @xmath39 only .
after the treatment of this special case , we generalize our result to arbitrary states and click povm elements .
the considered mixing process is described by the squeezing transformation @xcite @xmath70 where we can assume that @xmath71 is a real and positive number .
the input fields are transformed as @xmath72 with @xmath73 , @xmath74 , and @xmath75 .
the coherent input state may be written as @xmath76 this yields an output state in the form : @xmath77}{\rm e}^{\frac{\nu}{\mu}\hat a^\dagger\hat b^\dagger}|0,0\rangle .
\end{aligned}\ ] ] here it is useful to apply the well - known baker - campbell - hausdorff formula : @xmath78}{\rm e}^{\hat y}{\rm e}^{\hat x}$ ] , which is valid if @xmath79,\hat x]=0 $ ] and @xmath79,\hat y]=0 $ ] .
hence , for the choice @xmath80 , @xmath81 , and @xmath82=-\frac{\nu^2}{\mu}\alpha\hat a^\dagger$ ] , the output state can be written as @xmath83 now , we perform a partial trace of the second mode with @xmath39 , cf .
. the resulting output can be computed as @xmath84\right ) \\\nonumber = & \frac{{\rm e}^{-|\alpha|^2}}{\mu^2}{\rm e}^{\frac{1}{\mu}\alpha\hat a^\dagger}\left[\sum_{m=0}^\infty\left((1-\lambda)\frac{\nu^2}{\mu^2}\right)|m\rangle\langle m|\right]{\rm e}^{\frac{1}{\mu}\alpha^\ast\hat a } \\\nonumber = & \frac{{\rm e}^{-|\alpha|^2}}{\mu^2}{\rm e}^{\frac{1}{\mu}\alpha\hat a^\dagger}{:}{\rm e}^{\left[(1-\lambda)\frac{\nu^2}{\mu^2}-1\right]\hat a^\dagger\hat a}{:}{\rm e}^{\frac{1}{\mu}\alpha^\ast\hat a } \\ = & \frac{1}{\mu^2}{:}{\rm e}^{-\frac{1}{\mu^2}(\hat a-\mu\alpha)^\dagger(\hat a-\mu\alpha)}{\rm e}^{-\lambda\frac{\nu^2}{\mu^2}\hat a^\dagger\hat a } { : } \label{eq : addno } \\\nonumber = & \sum_{m=0}^\infty\frac{1}{m!}\left(-\lambda\frac{\nu^2}{\mu^2}\right)^m\hat a^\dagger{}^m{:}\frac{1}{\mu^2}{\rm e}^{-\frac{1}{\mu^2}(\hat a-\mu\alpha)^\dagger(\hat a-\mu\alpha)}{:}\hat a{}^m .
\end{aligned}\ ] ] for further considerations , it is useful to have a closer look at the normally ordered term .
it turns out that this expression represents a displaced thermal state @xmath85\hat d(\mu\alpha)^\dagger \\=&\int{\rm d}^2\alpha'\,\frac{1}{\pi ( \mu^2 - 1)}{\rm e}^{-\frac{|\alpha'-\mu\alpha|^2}{\mu^2 - 1}}|\alpha'\rangle\langle\alpha'| .
\end{aligned}\ ] ] hence , the @xmath45 function of a general input state has to be convoluted with this noise prior to the addition process itself , @xmath86 is is also worth mentioning that the coherent amplitude of the input state is amplified by the factor @xmath87 . for the particular measurement of @xmath22 in the form of eq .
, we get the output state as @xmath88\frac{\nu^2}{\mu^2}\right)^m\right ] \hat a^\dagger{}^m\lambda_{\mu}^{\rm ( noise)}(\hat\rho_{\rm in})\hat a{}^m \\=&\sum_{m = k}^\infty \frac{1}{m!}\mathcal d_{k , m}^{-\eta',\eta'}\hat a^\dagger{}^m\lambda^{\rm ( noise)}_{\mu}(\hat\rho_{\rm in})\hat a{}^m , \label{eq : additionkrauss } \end{aligned}\ ] ] with a resulting efficiency of @xmath89 the general input - output relation in eq .
is the quantum description of the process in fig .
[ fig : additionscheme ] .
again , the output state is normalized to the probability @xmath53 to have @xmath2 clicks at the detector system .
the attenuation given by @xmath90 is due to a mixing of the input field with vacuum in the non - linear medium , and not the result of detector imperfections .
let us also mention the similarity to the subtraction process .
the input - output relations in eqs . and
share the same formal structure , except for an exchange of annihilation and creation operators as well as an exchange of a loss process with an amplification introducing thermal noise .
moreover , even the no - click count event can be of some interest , cf . @xcite . for @xmath91
the process in eq .
maps a coherent input up to a normalization constant to a displaced thermal state , see appendix [ app : dths ] . from the variance @xmath92 of the thermalized output @xmath45 function
, one can directly compute the squeezing parameter @xmath71 , @xmath93 as an example for the addition protocol under study , we will show the output for the thermal input state in eq . .
the noise convolution yields as a convolution of two gaussian functions an effective noise contribution of @xmath94 . the @xmath2 click conditioned state in eq .
may be written in normally ordered form as @xmath95 see also eq . or appendix [ app : dths ] , where the more general example of a displaced thermal state is presented . the @xmath45 function can be easily extracted from an expansion of the @xmath2th power and the relation ( @xmath96 ) : @xmath97 for different numbers of clicks , @xmath2 , we observe different numbers of oscillations within the output @xmath45 function between negative and positive values ( fig . [
fig : additionexample ] ) .
hence , we directly verify the nonclassical features of the considered process .
the quantum correlated output states introduce a new class of nonclassical states with a regular @xmath45 function .
similar to the prominent single-/multi - photon added thermal states @xcite , these states may be denoted as @xmath2-click conditioned thermal states .
function of a @xmath2-click conditioned thermal state generated by a click counting detector with @xmath61 on - off diodes and a quantum efficiency @xmath66 .
the considered squeezing corresponds to @xmath98 .
the mean thermal photon number of the input state is @xmath68 .
, width=317 ]
so far , we studied processes conditioned to a measurement of @xmath2 clicks with systems of avalanche diodes .
now , we can study a manifold of combinations of these individual procedures .
a prominent example is an operation which applies one addition followed by a subtraction : @xmath99}. \end{aligned}\ ] ] it represents a noiseless amplification protocol as it has been experimentally realized in ref .
@xcite . in the limit of small coherent amplitudes ,
@xmath100 , this process acts like @xmath101 i.e. , we have a gain of two in the coherent amplitude without the addition of noise .
for a recent study on quantum limits of amplification protocols , we refer to ref .
@xcite .
let us consider the generalization of this process to a combination of a @xmath102-photon addition followed by a @xmath103-photon subtraction . applying click counting detectors
means that the scheme in fig .
[ fig : additionscheme ] is combined with the setup in fig .
[ fig : subtractionscheme ] .
this leads to the process depicted in fig .
[ fig : noiselessamplificationscheme ] .
we are going to study the output states which correspond to @xmath102 clicks of the first and @xmath103 clicks of the second detector .
consists of @xmath104 diodes with a quantum efficiency of @xmath105 .
the second detector d@xmath106 consists of @xmath107 diodes with a quantum efficiency of @xmath108 .
the final outcome is conditioned to @xmath102 clicks of the first and @xmath103 clicks of the second detector .
, width=317 ] let us study the scenario of quantum state engineering based on the full process description , as a generalization of the noiseless amplification process . in secs .
[ sec : addition ] and [ sec : subtraction ] , we gave the corresponding input - output equations of the individual processes .
hence , we obtain after some algebra that a coherent input state @xmath109 is mapped to the output @xmath45 function : @xmath110 which is normalized to the probability that @xmath102 addition and @xmath103 absorption processes have been realized simultaneously
. the occurring coefficients are @xmath111 ) } , \\\nonumber \lambda_{2;j_1,j_2}{=}&\frac{1+\eta_1\nu^2[1{-}j_1/n_1]}{t^2\nu^2(1-\eta_1[1{-}j_1/n_1])}+\frac{\eta_2 r^2[1{-}j_2/n_2]}{t^2 } , \\\nonumber \lambda_{1;j_1,j_2}{=}&\frac{\mu}{t\nu^2(1-\eta_1[1{-}j_1/n_1 ] ) } ,
\\\lambda_{0;j_1,j_2}{=}&1+\frac{1}{\nu^2(1-\eta_1[1{-}j_1/n_1])}. \end{aligned}\ ] ] we therefore get for a general input state , @xmath112 , the output state as @xmath113|\alpha\rangle\langle\alpha| .
\end{aligned}\ ] ] note that some terms in eq . can lead to @xmath114-shaped contributions , if the denominator of the coefficients become zero .
however , this only occurs in the unphysical case @xmath115 ( for @xmath116 ) . in fig .
[ fig : noiselessamplificationexample ] , the output @xmath45 function for a coherent input state is given .
the first row corresponds to @xmath117 click and the last row to @xmath118 clicks of the click addition part .
the first column represents @xmath119 clicks and the last column @xmath120 clicks of the subtraction .
it can be directly observed that all these engineered phase - space distributions exhibit nonclassical features ( orange , negative contributions ) .
the higher the number of clicks @xmath102 the more negative interference fringes appear .
in contrast to the nonclassical click addition , the subtraction diminishes these quantum features together with a deformation of the positive ( gray ) contributions .
the given examples clearly show that the considered setup renders it possible to prepare various nonclassical states with different types of quantum interference effects .
the probabilities for the realization of the individual click combinations are listed in table [ tab : noiselessamplificationexample ] . .
probability for the realization of a @xmath121 conditioned output state .
the rows are numbered by the addition clicks , @xmath102 , and the columns by the subtraction clicks , @xmath103 .
the measurement is done for the same parameters as in fig .
[ fig : noiselessamplificationexample ] . [ cols="^,^,^,^,^,^",options="header " , ] our general theoretical treatment enables experimentalists to predict and generate various nonclassical states in arbitrary ranges of parameters .
as shown , the same experimental setup can lead to many different forms of nonclassical correlations even with asymmetric phase - space distributions . for an easily accessible , coherent input state the process led to quantum states with a regular @xmath45 function ,
contrary to the idealized noiseless amplification process in eq . .
this allows a direct sampling of the @xmath45 function from experimental data which have been measured , for example , by balanced homodyne detection , cf .
let us also note that the plots in fig .
[ fig : noiselessamplificationexample ] have been obtained for a quantum efficiency of only @xmath122 . in the case of the photoelectric counting theory ,
the low efficiency domain can be used for the realization of quantum - mechanical weak values of observables @xcite .
finally , let us outline possible generalizations of the presented schemes .
first , various combinations of heralding , addition , and subtraction scenarios with click counting detectors could be combined .
second , one can condition the output not to single click events , but to multiple ones .
the resulting state is a mixing of the individual output states for different numbers of clicks .
similarly , one can withdraw only a certain percentage of states which are realized .
this would allow one to steer the mixing ratio between different numbers of clicks .
third , at one input port of the addition and subtraction protocols vacuum was considered .
a straight - forward extension of the presented approach could use this free port , e.g. , for ancilla states as they are applied in some amplification scenarios .
fourth , additional features may appear when generalized detection processes are involved , e.g. , two - photon absorption within the diodes , cf .
similarly , arbitrary noise models may be included at any stage of the protocol description .
moreover , a propagation in non - linear media , e.g. , in fibers , or higher order wave mixing may further enhance the applicability of click detector systems for quantum state engineering .
in conclusion , we have derived analytical input - output relations for the application of on - off detector systems in quantum state engineering .
this includes the use of a single avalanche photo diode up to any number of diodes in detector arrays or multiplexing schemes .
based on the click counting statistics of such detector systems , we could derive a proper description of several protocols and arbitrary input states .
first , we considered the heralded photon generation describing the fundamentals of our approach , e.g. , the positive operator valued measure for click counting .
second , we studied two prominent processes : photon addition and photon subtraction .
the click triggered outcome of these processes have been derived for arbitrary input states in terms of input - output relations .
as explicit examples , we studied displaced thermal states serving as an input .
this led to another class of click - conditioned displaced thermal states .
as a combination of addition and subtraction , we considered a noiseless amplification protocol .
we showed for this example a manifold of new quantum states having nonclassical features which can be directly generated by a post - selection to particular click numbers . additionally , we computed the probabilities for the occurrence of the desired output states of these non - deterministic processes . throughout the present work
, we consistently included the quantum efficiency of the detection process .
the occurring efficiencies of the diodes have been studied in high and low efficiency regimes .
the approach also revealed unavoidable attenuation of the process which has to be taken into account .
our results present a useful scheme for the choice of experimental conditions in order to realize tailor - made nonclassical states with systems of on - off detectors .
this work was supported by the deutsche forschungsgemeinschaft through sfb 652 .
j.s . gratefully acknowledges financial support from the oklahoma state university .
the symbol @xmath124 is defined as @xmath125 .
\end{aligned}\ ] ] expanding the @xmath2th power , we get the representation : @xmath126 from the computational point of view , a recursion relation is a practicable tool .
the initial values can be readily obtained as @xmath127 the recursion relation is derived as @xmath128 \\=&\left[\tau+\sigma\frac{k}{n}\right]{\mathcal d}^{\tau,\sigma}_{k , m{-}1}+\sigma\frac{n - k+1}{n}{\mathcal d}^{\tau,\sigma}_{k{-}1,m{-}1}. \end{aligned}\ ] ] note that this relation and the initial values imply that @xmath129 using the definition @xmath130 for @xmath131 , we additionally get @xmath132 for @xmath131
. moreover , the recursion yields the diagonal elements as @xmath133
we are going to derive an upper bound to the error one would obtain if a poissonian measurement is applied instead of the binomial one . the standard photoelectric counting theory is based on the povm elements @xmath134 for @xmath11 and the quantum efficiency @xmath5 .
the deviation between the expectation value @xmath135 for @xmath2 clicks , cf .
eq . , and @xmath136 is @xmath137)|\leq \|\hat\rho\|_{\rm tr } \|\hat p_k-\hat\pi_k\|_{\rm op } , \end{aligned}\ ] ] where we applied hlder s inequality .
since the trace norm is @xmath138 , we get an upper bound of the deviation by the operator norm @xmath139 .
this norm can be calculated by @xmath140 because both , @xmath141 and @xmath22 , are diagonal in a common basis .
note that @xmath142 for @xmath12 , see @xcite for the corresponding statistics , which implies that @xmath143 . for @xmath144 , we would additionally get the @xmath2-photon projector : @xmath145 .
we consider as input states for the addition and subtraction the class of displaced thermal states , which can be given by some equivalent representations : @xmath146\hat d(\alpha_0)^\dagger \\=&\frac{1}{\bar
n+1}{:}{\rm e}^{-(\hat a-\alpha_0)^\dagger(\hat a-\alpha_0)/(\bar n+1 ) } { : } , \end{aligned}\ ] ] where @xmath147 denotes the displacement , @xmath56 denotes the mean thermal photon number ( for zero displacement ) , and @xmath148 is the displacement operator .
note that @xmath149 yields the coherent state @xmath150 . following the multi - photon subtraction protocol ,
we observe that the state @xmath37 is mapped to @xmath151 hence , we obtain for the displaced thermal state the output state of a @xmath2 click subtraction as @xmath152|\alpha\rangle\langle \alpha| , \end{aligned}\ ] ] including the substitution @xmath153 .
the probability for obtaining this output state is @xmath154 let us now focus on the addition protocol .
we conclude from eq . that the state @xmath37 is mapped to @xmath155 later on ,
we give the analytical @xmath45 function for such a normally ordered expression .
convolving this expression with the displaced thermal input , we get @xmath156 the normalization is @xmath157 for the computation of the results , we used the following simple relations : the relation between the normal ordered representation and the @xmath45 function of thermal states ( @xmath158 ) , @xmath159 the gaussian integral formula ( @xmath160 ) , @xmath161 and as a combination of the previous formulas the following trace ( @xmath158 ) : @xmath162 it is also useful to recall the displacement operation @xmath163=@xmath164 and the representation @xmath165=@xmath166 .
99 p. kok , w. j. munro , k. nemoto , t. c. ralph , j. p. dowling , and g. j. milburn , rev .
* 79 * , 135 ( 2007 ) . c. k. hong and l. mandel , phys .
lett . * 56 * , 58 ( 1986 ) .
n. jain , s. r. huisman , e. bimbard , and a. i. lvovsky , opt .
express * 18 * , 18254 ( 2010 ) . c. wagenknecht ,
li , a. reingruber , x .- h .
bao , a. goebel , y .- a .
chen , q. zhang , k. chen , and j .- w .
pan , nature photon . * 4 * , 549 ( 2010 ) . c. sller , o. cohen , b. j. smith , i. a. walmsley , and ch .
silberhorn , phys .
a * 83 * , 031806(r ) ( 2011 ) . v. dauria , o. morin , c. fabre , and j. laurat , eur .
j. d * 66 * , 249 ( 2012 ) .
t. j. bartley , g. donati , j. b. spring , x .-
jin , m. barbieri , a. datta , b. j. smith , and i. a. walmsley , phys .
a * 86 * , 043820 ( 2012 ) .
m. j. collins , c. xiong , i. h. rey , t. d. vo , j. he , s. shahnia , c. reardon , t. f. krauss , m. j. steel , a. s. clark , and b. j. eggleton , nature commun . * 4 * , 2582 ( 2013 ) .
m. frtsch , j. u. frst , c. wittmann , d. strekalov , a. aiello , m. v. chekhova , ch .
silberhorn , g. leuchs , and ch .
marquardt , nature commun . * 4 * , 1818 ( 2013 ) .
g. s. buller and r. j. collins , meas .
* 21 * , 012002 ( 2010 ) .
h. shibata , k. shimizu , h. takesue , and y. tokura , appl .
express * 6 * , 072801 ( 2013 ) .
f. marsili , v. b. verma , j. a. stern , s. harrington , a. e. lita , t. gerrits , i. vayshenker , b. baek , m. d. shaw , r. p. mirin , and s. w. nam , nature photon . * 7 * , 210 ( 2013 ) .
a. k. nowak s. l. portalupi , v. giesz , o. gazzano , c. dal savio , p .- f .
braun , k. karrai , c. arnold , l. lanco , i. sagnes , a. lematre , and p. senellart , nature commun . * 5 * , 3240 ( 2014 ) .
a. allevi , m. bondani , and a. andreoni , opt
. lett . * 35 * , 1707 ( 2010 ) .
j. f. dynes , z. l. yuan , a. w. sharpe , o. thomas , and a. j. shields , opt .
express * 19 * , 13268 ( 2011 ) .
d. achilles , ch .
silberhorn , c. liwa , k. banaszek , and i. a. walmsley , opt . lett .
* 28 * , 2387 ( 2003 ) .
m. j. fitch , b. c. jacobs , t. b. pittman , and j. d. franson , phys .
a * 68 * , 043814 ( 2003 ) .
g. zambra , a. andreoni , m. bondani , m. gramegna , m. genovese , g. brida , a. rossi , and m. g. a. paris , phys .
lett . * 95 * , 063602 ( 2005 ) .
a. feito , j. s. lundeen , h. coldenstrodt - ronge , j. eisert , m. b. plenio , and i. a. walmsley , new j. phys .
* 11 * , 093038 ( 2009 ) .
e. waks , e. diamanti , b. c. sanders , s. d. bartlett , and y. yamamoto , phys .
lett . * 92 * , 113602 ( 2004 ) . l. a. jiang , e. a. dauler , and j. t. chang , phys . rev .
a * 75 * , 062325 ( 2007 ) .
l. dovrat , m. bakstein , d. istrati , e. megidish , a. halevy , l. cohen , and h. s. eisenberg , phys .
rev . a * 87 * , 053813 ( 2013 ) .
j. sperling , w. vogel , and g. s. agarwal , phys .
a * 85 * , 023820 ( 2012 ) .
j. sperling , w. vogel , and g. s. agarwal , phys .
* 109 * , 093601 ( 2012 ) .
t. j. bartley , g. donati , xian - min jin , a. datta , m. barbieri , and i. a. walmsley , phys .
rev . lett . *
110 * , 173602 ( 2013 ) .
blanchet , f. devaux , l. furfaro , and e. lantz , phys .
lett . 101 , 233604 ( 2008 ) .
moreau , j. mougin - sisini , f. devaux , and e. lantz , phys .
a * 86 * , 010101(r ) ( 2012 ) .
j. sperling , w. vogel , and g. s. agarwal , phys .
a * 88 * , 043821 ( 2013 ) .
j. peina , jr .
, o. haderka , v. michlek , and m. hamar , opt
. lett . * 37 * , 2475 ( 2012 ) .
j. peina , jr .
, o. haderka , v. michlek , and m. hamar , phys .
a * 87 * , 022108 ( 2013 ) .
g. s. agarwal and k. tara , phys .
a * 46 * , 485 ( 1992 ) .
a. zavatta , s. viciani , and m. bellini , science * 306 * , 660 ( 2004 ) .
a. biswas and g. s. agarwal , phys .
a * 75 * , 032104 ( 2007 ) .
t. kiesel , w. vogel , v. parigi , a. zavatta , and m. bellini , phys . rev . a * 78 * , 021804(r ) ( 2008 ) . a. v. dodonov and s. s. mizrahi , phys
a * 79 * , 023821 ( 2009 ) .
t. kiesel , w. vogel , m. bellini , and a. zavatta , phys .
a * 83 * , 032116 ( 2011 ) .
lee , j. lee , h. nha , and h. jeong , phys .
a * 85 * , 063815 ( 2012 ) .
s. n. filippov , v. i. manko , a. s. coelho , a zavatta , and m bellini , phys .
scr . * t153 * , 014025 ( 2013 ) .
s. rahimi - keshari , t. kiesel , w. vogel , s. grandi , a. zavatta , and m. bellini , phys .
. lett . * 110 * , 160401 ( 2013 ) .
p. marek and r. filip , phys .
a * 81 * , 022302 ( 2010 ) .
j. fiurek , phys .
a * 80 * , 053822 ( 2009 ) .
v. parigi , a. zavatta , m. kim , and m. bellini , science * 317 * , 1890 ( 2007 ) . c. m. caves , phys .
d * 26 * , 1817 ( 1982 ) .
t. c. ralph and a. p. lund , in _ quantum communication measurement and computing proceedings of 9th international conference _ , edited by a. lvovsky ( aip , new york , 2009 ) . f. ferreyrol , m. barbieri , r. blandino , s. fossier , r. tualle - brouri , and p. grangier , phys .
lett . * 104 * , 123603 ( 2010 ) .
a. zavatta , j. fiurek , and m. bellini , nature photon .
* 5 * , 52 ( 2011 ) .
n. bruno , v. pini , a. martin , and r. t. thew , new j. phys . * 15 * , 093002 ( 2013 ) .
s. s. mizrahi and v. v. dodonov , j. phys .
gen . * 35 * , 8847 ( 2002 ) .
s. barz , g. cronenberg , a. zeilinger , and p. walther , nature photon . * 4 * , 553 ( 2010 ) .
h. hbel , d. r. hamel , a. fedrizzi , s. ramelow , k. j. resch , and t. jennewein , nature * 466 * , 601 ( 2010 ) .
g. y. xiang , t. c. ralph , a. p. lund , n. walk , and g. j. pryde , nature photon . * 4 * , 316 ( 2010 ) . s .- y . lee and h. nha , phys .
a * 85 * , 043816 ( 2012 ) . c. navarrete - benlloch , r. garca - patrn , j. h. shapiro , and n. j. cerf , phys .
a * 86 * , 012328 ( 2012 ) . c. i. osorio , n. bruno , n. sangouard , h. zbinden , n. gisin , and r. t. thew , phys
a * 86 * , 023815 ( 2012 ) .
u. l. andersen and j. s. neergaard - nielsen , phys .
a * 88 * , 022337 ( 2013 ) .
r. ghobadi , a. lvovsky , and c. simon , phys .
lett . * 110 * , 170406 ( 2013 ) .
o. morin , j .- d .
bancal , m. ho , p. sekatski , v. dauria , n. gisin , j. laurat , and n. sangouard , phys .
lett . * 110 * , 130401 ( 2013 ) .
l. slodicka , g. htet , n. rck , p. schindler , m. hennrich , and r. blatt , phys .
* 110 * , 083603 ( 2013 ) . w. vogel and d .-
welsch , _ quantum optics _ , ( wiley - vch , weinheim , 2006 )
. g. s. agarwal , _ quantum optics _ , ( cambridge university press , cambridge , 2013 ) .
e. agudelo , j. sperling , and w. vogel , phys .
a * 87 * , 033811 ( 2013 ) . c. gehrke , j. sperling , and w. vogel , phys .
a * 86 * , 052118 ( 2012 ) .
r. j. glauber , phys .
lett . * 10 * , 84 ( 1963 ) .
e. c. g. sudarshan , phys .
* 10 * , 277 ( 1963 ) .
g. s. agarwal , m. o. scully , and h. walther , phys .
t48 * , 128 ( 1993 ) . s. pandey , z. jiang , j. combes , and c. m. caves , phys .
a * 88 * , 033852 ( 2013 ) . g. s. agarwal and p. k. pathak , phys .
a * 75 * , 032108 ( 2007 ) . | we derive an analytical description for quantum state preparation using systems of on - off detectors .
our method will apply the true click statistics of such detector systems .
in particular , we consider heralded quantum state preparation using correlated light fields , photon addition , and photon subtraction processes . using a post - selection procedure to a particular number of clicks of the detector system , the output states reveal a variety of quantum features .
the rigorous description allows the identification and characterization of fundamentally unavoidable attenuations within given processes .
we also generalize a known scenario of noiseless amplification with click detectors for the purpose of the preparation of various types of nonclassical states of light .
our exact results are useful for a choice of experimental parameters to realize a target state . |
the t - j model @xcite is commonly regarded as a minimal model capable of describing correctly the essential physics of the cuprate high - temperature superconductors . in its simplest form ,
used in the present paper , it is expressed by the following hamiltonian @xmath6 here @xmath7 is a gutzwiller projection operator , ensuring that no doubly occupied sites are present .
unfortunately , the rigorous treatment of the t - j model is limited to very special cases .
this is both due to the interaction term ( @xmath8 ) , as well as to the presence of @xmath9 operators . even in the @xmath10 limit ,
hamiltonian ( [ t - j exact ] ) is not an independent - particle one . to proceed further , ( [ t - j exact ] )
may be treated within the mean - field approximation .
this procedure yields an single - particle hamiltonian @xmath11 , being the mean - field analogue of @xmath12 ( [ t - j exact ] ) .
the resulting effective description is termed the _
renormalized mean - field theory _ ( rmft ) @xcite .
the crucial point in the construction of rmft is an approximate treatment of the gutzwiller projection . in a broader perspective
, this problem is closely related to an analytic evaluation of the expectation values of operators with respect to the following variational state @xmath13 in the above , @xmath14 was replaced by a more general gutzwiller correlator @xmath15 , differing from @xmath14 by the presence of the so - called fugacity factors @xmath16 @xcite .
for @xmath17 we recover @xmath14 .
@xmath18 is an uncorrelated single - particle state , which , within the framework of rmft , is chosen as the eigenstate of the @xmath11 .
explicitly , we are interested in evaluation of the expressions like @xmath19 @xmath20 is an arbitrary operator , @xmath21 @xmath22 are the relevant mean - fields , i.e. the expectation values of the corresponding single - particle operators @xmath23 , and both @xmath11 and @xmath24 are usually @xmath25- dependent .
consequently , @xmath26 is a @xmath27-valued function of the mean - fields .
each prescription of the form ( [ generalnog ] ) will be termed the _ renormalization scheme _ ( rs ) .
note , that usually the rs ( [ generalnog ] ) may be given a more specific form @xmath28 in the above , @xmath29 is termed _
renormalization factor_.
there are many different renormalization schemes ( [ generalnog ] ) proposed in the literature .
the simplest take into account only on the local ( site - dependent ) mean fields ( e.g. local charge density or magnetization ) , @xcite-@xcite the more advanced include also the inter - site quantities ( mean fields defined on bonds , e.g. pairing amplitude ) , cf .
@xcite , @xcite-@xcite .
in particular , the scheme of ref .
@xcite has been used in refs . @xcite and @xcite and also in our previous work , @xcite . in this paper
we focus our attention mainly on the renormalization scheme of .
@xcite , which , in our opinion , is the most promising one devised so far .
first , due to the presence of fugacity factors in ( [ correlated state ] ) , the gutzwiller projection does not change the local densities of spin up ( down ) electrons , i.e. @xmath30 .
next , this formalism may , in principle , be systematically extended beyond the second order in the inter - site quantities ( formulas of ref .
@xcite are provided up to this order ) .
also , it can be relatively easily applied to the extensions of hamiltonian ( [ t - j exact ] ) , or to the more complicated symmetry - breaking situations .
suppose , that we have constructed the rmft specifying appropriate single - particle mean - field hamiltonian @xmath11 .
next , we have to determine the optimal values of mean - fields appearing in a model .
one possibility is an application of the non - variational self - consistent approach based on the bogoliubov - de gennes ( bdg ) equations .
however , this route , apart from its other drawbacks , @xcite , encounters a serious difficulty in the following sense .
namely , the renormalization scheme of that ref .
@xcite is in general not of the form ( [ g o general ] ) .
this feature does not allow for an unambiguous identification of the renormalization factors and hence the construction of the effective mf hamiltonian .
namely , there is no clear way to ascribe the corresponding single - particle operator expression to the expectation values ( [ generalnog ] ) .
even if some way of replacing @xmath31 ( [ generalnog ] ) by the operator counterpart may seem more natural than others , there is in fact no unique way of carrying out such procedure .
also , different such ways obviously yield different versions of rmft hamiltonians and consequently different bdg self - consistent equations , hence different predictions of the model .
this lack of uniqueness does not appear within the variational method proposed by us recently @xcite , as will be discussed in detail below .
the main aim of the present paper is twofold .
first , we compare different renormalization schemes within the rmft effective single - particle picture for the t - j model .
moreover , as in to ref .
@xcite , we compare two distinct methods of solving such mf models : the present variational approach ( labeled as _ var _ ) and a non - variational one , based on the bogoliubov - de gennes self - consistent equations ( labeled as _ s - c _ ) . the paper is organized as follows . in section [ formalism ]
we present those parts of our approach which are relevant to the present discussion . in section [ results ]
we present our numerical results , first for the d - wave superconducting ( drvb ) , ( [ drvb ] and [ chc ] ) , and then for the staggered - flux ( sf ) solutions ( [ sf ] ) , respectively . in appendix
a we provide some of the technical details omitted in main text .
section [ summary ] contains a concluding remarks .
the most of standard mean - field models may be solved by employing two methods .
first , we may invoke the variational procedure , i.e. minimization of the mf grand potential or ground state energy , with respect to the values of the order parameters ( mean fields ) . secondly , by using the self - consistency ( bogoliubov - de gennes ) equations , expressing the basic fact that the mean - fields are averages of the corresponding operators .
however , those two routes are equivalent only for mf hamiltonians of the hartree - fock type ( e.g. the bcs hamiltonian ) @xcite . in the case of rmft for the t - j model
, the self - consistency of the mf formalism may be spoiled by unwary application of such variational procedure , for both nonzero temperatures and for @xmath32 .
this due to the non - hartree character of the rmft hamiltonian , caused by mf treatment of the gutzwiller projection .
in such situation , the non variational self - consistent method based on the bogoliubov - de gennes ( bdg ) equations is usually applied to those advanced versions of rmft , @xcite . however
, this route suffers from serious drawbacks @xcite .
the other solution is the appropriate modification of the variational method .
examples of such self - consistent variational mf approach may be found in refs .
@xcite , @xcite-@xcite .
the approaches of those references exhibit various degree of generality , and differ from each other with respect to technical details .
yet , all are based on the variational principle of quantum mechanics ( minimization of the expectation value of the hamiltonian ) , applied to the mf case . as such , they are applicable to @xmath32 situation only .
our point of departure is different .
we base our approach ( @xcite ) on the _ maximum entropy principle _
( maxent ) @xcite , which is the basis of bayesian mathematical statistics .
in particular , it may be viewed also as the basis of the standard ( non mean - field ) statistical mechanics @xcite . the application of maxent inference to the non - standard case of mf statistical - mechanical description is a natural extension of this fundamental principle ( @xcite , c.f .
also @xcite ) . to ensure the self consistency of mf description ,
it is necessary to introduce additional constraints , not present in the standard statistical mechanics .
this could be achieved in the most natural manner by the lagrange multiplier method .
explicitly , starting from an arbitrary mf hamiltonian @xmath33 , we define yet another mf hamilton operator @xmath34 according to appearing in ( [ gmf hamiltonian with lambdas ] ) should not be confused with the fugacity factors @xmath16 appearing in eqn .
( [ correlated state ] ) ] @xmath35 the correct grand - canonical mf density operator is then given by @xmath36 .
\label{correct mf density matrix}\ ] ] next , the generalized grand - potential landau functional is defined as @xmath37 ) = -\beta^{-1 } \ln(\mathcal{z}_{\lambda } ) , \label{mathcalf tj bcs like}\end{aligned}\ ] ] with the inverse temperature @xmath38 .
the equilibrium values of @xmath39 , @xmath40 are the solution of the set of equations @xmath41 for which @xmath42 has the lowest value . by taking the derivatives with respect to @xmath43 only , and subsequently putting @xmath44 , @xmath45 we obtain standard bogoliubov - de genes ( bdg ) self - consistent equations .
in such a case we denote the chemical potential @xmath46 , which corresponds to @xmath47 for the _ var _ method .
the thermodynamical grand potential @xmath48 and the free energy @xmath49 are defined respectively as @xmath50 for _ var _ , and similarly ( with @xmath51 , @xmath52 ) for _ s - c _ method .
obviously , the present formalism is valid for any non - zero temperature , but not for @xmath32 .
consequently , we should replace pure states @xmath53 in eqs .
( [ correlated state ] ) - ( [ g o general ] ) by mixed ones ( [ correct mf density matrix ] ) , @xmath54 . on the other hand , the gutzwiller approximation in a form ( [ generalnog ] ) or ( [ g o general ] ) is devised for a ground state , hence for temperature @xmath32 .
however , the solutions of the mean field models ( and the rmft of the t - j model in particular ) , obtained for non - zero , but sufficiently low @xmath55 are practically identical to those of the real @xmath32 analysis .
consequently , @xmath54 is practically indistinguishable from @xmath53 , and thus the application of the finite - temperature formalism in the @xmath56 limit is fully justified .
reader interested in details of our method may consult ref .
below we recall some of its features , which are important from the point of view of the present application .
they are justified in the appendix a. first , the @xmath25 - dependence of the nontrivial part of the hamiltonian @xmath57 ( [ gmf hamiltonian with lambdas ] ) is completely determined by the set of the single - particle operators out of which @xmath57 and @xmath33 are composed
. moreover , two mf hamiltonians @xmath58 and @xmath59 ( or @xmath60 and @xmath61 ) , having the same nontrivial operator part and the same @xmath25 -dependence of the expectation value , i.e. @xmath62 , are equivalent .
namely , they yield identical equilibrium values of mean fields , quasi - particle energies , thermodynamic potentials etc .
this feature guarantees that , when constructing rmft hamiltonian , different assignments of the operator expression to the expectation values ( [ aver hopping fuk ] ) lead to identical results .
parenthetically , the values of the lagrange multipliers @xmath43 may differ between two such equivalent mf hamiltonians @xmath63 , @xmath64 .
this indicates , that @xmath43 have no unambiguous physical interpretation by themselves , but only in the certain combinations with the mean fields .
only the lagrange multipliers related to the quantities of _ a priori _ known average values , e.g. particle number @xmath65 , are identical for all such equivalent hamiltonians . for the rs that can be given the form ( [ g o general ] ) , with @xmath17 in ( [ correlated state ] ) , and in the absence of the long - range antiferromagnetic order ,
the mean - field hamiltonian @xmath66 may be taken in a form @xcite @xmath67 in the above expression , @xmath68 ( @xmath69 ) are ordinary fermion creation ( annihilation ) operators , @xmath70 , and @xmath71 are respectively , the hopping amplitude ( bond - parameter ) and the rvb gap parameter , both taken for nearest neighbors @xmath72 . @xmath73 denotes the average value of the operator @xmath74 evaluated with the help of a mf state in notation of ref .
@xcite ] ( [ correlated state ] ) . in ( [ ren tj ham ] ) , the interaction term ( @xmath75 ) of hamiltonian ( [ t - j exact ] ) has been treated within the hartree - fock decoupling in both the particle - particle ( @xmath76 ) and particle - hole ( @xmath77 ) channels .
the renormalization factors @xmath78 and @xmath79 , are given by ( [ g o general ] ) , with @xmath80 and @xmath81 , respectively .
their explicit form depends on the approximation used to obtain r.h.s . of ( [ g o general ] ) . as mentioned above
, the non - hartree - fock character of rmft hamiltonians @xmath11 is due to mf treatment of the gutzwiller projection .
consequently , in order to obtain @xmath34 , we must add to @xmath66 the constraints corresponding to the mean fields appearing in @xmath82 and @xmath83 .
the simplest renormalization factors @xcite-@xcite depend solely on local hole densities @xmath84 , i.e. @xmath85 in the framework of our method , in the homogeneous case @xmath86 , the additional lagrange multiplier coupled to the total particle number is introduced . for a non - homogeneous case ,
local chemical potentials should be introduced , which may be found in refs .
@xcite . in the case of more complicated form of @xmath87 , @xmath88 factors , depending apart form @xmath89 also on the inter - site mean - field variables ( @xcite ) ,
the mf hamiltonian has to be modified further .
the lagrange multipliers related to the average hopping ( bond order ) and superconducting order parameters acquire non - zero values in the equilibrium situation . as pointed out previously , the form of gutzwiller approximation of ref .
@xcite does not reduce to multiplication by renormalization factors , ( [ g o general ] ) .
e.g. , for the hopping amplitude , we have @xmath90 ( eqn .
( 15 ) of ref .
@xcite , but with different notation ) .
this , in general , does not allow for identification of renormalization factor @xmath87 .
however , for completeness , we want to compare the variational approach with the non - variational treatment of rmft based on ref .
@xcite . hence , using ( [ aver hopping fuk ] ) we may quite reasonably define those factors for a simple homogeneous non - magnetic states ( @xmath91 , @xmath92 ) as @xmath93 @xmath94 the prescription ( [ gj ren fac x chi delta fuk ] ) will be referred to as a renormalization scheme ( i ) .
this will be confronted with @xmath95-factors taken from ref .
@xcite @xmath96 @xmath97 referred to as renormalization scheme ( ii ) . usually within the rmft , the physical ( renormalized ) superconducting parameter is defined as @xmath98 instead of bare @xmath99 itself , @xcite
this is also the case for scheme ( ii ) . on the other hand , within the scheme ( i )
, there appear a separate expression for renormalized value of @xmath100 , cf .
( 18 ) of @xcite . again , for simple homogeneous , non - magnetic states , the corresponding @xmath101 factor can be quite reasonably identified as @xmath102 note , that ( [ gdelta ren fac x chi delta fuk ] ) differs from ( [ gt ren fac x chi delta fuk ] ) .
below we present our numerical results , first for the plain d - wave superconducting state with no magnetic order ( drvb ) , and next for the staggered - flux ( sf ) normal solution . even for those simplest mf states ,
the results obtained within various renormalization schemes and/or the solving methods differ remarkably .
the solution analyzed here is constructed to possess full symmetry of the underlying square lattice , only the superconducting order parameter is assumed to have @xmath104 symmetry .
consequently , we are left with three independent mean fields @xmath105 , where @xmath106 , @xmath107 , and the same number of the corresponding lagrange multipliers , @xmath108 , where @xmath109 , @xmath110 , and @xmath111 , @xmath112 , @xmath113 for the bonds in @xmath114 ( @xmath115 ) directions , respectively .
also , @xmath116 , and all the above quantities are taken as real .
diagonalization of ( [ gmf hamiltonian with lambdas ] ) in the present case yields @xmath117 with @xmath118 @xmath119 also , @xmath120 @xmath121 the gutzwiller renormalization factors read now , respectively @xmath122 @xmath123 @xmath124 @xmath125 note , that for ( ii ) @xmath126 .
the generalized landau functional is given by @xmath127 from which the explicit form of eqs .
( [ derivative of mathcalf a , lambda ] ) can be easily obtained , but are not presented here in an explicit form .
_ numerical results .
_ we begin with the analysis of drvb solution for the magic doping , @xmath128 .
the equations ( [ derivative of mathcalf a , lambda ] ) and ( [ derivative of mathcalf lambda ] ) are solved for the lattice of @xmath129 sites , @xmath130 , with @xmath131 @xmath132 and for low temperature , @xmath133 . both drvb solution , as well as isotropic normal state ( fermi sea , ( fs ) ,
not discussed explicitly here ) are present .
the staggered flux state , expected for lower doping , have not been found ( _ var _ ) or is not stable against fs ( _ s - c _ ) for @xmath134 . in tables
i. and ii .
we give the values of the thermodynamic potentials , mean - fields and molecular fields for drvb solutions obtained within both methods ( _ var _ , _ s - c _ ) and both renormalization schemes ( ( i ) , ( ii ) ) .
table i. values of the thermodynamic potentials ( per site ) for drvb solutions .
@xmath135 ( @xmath49 ) stands for @xmath136 ( @xmath137 ) for _ var _ and @xmath138 ( @xmath139 ) for _ s - c _ methods , respectively . + [
cols="^,^,^,^,^",options="header " , ] within the standard mean - field approach ( e.g. @xcite ) the ( fictitious ) flux is defined as @xmath140 , and @xmath141 denotes plaquette composed from four bonds .
also , for _ s - c _ method , i.e. for @xmath142 , we have @xmath143 the last equality follows from ( [ t tau 1 sf ] ) .
interestingly , this equality holds also for the variational approach with @xmath144 , @xmath145 .
it can be shown analytically , that @xmath146 , which together with ( [ t tau 1 sf ] ) yields ( [ mf s - c flux ii equal i ] ) . thus the two possible and _ a priori _ different definitions of @xmath147 within _ var _ method turn out to be equivalent .
antiferromagnetic correlations are defined on the mf level as @xmath148 , @xcite .
in contrast to the drvb case , now the differences between renormalization schemes ( ( i ) vs ( ii ) ) within each ( _ var _ , _ s - c _ ) method are small . this is because in the absence of @xmath76 , the @xmath77 -dependences of @xmath149 and @xmath150 are quite similar , and the @xmath77 -dependence of @xmath151 is weak for doping @xmath152 , thus causing no qualitative and only minor quantitative differences between the renormalization schemes . on the other hand , the generic modifications introduced by the variational approach within each rs are more significant .
namely , from tab .
iv we see , that _ s - c _ method favors sf more strongly than _ var _ method , which is indicated by the values of @xmath147 and @xmath153 . also , @xmath154 and @xmath155 , determining quasi - particle spectra , are smaller within _ var _ method , and so are the quasi - particle energies , c.f .
[ dispertion for 512 sf ] .
in summary , in this work we have compared , within two methods of approach , the two gutzwiller renormalization schemes for the renormalized mean - field theory ( rmft ) of t - j model in its simplest form .
we emphasize the advantages of the combination of the renormalization scheme of ref .
@xcite with the variational method proposed in ref .
first , a number of theoretical arguments strongly favor this choice .
moreover , in contrast to the other cases investigated by us ( e.g. either non - variational method or renormalization scheme of ref .
@xcite ) , the former approach quite correctly predicts the upper critical doping @xmath156 for a disappearance of the sc order . also , the value of the optimal doping @xmath157 is quite correctly predicted . in our opinion , the formalism of ref . @xcite
augmented with the self - consistent variational treatment , gives a chance for the complete and consistent one - particle description ( in the form of rmft ) for a t - j model . such description , however , must encompass the t - j model in its complete form @xcite , and include also more complicated symmetry breaking patterns .
this is the subject of our current investigation .
[ [ appendix - a - equivalence - relation - for - mean - field - hamiltonians - appendixa ] ] appendix a : equivalence relation for mean - field hamiltonians [ appendixa ] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ below we present some details of our formalism , which are necessary for the present disscusion
. we also comment on the relationship between our method and the formalisms of ref .
we start from the mf hamiltonian of the form @xmath158 in the above , @xmath159 is an @xmath25-independent part , @xmath160 , @xmath161 and @xmath162 are some complex - valued functions of mean - fields @xmath25 .
operators @xmath163 are those , which average values are not present in @xmath33 .
we also assume that all the operators appearing above are bilinear in creation and/or annihilation operators . from ( [ mf h specyfic ] ) we have @xmath164 with @xmath165 , @xmath166 .
for a given value of @xmath25 , the self - consistency equations ( second half of the @xmath167 equations ( [ derivative of mathcalf a , lambda ] ) ) may be written as @xmath168 \nonumber \\ & = & \frac{\text{tr}[\hat{a}_{t } \exp\big(-\beta ( \hat{k}_{e0 } + \sum_{s=1}^{m } c^{\lambda}_{s}(\vec{a})\hat{a}_{s } + \sum_{w=1}^{m^{\prime } } g_{w}(\vec{a})\hat{b}_{w } ) \big)]}{\text{tr } [ \exp\big(-\beta ( \hat{k}_{e0 } + \sum_{s=1}^{m } c^{\lambda}_{s}(\vec{a})\hat{a}_{s}+ \sum_{w=1}^{m^{\prime } } g_{w}(\vec{a})\hat{b}_{w}\big)]}. \label{the return of the self consistent equations}\end{aligned}\ ] ] in above , the @xmath169 term canceled , and @xmath170 .
( [ the return of the self consistent equations ] ) may be formally solved ( which usually can not be achieved in an analytic fashion ) for @xmath43 , and then @xmath171 , thus for a given @xmath25 , eqs .
( [ the return of the self consistent equations ] ) determine @xmath172 . as a result
, the @xmath25-dependence of the coefficients @xmath172 is determined by the choice of operators @xmath159 , @xmath173 and @xmath174
. if we change the original form of the coupling of @xmath173 operators to the mean fields , according to @xmath175 but without changing @xmath162 functions , then the lagrange multipliers change in a way that @xmath172 is unchanged , in order to fulfill ( [ the return of the self consistent equations ] ) .
the only @xmath25-dependent part which may be non - trivially modified by ( [ redefinition of couplings ] ) is @xmath176 .
obviously , @xmath177 , defined as ( c.f .
@xcite ) @xmath178 \label{entropy mf}\ ] ] and @xmath179 may be also modified . however , the density operator @xmath180 is invariant under ( [ redefinition of couplings ] ) , so are , for given @xmath25 , all the averages , also those of @xmath159 and @xmath163 operators . moreover ,
if the transformations ( [ redefinition of couplings ] ) are such that @xmath181 remains unchanged , the value of @xmath182 is not modified .
consequently , in such a situation , the equilibrium values of mean fields , as well as of the quantities @xmath172 ( e.g. @xmath183 of eqn .
( [ t tau ] ) or @xmath184 of eqn .
( [ t tau sf ] ) ) are also invariants of ( [ redefinition of couplings ] ) . summarizing , all mf hamiltonians constructed from the same set of operators @xmath159 , @xmath173 and @xmath174 and having the same @xmath25-dependence of the expectation value , are equivalent , and belong to the same equivalence class . the transformations ( [ redefinition of couplings ] ) may be viewed in analogy to gauge transformations , not changing the physical content of the model .
now , we choose a transformation ( [ redefinition of couplings ] ) of a specific form , @xmath188 @xmath189 this yields @xmath190 with @xmath191 , @xmath192 . using ( [ mean value of h gmf with lambda , d / da ] ) we may rewrite ( [ simples h lambda ] ) as @xmath193 because @xmath194 .
the form ( [ simples h lambda wyrazony eksplicite ] ) is the most convenient , as half of the variables ( @xmath195 ) are eliminated , which reduces the number of equations to be solved numerically . also , in the limit ( @xmath56 ) the results of finite - temperature formalism are essentially identical to those the true @xmath32 analysis , and we may compare them with those of ref .
hamiltonian of that reference reads in our notation @xmath196 and differs from ( [ simples h lambda ] ) only by the constant term @xmath197 . consequently , both ( [ simples h lambda ] ) and ( [ simples h lambda kai ] ) have the same eigenvalues and eigenvectors ( the presence of @xmath198 obviously does not change the above arguments ) . please note , that the analytical evaluation of lagrange multipliers through ( [ mean value of h gmf with lambda , d / da ] ) , and hence the application of the method of ref .
@xcite is possible only if the conditions ( [ complete h ] ) are fulfilled , but not for the general form ( [ mf h specyfic ] ) of the mf hamiltonian .
an invaluable technical help that one of the authors ( jj ) received from andrzej kapanowski , jan kaczmarczyk , gosia kaliszan and micha kos is warmly acknowledged .
all the numerical computation was performed using gsl ( gnu scientific library ) efficient procedures .
the authors acknowledge the grant from ministry of science and higher education .
j. spaek , a. m. ole , jagiellonian university preprint sspju-6/76 , ( 1976 ) ; k. a. chao , j. spaek , a. m. ole , j. phys .
c * 10 * , l271 ( 1977 ) ; for didactical review , see : j. spaek , acta phys .
a * 111 * , 409 ( 2007 ) . | recently , fukushima [ phys . rev .
b * 78 * 115105 ( 2008 ) ] proposed a systematic derivation of the gutzwiller approximation for the t - j model . in the present paper , using this approach we construct an effective single - particle hamiltonian , which leads to a renormalized mean - field theory ( rmft ) .
we also use the method proposed by us recently and based on the maximum entropy principle ( maxent ) , which in turn , yields a consistent statistical description of the problem . on the examples of
non - magnetic superconducting d - wave resonating valence bond ( drvb ) and normal staggered - flux ( sf ) solutions , we compare two selections of the gutzwiller renormalization schemes , i.e. the one proposed by fukushima with that used earlier by sigrist et al .
[ phys .
rev .
b * 49 * , 12 058 ( 1994 ) ] .
we also confront the results coming from our variational solutions with the self - consistency conditions build in , with those of the non - variational approach based on the bogoliubov - de gennes self - consistent equations .
combination of the present variational approach with the new renormalization scheme ( taken from fukushima s work ) provides , for @xmath0 , an upper critical hole concentration @xmath1 for the disappearance of the d - wave superconductivity .
also , the hole concentration @xmath2 is obtained for the optimal doping .
these results are in rough accordance with experimental results for high-@xmath3 superconducting cuprates . a statistically consistent variational approach to the renormalized mean - field theory of the t - j model : critical hole concentrations for a paired state + jakub jdrak@xmath4 and jozef spaek@xmath5 + marian smoluchowski institute of physics , jagiellonian university , + reymonta 4 , 30 - 059 krakw , poland + * pacs : * 05.30.-d , 71.10.fd , 75.10.jm . |
the ultrarelativistic heavy - ion collider facilities like rhic at bnl and lhc at cern strive to produce a new form of hot qcd matter .
the experiments show @xcite that it has very intricate properties and presents a big challenge especially for theoretical understanding . while above the ( pseudo)critical temperature @xmath5 mev this matter is often called the quark - gluon plasma ( qgp ) , it can not be a perturbatively interacting quark - gluon gas ( as widely expected before rhic results @xcite ) until significantly higher temperatures @xmath6 . instead
, the interactions and correlations in the hot qcd matter are still strong ( e.g. , see refs .
@xcite ) so that its more recent and more precise name is _ strongly coupled qgp _ ( sqgp ) @xcite .
one of its peculiarities seems to be that strong correlations in the form of quark - antiquark ( @xmath7 ) bound states and resonances still exist @xcite in the sqgp well above @xmath8 . in the old qgp paradigm ,
even deeply bound charmonium ( @xmath9 ) states such as @xmath10 and @xmath11 were expected to unbind at @xmath12 , but lattice qcd simulations of mesonic correlators now indicate they persist till around @xmath13 @xcite or even above @xcite .
similar indications for light - quark mesonic bound states are also accumulating from lattice qcd @xcite and from other methods @xcite .
this agrees well with the findings on the lattice ( e.g. , see ref .
@xcite for a review ) that for realistic explicit chiral symmetry breaking ( chsb ) , i.e. , for the physical values of the current quark masses , the transition between the hadron phase and the phase dominated by quarks and gluons , is not an abrupt , singular phase transition but a smooth , analytic crossover around the _
pseudo_critical temperature @xmath8 .
it is thus not too surprising that a clear experimental signal of , e.g. , deconfinement , is still hard to find and identify unambiguously . the most compelling signal for production of a new form of qcd matter , i.e. , sqgp , would be a restoration - in hot and/or dense matter - of the symmetries of the qcd lagrangian which are broken in the vacuum .
one of them is the [ su@xmath14(@xmath15 ) flavor ] chiral symmetry , whose dynamical breaking results in light , ( almost-)goldstone pseudoscalar ( @xmath16 ) mesons namely the octet @xmath17 , as we consider all three light - quark flavors , @xmath18 . the second one is the u@xmath14(1 ) symmetry .
its breaking by the non - abelian axial adler - bell - jackiw anomaly ( ` gluon anomaly ' for short ) makes the remaining pseudoscalar meson of the light - quark sector , the @xmath0 , much heavier , preventing its appearance as the ninth ( almost-)goldstone boson of dynamical chiral symmetry breaking ( dchsb ) in qcd .
the first experimental signature of a partial restoration of the u@xmath14(1 ) symmetry seems to have been found in the @xmath19 gev central au+au reactions at rhic .
namely , csrg _ et al . _
@xcite analyzed combined data of phenix @xcite and star @xcite collaborations very robustly , through six popular models for hadron multiplicities , and found that at 99.9% confidence level , the @xmath0 mass , which in the vacuum is @xmath20 mev , is reduced by at least 200 mev inside the fireball .
it is the sign of the disappearing contribution of the gluon axial anomaly to the @xmath0 mass , which would drop to a value readily understood together with the ( flavor - symmetry - broken ) octet of @xmath21 ( @xmath22 ) pseudoscalar mesons .
this is the issue of the return of the prodigal goldstone boson " predicted @xcite as a signal of the u@xmath14(1 ) symmetry restoration . another related but less obvious issue to which we want to draw attention , concerns the status , at @xmath3 , of the famous witten - veneziano relation ( wvr ) @xcite @xmath23 between the @xmath0 , @xmath24 and @xmath25-meson masses @xmath26 , pion decay constant @xmath27 , and yang - mills ( ym ) topological susceptibility @xmath28 .
wvr was obtained in the limit of large number of colors @xmath29 @xcite .
it is well satisfied at @xmath4 for @xmath28 obtained by lattice calculations ( e.g. , @xcite ) . nevertheless , the @xmath1-dependence of @xmath28 is such @xcite that the straightforward extension of eq .
( [ wittenvenez ] ) to @xmath3 @xcite , i.e. , replacement of all quantities unless their @xmath1-dependence is specifically indicated in formulas or in the text . ] therein by their respective @xmath1-dependent versions @xmath30 , @xmath31 , @xmath32 , @xmath33 _ and _ @xmath34 , leads to a conflict with experiment @xcite .
since this extension of eq .
( [ wittenvenez ] ) to @xmath3 was studied in ref .
@xcite before the pertinent experimental analysis @xcite , one of the purposes of this paper is to revisit the implications of the results of ref .
@xcite for wvr at @xmath3 , and demonstrate explicitly that they are practically model - independent .
the other , more important purpose is to propose a mechanism which can enable wvr to agree with experiment at @xmath3 .
both issues pointed out before eq .
( [ wittenvenez ] ) and around it , are best understood in a model - independent way if one starts from the chiral limit of vanishing current quark masses ( @xmath35 ) for all three light flavors , @xmath36 .
then not only pions and kaons are massless , but is also @xmath24 , which is then ( since the situation is also su(3)-flavor - symmetric ) a purely su(3)-octet state , @xmath37 .
in contrast , @xmath0 is then purely singlet , @xmath38 ; since the divergence of the singlet axial quark current @xmath39 is nonvanishing even for @xmath35 due to the gluon anomaly , the @xmath0 mass squared receives the anomalous contribution @xmath40 ( @xmath41 in the notation of ref .
@xcite ) which is nonvanishing even in the chiral limit : @xmath42 however , @xmath43 and @xmath44 are known accurately decay constant " @xmath44 is , strictly speaking , not a well - defined quantity , as _ two _ @xmath0 decay constants are actually needed : the singlet one , @xmath45 , and the octet one , @xmath46 ; _ e.g. _ , see an extensive review @xcite or the short appendix of ref . @xcite .
] only in the large @xmath29 limit . there
, in the leading order in @xmath47 , @xmath48 is given by the ym ( i.e. , pure glue " ) topological susceptibility @xmath28 times @xmath49 @xcite , and the @xmath0 decay constant " @xmath44 is the same as @xmath50 @xcite .
thus , keeping only the leading order in @xmath47 , the last equality is wvr in the chiral limit .
the consequences of eq .
( [ wv_chlim ] ) remain qualitatively the same realistically away from the chiral limit
. this will soon become clear on the basis of , e.g. , eq .
( [ m2_prop_m ] ) below .
namely , due to dchsb in qcd , for relatively light current quark masses @xmath51 @xmath52 , the @xmath21 bound - state pseudoscalar meson masses ( including the _ nonanomalous _ parts of the @xmath0 and @xmath24 masses ) behave as @xmath53 the pseudoscalar mesons ( including @xmath0 ) thus obtain relatively light nonanomalous contributions @xmath54 to their masses @xmath55 , allowing them to reach the empirical values .
that is , instead of the eight strictly massless goldstone bosons , @xmath56 and @xmath24 are relatively light almost - goldstones . among them , in the limit of isospin symmetry ( @xmath57 ) , only @xmath24 now receives also the gluon - anomaly contribution since the explicit su(3 ) flavor breaking between the nonstrange ( @xmath58 ) @xmath59-quarks and @xmath60-quarks causes the mixing between the isoscalars @xmath24 and @xmath0 . for @xmath61 , eq .
( [ wv_chlim ] ) is replaced by the usual wvr ( [ wittenvenez ] ) containing also the nonanomalous contributions to meson masses .
nevertheless , these contributions largely cancel due to the approximate su(3 ) flavor symmetry and to dchsb [ i.e. , eq .
( [ m2_prop_m ] ) ] .
this can be seen assuming the usual su(3 ) @xmath7 content of the pseudoscalar meson nonet with well - defined isospin opens the possibility of studying the interesting scenario of maximal isospin violation at high @xmath1 @xcite , but as the effects of the small difference between @xmath62 and @xmath63 are not important for the present considerations , we stick to the isospin limit throughout the present paper . ] quantum numbers , in particular the isoscalar @xmath64 octet and singlet etas , @xmath65 , @xmath66 , whose mixing yields the physical particles @xmath24 and @xmath0 . since the _ nonanomalous _ parts of the @xmath67 and @xmath68 masses squared , @xmath69 and @xmath70 , are respectively @xmath71 and @xmath72 ( see , e.g. , ref .
@xcite ) , and since @xmath73 , the nonanomalous parts of the @xmath24 and @xmath0 masses are canceled by @xmath74 in wvr ( [ wittenvenez ] ) . another way of seeing this is expressing the nonanomalous parts of @xmath75 by eq .
( [ m2_prop_m ] ) .
thus again @xmath76 , showing again that already wvr s chiral - limit - nonvanishing part ( [ wv_chlim ] ) reveals the essence of the influence of the gluon anomaly on the masses in the @xmath0-@xmath24 complex .
this is important also for the presently pertinent finite-@xmath1 context because thanks to this , below it will be shown model - independently that wvr ( [ wittenvenez ] ) containing the ym topological susceptibility @xmath28 implies @xmath1-dependence of @xmath0 mass in conflict with the recent experimental results @xcite .
namely , the gluon - anomaly contribution ( [ wv_chlim ] ) is established at @xmath4 but it is not expected to persist at high temperatures . ultimately , @xmath0 should also become a massless goldstone boson at sufficiently high @xmath1 , where @xmath77 .
however , according to wvr , @xmath78 falls only for @xmath1 where @xmath79 does not fall faster than @xmath80 , as stressed in ref .
@xcite . the wvr s chiral - limit version ( [ wv_chlim ] ) manifestly points out the ratio @xmath81 as crucial for the anomalous @xmath0 mass , but the above discussion shows that this remains essentially the same away from the chiral limit . in the present context
, it is important for practical calculations to go realistically away from the chiral limit , in which the chiral restoration is a sharp phase transition at its critical temperature @xmath2 where the chiral - limit pion decay constant vanishes very steeply , i.e. , as steeply as the chiral quark condensate .
in contrast ,
for realistic explicit chsb , i.e. , @xmath62 and @xmath63 of several mev , this transition is a _ smooth crossover _ ( e.g. , see ref .
@xcite ) . for the pion decay
constant , this implies that @xmath33 still falls relatively steeply around _
pseudo_critical temperature @xmath2 , but less so than in the chiral case , and even remains finite , enabling the usage of wvr ( [ wittenvenez ] ) for the temperatures across the chiral and u@xmath14(1 ) symmetry restorations .
wvr is very remarkable because it connects two different theories : qcd with quarks and its pure - gauge , ym counterpart .
the latter , however , has much higher characteristic temperatures than qcd with quarks : the melting temperature " @xmath82 where @xmath34 starts to decrease appreciably was found on lattice to be , for example , @xmath83 mev @xcite or even higher , @xmath84 mev @xcite . in contrast , the pseudocritical temperatures for the chiral and deconfinement transitions in the full qcd are lower than @xmath82 by some 100 mev or more ( e.g. , see ref .
@xcite ) due to the presence of the quark degrees of freedom .
this difference in characteristic temperatures , in conjunction with @xmath34 in wvrs ( [ wittenvenez ] ) and ( [ wv_chlim ] ) would imply that the ( partial ) restoration of the u@xmath14(1 ) symmetry ( understood as the disappearance of the anomalous @xmath85 mass ) should happen well after the restoration of the chiral symmetry .
but , this contradicts the rhic experimental observations of the reduced @xmath0 mass @xcite _ if _ wvrs ( [ wittenvenez ] ) , ( [ wv_chlim ] ) hold unchanged also close to the qcd chiral restoration temperature @xmath2 , around which @xmath33 decreases still relatively steeply of the unphysical @xmath86 pseudoscalar with @xmath33 in fig .
[ chi1757_qq_f ] . ]
@xcite for realistic explicit chsb , thus leading to the increase of @xmath87 and consequently also of @xmath88 .
there is still more to the relatively high resistance of @xmath34 to temperature : not only does it start falling at rather high @xmath82 , but @xmath34 found on the lattice is falling with @xmath1 _ relatively _ slowly .
in some of the applications in the past ( e.g. , see refs .
@xcite ) , it was customary to simply rescale a temperature characterizing the pure - gauge , ym sector to a value characterizing qcd with quarks .
( for example , refs . @xcite rescaled @xmath89 mev found by ref .
@xcite to 150 mev ) .
however , even if we rescale the critical temperature for melting of the topological susceptibility @xmath34 from @xmath82 down to @xmath2 , the value of @xmath87 still increases a lot @xcite for the pertinent temperature interval starting already below @xmath2 .
this happens because @xmath34 falls with @xmath1 more slowly than @xmath79 .
( it was found @xcite that the rescaling of @xmath82 would have to be totally unrealistic , to less than 70% of @xmath2 , in order to achieve sufficiently fast drop of the anomalous contribution that would allow the observed enhancement in the @xmath0 multiplicity . )
these wvr - induced enhancements of the @xmath0 mass for @xmath90 were first noticed in ref .
this reference used a concrete dynamical model ( with an effective , rank-2 separable interaction , convenient for computations at @xmath91 ) @xcite of low energy , nonperturbative qcd to obtain mesons as @xmath21 bound states in dyson - schwinger ( ds ) approach @xcite , which is a bound - state approach with the correct chiral behavior ( [ m2_prop_m ] ) of qcd .
nevertheless , this concrete dynamical ds model was used in ref .
@xcite to get concrete values for only the nonanomalous parts of the meson masses , but was essentially _ not _ used to get model predictions for the mass contributions from the gluon anomaly , in particular @xmath34 . on the contrary , the anomalous mass contribution was included , in the spirit of @xmath47 expansion , through wvr ( [ wittenvenez ] ) .
thus , the @xmath1-evolution of the @xmath0-@xmath24 complex in ref . @xcite was not dominated by dynamical model details , but by wvr , i.e. , the ratio @xmath87 . admittedly , @xmath33 was also calculated within this model , causing some _ quantitative _ model dependence of the anomalous mass in wvr , but this can not change the qualitative observations of ref .
@xcite on the @xmath0 mass enhancement .
namely , our model @xmath33 , depicted as the dash - dotted curve in fig .
[ chi1757_qq_f ] , obviously has the right crossover features @xcite .
it also agrees qualitatively with @xmath33 s calculated in other realistic dynamical models @xcite .
various modifications were tried in ref .
@xcite but could not reduce much the @xmath0 mass enhancement caused by this ratio , let alone bring about the significant @xmath0 mass reduction found in the rhic experiments @xcite . , of @xmath92 , @xmath93 , @xmath27 and @xmath94 , i.e. , the @xmath95-dependences of the quantities entering in the anomalous contributions to various masses in the @xmath0-@xmath24 complex see eq .
( [ metaprime ] ) and formulas below it .
the solid curve depicts @xmath92 for @xmath96 in eq .
( [ chitildechicond ] ) , and the short - dashed curve is @xmath92 for @xmath97 . at @xmath4 ,
the both @xmath98 s are equal to @xmath99 , the weighted average @xcite of various lattice results for @xmath28 .
the dotted ( red ) curve depicts @xmath100 , the dash - dotted ( blue ) curve is @xmath27 , and the long - dashed ( blue ) curve is @xmath94 . ( colors online . ) , width=302 ] one must therefore conclude that either wvr breaks down as soon as @xmath1 approaches @xmath2 , or that the @xmath1-dependence of its anomalous contribution is different from the pure - gauge @xmath34 .
we will show that the latter alternative is possible , since wvr can be reconciled with experiment thanks to the existence of another relation which , similarly to wvr , connects the ym theory with full qcd .
namely , using large-@xmath29 arguments , leutwyler and smilga derived @xcite , at @xmath4 , @xmath101 the relation ( in our notation ) between the ym topological susceptibility @xmath28 , and the full - qcd topological susceptibility @xmath102 , the _ chiral - limit _ quark condensate @xmath103 , and @xmath104 , the harmonic average of @xmath15 current quark masses @xmath51 .
that is , @xmath104 is @xmath15 times the reduced mass . in the present case of @xmath18 , @xmath105 , so that @xmath106 equation ( [ chitilde ] ) is a remarkable relation between the two pertinent theories . for example , in the limit of all very heavy quarks ( @xmath107 ) , it correctly leads to the result that @xmath28 is equal to the value of the topological susceptibility in _ quenched _ qcd , @xmath108 .
this holds because @xmath102 is by definition the vacuum expectation value of a gluonic operator , so that the absence of quark loops would leave only the pure - gauge , ym contribution .
however , the leutwyler - smilga relation ( [ chitilde ] ) also holds in the opposite ( and presently pertinent ) limit of light quarks .
this limit still presents a problem for getting the full - qcd topological susceptibility @xmath102 on the lattice @xcite , but we can use the light - quark - sector result @xcite @xmath109 where @xmath110 stands for corrections of higher orders in small @xmath51 , and thus of small magnitude . the leading term is positive ( as @xmath111 ) , but @xmath110 is negative , since eq .
( [ chitilde ] ) shows that @xmath112 .
although small , @xmath110 should not be neglected , since @xmath113 would imply , through eq .
( [ chitilde ] ) , that @xmath114 . instead , its value ( at @xmath4 ) is fixed by eq .
( [ chitilde ] ) : @xmath115 all this starting from eq .
( [ chitilde ] ) has so far been at @xmath4 . if the left- and right - hand side of eq . ( [ chitilde ] ) are extended to @xmath3 , it is obvious that the equality can not hold at arbitrary temperature @xmath3 .
the relation ( [ chitilde ] ) must break down somewhere close to the ( pseudo)critical temperatures of full qcd ( @xmath116 ) since the pure - gauge quantity @xmath28 is much more temperature - resistant than the right - hand side , abbreviated as @xmath117 .
the quantity @xmath117 , which may be called the effective susceptibility , consists of the full - qcd quantities @xmath102 and @xmath103 , the quantities of full qcd with quarks , characterized by @xmath2 , just as @xmath33 . as @xmath118 , the chiral quark condensate @xmath119 drops faster than the other dchsb parameter in the present problem , namely @xmath33 for realistically small explicit chsb .
( see fig .
[ chi1757_qq_f ] for the results of the dynamical model adopted here from ref .
@xcite , and , e.g. , refs .
@xcite for analogous results of different ds models ) .
thus , the troublesome mismatch in @xmath1-dependences of @xmath33 and the pure - gauge quantity @xmath34 , which causes the conflict of the temperature - extended wvr with experiment at @xmath120 , is expected to disappear if @xmath34 is replaced by @xmath121 , the temperature - extended effective susceptibility .
the successful zero - temperature wvr ( [ wittenvenez ] ) is , however , retained , since @xmath122 at @xmath4 .
extending eq .
( [ chi_small_m ] ) to @xmath3 is something of a guesswork as there is no guidance from the lattice for @xmath123 [ unlike @xmath34 ] .
admittedly , the leading term is straightforward as it is plausible that its @xmath1-dependence will simply be that of @xmath119 .
nevertheless , for the correction term @xmath110 such a plausible assumption about the form of @xmath1-dependence can not be made and eq .
( [ cat0 ] ) , which relates ym and qcd quantities , only gives its value at @xmath4 .
we will therefore explore the @xmath1-dependence of the anomalous masses using the following ansatz for the @xmath91 generalization of eq .
( [ chi_small_m ] ) : @xmath125^{\delta } \ , , \label{chi_small_mt}\ ] ] where the correction - term @xmath1-dependence is parametrized through the power @xmath126 of the presently fastest - vanishing ( as @xmath127 ) chiral order parameter @xmath119 .
the @xmath91 extension ( [ chi_small_mt ] ) of the light - quark @xmath102 , eq . ( [ chi_small_m ] ) , leads to the @xmath91 extension of @xmath117 : @xmath128^{\delta } \right ) .
\label{chitildechicond}\ ] ] we now use @xmath129 in wvr instead of @xmath34 used by ref . @xcite .
this gives us the temperature dependences of the masses in the @xmath24-@xmath0 complex , such as those in fig .
[ m_pchi1757 ] illustrating the cases @xmath96 and @xmath97 .
it is clear that @xmath129 ( [ chitildechicond ] ) blows up as @xmath127 if the correction term there vanishes faster than @xmath119 _ squared_. thus , varying @xmath126 between 0 and 2 covers the cases from the @xmath1-independent correction term , to ( already experimentally excluded ) enhanced anomalous masses for @xmath126 noticeably above 1 , to even sharper mass blow - ups for @xmath130 when @xmath127 . on the other hand
, it does not seem natural that the correction term vanishes faster than the fastest - vanishing order parameter @xmath119 . indeed , already for the same rate of vanishing of the both terms ( @xmath97 )
, one can notice in fig .
[ m_pchi1757 ] the start of the precursors of the blow - up of various masses in the @xmath0-@xmath24 complex as @xmath127 although these small mass bumps are still experimentally acceptable .
thus , in fig . [ m_pchi1757 ] we depict the @xmath97 case , and @xmath96 ( @xmath1-independent correction term ) as the other acceptable extreme .
since they turn out to be not only qualitatively , but also quantitatively so similar that the present era experiments can not discriminate between them , there is no need to present any ` in - between results ' , for @xmath131 .
next we turn to completing the explanation how the above - mentioned results in fig .
[ m_pchi1757 ] were obtained .
, of the pseudoscalar meson masses for two @xmath129 , namely eq .
( [ chitildechicond ] ) with @xmath96 ( upper panel ) and with @xmath97 ( lower panel ) .
the meaning of all symbols is the same on the both panels : the masses of @xmath0 and @xmath24 are , respectively , the upper and lower solid curve , those of the pion and nonanomalous @xmath132 pseudoscalar are , respectively , the lower and upper dash - dotted curve , @xmath133 and @xmath134 are , respectively , the short - dashed ( red ) and long - dashed ( red ) curve , @xmath135 is the medium - dashed ( blue ) , and @xmath136 is the dotted ( blue ) curve .
( colors online . )
the straight line @xmath137 is twice the lowest matsubara frequency.,width=302 ] , of the pseudoscalar meson masses for two @xmath129 , namely eq .
( [ chitildechicond ] ) with @xmath96 ( upper panel ) and with @xmath97 ( lower panel ) .
the meaning of all symbols is the same on the both panels : the masses of @xmath0 and @xmath24 are , respectively , the upper and lower solid curve , those of the pion and nonanomalous @xmath132 pseudoscalar are , respectively , the lower and upper dash - dotted curve , @xmath133 and @xmath134 are , respectively , the short - dashed ( red ) and long - dashed ( red ) curve , @xmath135 is the medium - dashed ( blue ) , and @xmath136 is the dotted ( blue ) curve .
( colors online . )
the straight line @xmath137 is twice the lowest matsubara frequency.,width=302 ] using @xmath129 in wvr instead of @xmath34 used by ref .
@xcite , does not change anything at @xmath4 , where @xmath138 , which remains an excellent approximation even well beyond @xmath4 .
nevertheless , this changes drastically as @xmath1 approaches @xmath2 . for @xmath90 , the behavior of @xmath129
is dominated by the @xmath1-dependence of the chiral condensate , tying the restoration of the u@xmath14(1 ) symmetry to the chiral symmetry restoration . as for the nonanomalous contributions to the meson masses , we use the same ds model ( and parameter values ) as in ref .
@xcite , since it includes both dchsb and correct qcd chiral behavior as well as realistic explicit chsb .
that is , all nonanomalous results ( @xmath139 , as well as @xmath140 and @xmath94 , the mass and decay constant of the unphysical @xmath132 pseudoscalar meson ) in the present paper are , for all @xmath1 , taken over from ref .
we used this same model also for computing the chiral quark condensate @xmath103 , including its @xmath1-dependence displayed in fig .
[ chi1757_qq_f ] .
this defines completely how the results displayed in fig .
[ m_pchi1757 ] were generated .
for details , see ref .
@xcite ( and ref .
@xcite for @xmath133 and @xmath134 ) .
here we list only the formulas which , in conjunction with fig .
[ chi1757_qq_f ] , enable the reader to understand easily the @xmath1-dependences of the masses in fig .
[ m_pchi1757 ] : the theoretical @xmath0 and @xmath24 mass eigenvalues are @xmath141 , \quad \label{metaprime } \\
\label{meta } m_{\eta}^2(t ) & = & \frac{1}{2 } \left [ m_{\eta_{ns}}^2(t ) + m_{\eta_{s}}^2(t ) - \delta_{\eta \eta'}(t ) \right ] , \quad\end{aligned}\ ] ] where @xmath142 ^ 2 + 8 \beta^2 x^2}$ ] , @xmath143 @xmath144 @xmath145 @xmath146 in all expressions after eq .
( [ meta ] ) , the @xmath1-dependence is understood . in both cases considered for the topological susceptibility ( [ chi_small_mt ] ) [ @xmath96 ,
i.e. , the constant correction term , and @xmath97 , i.e. , the strong @xmath1-dependence @xmath147 of both the leading and correction terms in @xmath123 ] , the results are consistent with the experimental findings on the decrease of the @xmath0 mass of csrg _ et al .
in the light of the recent experimental results on the @xmath0 multiplicity in heavy - ion collisions @xcite , we revisited the earlier theoretical work @xcite concerning the thermal behavior of the @xmath0-@xmath24 complex following from wvr straightforwardly extended to @xmath3 .
we have confirmed the results of ref .
@xcite on wvr where the ratio @xmath81 dominates the @xmath1-dependence , and clarified that these results are practically model - independent .
it is important to note the difference between our approach and those that attempt to give model predictions for topological susceptibility , such as refs . @xcite .
by contrast , in refs . @xcite and here , as well as earlier works @xcite at @xmath4 , a ds dynamical model is used ( as far as masses are concerned ) to obtain only the nonanomalous part of the light pseudoscalar meson masses ( where the model dependence is however dominated by their almost - goldstone character ) , while the anomalous part of the masses in the @xmath0-@xmath24 complex is , through wvr , dictated by @xmath149 . in this ratio ,
@xmath33 is admittedly model - dependent in quantitative sense , but other realistic models yield qualitatively similar crossover behaviors @xcite of @xmath33 for @xmath150 , as exemplified by our fig .
[ chi1757_qq_f ] , and fig . 2 in ref .
@xcite , and by fig . 6 in ref . @xcite .
such @xmath33 behaviors are also in agreement with the @xmath1-dependence expected of the dchsb order parameter on general grounds : a pronounced fall - off around @xmath2 but exhibiting , in agreement with lattice @xcite , a smooth crossover pattern for nonvanishing explicit chsb , a crossover which gets slower with growing @xmath51 [ e.g. , compare @xmath33 with @xmath151 in fig .
[ chi1757_qq_f ] ] .
in contrast to the qcd topological susceptibility @xmath102 , the ym topological susceptibility and even its @xmath1-dependence @xmath34 , including its melting " temperature @xmath152 , can be extracted @xcite reasonably reliably from the lattice @xcite .
thus , it was not modeled in ref .
hence our assertion that the results of ref .
@xcite unavoidably imply that the straightforward extension of wvr to @xmath3 is falsified by experiment @xcite , especially if one recalls that even the sizeable @xmath1-rescaling @xcite @xmath153 was among the attempts to control the @xmath0 mass enhancement @xcite .
nevertheless , we have also shown that there is a plausible way to avoid these problems of the straightforward , naive extension of wvr to @xmath3 , and this is the main result of the present paper . thanks to the existence of another relation , eq .
( [ chitilde ] ) , connecting the ym quantity @xmath28 with qcd quantities @xmath102 and @xmath154 , it is possible to define a quantity , @xmath117 , which can meaningfully replace @xmath34 in finite-@xmath1 wvr .
it remains practically equal to @xmath28 up to some 70% of @xmath2 , but beyond this , it changes following the @xmath1-dependence of @xmath155 . in this way , the successful zero - temperature wvr is retained , but the ( partial ) restoration of u@xmath14(1 ) symmetry [ understood as the disappearing contribution of the gluon anomaly to the @xmath0 ( @xmath67 ) mass ] is naturally tied to the restoration of the su@xmath14(3 ) flavor chiral symmetry and to its characteristic temperature @xmath2 , instead of @xmath152 .
it is very pleasing that this fits in nicely with the recent _ ab initio _ theoretical analysis using functional methods @xcite , which finds that the anomalous breaking of u@xmath14(1 ) symmetry is related to dchsb ( and confinement ) in a self - consistent manner , so that one can not have one of these phenomena without the other . of course
, the most important thing is that this version of the finite-@xmath1 wvr , obtained by @xmath156 , is consistent with experiment @xcite for all reasonable strengths of @xmath1-dependence [ @xmath157 in eq .
( [ chi_small_mt ] ) ] .
namely , the both tablets in fig .
[ m_pchi1757 ] show , first , that @xmath0 mass close to @xmath2 suffers the drop of more than 200 mev with respect to its vacuum value .
this satisfies the minimal experimental requirement abundantly .
second , fig .
[ m_pchi1757 ] shows an even larger drop of the @xmath67 mass , to some 400 mev , close to the best " value of the in - medium @xmath0 mass ( 340 mev , albeit with large errors ) obtained by csrg _ et al .
this should be noted because the @xmath67 mass inside the fireball is possibly even more relevant .
namely , although it is , strictly speaking , not a physical meson , @xmath67 is the state with the @xmath7 content closest to the @xmath7 content of the physical @xmath0 _ in the vacuum_. thus , among the isoscalar @xmath7 states inside the fireball , @xmath67 has the largest projection on , and thus the largest amplitude to evolve , by fireball dissipation , into an @xmath0 in the vacuum .
d. h. , d. kl . and
s. b. were supported through the project no .
0119 - 0982930 - 1016 , and d. ke .
through the project no .
098 - 0982887 - 2872 of the ministry of science , education and sports of croatia .
d. h. and d. kl .
acknowledge discussions with r. alkofer .
the support by compstar network is also acknowledged .
e. v. shuryak and i. zahed , phys
* d70 * , 054507 ( 2004 ) , [ hep - ph/0403127 ] . m. a. stephanov , pos * lat2006 * , 024 ( 2006 ) [ arxiv : hep - lat/0701002 ] .
d. b. blaschke and k. a. bugaev , fizika * b13 * , 491 ( 2004 ) , [ nucl - th/0311021 ] .
s. datta , f. karsch , p. petreczky and i. wetzorke , nucl .
. suppl . * 119 * , 487 ( 2003 ) , [ hep - lat/0208012 ] .
m. asakawa and t. hatsuda , phys .
lett . * 92 * , 012001 ( 2004 ) , [ hep - lat/0308034 ] .
e. shuryak , nucl . phys . *
a774 * , 387 ( 2006 ) , [ hep - ph/0510123 ] .
f. karsch _ et al .
_ , nucl . phys . *
a715 * , 701 ( 2003 ) , [ hep - ph/0209028 ] .
p. maris , c. d. roberts , s. m. schmidt and p. c. tandy , phys .
c * 63 * , 025202 ( 2001 ) [ arxiv : nucl - th/0001064 ] .
m. mannarelli and r. rapp , phys . rev . *
c72 * , 064905 ( 2005 ) , [ hep - ph/0505080 ] .
b. lucini , m. teper and u. wenger , nucl . phys . *
b715 * , 461 ( 2005 ) , [ hep - lat/0401028 ] .
l. del debbio , l. giusti and c. pica , phys .
rev . lett . * 94 * , 032003 ( 2005 ) , [ hep - th/0407052 ] . b. alles , m. delia and a. di giacomo , phys .
* d71 * , 034503 ( 2005 ) , [ hep - lat/0411035 ] . s. drr , z. fodor , c. hoelbling and t. kurth , jhep * 0704 * , 055 ( 2007 ) [ arxiv : hep - lat/0612021 ] . | we discuss and propose the minimal generalization of the witten - veneziano relation to finite temperatures , prompted by star and phenix experimental results on the multiplicity of @xmath0 mesons . after explaining why these results show that the zero - temperature witten - veneziano relation can not be straightforwardly extended to temperatures @xmath1 too close to the chiral restoration temperature @xmath2 and beyond , we find the quantity which should replace , at @xmath3 , the yang - mills topological susceptibility appearing in the @xmath4 witten - veneziano relation , in order to avoid the conflict with experiment at @xmath3 .
this is illustrated through concrete @xmath1-dependences of pseudoscalar meson masses in a chirally well - behaved , dyson - schwinger approach , but our results and conclusions are of a more general nature and , essentially , model - independent . |
parametric x - rays ( pxr ) produced by a relativistic charged particle uniformly moving through a single crystal were declared as a very attractive radiation mechanism for x - ray source since its theoretical prediction in 1972 @xcite - @xcite and first experimental observation in 1985 @xcite . actually , it has a number of indisputable features , namely , high monochromaticity , energy tuneability , polarization , directivity , and possibility to be emitted to large angles relative to beam direction , i.e. provide monochromatic x - rays virtually free of background .
pxr properties were studied at a number of accelerator facilities worldwide but a pxr - based x - ray source developments are still in progress . evidently , this is connected with the fact that pxr research is performed at existing linear and annular accelerators and limited by their technical specifications .
so , we should recognize how to design a pxr - based source that will meet the consumer s requirements for its use in industry and medicine .
industry and science , as consumers , essentially yield to medicine , which represents the huge market for devices and methods for x - ray diagnostics . statistically
, x - ray devices count for about 60% of medical diagnostics devices @xcite . despite the development of alternative diagnostics methods , e.g. ultrasound and magnetic resonance imaging
, the number of studies in the field of x - ray diagnostics remains high .
first of all , medicine demands decreased irradiation doses taken together with increased resolution and contrast .
these requirements can be met if the customer will be offered a monochromatic tuneable source , providing x - radiation with exactly the energy required for examination of a specific organ .
for example , in mammography the maximum spectral density lies in the range of 17 - 20 kev ; in radiography of the chest , extremities and head , the range of 40 - 50 kev is optimal ; while the 50 - 70 kev range is used for abdomen and pelvis radiography .
precise selection of monochromatic radiation energy will allow a reduction of a patient dose of tens times , even in comparison with digital radiography and computer tomography , with improvement of x - ray image quality .
however , in comparison with scientific and industrial applications where signal accumulation is possible , in the case of medicine the source should emit number of quanta sufficient to provide necessary contrast of an image at the specific value of the signal - to - noise ratio .
estimations @xcite give minimal value of x - ray quanta necessary for high - quality image equal to @xmath010@xmath1 photons / mm@xmath2 .
pxr quantum yield in a thin crystal target even at high enough ( e.g. 900 mev ) beam energy is not very large and amounts only @xmath010@xmath3 - 10@xmath4 photons / e@xmath5sr @xcite . assuming the target to be imaged is 30 cm diameter at one meter distance from a source , it is necessary to apply @xmath0 0.1 a current to get a required x - ray flux .
such currents can be achieved only in low energy accelerators .
though theoretically pxr may be emitted at any energy of charged particles @xcite , there is a factor limiting the minimal value of beam energy for source under discussion . assuming x - ray energy necessary for subtractive coronary angiography @xmath6=33 kev then the angle of radiation @xmath7 in a diamond target will come to @xmath8 for ( 111 ) reflex and @xmath9 for ( 220 ) reflex that it is enough for arrangement of a patient at 1 - 2 meter distance from the target .
now let us estimate energy @xmath10 of an electron beam providing angular width of reflex @xcite @xmath11 ( @xmath12 is the fourier component of a crystal dielectric constant , @xmath13 , @xmath14 is the mean - square angle of multiple scattering ) equal @xmath15 , i.e. energy at which x - ray reflex and forward background are still separated from each other and the radiation in which we are interested keeps all the properties necessary for a medical source .
the dominant component here is the mean - square angle of multiple scattering @xmath16 and lowest estimation of beam energy gives a value @xmath017 mev . to provide more real evaluations for pxr source based on low - energy electron beam
, we should look into multiple scattering ( ms ) of electrons more precisely , because ms appeared to be the most significant factor affecting pxr intensity and angular distribution at relatively low energies and/or thick targets @xcite . in low energy cases
we should nt use kinematics formulae @xcite , where ms is considered phenomenologically .
let us describe the multiple scattering influence on x - ray reflex characteristics in more detail following @xcite .
to get an expression for spectral - angular density of radiation with ms account , it is necessary to average the squared modulus of all possible particle trajectories in a crystal in the following expression @xcite @xmath17 the obtained expression will describe all mechanisms of the radiation : pxr , _ bremsstrahlung _ , and transition radiation in a single manner . the velocity vector is presented in the form : @xmath18 ( where @xmath19 determines the direction of the initial velocity of a charged particle , @xmath20 is the the angle of multiple scattering of a particle ( @xmath21 ) , @xmath22 is the two - dimensional vector , @xmath23 and @xmath24 at @xmath25 ) .
averaging is followed @xcite , where the procedure was performed with the help of dispersion function , satisfying the fokker - planck kinetic equation in the amorphous medium .
averaging of the right part of the equation ( [ 1 ] ) was performed and the integral expressions for spectral - angular distributions of the radiation , taking into account multiple scattering for the lateral and forward pxr maxima were derived . selected functions of the particle distributions on coordinates and angles
are usually applied in calculations of _ bremsstrahlung _ intensity of ultra - relativistic electrons in amorphous media .
it is applicable if direction of vector of a charged particle initial velocity does not coincide with directions of the basic crystallographic axes or planes .
also , one can ignore the influence of crystal structure on charged particles trajectories and apply dispersion functions of amorphous medium when value of initial angular divergence of a particle beam , falling on the crystal along the direction of main crystallographic axes or planes , satisfies the inequality @xmath26 , where @xmath27 is the potential of atomic axis or plane , @xmath28 is the particle energy . in this case during the passage of charged particles through the crystal , orientational effects will be expressed less strongly than in the case of a well - collimated beam , and the averaging of ( [ 1 ] ) over simple dispersion functions ensures a satisfactory approach .
it is necessary to note , that in the case , when the value @xmath29 ( @xmath30 is the bragg frequency , @xmath31 , @xmath32 is the angle between the particle velocity @xmath19 and planes corresponding to the vector @xmath33 , @xmath34 is the langmuir frequency of a crystal ) , it is not required to make additional averaging in ( [ 1 ] ) on the initial divergence of the vector direction @xmath19 .
there are two factors limiting the longitudinal size of quasi - cherenkov radiation formation area : ms of charged particles on atoms of substance and x - ray absorption in the medium . as in case of radiation generation in an amorphous medium , the pxr characteristics essentially depend on the relation between crystal thickness along the direction of the charged particle movement @xmath35 * * * * and _ bremsstrahlung _ coherent length @xmath36 . the theory of pxr with the phenomenological account of ms influence on particle energy range exceeding threshold energy @xmath37 is in good agreement with experimental results only at crystal thickness @xmath38 . in this case
ms influence appears only as a small addition in the intensity of generated x - ray caused by _
bremsstrahlung_. also ms influence on pxr phase is taken into account and interference between pxr and _ bremsstrahlung _ is neglected . in the inverse case
ms essentially changes the parameters of the pxr itself .
to demonstrate ms effect on pxr characteristics , numerical calculations were performed for following conditions : target si ( 220 ) in laue geometry , @xmath39 , electron beam energy 900 mev .
pxr spectral - angular density as a function of crystal target thickness is shown in fig.[fig1 ] .
= 8 cm it is evident , that already at @xmath40 presence of ms results in pxr intensity distinctly less than intensity determined without ms account . during further increase of the crystal thickness
, spectral - angular density achieves saturation at thickness of @xmath41 and then goes down .
it is seen from this figure that pxr spectral - angular density without ms ( curve 1 ) continues to grow with crystal thickness increase even up to @xmath42 .
the tendency shows that absorption still does not plays an essential role though @xmath35 becomes of the order of @xmath43 . simultaneously with the variation of pxr spectral - angular density with crystal thickness increase , change of its spectral width appears ( see fig.[fig2 ] ) .
= 8 cm as long as the thickness of the crystal target does not exceed @xmath44 , spectral - angular density and spectral width of pxr reflex determined with ms account practically coincide with the values received without ms consideration . at the same time
, spectral width of pxr peak decreases with @xmath35 increase . the width of pxr maximum is inversely proportional to radiation coherent length . as the target thickness increase ( on thickness of @xmath45 and larger )
pxr coherent length begins to be limited by @xmath36 .
so , pxr peak width without ms continues to decrease with @xmath46 * * * * growth , but ms account results in stabilization of peak width at @xmath40 ( fig.[fig2 ] ) .
thus , calculation results , demonstrated in the figures [ fig1 ] and [ fig2 ] , show that ms of a charged particle results not only in radiation phase shift , but also in essential decrease of pxr spectral - angular density at angles of @xmath47 as well as in change of reflex width , even in case of @xmath40 . as a result of joint processes of pxr spectral narrowing and spectral - angular density decrease
, pxr integral characteristics do not vary essentially . as an example , in fig.[fig3 ] there are given dependence of pxr angular distributions ( at @xmath48rad ) on the crystal thickness without ms ( 1 ) and in the ms presence ( 2 ) .
one can see that for the target thickness @xmath49 the maximal value of spectral - angular distribution ( at @xmath50 rad ) calculated without ms account exceeds the the maximal value obtained with ms account more than two orders of magnitude ( compare with fig.[fig1 ] ) , while amplitudes of angular distributions ( for the same angle ) differ less than by one order ( fig.[fig3 ] ) .
ms also results in pxr angular distribution spreading , therefore if the angular aperture of a detector is big enough ( @xmath51 ) , the difference in results for integrated number of the quanta calculated with the ms account and without taking it into account will be even less , than for angular distributions .
thus , the most significant difference of the results obtained with and without ms account is observed for differential distributions .
as well as for cherenkov radiation in homogeneous media , pxr spectral - angular density has threshold behavior . however , due to essential dispersion of a crystal refraction index , dependence of integral pxr intensity on particle energy is relatively smooth
it is differs in @xmath52 and @xmath53 energy regions , where threshold energy @xmath54 , @xmath55 .
so , pxr quantum yield at particle energy @xmath52 * * is @xmath56 and @xmath57 at @xmath53 .
note , such dependences of pxr intensity on energy were calculated without taking into account of ms influence on radiation characteristics .
the _ bremsstrahlung _ coherent length is directly proportional to charged particle energy : @xmath58 , where @xmath59 , _ _
e__@xmath60= 21.2 mev - scale energy , _
_ l__@xmath61 - radiation length .
therefore , by fixing the thickness of a crystal target and changing the energy of charged particles ( thus changing @xmath36 ) , it is possible to calculate dependences of spectral - angular density and angular distribution , and dependence of spectral maximum width on the @xmath62 ratio ( similar to shown in fig.[fig1 ] - fig.[fig3 ] ) . certainly , the nature of these dependences is completely different . a change of a crystal target thickness results in the change of a trajectory length where particle radiates coherently . in fact
, this is its cut - off because the pxr coherent length is equal to infinity .
besides , on particle energy decrease the pxr generation is realized at increasing values of @xmath63 ( detuning parameter from the exact value of the bragg angle ) that in turn results in decrease of radiation intensity .
pxr spectral - angular distributions calculated for the same geometry as fig.[fig1 ] and fig.[fig2 ] were plotted in fig.[fig4 ] .
= 8 cm crystal thickness was chosen equal to 0.01 cm and polar angle of radiation was @xmath64=1.8 mrad .
it is evident from fig.[fig4 ] , which for 9 gev electron beam energy distributions calculated without ( curve 3 ) and with ms account ( curve 4 ) , practically coincides . in this case
@xmath65 cm , i.e. @xmath66 * * , * * so a situation of weak ms is realized here .
change of energy from 9 gev down to 900 mev yields in change of _ bremsstrahlung _ coherent length to the order of magnitude , @xmath67cm*. * in this case target thickness along direction of charged particle movement @xmath35 becomes of one order with @xmath36 .
presence of ms results in essential decrease of height and spreading of pxr peak ( curve 2 ) in comparison with the calculations , without taking ms into account ( curve 1 ) .
the spectrum corresponding to 900 mev appeared shifted down for @xmath68 along frequency axis relative to spectrum corresponding to 9 gev .
in the same time , shift of curve 2 relative to curve 1 as a result of ms influence for 900 mev appeared to be just @xmath69 .
amplitudes of spectral - angular distributions without ms ( curves 1 and 3 ) has decreased less than 1.5 times , while ms account decrease it more than four times with energy decrease from 9 gev down to 900 mev .
comparing spectral - angular distributions for 900 mev energy , it is possible to see that ms account results in decrease of height of spectral - angular distribution approximately 3.2 times and its spreading is @xmath702.5 times .
thus , ms presence results in decrease of pxr angular intensity .
pxr angular distributions with ms account for the geometry described above in a crystal with thickness of 0.13 cm are plotted in fig.[fig5 ] = 8 cm detector resolution was considered equal to @xmath71 .
it is evident from the figure that energy decrease to three times ( from 900 down to 300 mev ) results in a decrease of maximum intensity almost five times . at the same time
there is a shift of angular distribution maximum to the bigger angles from 2.9 mrad up to 4.8 mrad and spreading almost to one and half times more .
thus ms presence results in noticeable decrease of angular intensity , angular distribution spreading and to a shift of maximum in angular distribution towards the bigger angles in comparison with calculations without ms account .
for example , for 300 mev electron energy the maximum in pxr angular distribution without ms account is achieved at 2.3 mrad angle .
study of radiation yield energy dependence is important for correct understanding of the pxr generation mechanism .
this dependence , as mentioned above , has a threshold nature and its specific shape at experimental measurement in many respects is determined by the angular size of the detector . as energy decreases , the effective angle of radiation emission @xmath72 increases .
value of _ _ l__@xmath73 depends on crystal thickness , absorption length , bragg angle , and radiation geometry @xcite .
= 8 cm results of pxr quantum yield calculations for electron energy in a range from 300 up to 900 mev are plotted in fig.[fig6 ] ; angular aperture of detector is 9.5@xmath7410@xmath75 rad .
the angular aperture of detector is @xmath76 even for 300 mev , that is why the detector collects practically all generated radiation and we have behaviour of quantum yield energy dependence that is `` natural '' for proper pxr .
presence of ms results in a steep decrease of pxr quantum yield .
for example , for 700 mev electron energy pxr quantum yield without ms exceeds the result received with ms consideration by up to 1.5 times , for 300 mev this value already amounts @xmath702.9 times .
summing up , multiple scattering of charged particles in a crystal target changes considerably the characteristics of x - rays originating due to a particle passage through a single crystal .
ms effect is bigger at low energies and/or thick targets and demonstrates itself in x - rays distributions spreading , monochromaticity , and quantum yield decrease .
nevertheless , pxr at low energies can provide monochromaticity of 10@xmath75 - 10@xmath77 which is still applicable in medical imaging . using the approach described above ,
let us estimate the characteristics of pxr source , which can be produced on real medical accelerators @xcite . as a rule ,
maximal energy of electron beam in these accelerators is equal 20 - 28 mev .
we have evaluated angular density of radiation for next conditions : e@xmath78= 25 mev , 33 kev x - rays , symmetrical laue case for ( 111 ) , ( 220 ) , and ( 400 ) .
silicon target of l=0.01 cm was chosen which is somewhat bigger , but more real than optimal for these conditions ( about ( 1@xmath7910)@xmath80 , i.e. @xmath81 microns ) target thickness .
angles between electron velocity direction and direction to diffraction reflex are @xmath706.9 , 11.2 , and 15.9 degrees , respectively .
angular densities for discussed cases are plotted in fig.[fig7 ] .
= 8 cm integration on angular density gives us total amount of x - ray quanta on @xmath82 cm area located at 1.5 m distance from the target .
appropriate amounts appeared equal to @xmath83 , @xmath84 , and @xmath85 photons / e@xmath86 for ( 111 ) , ( 220 ) , and ( 400 ) reflexes , respectively .
it is evident , that despite a decrease of radiation quantum yield at low energies , the integral amount of quanta do not decrease very drastically and the number of x - ray quanta needful for quality image can still be achieved at 0.1 - 0.2 a beam current .
however , to provide some margin of safety for pxr application for a medical x - ray source , it is necessary to find opportunities to increase its radiation yield .
let us consider some variants of radiation yield increase .
one of the ways to increase x - rays spectral - angular density may be the generation of radiation in a multi - wave mode , in other words , when pxr emission goes simultaneously on several systems of the crystallographic planes .
in @xcite measurements of angular distributions of pxr generated in a gaas crystal by relativistic electrons with energies 500 and 900 mev were reported . in these experiments the anomalies ( i.e. significant intensity increase in narrow angular range ) , which can not be explained by the pxr two - wave theory @xcite , were observed .
the analysis of geometry has shown that conditions of multi - wave diffraction were realized for emitted photons .
thorough theoretical description for experiments @xcite was performed in @xcite . in @xcite demonstration of multi - wave effects in low - energy range
was numerically analysed for pxr generated by 7 mev electrons in conditions of eight - wave diffraction .
calculations have shown that , despite of strong ms of electrons at low energies , multi - wave effects may appear in pxr angular distribution and also result in the formation of a strong narrow peak in the centre of two - wave angular distribution .
next a rather promising way for radiation spectral - angular density increase is the application of stratified ( multi - layered ) crystal targets @xcite .
pxr intensity increases up to 7 - 8 times in 10-layer target made of relatively thin silicon crystal foils in comparison with monolithic target of equivalent thickness have been reported in the paper . at last ,
targets made of mosaic crystals should be mentioned ( for example , @xcite ) .
they also can provide an increase in the intensity of x - rays .
research of the behaviour of sophisticated targets at low energies , preferably at conditions that can provide existing medical accelerators , and determination of their optimal characteristics in this range , may lead to a practical design for a real medical monochromatic x - ray source . at the end
, our results were obtained for laue geometry , but for bragg geometry ms influence can be somewhat weaker .
bragg geometry of pxr radiation can also provide more intensive x - ray yield .
it is intended to be a subject of our further research .
authors are very grateful to dr .
o. missevich for fruitful discussions on x - ray imaging issues . | prospects of parametric x - rays ( pxr ) application for the development of a tuneable quasi - monochromatic x - ray source for medical imaging are discussed .
analysis of basic requirements for electron accelerator shows that it must be relatively low - energy and high - current linac . in comparison with known ultra - relativistic cases , at low energies pxr properties
will be modified to a great extent by multiple scattering of the electrons .
pxr intensity dependence on target thickness and beam energy are calculated taking multiple scattering into account . it is concluded that pxr source based on real medical accelerators is feasible and can provide x - ray flux needful for obtaining high quality medical images . *
_ keywords _ : * parametric x - rays ; pxr ; monochromatic tuneable x - ray source ; medical imaging * * _ pacs _ * * _ _ : _ _ 78.70.-g ; 11.80.la ; 87.56.by |
if low - energy supersymmetry ( susy ) is realised in nature it should show up at the large hadron collider ( lhc ) .
strongly interacting particles as gluinos and squarks will be most likely the first susy particles to be seen at lhc .
gluinos of mass less than @xmath20 tev and squarks of mass less than @xmath21 tev [ 1 - 3 ] can be detected , covering in such a way the whole theoretically motivated parameter space .
lhc is also a good laboratory for the search of electro - weakly interacting particles , e.g. sleptons [ 4,5 ] . in a recent paper [ 6 ] it was shown within the minimal supergravity ( msugra ) [ 7 ] model that sleptons in the mass range of @xmath22 to 400 gev can be detected at lhc by investigating the signature @xmath23 .
however , this final state where direct production ( drell - yan ) of sleptons predominates requires high luminosity , @xmath24 .
sleptons can be also produced indirectly in the decays of charginos and neutralinos , especially in @xmath25 decays .
the charginos and neutralinos , can in turn be produced directly or come from gluinos and/or squarks .
this leads to final states with @xmath26 . actually , this indirect slepton production through @xmath27 decays has the largest cross - section in a sizable region of the parameter space accessible at lhc and could allow sleptons to be already revealed at @xmath28 , i.e. simultaneously with strongly interacting sparticles .
thus , indirect production of sleptons can be more important for observing a slepton signal than direct one [ 8 ] .
moreover , in such a way the mass reach for sleptons search can be extended up to @xmath29 740 gev .
having evidence for susy at lhc , one of the next tasks will be to find out the underlying model and to determine the model parameters . in this paper
we work out a method to determine susy parameters and suggest a strategy for getting information on masses of susy particles by means of the signature @xmath30 .
our study is made within the framework of the minimal supergravity model ( msugra ) [ 7 ] . in this model
all scalar particles ( sfermions and higgs bosons ) have a common mass @xmath2 at @xmath31 gev .
the gaugino masses @xmath32 ( corresponding to u(1 ) , su(2 ) , and su(3 ) , respectively ) unify to a common gaugino mass @xmath3 , and all trilinear coupling parameters @xmath33 have the same value @xmath34 at @xmath35 .
one also has unification of the electroweak and strong coupling parameters @xmath36 [ 9 ] .
a further reduction of the parameters is given by invoking radiative symmetry breaking. as a consequence , one has only the following input parameters : @xmath37 . here
@xmath38 , the ratio of the two vacuum expectation values of the two higgs doublets , and @xmath39 is the higgsino mass parameter .
the whole susy particle spectrum can then be calculated by making use of renormalization group equations ( rge ) .
this model is also incorporated in the monte
carlo generator isasusy [ 10 ] which is used in our analysis .
this paper is aimed at determining the parameters @xmath2 and @xmath3 with fixed @xmath40 . knowing these parameters
, we can calculate the masses of the superpartners using rge . for this purpose
, we study in detail the leptonic decays of @xmath41 which have some useful features . within the msugra model
, @xmath41 has two - body decays , @xmath42 , in the region @xmath43 of the parameter space , whereas in the region @xmath44 , @xmath45 gev the @xmath41 has three - body decays , @xmath46 . in both regions the invariant dilepton mass spectrum ( @xmath47 ) has a maximum @xmath48 , and therefore a pronounced structure with a sharp edge can be seen at the kinematical endpoint .
this property was discussed first in ref . [
11 ] in the case of three - body decays of @xmath41 and then in ref .
[ 12 ] in the case of two - body decays .
the generality of this feature , i.e. the observability of an edge in the @xmath47 spectrum even in inclusive @xmath49 and @xmath50 final states in a large part of the parameter space was shown in [ 13 ]
. we will show how much the parameters are constrained by a measurement of the @xmath48 value of the dilepton mass spectra .
moreover , we will discuss a method , based on the analysis of the @xmath47 spectrum , to find out whether the observed edge is due to the two - body or three - body decays of @xmath41 .
within the minimal supersymmetric standard model ( mssm ) the masses of the neutralinos are determined by the parameters @xmath51 , @xmath39 , and @xmath40 using @xmath52 ( @xmath53 being the u(1 ) gaugino mass ) . in the following , we fix @xmath54 ( we assume that @xmath40 could be known from previous experiments ) and
take @xmath55 . in msugra
@xmath56 quite generally turns out to be @xmath57 , so that in this case @xmath58 .
both @xmath59 and @xmath41 are gaugino like , @xmath59 is almost a pure @xmath60 , and @xmath41 almost a pure @xmath61 . in supergravity
the slepton masses are given by [ 14 ] : @xmath62 @xmath63 @xmath64 analogous equations exist for squarks [ 14 ] .
therefore , when the parameters @xmath65 and @xmath66 are known we can calculate all sparticle masses . a special case is the third generation of squarks and sleptons , where l@xmath67r mixing plays a crucial rle .
neutralinos @xmath41 can be produced at the lhc through a drell - yan mechanism ( direct production ) , in association with strongly interacting sparticles , or in the decay chain of gluinos and squarks ( indirect production ) .
gluino and squark pair production processes are the dominant source of @xmath41 s because of large strong interaction cross - sections .
the branching ratios of gluino and squark decays into @xmath41 are also sizable and are shown in fig.1 as a function of the model parameters @xmath2 and @xmath3 .
one can see that the regions in ( @xmath2,@xmath3 ) plane where the decays @xmath68 and @xmath69 are open , are complementary . in the region
@xmath70 gluinos are lighter than squarks and can decay into @xmath41 , while squarks prefer to decay into gluinos . in the region
@xmath71 squarks are lighter than gluinos and can decay into @xmath41 , but gluinos decay into squarks .
hence the decays @xmath68 ( @xmath69 ) and @xmath72 ( @xmath73 ) can coexist ( see also fig.1 ) .
fig.2(a , b ) shows @xmath74 for indirect @xmath41 production from gluinos and squarks as a function of @xmath2 and @xmath3 .
there are three different leptonic decays of @xmath41 interesting for our study : @xmath75 , @xmath76 and @xmath77 .
when decays of @xmath41 to sleptons are allowed , sleptons decay directly into the lightest supersymmetric particle ( lsp@xmath78 ) with br(@xmath79 .
branching ratios of @xmath41 decays into leptons directly and via sleptons are shown in fig.3 as a function of @xmath2 and @xmath3 .
the regions where these decays are kinematically allowed are complementary in the parameter plane , depending on whether @xmath41 is lighter or heavier than @xmath10 .
thus , one can distinguish three domains in the ( @xmath2,@xmath3 ) plane , which are ( also see fig.4 ) : domain ii ( @xmath80 @xmath81 = @xmath82 = @xmath67 = @xmath83 domain i ( @xmath84 @xmath85 @xmath86 with @xmath87 + + domain ii ( @xmath88 @xmath81 @xmath89 @xmath86 with @xmath90 + + domain iii ( @xmath91 @xmath92 @xmath86 with @xmath93 + in domain iii , the decay @xmath94 would also be kinematically allowed , but since the @xmath60 component of @xmath41 is very small , the coupling to @xmath95 is also small .
therefore , the decay @xmath96 is very much suppressed in the whole domain . in fig.5
we show the regions for @xmath97 and 0.1 pb , @xmath98 , 0.1 pb and @xmath99 , 0.002 pb in the ( @xmath100 ) plane from indirect and associated @xmath41 production followed by decays to @xmath101 final states directly or via sleptons . one can see that there are regions in domains ii and iii where the mentioned decays coexist .
finally , in fig.6 we show @xmath74 for indirect and associated production of @xmath41 decaying into leptons directly or via sleptons .
in order to determine the parameters @xmath2 and @xmath3 we will exploit in the following the kinematical features of the two- and three - body leptonic decays of @xmath41 . as pointed out in
[ 11 ] , the decay @xmath103 ( domain i ) has the useful kinematical property that the invariant mass of the two leptons @xmath47 has a maximum at @xmath104 whereas for the decays @xmath105 ( domain ii and iii ) the maximum of @xmath47 is given by [ 12 ] : @xmath106 thus , the @xmath47 distribution has a very characteristic shape with a sharp edge at the kinematical endpoint @xmath107 .
as the main source of @xmath41 s is their indirect production in gluino and squark decays , the most suitable signature for selecting the @xmath6 decays is provided by the topology with two same flavour opposite sign leptons accompanied by large missing transverse energy and usually accompanied by a high multiplicity of jets . in this paper
we thus concentrate on the @xmath108@xmath109@xmath110 channel , where the final state leptons are electrons and muons .
the simulations are done at the particle level , with parametrised detector responses based on detailed detector simulations .
these parametrisations are adequate for the level of detector properties we want to investigate , and are the only practical ones in view of the multiplicity and complexity of the final state signal and background channels investigated .
the essential ingredients for the investigation of susy channels are the response to jets , @xmath111 , the lepton identification and isolation capabilities of the detector , and the capability to tag b - jets .
the cms detector simulation program cmsjet 3.2 [ 15 ] is used .
it incorporates the full electro - magnetic ( ecal ) and hadronic ( hcal ) calorimeter granularity , and includes main calorimeter system cracks in rapidity and azimuth .
the energy resolutions for muons , electrons ( photons ) , hadrons and jets are parametrised .
transverse and longitudinal shower profiles are also included through appropriate parametrisations .
the main detector features incorporated in the monte carlo description are : @xmath112 hadronic tracks , muons and electrons are measured up to @xmath113=2.4 @xmath112 deflection of charged particles due to the 4 t magnetic field is included .
@xmath112 the resolution for the muon system is parametrised according to [ 16 ] .
@xmath112 the calorimetric coverage goes up to at @xmath113=5 for the hcal and @xmath113=2.6 for the ecal .
@xmath112 ecal energy resolution parametrized as : @xmath114 @xmath112 hcal energy resolution is parametrised according to [ 17 ] as a function of @xmath115 ; a typical hadron resolution is : @xmath116 @xmath112 energy resolution for very forward calorimeter ( vfcal ) , in the parallel plate chambers option : @xmath117 @xmath112 granularity of calorimeters : [ cols="^,^,^",options="header " , ] one should first notice that the observation of two edges at @xmath11 pb@xmath12 would determine the ( @xmath100 ) point uniquely .
this is due to the fact that the set of the edge position values in the @xmath47 spectrum is different at each point of the parameter region ( _ case 1 _ ) , where two edges are expected to be observed at @xmath118 pb@xmath12 , see figs.10 and 14 . with a luminosity @xmath119 pb@xmath12
the positions of the two edges will fix two ( @xmath100 ) points , belonging to domains ii and iii and corresponding to _ case 1 _ and _ case 2 _ , respectively . at high
luminosity , @xmath120 pb@xmath12 , the observation of two edges can give up to three possible ( @xmath100 ) points .
one of them is from domain ii , corresponding to _ case 1_. the lines @xmath121 corresponding to @xmath122 decays have the form of an ellipse and can cross the @xmath121 lines corresponding to @xmath123 decays twice .
hence , two points with the same set of edge positions in the @xmath47 spectrum can be found in domain iii corresponding to _ case 2_. a discrimination between these points is possible on basis of the event kinematics , and/or by an analysis of the total event rate and the relative number of events corresponding to the two peaks , ( see figs.5,10 - 12 )
. for a definite value of the edge position @xmath48 one expects a different shape of the @xmath47 distributions in two- and three - body decays ( see fig.13 , where the first peak is due to a two - body decay of @xmath41 and the second one due to a three - body decay ) .
as we have seen from figs.7 and 8 the signal events contribute in the interval @xmath124 .
in the following we only consider events in this mass region .
the average value @xmath125@xmath47@xmath126 of signal and background events of this mass region is shown in fig.15 as a function of @xmath2 .
the errors are calculated by taking into account the statistical error , a systematic error in the measurement of the edge position , and a systematic error of 30 @xmath127 for background uncertainty ( the main background is @xmath128 ) .
one clearly sees that @xmath125@xmath47@xmath126 is significantly smaller in the case of a direct three - body decay @xmath129 .
thus the shape of the dilepton mass spectrum already allows one to decide whether @xmath41 decays into a slepton or not . in order to distinguish between domains ii and iii ,
we suggest to use the fact that in general the contour lines with the same @xmath48 for right and left sleptons have no overlap in @xmath3 in regions of parameter space which are accessible at a given luminosity , see figs.9 and 14 as an example .
it means that the masses of @xmath59 s are different for these two lines , and hence @xmath111 is expected to be different . in fig.16
we show the @xmath125@xmath111@xmath126 values for events with @xmath130 decays after the cuts @xmath131 gev and @xmath132 gev , @xmath133 .
the errors are calculated by taking into account the statistical error and a systematic error in the measurement of the edge position . as can be seen from fig.16 ,
@xmath134@xmath111@xmath135 is larger in the case of @xmath136 than in the case of @xmath137 as expected . when the correct @xmath48 line is chosen , the last step is to find the point ( @xmath100 ) on this line . in general
the cross section falls with increasing @xmath3 and @xmath2 .
thus , we study the event rate along the corresponding @xmath48 line .
we first discuss the domain iii , where the situation is simpler .
for the event rate analysis at @xmath11 pb@xmath12 , to reduce the uncertainties due to background , we use a harder cut on @xmath111 , @xmath138 gev .
the dependence of the expected event rate on @xmath2 is shown in fig.17 .
the errors are calculated by taking into account the statistical error and a systematic error of 30 @xmath127 for background uncertainty . a systematic error due to the precision of the edge position measurement
is also taken into account . from the observed event rate
we can then determine @xmath2 with a good accuracy , @xmath139 gev .
the parameter @xmath3 is then given by the @xmath140line in the @xmath141 plane .
the precision obtained in such a way is @xmath142 gev
. in domain ii , the event rate along a line of definite @xmath48 is first increasing and then decreasing with @xmath2 , see fig.18 .
this is mainly due to the change in the branching ratios ( see fig.4 ) .
the dependence of the event rate is , however , such that @xmath2 can not be determined uniquely . to a given event rate there
correspond in general two @xmath2 values .
the ambiguity can , however , be solved at high luminosity @xmath120 pb@xmath12 , when two edges in the @xmath47 distribution can be observed .
for domain i , the @xmath2 dependence of the event rate is shown in fig.19a , again for @xmath143 gev .
notice the steep increase of the rate at @xmath144 gev .
this is due to the fact that the decay channel @xmath129 is just opening in this region .
as can be seen from the curve in fig.19a , there is an ambiguity in the determination of @xmath2 if the event rate is in the region @xmath145 or 120 gev @xmath146 240 gev . here
it helps if we look at the average number of jets @xmath125@xmath147@xmath126 in the events under study .
fig.19b shows @xmath125@xmath147@xmath126 as a function of @xmath2 .
@xmath125@xmath147@xmath126 is rising with @xmath2 as more jets are produced as the squarks become heavier . with the measured @xmath125@xmath147@xmath126
we can resolve the ambiguity in the mentioned region 120 gev @xmath146 240 gev and thus determine @xmath2 with @xmath148 gev .
in this paper we have performed a detailed analysis of events with the signature @xmath30 to be expected in pp collisions at lhc .
our aim has been to determine the parameters @xmath2 and @xmath3 of the minimal supergravity model and to get information on the mass spectrum of susy particles , assuming knowledge of @xmath149 from previous experiments .
we have exploited the property of the @xmath41 leptonic decays @xmath150 , @xmath151 that the invariant mass of the two final leptons has a maximum , @xmath48 , clearly visible even in inclusive production . we have determined for different luminosities the regions in the ( @xmath100 ) parameter plane where one or two edges can be observed in the invariant dilepton mass distributions .
these regions already give preliminary information about the model parameters .
the appearance of the edges in the @xmath47 distributions can be already seen with a luminosity @xmath11 pb@xmath12 .
therefore we have concentrated on a low luminosity study . on the other hand , in case no such observation will be made at this luminosity , the corresponding parameter region can be excluded , and the same analysis can be done at higher luminosity .
we have shown that a measurement of the @xmath48 value constrains the parameters mainly to three lines in the ( @xmath100 ) parameter plane .
the lines correspond to the decay modes @xmath152 , @xmath153 , @xmath154 respectively .
we have worked out a method to discriminate the three - body from the two - body @xmath41 decays . in the case of three - body @xmath41 decays the parameter @xmath3 can be determined by the measured value of @xmath48 with a precision of @xmath155 gev .
the parameter @xmath2 can then be determined from the observed event rate with a precision of 7 - 3 gev . in the case of two - body @xmath41
decays , a measurement of the missing transverse energy can allow one to distinguish between the two possible decays @xmath156 and @xmath157 , but a more detailed study is needed . by an event rate analysis along the corresponding line in the ( @xmath100 ) plane
we can determine @xmath2 and @xmath3 , @xmath158 gev .
knowing @xmath2 , @xmath3 and @xmath149 , the masses of all susy particles ( except for the 3rd generation of squarks and sleptons ) are calculated by rge .
the precisions which can be achieved are @xmath159 gev .
in such a way it is possible to obtain information about susy particle masses already with low luminosity ( @xmath160 pb@xmath12 ) even without having direct experimental evidence for their existence . this is especially important for sleptons in a parameter region where high luminosity would be necessary to detect them through direct production .
this study has been performed for @xmath161 , but it is also possible for high values of @xmath149 .
most likely , for large @xmath162 a higher luminosity will be needed because of smaller branching ratios of the @xmath41 leptonic decays .
let us mention some further interesting aspects of this work .
selecting the two - body @xmath41 leptonic decays by our method represents an indirect evidence for sleptons in the framework of msugra .
in this way it is possible to probe slepton masses up to @xmath163 gev well beyond what is possible in direct [ 5,6 ] searches .
as it has been shown in this study , the edge in the invariant dilepton mass distributions is expected to appear at @xmath164 gev , being quite generally a signal for a two- or three - body decay of some abundantly produced heavy object .
hence such an observation may serve as a first evidence for physics beyond the standard model , and if observed with significant @xmath111 it would be a clear evidence for susy , more specifically for @xmath41 production .
we especially thank howard baer for answering many questions and for very useful discussions .
we would like to thank alfred bartl for his interest in this work and valuable discussions .
thanks for financial support by the austrian academy of sciences .
h. baer , f. paige , s. protopopescu and x. tata , in _ proceedings of the workshop on physics at current accelerators and superolliders _ , ed .
j. hewett , a. white and d. zeppenfeld ( argonne national laboratory , 1993 ) . | we analyse the events with two same flavour , opposite sign leptons + @xmath0 + ( jets ) as expected in @xmath1 collisions at lhc within the framework of the minimal supergravity model .
the objective is the determination of the parameters @xmath2 and @xmath3 of this model ( for a given value of @xmath4 ) .
the signature @xmath5 + ( jets ) selects the leptonic decays of @xmath6 , @xmath7 , @xmath8 .
we exploit the fact that the invariant dilepton mass distribution has a pronounced structure with a sharp edge at the kinematical endpoint even in such an inclusive final state over a significant part of parameter space .
we determine the domain of parameter space where the edge is expected to be visible .
we show that a measurement of this edge already constrains the model parameters essentially to three lines in the ( @xmath9 ) parameter plane .
we work out a strategy to discriminate between the three body leptonic decays of @xmath6 and the decays into sleptons @xmath10 .
this procedure may make it possible to get information on susy particle masses already with low luminosity , @xmath11 pb@xmath12 .
# 1#23.6pt cms note 1997/094
hephy - pub 678/97 * determining the parameters of the minimal supergravity model from final states at lhc . * d. denegri@xmath13 , w.majerotto@xmath14 , and l. rurua@xmath15 @xmath13_centre detudes nucle@xmath16ire de saclay , gif - sur - yvette , france
_ @xmath14_institut f@xmath17r hochenergiephysik , @xmath18sterreichische akademie d.wissenschaften , vienna , austria _
@xmath19_institute of physics , tbilisi , georgia _ |
the density matrix renormalization group ( dmrg ) @xcite has been widely applied to one - dimensional ( 1d ) quantum systems and two - dimensional ( 2d ) classical systems @xcite .
a frontier in dmrg is to extend its numerical algorithm to higher dimensional systems , chiefly for 2d quantum and 3d classical systems .
as far as the finite system algorithm is concerned , decomposition of higher - dimensional clusters to 1d chains proposed by liang and pang works efficiently @xcite . on the other hand ,
we have not obtained any satisfactory answer to extend dmrg toward infinite - size systems in higher dimension .
nishino and okunishi proposed a way of extending dmrg to 3d classical systems , which they call ` the corner tensor renormalization group ( cttrg ) ' @xcite , as a 3d generalization of both the transfer matrix dmrg @xcite and the corner transfer matrix renormalization group ( ctmrg ) @xcite for 2d classical systems .
two major problems are found in cttrg when it is applied to the 3d ising model .
one is that the calculated transition temperature @xmath0 is much higher than one of the most reliable @xmath0 obtained by the monte carlo ( mc ) simulations @xcite .
the other problem is the very slow decay of the the density - matrix eigenvalues @xcite , that spoils the numerical efficiency of the block - spin transformation .
a way of generalizing dmrg to higher dimensions is to investigate the variational structure of dmrg , where the variational state for the transfer matrix or the hamiltonian is written in a product of orthogonal matrices @xcite .
two types of 2d tensor product states have been considered as higher - dimensional extensions of this matrix product state .
one is ` the interaction round a face ( irf ) ' type product states shown in figure ( [ fig:1](a ) ) @xcite , and the other is the vertex type one in figure ( [ fig:1](b ) ) @xcite . for a 2d tensor product state @xmath1 , the variational energy and the variational partition function , respectively , is written as @xmath2 where @xmath3 and @xmath4 denotes a hamiltonian for a 2d quantum system and a transfer matrix for a 3d classical system .
let us call such a variational estimation as ` the tensor product variational approximation ( tpva ) ' in the following .
calculation of @xmath5 have been performed by mc simulation @xcite , product wave function renormalization group ( pwfrg ) @xcite or ctmrg @xcite . a key point in tpva
is to find a good variational state @xmath1 .
so far , they assumed a specific form of @xmath1 , which contains several variational parameters , and tried to find out the best @xmath1 by way of the parameter sweep .
for example , okunishi and nishino @xcite investigated tpva for the 3d ising model , assuming @xmath1 in the form of the kramers - wannier ( kw ) approximation @xcite .
their variational state contains two adjustable parameters , and the best @xmath1 is obtained through a two - parameter sweep .
such an intuitive construction of @xmath1 is , however , not always applicable ; for example , we do nt know how to extend the kw approximation for the 3d potts models @xcite .
how can we obtain the best @xmath1 automatically for higher - dimensional systems ?
we find an answer for 3d classical systems . in this paper
we propose a self - consistent improvement for the tensor product state @xmath1 by way of ctmrg .
we choose the 3d ising model as an example of the 3d classical systems , and formulate our self - consistent method in terms of the ising model . in the next section , we introduce the simplest 2d tensor product state , and give the formal expression of the variational partition function @xmath5 in eq.([varf ] ) .
we then obtain the self - consistent equation for @xmath1 in 3 , considering the variation @xmath6 . in
4 we propose a numerical algorithm to solve the self - consistent equation . in 5 we check the numerical efficiency and stability of this algorithm when it is applied to the 3d ising model and the 3-state ( @xmath7 ) 3d potts model .
conclusions are summarized in 6 .
we briefly review the variational formulation of tpva that was used for the kw approximation of the 3d ising model @xcite .
let us consider the 3d ising model on the simple cubic lattice of the size @xmath8 to @xmath9 , @xmath10 and @xmath11 directions , respectively , where open ( or fixed ) boundary conditions are assumed for both @xmath9 and @xmath10 directions .
we are interested in the bulk property of this model , and therefore suppose that the system size @xmath12 is sufficiently large .
suppose that the neighboring ising spins @xmath13 and @xmath14 have ferromagnetic interaction @xmath15 .
the transfer matrix @xmath4 from a @xmath16 spin layer @xmath17 \equiv \left [ \begin{array}{ccccc } \sigma_{1\,\,\,1 } & \ldots & \sigma_{1\,\,n } & \ldots & \sigma_{1\,\,\,2n } \\ \vdots & \ddots & \vdots & \ddots & \vdots \\ \sigma_{n\,\,1 } & \ldots & \sigma_{n\,\,n } & \ldots & \sigma_{n\,2n } \\ \vdots & \ddots & \vdots & \ddots & \vdots \\ \sigma_{2n\,1 } & \ldots & \sigma_{2n\,n } & \ldots & \sigma_{2n\,2n } \end{array } \right ] \label{lspin}\ ] ] to the next layer @xmath18 $ ] is then expressed as a product of local factors @xmath19 = \prod_{i=1}^{~2n-1~ } \prod_{j=1}^{~2n-1~ } x_{ij}^{~ } \equiv \prod_{ij } x_{ij}^{~ } \
, , \label{trm}\ ] ] where @xmath20 represents the boltzmann factor for a local cube @xmath21 parameterized by @xmath22 .
we have used the notation @xmath23 and @xmath24 .
( see figure ( [ fig:2 ] ) . )
we define @xmath25 $ ] so that it is symmetric , because the symmetry simplifies the following formulation .
the variational state in tpva is a uniform product of local tensors . in this paper
, we focus on the simplest construction of the tensor product state @xmath26 = \prod_{ij}^{~ } w_{ij}^{~ } = \prod_{ij } w\left ( \begin{array}{cc } \sigma_{i\,j}^{~ } & \sigma_{i\,j'}^{~ } \\
\sigma_{i'\!j}^{~ } & \sigma_{i'\!j'}^{~ } \end{array } \right ) \ , , \label{prods}\ ] ] where the local tensor @xmath27 does not contains auxiliary variables , which are shown by black squares in figure ( [ fig:1](a ) ) @xcite ; to include the auxiliary variables is straightforward , but makes the following equations rather lengthy .
the tensor product state @xmath28 $ ] is uniform in the sense that @xmath27 is position independent .
the local tensor @xmath27 has 16 parameters , but not all of them are physically independent @xcite . for the book keeping , let us use the notation @xmath29 for the plaquett spin , and write the local tensor @xmath27 simply as @xmath30 . in the same manner ,
let us write @xmath20 as @xmath31 .
( see figure ( [ fig:2 ] ) . )
using @xmath32 $ ] and @xmath28 $ ] thus defined , the variational partition function per layer is expressed as @xmath33 [ \sigma ] } ^{~ } v [ \bar\sigma ] \ ,
t [ \bar\sigma | \sigma ] \ , v [ \sigma ] } { \displaystyle \sum _ { [ \sigma ] } ^{~ } \left ( v [ \sigma ] \right)^2 } = \frac{\displaystyle \sum _ { [ \bar\sigma ] [ \sigma ] } ^{~ } \prod_{ij } w\ { \bar\sigma_{i\,j}^{~ } \ } \ , x\ { \bar\sigma_{ij } | \ , \sigma_{ij}^{~ } \ } \ , w\ { \sigma_{i\,j}^{~ } \}}{\displaystyle \sum _ { [ \sigma ] } ^{~ } \prod_{ij } \left ( w\ { \sigma_{i\,j}^{~ } \ } \right)^2 } \nonumber\\ & = & \frac{\displaystyle \sum _ { [ \bar\sigma ] [ \sigma ] } ^{~ } \prod_{ij } g^1_{~}\ { \bar\sigma_{ij } | \ , \sigma_{ij}^{~ } \ } } { \displaystyle \sum _ { [ \sigma ] } ^{~ } \prod_{ij } g^0_{~}\ { \sigma_{ij}^{~ } \ } } \equiv \frac{z^1_{~}}{z^0_{~ } } \ , , \label{varf2}\end{aligned}\ ] ] where we have defined @xmath34 and @xmath35 as @xmath36 it should be noted that @xmath37 is a partition function of an irf model @xcite on @xmath16 square , and @xmath38 is that of a 2-layer lattice model of the same size .
now we explain the self - consistent equation for the variational state @xmath39 $ ] , the equation which is satisfied when @xmath5 in eq.([varf2 ] ) is maximized .
let us consider the variation of @xmath5 with respect to the variations of local tensors @xmath40 under the condition that the system size @xmath12 is sufficiently large and the boundary effect is negligible . then most of the terms in the r.h.s .
are almost the same , and it is sufficient to consider the variation of @xmath5 with respect to the local change @xmath41 @xmath42 at the center of the system , where @xmath41 represents the local tensor at the center .
( see eqs.([lspin ] ) and ( [ prods ] ) . )
the variation @xmath43 can be explicitly written down by use of two matrices .
one is the diagonal matrix @xmath44'}^{~ } \prod_{(ij)\neq(nn)}^{~ } \!\!\!\ ! g^0_{~}\ { \sigma_{ij}^{~ } \ } \ , , \label{mata}\ ] ] where @xmath45'}^{~}$ ] denotes spin configuration sum for all the spins in the layer @xmath46 $ ] except for the spin plaquett @xmath47 at the center ; we interpret @xmath48 as a 16-dimensional matrix @xmath49 where @xmath50 and is zero when @xmath51 . from the definition ,
@xmath37 in eq.([varf2 ] ) is equal to @xmath52 .
the other matrix is @xmath53 ' [ \sigma ] ' } ^{~ } \prod_{(ij)\neq(nn)}^{~ } \!\!\!\ ! g^1_{~}\ { \bar\sigma_{ij } | \ , \sigma_{ij}^{~ } \ } \ , , \label{matb}\ ] ] which is related to @xmath38 as @xmath54 . by use of @xmath48 and latexmath:[$b\ { \bar\sigma_{nn}^{~ } @xmath5 as @xmath56 where we have dropped the subscripts from @xmath47 and @xmath57 for book keeping .
the condition @xmath58 @xmath59 draws the eigenvalue problem @xmath60 between the matrix @xmath61 and @xmath62 ; here we regard @xmath63 as a 16-dimensional vector .
this is the self - consistent equation that an optimized tensor product state @xmath28 $ ] should satisfy .
to use the self - consistent relation eq.([selfc ] ) , we have to obtain @xmath64 and @xmath65 for very large @xmath66 .
though it is impossible to obtain @xmath64 and @xmath65 exactly , the ctmrg @xcite enables us to numerically obtain them very accurately .
let us introduce a new notation @xmath67 which groups a pair of adjacent spins @xmath68 and @xmath69 .
using @xmath70 , we can rewrite the stack of two plaquett spins @xmath71 as @xmath72 @xmath31 as @xmath73 , and @xmath74 as @xmath75 .
( see figure ( [ fig:2 ] ) ) we drop the subscripts from @xmath76 to write it as @xmath77 when its position is apparent .
the matrices @xmath64 and @xmath78 can be expressed as a combination of the corner transfer matrices ( ctms ) and the half - row transfer matrices ( hrtms ) , that appears when we apply ctmrg to both the denominator and the numerator of eq.([varf2 ] ) to obtain @xmath37 and @xmath38 @xcite .
let us write the ctm used for the calculation of @xmath37 and @xmath38 , respectively , as @xmath79 and @xmath80 , where @xmath81 , @xmath82 , @xmath83 , and @xmath84 are @xmath85-state block spin variables .
also let us write hrtm as @xmath86 and @xmath87 in the same manner .
note that @xmath79 and @xmath86 are created from @xmath88 , and @xmath80 and @xmath87 are from @xmath89 . combining these ctms and hrtms , @xmath64 and @xmath90
are constructed as @xmath91 where the positions of the spin variables are shown in figure ( [ fig:3 ] ) . in principle , we can use @xmath64 and @xmath90 thus constructed to solve the self - consistent eq.([selfc ] ) . to make the self - consistent improvement for @xmath63 more efficiently ,
we employ a numerical algorithm that simultaneously performs the extension of the system size in ctmrg and the self - consistent improvement by eq.([selfc ] ) .
the numerical procedures are as follows : * create @xmath92 and @xmath93 from @xmath94 defined in eq.([cube ] ) and the initial @xmath63 : @xmath95 the choice of the initial @xmath63 is not so relevant , since it is improved afterward . *
create the initial @xmath79 and the initial @xmath96 from @xmath92 , following the standard initialization procedure in ctmrg @xcite . also create @xmath97 and @xmath98 from @xmath93 in the same way . *
obtain the matrices @xmath64 and @xmath78 using eqs.([crctm ] ) .
* improve @xmath63 by multiplying @xmath99 @xmath100 and normalize @xmath101 so that @xmath102 is satisfied . * recreate @xmath92 and @xmath93 by substituting @xmath101 into eqs.([creg ] ) . *
extend @xmath103 and @xmath104 to obtain @xmath105 and @xmath106 , respectively , by joining the recreated @xmath34 and @xmath35 as shown in figure ( [ fig:4 ] ) ; the numerical details are shown in ref . @xcite . also extend @xmath107 and @xmath108 to obtain @xmath109 and @xmath110 .
* create density matrices from the extended ctms , and diagonalizing them to obtain the rg transformations @xmath111 and @xmath112 , where @xmath81 and @xmath83 are @xmath85-state block spins .
then apply the rg transformations to @xmath105 , @xmath106 @xmath109 and @xmath110 .
* goto ( c ) , and repeat ( c)-(g ) to improve @xmath63 iteratively , and stop when @xmath63 reaches its fixed point .
to summarize , we put three additional steps ( a ) , ( c ) and ( d ) to the standard ctmrg algorithm .
let us check the numerical efficiency and stability of the self - consistent tpva through trial applications to the 3d ising model and the ferromagnetic @xmath7 potts model . figure ( [ fig:5 ] ) shows the spontaneous magnetization @xmath113 at the center of the @xmath114 system , where the curve , cross marks , and triangles , respectively , represent the result of the mc simulation by tarpov and blte @xcite , kw approximation @xcite , and the self - consistent tpva .
we calculate @xmath115 after repeating the iteration ( c)-(g ) in the last section for @xmath116 times at most , keeping @xmath117 to @xmath118 states for the block spin variables ; the convergence with respect to @xmath85 is very fast , where we obtain almost the same @xmath115 for the cases @xmath119 and @xmath120 .
the self - consistent improvement by eq.([improv ] ) is monotonous in the whole parameter range , and no oscillatory instability is observed .
the calculated transition point @xmath121 is about 1.3% smaller than the mc result @xmath122 .
it turns out that the spontaneous magnetization calculated by kw approximation , which gives the transition point @xmath123 , is quite close to the result of the self - consistent ( sc ) tpva .
this means that the intuitive choice of the variational state in kw approximation is actually very good , within the simplest product state defined in eq.([prods ] ) .
inclusion of auxiliary variables to the tensor product state is necessary for the further improvement of the tensor product variational state .
note that the computational time required for the kw approximation is several times larger than the self - consistent tpva , because the former finds the partition function extremum via 2-parameter sweep .
figure ( [ fig:6 ] ) shows the energy per bond @xmath124 of the ferromagnetic 3d @xmath7 potts model , which is calculated by tpva keeping @xmath85 up to 15 .
the self - consistent improvement by eq.([improv ] ) is again monotonous , and @xmath125 is sufficient to get the converged data ; we need smaller @xmath66 for the potts model than ising model , because the phase transition of the potts model is first order . the calculated energy per bond jumps from @xmath126 to @xmath127 at the calculated transition point @xmath128 , where the calculated free energy of the disordered phase coincides with that of the ordered phase .
the calculated transition point is about 0.18% smaller than one of the most reliable mc result @xmath129 @xcite .
the latent heat @xmath130 is about 41% larger than the mc result @xmath131 @xcite .
we have proposed a self - consistent tpva , which gives the optimized tensor product state for 3d classical systems , by way of the self - consistent improvement of the local tensors .
since the method finds out the best variational state without using a priori knowledge of the system , the self - consistent tpva is applicable for various 3d models described by short range interactions . to generalize the self - consistent tpva to 2d quantum systems
is a next subject that one might consider .
this generalization is not trivial , since we have used the specific property of 3d classical systems when we obtain the self - consistent equation .
t. n. thank to g. sierra and m.a .
martn - delgado for the discussion about the tensor product state at csic .
k. o. is supported by jsps research fellowships for young scientists .
this work was partially supported by the `` research for the future '' program from the japan society for the promotion of science ( jsps - rftf97p00201 ) and by the grant - in - aid for scientific research from ministry of education , science , sports and culture ( no . 09640462 and no .
11640376 ) .
most of the numerical calculations were done by compaq fortran on the hpc alpha21264 linux workstation .
white , _ phys .
_ * 69 * ( 1992 ) 2863 .
white , _ phys .
rev . _ * b48 * ( 1993 ) 10345 .
lecture notes in physics , eds .
i. peschel , x. wang , m. kaulke , and k. hallberg ( springer verlag , 1999 ) .
k. hallberg , cond - mat/9910082 .
s. liang and h. pang , _ phys .
rev . _ * b49 * ( 1994 ) 9214 .
t. nishino and k. okunishi , _ j. phys
* 68 * ( 1999 ) 3066 .
t. nishino , _ j. phys .
_ 64 ( 1995 ) 3598
. t. nishino and k. okunishi , _
* 65 * ( 1996 ) 891 .
t. nishino and k. okunishi , _ j. phys
* 66 * ( 1997 ) 3040 .
talapov and h.w.j .
blte , _ j. phys
. a , math . gen . _ * 29 * ( 1996 ) 5727 .
a.m. ferrenberg and d.p .
landau , _ phys .
_ * b44 * ( 1991 ) 5081 .
i. peschel and m .- c .
chung , cond - mat/9906224 .
s. stlund and s. rommer , _ phys .
_ * 75 * ( 1995 ) 3537 ; s. rommer and s. stlund , _ phys .
* b55 * ( 1997 ) 2164 .
g. sierra and m.a .
martn - delgado , cond - mat/9811170 .
h. niggemann , a. klmper and j. zittartz , _ z. phys .
_ * b104 * ( 1997 ) 103 .
y. hieida , k. okunishi and y. akutsu , _ new j. phys . _
* 1 * ( 1999 ) 7 .
kramers and g.h .
wannier , _ phys .
_ * 60 * ( 1941 ) 263 .
k. okunishi and t. nishino , cond - mat/9909097 .
r. b. potts , _ proc .
soc . _ * 48 * 106 .
baxter , _ exactly solved models in statistical mechanics _
( academic press , london , 1982 ) .
w. janke and r. villanova , _ nucl .
phys . _ * b489 * ( 1997 ) 679 . | we propose a numerical variational method for three - dimensional ( 3d ) classical lattice models .
we construct the variational state as a product of local tensors , and improve it by use of the corner transfer matrix renormalization group ( ctmrg ) , which is a variant of the density matrix renormalization group ( dmrg ) applied to 2d classical systems .
numerical efficiency of this approximation is investigated through trial applications to the 3d ising model and the 3d 3-state potts model .
dmrg ; ctmrg ; variational formulation |
a system of two coupled nonlinear schrdinger ( nls ) equations represents the basic model of propagation of weakly dispersive waves having vectorial nature , i.e. , characterized by two components , below designated by @xmath3 and @xmath4 .
this model is relevant for almost all applications of the nonlinear physics .
when the system is supplied by balanced gain and losses , as well as by the linear coupling , characterized respectively by positive coefficients @xmath5 and @xmath6 , it reads @xmath7 where all the nonlinear coefficients , i.e. , @xmath8 , @xmath9 and @xmath10 are real .
model ( [ manakov ] ) describes wave guiding in structures obeying parity - time ( @xmath0 ) symmetry @xcite . here
the parity symmetry is defined by the mapping @xmath11 .
time reversal operator @xmath12 is defined by the map @xmath13 this means that if @xmath14 , then the nonlinearity is also @xmath0 symmetric , i.e. , for any solution @xmath15 defined in the symmetric interval @xmath16 $ ] , there also exists another solution @xmath17 hereafter an overbar stands for the complex conjugation . notice that the nonlinear coupling coefficients @xmath10 in ( [ manakov ] ) are considered to be equal in the both equations to ensure that the system is hamiltonian in the absence of gain and loss with @xmath18 . coupled nls equations ( [ manakov ] ) with @xmath0-symmetric nonlinearity @xmath14 and with @xmath19 , introduced independently in @xcite for the constant gain - and - loss coefficient and in @xcite for gain and losses localized along the waveguide , can be viewed as an extension of the nonlinear dimer model @xcite . in the subsequent works @xcite more general nonlinearities , including ones with @xmath20 , were thoroughly studied .
it was found that the model supports bright solitons @xcite , dark solitons @xcite , peregrin solitons and rogue waves @xcite , and breathers @xcite .
for all refereed cases it is important that the nonlinearity obeys @xmath0 symmetry ( for the relevance of @xmath0 symmetry of the nonlinearity , see @xcite ) . system ( [ manakov ] ) is integrable if @xmath21 and @xmath22 , representing the well - known manakov model @xcite ( see also @xcite for review of these integrable systems ) .
it turns out , that this particular type of ( all equal ) nonlinearities is peculiar also for the @xmath0-symmetric model which accounts for gain and losses , as well as for the linear coupling .
more generally , the role of the type of nonlinearity ( i.e. , non @xmath0-symmetric , @xmath0-symmetric , and manakov type ) can be crucial for the long time behavior of the solution . indication on such possibility stems form the recent results on the stability of trajectories of the nonlinear @xmath0-symmetric dimer .
namely , the nonlinear dimer @xmath23 has trajectories escaping to infinity ( as @xmath24 ) @xcite , while the @xmath25-symmetric manakov dimer , i.e. , the system @xmath26 has all trajectories bounded @xcite , provided that @xmath27 .
note that the later constraint is related to linear stability of the zero equilibrium in systems ( [ manakov ] ) , ( [ dimer ] ) , and ( [ dimerm ] ) .
in the theory of @xmath0-symmetric systems , it is referred to as a condition of unbroken @xmath0 symmetry @xcite . in the present paper
, we address the existence and boundedness of solutions to the cauchy problem for the generalized manakov system ( [ manakov ] ) with different types of nonlinear coefficients . for the situations where analytical results are not available
, we perform numerical simulations .
we also extend the global existence results to the case of two spatial dimensions .
the paper is organized as follows .
main results in one dimension are presented in section 2 .
they consist of two theorems , the first one about global well - posedness of the general model ( [ manakov ] ) and the second one is about bounds on the @xmath2-norm of the solution in the @xmath0-symmetric manakov case .
the two theorems are then proved in sections 3 and 4 .
section 5 reports numerical results illustrating that the @xmath1 norm is also bounded in the @xmath0-symmetric manakov case , but it may grow to infinity in the @xmath0-symmetric non - manakov case .
finally , extensions of these results to the @xmath0-symmetric manakov system in two spatial dimensions are considered in the concluding section 6 .
because system ( [ manakov ] ) is semi - linear , it can be rewritten in an integral form by using duhamel s principle . applying the well - known method based on the contraction mapping principle
( see @xcite for details ) , one can immediately establish existence and uniqueness of local solutions of the cauchy problem for system ( [ manakov ] ) in sobolev spaces @xmath28 for any @xmath29 . in what follows ,
we are interested in the behavior of solutions in the energy space @xmath30 , which corresponds to @xmath31 and appears to be most relevant for physical applications .
our first result establishes existence of global solutions in the energy space for arbitrary choice of coefficients in the generalized manakov system ( [ manakov ] ) . for any @xmath32
, there exists a unique global solution @xmath33 of the generalized manakov system ( [ manakov ] ) such that @xmath34 .
[ theorem-1 ] we remark , that explicit examples of global solutions , including bright solitons , can be readily obtained in the case @xmath35 and @xmath27 , since the substitution @xcite @xmath36 with @xmath37 defined by @xmath38 , reduces system ( [ manakov ] ) to the integrable nonlinear schrdinger equation @xmath39 the @xmath1-norm of a global solution to the generalized manakov system ( [ manakov ] ) remains finite but may grow as @xmath24 .
this growth happens typically in the systems with gains and losses , at least for sufficiently large initial data , see e.g. @xcite .
it is then quite surprising that the manakov system ( [ manakov ] ) with @xmath40 has bounded solutions in the @xmath2-norm if @xmath27 ( i.e. under the condition of the unbroken @xmath0 symmetry ) .
the next theorem formulates the corresponding result .
assume @xmath22 and @xmath27 .
for any global solution @xmath33 of theorem [ theorem-1 ] , there exists a constant @xmath41 such that @xmath42 for @xmath43 , there exists a global solution of theorem [ theorem-1 ] such that @xmath44 [ theorem-2 ] note that theorem [ theorem-2 ] generalizes the result of our previous work @xcite devoted to the integrable @xmath25-symmetric dimer ( [ dimer ] ) that corresponds to the @xmath45-independent solutions of system ( [ manakov ] ) . in comparison ,
the result of theorem [ theorem-1 ] is novel and the proof is based on deriving an apriori energy estimates for the @xmath1-norm of the local solutions and applying gronwall s inequality .
also note that the results of theorems [ theorem-1 ] and [ theorem-2 ] do not clarify if the @xmath1-norm of the solution is also globally bounded in the case @xmath22 and @xmath46 .
numerical simulations reported below in sec . [ sec : num ] support the conjecture that the @xmath1-norm does remain bounded in the dynamics of the @xmath0-symmetric manakov system but may grow to infinity in the @xmath0-symmetric non - manakov case .
consider a local solution @xmath47,h^{1}(\mathbb{r } ) \times h^{1}(\mathbb{r } ) ) \ ] ] to the generalized manakov system ( [ manakov ] ) in the energy space , which exists for some @xmath48 due to the standard contraction mapping principle @xcite .
let us recall the conserved quantities for the hamiltonian version of the generalized manakov equations ( [ manakov ] ) with @xmath18 , which are well defined in the energy space : @xmath49 and @xmath50 if @xmath51 , these integral quantities are no longer constant in time .
therefore , we shall first establish the balance equations for the rate of change of these quantities . in what follows , we use notations
@xmath52 for the @xmath53-norm of @xmath54 .
if @xmath3 also depends on @xmath55 , we do not usually write the variable @xmath55 explicitly . from system ( [ manakov ] )
, it is clear that the local solution ( [ loc - sol ] ) satisfies @xmath56,h^{-1}(\mathbb{r } ) \times h^{-1}(\mathbb{r}))$ ] . using the duality of sobolev spaces @xmath30 and @xmath57
, we obtain @xmath58 using gronwall s inequality for ( [ eq : ds0 ] ) , we deduce that @xmath59 does not blow up in a finite time and grows not faster than an exponential function : @xmath60.\ ] ] the balance equation for the energy @xmath61 is not defined if we only work with the local solution ( [ loc - sol ] ) . however , we can approximate @xmath1-solutions of system ( [ manakov ] ) by sequences of local solutions of system ( [ manakov ] ) in @xmath62 , for which the energy balance equation is derived using integration by parts . then
, we take the limits for the sequences and obtain the following balance equation for the local solutions in @xmath30 : @xmath63 we shall now control the homogeneous @xmath30 norm of the solution by using the integral quantity @xmath64 note that @xmath65 is a part of @xmath61 but it does not satisfy a nice balance equation , for which the nonlinear terms cancel out . in other words , the energy balance equation for @xmath65 can not be estimated in terms of @xmath65 only .
for this purpose , we use @xmath61 and , therefore , we shall first establish a correspondence between @xmath65 and @xmath61 .
let us recall the gagliardo nirenberg inequality @xcite .
there exists a positive constant @xmath66 such that for every @xmath67 , we have @xmath68 using the gagliardo nirenberg inequality ( [ gninequality ] ) , we estimate @xmath65 from above : @xmath69 where @xmath70 . recall that both @xmath71 and @xmath72 are considered to be non - negative for simplicity . from inequality ( [ eq : d2 ] ) , we obtain @xmath73 the energy balance equation ( [ energy - balance ] ) can now be estimated again with the use of the gagliardo nirenberg inequality ( [ gninequality ] ) : @xmath74 and using ( [ eq : d2 ] ) , ( [ eq : d ] ) , and the elementary inequality @xmath75 in order to bound @xmath65 , we obtain @xmath76 using that @xmath77 and integrating ( [ eq : e2 ] ) , we obtain the integral inequality @xmath78 notice that because of the bound ( [ bound - l2 ] ) , function @xmath79 is defined for all @xmath80 $ ] and does not grow faster than an exponential function . using gronwall s inequality for ( [ eq : e3 ] ) , we obtain @xmath81 does not blow up in a finite time and grows not faster than an exponential function . the same is true for @xmath65 thanks to the bound ( [ eq : d ] ) . extending the local solution of the generalized manakov system ( [ manakov ] ) in @xmath30 to larger values of @xmath82 by using a priori energy inequalities ( [ bound - l2 ] ) , ( [ eq : d ] ) , and ( [ eq - e4 ] ) and applying the standard continuation arguments
, we conclude the proof of theorem [ theorem-1 ] on the existence of the unique global solution of the generalized manakov system ( [ manakov ] ) in @xmath30 . for the global solution of theorem [ theorem-1 ] , we can introduce the integral stokes variables @xmath83 let us consider the case @xmath22 .
we compute temporal derivatives of the integral stokes variables and obtain @xmath84 where @xmath85 is defined by ( [ power ] ) and satisfies the balance equation ( [ eq : ds0 ] ) .
note that all nonlinear terms cancel out in the system of evolution equations ( [ stokes - system ] ) for the integral stokes variables .
it follows from system ( [ stokes - system ] ) that @xmath86 and @xmath87 are constants of motion . excluding @xmath88 and @xmath89 from system ( [ stokes - system ] ) ,
we obtain the linear oscillator equation for @xmath59 @xmath90 where we have introduced @xmath91 the linear oscillator equation ( [ lin - osc ] ) immediately implies that if @xmath27 , then @xmath59 oscillates with the frequency @xmath92 and therefore , can be globally bounded for all @xmath93 by a constant @xmath94 . indeed , from the linear equation ( [ lin - osc ] ) with the initial conditions @xmath95 and @xmath96 , we compute @xmath97 where integration constants are given by @xmath98 hence @xmath99 , where the sharp upper bound @xmath100 is computed from the initial conditions by @xmath101 this argument proves bound ( [ bound - l2 - 1 ] ) of theorem [ theorem-2 ] .
if @xmath43 , there exists solutions to eq .
( [ lin - osc ] ) with @xmath102 such that @xmath59 grows unboundedly as @xmath24 .
therefore , the @xmath2-norm of global solutions of the @xmath0-symmetric manakov system with @xmath103 grows to infinity , according to bound ( [ bound - l2 - 2 ] ) .
this concludes the proof of theorem [ theorem-2 ] .
note that the linear oscillator equation ( [ lin - osc ] ) is derived in our previous work @xcite for the @xmath45-independent solutions by using the local stokes variables .
furthermore , we note here that for @xmath45-independent solutions , the @xmath25-symmetric manakov dimer ( [ dimerm ] ) can be reduced to the linear system . indeed , considering @xmath104 and setting @xmath105
, we obtain the system of linear equations @xmath106 which is stable for @xmath27 and unstable for @xmath43 .
although these results are rather trivial for @xmath45-independent solutions to the @xmath25-symmetric manakov dimer ( [ dimerm ] ) , we have shown here that these results can be extended to bound the @xmath2-norm of @xmath45-dependent solutions to the @xmath25-symmetric manakov system ( [ manakov ] ) with @xmath22 in the energy space .
the results of theorems [ theorem-1 ] and [ theorem-2 ] do not clarify if the @xmath1-norm of global solutions of the @xmath0-symmetric system ( [ manakov ] ) remains bounded for @xmath27 . in order to obtain additional insight into this issue
, we have undertaken a set of numerical simulations .
our numerical results indicate that the @xmath1-norm of the global solutions does remain bounded for the manakov system ( @xmath22 ) with unbroken @xmath0 symmetry ( @xmath27 ) . on the other hand , in the @xmath0-symmetric non - manakov case with @xmath107 , unbounded solutions exist even if @xmath27 .
the analytical proof of these conjectures remains opened for further studies .
more specifically , we run a set of numerical simulations of the dynamics governed by the generalized manakov system ( [ manakov ] ) which is solved in the finite domain @xmath108 $ ] with zero boundary conditions , @xmath109 , using a semi - implicit difference scheme .
the numerically obtained solution is used to evaluate quantities @xmath65 and @xmath59 which characterize the @xmath1-norm of the global solution . for the initial conditions at @xmath110
, we use gaussian beams @xmath111 truncated on the interval @xmath112 , where @xmath113 , @xmath114 , @xmath115 and @xmath116 are real constants .
notice that the functions @xmath117 and @xmath118 are normalized such that @xmath119 the width @xmath120 of the computational domain is sufficiently large such that the numerical value of the initial conditions ( [ ic ] ) nearly vanish at @xmath121 .
-symmetric system ( [ manakov ] ) .
three rows present results obtained for different initial conditions ( [ ic ] ) : upper , middle and the lower rows correspond to @xmath122 as follows : ( 3 , 1 , 1 , 0.5 ) , ( 3 , 5 , 1 , 2 ) , and ( 3 , 0.2 , 1 , 0.2 ) . panels from the left column show profiles of the initial conditions @xmath123 ( black lines ) and @xmath124 ( red lines ) .
panels from the central column show dependencies @xmath65 ( green lines , labeled as @xmath125 ) and @xmath59 ( blue lines , labeled as @xmath85 ) found for the manakov case @xmath126 .
panels from the right column show the same data , obtained from the same initial conditions as in the middle column , but for the non - manakov case @xmath127 and @xmath128 .
for all panels , @xmath129 and @xmath130 , i.e. , @xmath0 symmetry is unbroken . , scaledwidth=80.0% ] we solved the cauchy problem in the finite domain @xmath108 $ ] for several different sets of initials conditions @xmath122 , as well as for different combinations of @xmath72 and @xmath71 ( with @xmath27 ) and for different @xmath120 . in fig .
[ fig_conj ] we present results of numerical runs for three different initial conditions .
figure [ fig_conj ] addresses the case of the manakov nonlinearity with @xmath131 ( panels in the middle column ) and the non - manakov case with @xmath132 and @xmath133 ( panels in the right column ) . for all sets of the initial conditions ( [ ic ] ) and for
all @xmath120 we have tested , we observed that in the case @xmath134 , the numerically obtained @xmath1 norm remains bounded ( see panels in the middle columns of fig .
[ fig_conj ] ) .
this allows us to conjecture on boundness of the @xmath1-norm of all global solutions to the @xmath0-symmetric manakov system ( [ manakov ] ) with @xmath134 and @xmath27 . on the other hand ,
unbounded numerical solutions exist in the non - manakov case ( see the upper panel in the right column of fig . [ fig_conj ] ) . in all panels showing dynamics of the manakov system ,
we observe the oscillatory behavior of @xmath59 which is characterized by the period ( see ( [ lin - osc ] ) and ( [ omega ] ) ) , @xmath135 since our main interest is related to the existence ( or non - existence ) of globally bounded solutions , in fig .
[ fig_conj ] we show the results obtained for a situation when the input beam in the equation with gain has the intensity much larger than the intensity of the input beam for the nls equation with dissipation ( c.f .
black and red curves in the left columns of fig . [ fig_conj ] ) .
the opposite situation when the intensity of the input beam for the lossy equation is much larger than that for the equation with gain was also tested in our numerics with the same conclusions on boundedness and unboundedness of solutions .
in this concluding section , we generalize the proof of the global existence in the energy space for the manakov system , i.e. for ( [ manakov ] ) with all equal nonlinear coefficients , presented above , to the case of two spatial dimensions , @xmath136 . in this context
we note that the analysis of global existence and blow - up of solutions in a hamiltonian counterpart of this system , it corresponds to @xmath137 , was recently presented in @xcite .
respectively we shall rewrite the system ( [ manakov ] ) in two spatial dimensions with the manakov - type nonlinearity : @xmath138 here @xmath139 is the two - dimensional laplacian and all the nonlinear coefficients @xmath8 , @xmath9 and @xmath10 are set to unity .
we shall now formulate and prove the global existence result for the @xmath0-symmetric manakov system ( [ manakov2d ] ) by generalizing the ideas of theorems [ theorem-1 ] and [ theorem-2 ] .
this generalization is possible due to the fact that the @xmath2-norm of a global solution to system ( [ manakov2d ] ) is bounded for @xmath27 by the constant @xmath100 , which is determined in terms of the initial conditions in @xmath140 by the explicit expression ( [ qmax ] ) .
this global bound allows us to obtain an a priori bound for the energy @xmath141 note here that the local solution in the energy space @xmath142,h^1(\mathbb{r}^2 ) \times h^1(\mathbb{r}^2)),\ ] ] is not derived by the standard contraction mapping principle for the semi - linear equations .
nevertheless , a modified contraction method can be developed with the use of stritcharz estimates as in @xcite .
therefore , we use here the local solution ( [ loc - sol-2 ] ) obtained from this modified contraction method .
next , let us recall the gagliardo - nirenberg inequality in the space of two dimensions @xcite .
there exists a positive constant @xmath66 such that for every @xmath143 , we have @xmath144 moreover , it was shown by weinstein @xcite that @xmath66 can be found sharply in terms of the ground state solution @xmath145 of the stationary equation @xmath146 namely , @xmath147 . in physics ,
the solution @xmath148 is also known as the townes soliton @xcite .
now we can formulate a global existence result for the two - dimensional @xmath0-symmetric manakov system ( [ manakov2d ] ) .
consider the two - dimensional @xmath0-symmetric manakov system ( [ manakov2d ] ) with @xmath27 .
if the initial conditions @xmath149 satisfy the constraint @xmath150 where @xmath100 is defined by ( [ qmax ] ) , then there exists a unique global solution @xmath151 such that @xmath152 and @xmath153 . on the other hand ,
if @xmath154 , there exists initial conditions @xmath149 , such that local solutions ( [ loc - sol-2 ] ) blow up in a finite time , that is , there exists @xmath155 such that @xmath156 [ theorem-2d ] we shall now sketch the proof of theorem [ theorem-2d ] by following the ideas of the proof of theorem [ theorem-1 ] . defining @xmath65 as in ( [ def - d ] ) , we employ the gagliardo - nirenberg inequality ( [ gn2d ] ) and estimate @xmath65 from above as @xmath157 for all @xmath158 $ ] we have @xmath99 .
if the condition ( [ condition2d ] ) is met ( recall that @xmath159 ) , then @xmath65 is controlled by @xmath160 from above as @xmath161 using then the balance equation for @xmath162 as in ( [ energy - balance ] ) , we obtain for all @xmath158 $ ] @xmath163 integrating the latter inequality similarly to what is done in the derivation of ( [ eq : e3 ] ) and ( [ eq - e4 ] ) , we obtain a global ( exponentially growing ) bound on the @xmath1-norm of the local solution .
hence , the local solution ( [ loc - sol-2 ] ) can be extended to larger values of @xmath82 , and finally , globally for all @xmath93 . to show the other part of theorem [ theorem-2d ] concerning the finite - time blow - up ( [ blowup2d ] ) , we note the following trick , which is explored in the recent works @xcite . using the substitution @xmath164 with @xmath37 defined by @xmath165 ,
we reduce the @xmath0-symmetric manakov system ( [ manakov2d ] ) with @xmath27 to a scalar nls equation @xmath166 solutions of equation ( [ nls2d ] ) conserve the @xmath2-norm and may blow up in a finite time if @xmath167 @xcite . the corresponding blowing - up solution of the @xmath0-symmetric manakov system ( [ manakov2d ] ) also conserves the @xmath2 norm , which then satisfies @xmath168 .
this concludes the proof of theorem [ theorem-2d ] .
we note that theorem [ theorem-2d ] does not clarify if solutions to the @xmath0-symmetric manakov system ( [ manakov2d ] ) may blow up in a finite time if the initial data corresponds to @xmath100 fitting in the interval @xmath169 this question remains open for further studies .
the authors thank vadim vekslerchik for helpful discussion at an earlier stage of the project . the work of d.p .
is supported by the ministry of education and science of russian federation ( the base part of the state task no .
2014/133 ) .
daz and vvk acknowledge support of fct ( portugal ) under the grants pest - oe / fis / ui0618/2014 and ptdc / fis - opt/1918/2012 .
y. v. bludov , r. driben , v. v. konotop , and b. a. malomed , instabilities , solitons and rogue waves in coupled nonlinear waveguides , j. opt . , * 15 * , 064010 ( 2013 ) .
n. v. alexeeva , i. v. barashenkov , a. a. sukhorukov , and y. s. kivshar , optical solitons in @xmath0-symmetric nonlinear couplers with gain and loss , phys .
a , * 85 * , 063837 ( 2012 ) . | we study a system of two coupled nonlinear schrdinger equations , where one equation includes gain and the other one includes losses .
strengths of the gain and the loss are equal , i.e. , the resulting system is parity - time ( @xmath0 ) symmetric . the model includes both linear and nonlinear couplings , such that when all nonlinear coefficients are equal , the system represents the @xmath0-generalization of the manakov model . in the one - dimensional case , we prove the existence of a global solution to the cauchy problem in energy space @xmath1 , such that the @xmath1-norm of the global solution may grow in time . in the manakov case ,
we show analytically that the @xmath2-norm of the global solution is bounded for all times and numerically that the @xmath1-norm is also bounded . in the two - dimensional case
, we obtain a constraint on the @xmath2-norm of the initial data that ensures the existence of a global solution in the energy space @xmath1 . |
the last several years have seen significant developments in detection and understanding of energetic emission from pulsating neutron stars .
new surveys and detectors have not only discovered more sources but have discovered new and unexpected characteristics and behavior of these sources .
three types of neutron stars that are the subject of this review , rotation - powered pulsars , anomalous x - ray pulsars ( axps ) and soft gamma - ray repeaters ( sgrs ) , were thought to be separate and unrelated objects as recently as seven years ago .
today , we know that sgrs and axps are both neutron stars possessing magnetic fields with unprecedented strength of @xmath0 g , collectively called magnetars " , that show both steady x - ray pulsations as well as soft @xmath1-ray bursts .
their inferred steady x - ray luminosities are about one hundred times higher than their spin - down lumnosities , requiring a source of power beyond the magnetic dipole spin - down that powers radio pulsars . but
some rotation - powered radio pulsars have recently been discovered that have magnetic fields , as inferred from dipole spin - down , overlapping the range of magnetar fields @xcite .
new high - energy components discovered in the spectra of a number of axps and sgrs require non - thermal particle acceleration and look very similar to high - energy spectral components of young rotation - powered pulsars @xcite .
so these three types of source class , that were seemly unrelated , are now more intimately connected in ways that raise interesting questions about their true nature . at the present time
, we treat the radio pulsars and magnetars as separate classes , although they are both spinning - down neutron stars .
it is believed that all of the observed emission associated with radio pulsars is derived from their rotational energy loss or from their heat of formation .
this includes all pulsed radiation from radio ( representing a very small fraction of their spin - down luminosity ) to @xmath1-ray wavelengths , as well as surrounding emission from the pulsar wind nebula at all wavelengths .
on the other hand , even the average quiescent emission observed from magnetars can not be accounted for by rotational energy loss , requiring another source of power in sgrs and axps , one that is apparently either not available or not being used by the radio pulsars .
their source of power is not the only feature that puts rotation - powered pulsars ( rpps ) and magnetars in a separate class .
most of their observational characteristics are very different , although there are a few similarities .
their range of rotation periods , 1.5 ms - 8 s for radio pulsars and 5 - 11 s for magnetars , overlap somewhat and they are both observed to spin down more or less steadily with occasional glitches .
the measured period derivatives also overlap somewhat .
if plotted together on a @xmath2-@xmath3 diagram , their distributions occupy almost , but not quite , separate areas ( see figure 1 ) .
both have high energy emission and accelerate particles .
but rotation - powered pulsars and magnetars display more prominent differences than similarities .
rpps have never been observed to have @xmath1-ray bursts , as is characteristic of sgrs and also recently of axps .
magnetars have very luminous x - ray emission while the rpps that have magnetar - strength fields have very weak or undetectable x - ray emission @xcite . while all but a few known rpps are radio pulsars ,
no radio pulsations have been detected from any sgr or axp .
the one report of radio pulsations from sgr1900 + 14 @xcite was unconfirmed @xcite and probably confused with emission from the nearby radio pulsar psr j1907 + 0918 .
a @xmath2-@xmath3 plot of known rpps and magnetars with measured @xmath3 is shown in figure 1 .
there are presently about 1500 known radio pulsars , but that is expected to change significantly over the next few years , with several new surveys starting in the northern hemisphere .
about 30 radio pulsars are x - ray pulsars and about 9 of these are also @xmath1-ray pulsars .
the millisecond pulsars , believed to have been spun - up by accretion torques of a binary companion @xcite , have much lower surface magnetic fields in the range of @xmath4 g and occupy the lower left part of the @xmath2-@xmath3 diagram .
there are a handful of rotation - powered pulsars that were discovered at x - ray or @xmath1-ray wavelengths and do not have any detected radio pulsations .
it is not known whether these pulsars are truly radio silent , below sensitivity of radio searches , or have radio beams pointed in a different direction . in the magnetar class , there are 8 known axps and 5 sgrs ( only 3 of which have measured period derivatives ) .
our knowledge of high - energy emission from rpps has increased dramatically over the past ten years .
the compton gamma - ray observatory ( cgro ) discovered five new @xmath1-ray pulsars which , added to the crab and vela @xmath1-ray pulsars that were previously known , brings the number to seven that are known with high significance @xcite .
there are about five others that were detected with less significance , one of these being the ms pulsar psr 0218 + 4232 , which is also an x - ray pulsar .
although the number of @xmath1-ray pulsars is still small , some interesting patterns in their emission characteristics have emerged .
their luminosity ( really their flux times distance squared , since the solid angle is not known ) follows the relation @xcite , @xmath5 , where @xmath6 is the spin - down luminosity .
equivalently , the high energy luminosity is proportional to the voltage across the open field lines , or @xmath7 v and to the goldreich - julian flux , the maximum particle flux emerging over the polar caps .
their spectral index above 100 mev shows a decrease with characteristic age of the pulsar , with photon spectral indices ranging from -2.2 , in the case of the crab , to -1.2 in the case of psr b1055 - 52 .
all @xmath1-ray pulsar spectra show evidence for a high energy turnover , either in the egret band around a few gev or implied sharp breaks or turnovers between 10 and 100 gev from upper limits above 100 gev , from searches by air cherenkov telescopes @xcite .
no pulsed emission has been detected above 20 gev in any pulsar .
the @xmath1-ray light curves generally show two narrow peaks separated by 0.3 - 0.5 in phase , so that the profiles are very broad compared to radio pulse profiles , that usually occupy a small fraction of phase .
but in the case of the two shortest period @xmath1-ray pulsars , the crab and ms psr j0218 + 4232 , the radio profile is as broad as the @xmath1-ray profile and the peaks are correlated in phase . for the other @xmath1-ray pulsars ,
single narrow radio peaks lead the double @xmath1-ray pulses in phase .
this behavior possibly suggests a different relation between radio and high - energy emission sites in short period and longer period pulsars .
a much larger number of pulsars are visible in the x - ray band @xcite , and the population of x - ray pulsars is more diverse , containing many more middle - aged and older sources .
the spectra often show both non - thermal or power law components ( probably magnetospheric in origin ) and one or more thermal components ( originating from cooling of the whole neutron star surface and/or heated polar caps ) .
the non - thermal spectrum is harder in the x - ray band than in the @xmath1-ray band , so that the spectrum breaks in the soft @xmath1-ray band . in the young
, crab - like pulsars the magnetospheric emission significantly dominates over any surface thermal emission . in the middle - aged pulsars multiple thermal components
are often present , a cool blackbody with @xmath8 mk and a hot blackbody with @xmath9 mk . as the neutron star ages , the cooling component decreases rapidly , so that in old pulsars ( including the millisecond pulsars ) the heating component may dominate .
the luminosity in the x - ray band has a different dependence on spin - down luminosity @xcite , @xmath10 , although there is a very large scatter in the correlation .
since there are many more measured pulse profiles of x - ray pulsars , there is a larger data base to compare radio and high - energy profiles .
the trend of greater multiwavelength phase coherence in the profiles of fast pulsars , that we are beginning to see for @xmath1-ray pulsars , seems also true at x - ray wavelengths .
psr b0540 - 69 , a 50 ms pulsar in the lmc , and psr j1617 - 5055 , a 69 ms pulsar in rcw103 , show x - ray and radio pulses that are very similar and in phase .
there is also phase coherence of x - ray and radio pulses in ms pulsars psr j0218 + 4232 , psr b1821 - 24 , psr j0437 - 4715 and psr b1937 + 21 .
the broad - band spectra of pulsars show several trends @xcite .
the emission power peaks in hard x - rays for younger pulsars like the crab and psr b1509 - 58 , so that thermal components are not visible . for older pulsars the power peaks around several gev and
the non - thermal power in x - rays is much smaller so that the thermal components become quite prominent .
quite a few have optical pulsations whose spectra often connect to the non - thermal x - ray components .
the spectra of ms pulsars behave differently from either of these groups , with very hard x - ray spectra ( photon indices of -0.6
-1.2 ) and weak detections or upper limits above 100 kev that imply a power peak around 1 mev @xcite .
although egret significantly detected @xmath1-ray pulsations from only seven pulsars , it detected around 150 point sources that have no identified counterparts @xcite .
many of these lie near the galactic plane and may be pulsars . since the cgro mission ended in 2000 ,
there have been over 700 new pulsars discovered @xcite , and around 50 lie inside egret source error circles @xcite @xcite .
about 20 of these are plausible counterparts both on energetic grounds or from model estimates @xcite .
these sources unfortunately contain too few photons to perform any credible period searches of the egret archival data with present pulsar ephemerides , so identification must await more @xmath1-ray observations .
sgrs and axps were both discovered several decades ago , but they have only recently been recognized as similar objects ( for detailed review , see woods & thompson @xcite ) .
sgrs were first detected around 1979 as @xmath1-ray transients and were thought to be a type of classical @xmath1-ray burst .
they undergo repeated bursts with several tenths of second duration and average energy @xmath11 , and their bursting often occurs in episodes spaced years apart .
they more rarely undergo giant superflares of total energy @xmath12 , consisting of an initial spike of duration several hundred ms followed by a longer pulsating decay phase of duration several hundred seconds .
such superflares have been observed in three sgr sources , sgr0526 - 66 ( the famous 5th march 1979 event ) , sgr1900 + 14 @xcite and very recently in sgr1806 - 20 @xcite . in 1998 , sgr1806 - 20 was discovered to have 7.47 s pulsations in its quiescent x - ray emission @xcite and a large @xmath3 that implies a surface magnetic field of @xmath13 g if due to dipole spin - down .
quiescent periodicities of 8 s and 5.16 s and large @xmath3 were subsequently detected in sgr0526 - 66 and sgr1900 + 14 , implying similarly high surface magnetic fields . in all three sources ,
the quiescent periods are the same periods seen in the decay phases of their superflares .
the quiescent pulse profiles are very broad and undergo dramatic changes just before and after superflares .
the profiles are often more complex , with multiple peaks before flares , changing to more simple single peaks profiles following the flares .
all of the sgrs lie near the galactic plane and are thought to have distances around 10 - 15 kpc ( except for sgr0526 - 66 , which is in the lmc ) .
the first axps were discovered as pulsating x - ray sources in the early 1980s by exosat and were thought to be a strange ( anomalous ) type of accreting x - ray pulsar .
they are bright x - ray sources possessing luminosities ( in their highest states ) of @xmath14 , but show no sign of any companion or accretion disks that would be required to support the accretion hypothesis .
the axps have pulsation periods in a relatively narrow range of 5 - 11 s and are observed to be spinning down with large period derivatives @xcite .
their pulse profiles are broad and very similar to those of sgr sources .
the very high surface magnetic fields of @xmath0 g implied by dipole spin - down were originally controversial , but have come to be accepted after the quiescent periods were found in sgrs and especially following the recent discovery of sgr - like bursts from several axps @xcite .
it is now believed that sgrs and axps are two varieties of the same type of object , very strongly magnetized , isolated neutron stars possibly powered by magnetic field decay . in both sources ,
the high - state quiescent luminosities of @xmath15 are much higher than their spin - down luminosities of @xmath16 , demanding an alternative power source .
periodic monitoring of axps have revealed a significant amount of flux variability in the sources .
the luminosity of several axps has even changed by several orders of magnitude .
1e 1048.1 - 5937 has been observed to undergo pulsed flux flares lasting a few weeks @xcite , and the baseline luminosity between such flaring events is only @xmath17 which is close to the spin - down luminosity . since the sources were discovered in their high states , such extended low states
were not observed until recently .
the quiescent spectra of axps and sgrs consist of a thermal component fit by @xmath18 0.5 - 1 kev blackbodies and one or more non - thermal components . until recently ,
the non - thermal spectra below 10 kev were fit with steep power laws having indices
@xmath19 @xmath20 .
when _ integral _ and _
rxte _ recently measured the spectra above 10 kev for the first time hard , non - thermal components were discovered in three axps , and also sgr 1806 - 20 . in two of the axps ,
the differential spectra between 10 kev and 50 kev are extremely flat : 1e 1841 - 045 @xcite has a power - law index of @xmath21 and 4u 0142 + 61 @xcite displays an index of @xmath22 , both much flatter than the steep non - thermal components in the classic x - ray band .
rxs j1708 - 40 possesses a slightly steeper continuum with @xmath23 .
the non - thermal tail of quiescent emission in sgr 1806 - 20 is similarly pronounced @xcite @xcite , but somewhat steeper , with an index of @xmath24 @xmath25 extending to 100 kev .
such hard non - thermal components require continuous particle acceleration during the quiescent state .
the characteristics of emission from rotation - powered pulsars have been studied in great detail for many years .
even so , the origin of the pulsations is still not understood at any wavelength except to realize that the radio emission must be a coherent process requiring significant particle densities and probably electron - positron pairs @xcite , and the emission at higher energies comes from incoherent processes requiring acceleration of particles to energies of at least 10 tev .
the mechanisms and models for pulsed high - energy emission from rotation - powered pulsars was traditionally based on the physics of particle acceleration , which results from the strong electric fields ( @xmath26 ) that are induced along open magnetic field lines .
two accelerator sites have been studied in most detail and two main types of model have developed , although pulsed emission from the wind zone has also been proposed @xcite .
polar cap models @xcite are based on particle acceleration near the magnetic poles and outer gap models @xcite are based on acceleration in vacuum gaps in the outer magnetosphere , however the emission geometries of both models have significantly evolved in recent years .
polar cap models divide into two types that depend on how the goldreich - julian ( or corotation ) charge density , which would cause force - free conditions , is supplied and distributed .
vacuum gap accelerators @xcite assume that charges are trapped in the neutron star surface so that the full vacuum potential drop can develop above the surface .
such accelerators are unstable to break - down by electron positron pair creation and are therefore non - steady accelerators .
space charge - limited flow ( sclf ) accelerators @xcite assume that charge is freely supplied by the neutron star , so that the charge density at the stellar surface is equal to the corotation charge density .
although @xmath27 at the neutron - star surface in these models , the space charge along open field lines above the surface deviates from the corotation charge , due to the curvature of the field @xcite and to general relativistic inertial frame dragging @xcite , causing a steady @xmath26 that increases with height .
accelerated particles radiate @xmath1-rays that create electron - positron pairs and the @xmath26 is screened above a pair formation front ( pff ) by polarization of the pairs .
the potential drop is thus self - adjusted to give particle lorentz factors around @xmath28 , assuming a dipole magnetic field . above the pff
, force - free conditions could develop if the pair multiplicity is sufficient .
this is likely for relatively young pulsars , which can produce pairs through curvature radiation @xcite , but older pulsars that can produce pairs only through inverse compton emission are expected to be pair starved and their open magnetospheres would not achieve a force - free state @xcite . given the small radiation loss length scales for particles of these energies , the high energy radiation will occur within several stellar radii of the surface .
one thus expects that the radiation spectra will exhibit sharp high - energy cutoffs due to magnetic pair production attenuation at energies in the range 0.01 - 20 gev , depending inversely on field strength .
the radiation from electromagnetic cascades produces a hollow cone of emission around the magnetic pole , with opening angle determined by the polar cap half - angle , @xmath29 , at the radius of emission @xmath30 .
the characteristics of emission from this type of polar cap model @xcite have been successful in reproducing many features of @xmath1-ray pulsars , including the wide double - peaked pulses observed in @xmath1-ray pulsars like the crab , vela and geminga , and the phase - revolved spectra .
however , since the polar cap opening angle is very small ( a few degrees ) , the broad profiles which require beam opening angles of the order of the magnetic inclination angle , can not be produced unless the pulsar is nearly aligned .
more recent versions of the polar cap model @xcite have explored acceleration in the ` slot gap ' , a narrow region bordering the last open field line in which the electric field is unscreened .
near the open field line boundary , the @xmath26 is lower and a larger distance is required for the electrons to accelerate to the lorentz factor needed to radiate photons energetic enough to produce pairs .
the pff thus occurs at higher and higher altitudes and curves upward as the boundary is approached , approaching infinity and becoming asymptotically parallel to the last open field line .
if the @xmath26 is effectively screened above the pff , then a narrow slot surrounded by two conducting walls is formed .
pair cascades therefore do not take place near the neutron star surface in the slot gap , as do the pair cascades along field lines closer to the magnetic pole , but occur on the inner edge of the slot gap at much higher altitudes .
the high energy emission beam is a hollow cone with much larger opening angle than that of the lower altitude cascade emission .
morini @xcite first noted an interesting feature of the geometry of emission in the outer magnetosphere of a rotating dipole . if one assumes that photons are radiated tangent to the magnetic field from the polar cap to the light cylinder , then the relative phase shifts of photons emitted at different radii due to dipole curvature , aberration and time - of - flight nearly cancel on field lines on the trailing edge of the open region .
radiation along such trailing field lines bunches in phase , forming a sharp peak in the profile . on the other hand ,
photons emitted at different radii along leading field lines spreads out in phase .
the effect is most pronounced for large inclination of the magnetic axis to the rotation axis .
a plot of observer angle to the rotation axis versus phase , as shown in figure 2a and 2b , clearly displays the sharp lines of emission , or caustics , along the trailing field lines originating at radii @xmath31 , where @xmath32 is the light cylinder radius .
morini s model for the vela pulsar @xcite considered the caustic emission from only one magnetic pole .
dyks & rudak @xcite explored a purely geometrical model in which they assume emission along the last open field lines of both magnetic poles and discovered that such a two - pole caustic model ( see figure 2 ) can naturally reproduce the main features of @xmath1-ray pulsar profiles .
several different types of pulsar emission models , including the extended slot gap and outer gap , feature caustics .
the extended slot gap model is a consequence of the sclf polar cap models and originates from the fact that the potential in the slot gap is unscreened , so that electrons on field lines which thread the slot gap will continue accelerating to very high altitudes @xcite , maintaining lorentz factors @xmath33 .
the emission pattern exhibits both the hollow cone centered on the magnetic poles , from the low - altitude cascades , as well as caustics from high - altitude emission in the slot gap along the trailing field lines .
thus , the slot gap may provide a physical basis for the two - pole caustic model , although a full radiation model of the slot gap remains to be developed .
outer - gap models @xcite assume that acceleration occurs in vacuum gaps that develop in the outer magnetosphere , along the last open field line above the null charge surfaces , where the goldreich - julian charge density changes sign , and that high - energy emission results from photon - photon pair production - induced cascades .
the pair cascades screen the accelerating electric field and limit the size of the gap both along and across the magnetic field .
the geometry of the outer gaps prevents an observer from seeing gaps associated with both magnetic poles , putting them in the class of one - pole caustic models .
the high energy pulse profiles that are observed to have two widely separated , sharp peaks are formed by caustics , the leading peak originating from overlapping field lines at @xmath34 and the trailing peak originating from the caustic along trailing field lines at @xmath31 .
an example of outer gap profile formation is shown in figure 2c and 2d .
some recent outer gap models point out that in general the gap lower boundary should exist somewhat below the null charge surface , and the location of this boundary will depend on the external currents that flow into or out of the gap @xcite .
if the gap moves close enough to the neutron star , depending on inclination angle , it might be possible for an observer to see gaps from both magnetic poles .
although outer gap and slot gap models can produce similar pulse profiles , the phase plots of their emission show prominent differences .
as displayed in figure 2a , slot gap / two - pole caustic emission fills the entire sky and the emission can be viewed at a wide range of angles .
radiation will be visible at all pulse phases so that pulse profiles like the one shown in figure 2b will include ` off - pulse ' emission .
outer gap emission fills only a fraction of the sky , as shown in figure 2c , since outward - going radiation occurs only above the null charge surface in each hemisphere @xcite .
outer gap emission is therefore restricted to large inclination and viewing angles @xmath35 .
the peaks in outer gap profiles drop sharply at their outer edges and there is no off - pulse emission outside the peaks , as is evident in figure 2d .
magnetars were actually predicted by duncan & thompson ( dt ) @xcite before their existence was observationally verified . in this model , some neutron stars born with ms periods @xmath36 generate huge magnetic fields of @xmath37 by dynamo action soon after their birth .
such high fields have several properties that could distinguish behavior of magnetars from that of neutron stars having lower fields .
magnetar fields can decay on much shorter timescales , t_amb^ 10 ^ 5 yr _
( ) ^ - 2 due to ambipolar diffusion .
diffusion of magnetic flux out of the neutron star core on these timescales provides the power to magnetars in the dt model .
magnetar - strength fields also apply higher stresses to the stellar crust , so that the yield strain can exceed the crustal strength .
this property is responsible for the small sgr and axp bursts in the dt model @xcite .
if a toroidal component of the field @xmath38 g develops in the interior of the star , it can twist the external field @xcite .
such action can cause the superflares if the twisted field lines reconnect . finally , due to the much faster heat transport in very strong magnetic fields ,
there is a greater heat flux through the crust @xcite .
such a property may explain the much hotter surface temperatures of magnetars and the high quiescent x - ray emission .
magnetar fields also induce significant changes in radiative processes ( see @xcite and @xcite for reviews ) . in general , in magnetic fields approaching and exceeding the quantum critical field @xmath39 g , radiative processes such as compton scattering , cyclotron an synchrotron emission and absorption , and pair production and annihilation , classical descriptions are largely inaccurate and qed descriptions must be used .
in addition , new processes become possible in strong fields , such as one - photon pair production and annihilation , vacuum polarization and photon splitting , that can not take place in field - free environments .
these processes , in particular vacuum polarization @xcite , strongly influence the propagation of photons in the neutron star atmosphere and the spectrum of the emergent radiation . according to the dt model @xcite ,
the magnetar superflares result from reconnection of sheared or twisted external field lines , leading to particle acceleration and radiation of hard emission .
the estimated luminosity of such events , e_flare r^3 4 10 ^ 46 erg ( ) ^2 , is similar to observed luminosities of superflares .
the smaller bursts result from cracking of the crust , which is continually overstressed by diffusion of magnetic flux from the neutron star interior .
the shaking of magnetic footpoints then excites alfven waves that accelerate particles .
the energy radiated in such events would be e_sgr 10 ^ 41 erg ( ) ^ - 2 _ ( ) ^2 _ ( ) ^2 , where @xmath40 is the length scale of the displacements , @xmath41 is the crustal field and @xmath42 is the yield strain of the crust .
the quiescent emission in the dt model is powered by magnetic field decay through conduction of heat from the core .
the neutron star crust is heated to a temperature of @xmath43 where @xmath44 is the core temperature , providing a luminosity l_x 6 10 ^ 35 erg s^-1 ( ) ^4.4 .
thompson & beloborodov @xcite have proposed that the hard , non - thermal quiescent component is due to the creation of a strong @xmath26 induced by twisting of field in the closed region , producing synchrotron radiation from electron acceleration at high altitude .
an alternative model for magnetar activity and emission has been discussed by heyl & hernquist @xcite .
the burst and quiescent radiation are a result of shocks from fast - mode plasma waves .
the hard quiescent component is due to a pair - synchrotron cascade @xcite .
baring @xcite proposes that resonant compton upscattering of thermal x - rays by accelerated particles in the open field region produces the quiescent hard emission .
despite all the progress made in understanding the nature and working of rotation - powered pulsars , sgrs and axps over the last decade , there is still a wealth of questions remaining to be answered . among the most vexing
is the unknown fundamental difference that distinguishes rotation - powered pulsars from magnetars . among the most persistent and enduring unresolved issues
is the geometry of pulsar accelerators .
in addition , we would like to know how magnetars accelerate particles and if axps have giant flares like sgrs .
current observations by integral are beginning to uncover clues to magnetar particle acceleration .
it is hoped that swift may find the nature of the short @xmath1-ray bursts and their possible relation to sgr bursts .
future high - energy @xmath1-ray telescopes such as agile and glast will detect many more pulsars , possibly several hundred in the case of glast , many of them radio quiet . with this large increase in statistics , as well as more detailed phase - resolved spectroscopy
, we may be able to map out the geometry ( and possibly even uncover the physics ) of pulsar particle acceleration .
duncan , r. c. 2000 , fifth huntsville gamma - ray burst symposium , eds : r. marc kippen , robert s. mallozzi , gerald j. fishman .
526 ( american institute of physics , melville , new york ) p.830 [ astro - ph/0002442 ] .
kuiper , l. & hermsen , w. 2003 , proc . of 4th agile science workshop `` x - ray and gamma - ray astrophysics of galactic sources '' , ed .
m. tavani , a. pellizzoni & s. vercellone , ( aracne ) p. 47
[ astro - ph/0312204 ] . | i will review the latest developments in understanding the high - energy emission of rotation - powered pulsars and magnetically - powered anomalous x - ray pulsars ( axps ) and soft gamma - ray repeaters ( sgrs ) .
these fields have been extremely active in the last few years , both observationally and theoretically , driven partly by new x - ray data from chandra , xmm - newton and rxte . at the same time , the parkes multibeam survey has discovered over 700 new radio pulsars , some of them young and coincident with egret sources , and others having magnetar - strength magnetic fields .
these new observations are raising important questions about neutron star birth and evolution , as well as the properties of their high - energy emission .
address = nasa goddard space flight center , greenbelt , md 20771 , usa |
ultra high energy cosmic rays ( uhecrs ) are still an open problem in astro - particle physics .
the 11 akeno grand air shower array ( agasa ) events with energy larger than @xmath5 ev @xcite contradict the expected suppression of the uhecr spectrum due to the interaction with the cosmic microwave background ( cmb ) radiation , the graisen - zatsepin - kuzmin ( gzk ) cut - off @xcite . on the other hand the hires data seems to be consistent with the gzk cut - off picture @xcite .
if the uhecr primaries are protons and if they propagate rectilinearly , as the claimed correlation with bl - lacs at energy @xmath6 ev implies , than their sources must be seen in the direction of the highest energies events with energies up to @xmath7 ev detected by hires , fly s eye and agasa @xcite . at these energies
the proton attenuation length is only about @xmath8 mpc and no counterparts in any frequency band was observed in the direction of these uhecr events .
this is a strong indication that cr particles with energies larger than @xmath5 ev may have a different origin from those with lower energies . another important point of this discussion
is related to the low energy part of the cr spectrum , there are , infact , strong evidences that these cr with energies @xmath9 are extragalactic protons , most probably from active galactic nuclei ( agns ) .
this statement is based on some robust experimental and theoretical evidences : ( i ) extensive air shower ( eas ) data confirm protons as primaries , ( ii ) the dip seen with great accuracy in the data of agasa , hires , fly s eye and yakutsk , is a strong signature of the propagation of uhe protons in the extragalactic space , ( iii ) the beginning of the gzk cut - off seen in the spectra of agasa and hires . excluding the correlation of uhecr with bl lacs from the analysis becomes also possible the propagation of protons in very strong magnetic fields .
nevertheless , also in this case , the lack of a nearby source in the direction of the highest energy events ( e.g. at @xmath10 ev ) remains a problem .
in fact , for very strong field strengths @xmath11 ng : the deflection angle , @xmath12 given by the attenuation length @xmath13 and the larmor radius @xmath14 , is small and sources should be seen .
many ideas have been put forward aiming to explain the observed supergzk ( @xmath15 ev ) events : strongly interacting neutrinos and new light hadrons as unabsorbed signal carriers , @xmath16-bursts , lorentz - invariance violation , topological defects ( td ) and superheavy dark matter ( shdm ) ( see @xcite for reviews ) .
the two last models listed above , that represent the most promising top - down models , share a common feature : uhe particles are produced in the decay of superheavy ( sh ) particles or in their annihilation . in the case of td they are unstable and in the case of shdm long - lived particles .
we shall call them collectively @xmath17 particles .
annihilation takes place in the case of monopolonia , necklaces @xcite and shdm particles within some special models . from the point of view of elementary particle physics all these processes proceed in a way similar to @xmath18 annihilation into hadrons : two or more off - mass - shell quarks and gluons are produced and they initiate qcd cascades .
finally the partons are hadronized at the confinement radius .
most of the hadrons in the final state are pions and thus the typical prediction of all these models is the dominance of photons at the highest energies @xmath15 ev . let us now concentrate our attention to the computation of the uhecr spectrum produced in the decay of @xmath17 particles .
the spectrum of hadrons produced in the decay / annihilation of @xmath17 particles is another signature of models with superheavy @xmath17 particles .
the mass of the decaying particle , @xmath3 , or the energy of annihilation @xmath0 , is in the range @xmath19
@xmath1 gev .
the existing qcd mc codes become numerically unstable at much smaller energies , e.g. , at @xmath20 gev . moreover
, the computing time increases rapidly going to larger energies . in this work
we will review our results obtained , in the computation of the top - down spectrum of uhecr , using two different computational techniques : one based on a mc computation scheme @xcite and the other based on the dglap evolution equations @xcite . in both cases susy is included in the computation .
monte carlo simulations are the most physical approach for high energy calculations which allow to incorporate many important physical features as the presence of susy partons in the cascade and coherent branching .
the perturbative part of our mc simulation scheme is similar to other existing mc codes it also includes in a standard way susy and hence is reliable . for the non - perturbative hadronization part an original phenomenological approach is used in ref .
the fragmentation of a parton @xmath21 into an hadron @xmath22 is expressed through perturbative fragmentation function of partons @xmath23 , that represents the probability of fragmentation of a parton @xmath21 into a parton @xmath24 with momentum fraction @xmath25 , convoluted with the hadronization functions @xmath26 at scale @xmath27 , that is understood as the fragmentation function of the parton @xmath21 into the hadron @xmath22 at the hadronization scale @xmath28 gev @xcite . to obtain the fragmentation functions of hadrons
one has : d_i^h(x , m_x)= _ j = q , g_x^1d_i^j(,m_x)f_j^h(z , q_0 ) [ hfunc ] where the hadronization functions do not depend on the scale @xmath3 .
this important property of hadronization functions allows us to calculate @xmath29 from available lep data , @xmath30 at the scale @xmath31 , and then to use it for the calculation of fragmentation functions @xmath30 at any arbitrary scale @xmath3 .
our approach reduces the computing time compared to usual mc simulations and allows a fast calculation of hadron spectra for large @xmath3 up to @xmath32 . the perturbative part of the mc simulation in ref .
@xcite includes standard features such as angular ordering , which provides the coherent branching and the correct sudakov form factors , as well as susy partons .
taking into account susy partons results only in small corrections to the production of hadrons , and therefore a simplified spectrum of susy masses works with good accuracy .
the weak influence of supersymmetry is explained by the decay of susy partons , when the scale of the perturbative cascade reaches the susy scale @xmath33 .
most of the energy of susy partons remains in the cascade in the form of energy of ordinary partons , left after the decay of susy partons .
the qualitatively new effect caused by supersymmetry is the effective production of the lightest supersymmetric particles ( lsp ) , which could be neutralinos or gluinos .
the fragmentation functions @xmath30 at a high scale @xmath3 can be calculated also evolving them from a low scale , e.g. @xmath31 , where they are known experimentally or with great accuracy using the mc scheme .
this evolution is described by the dokshitzer - gribov - lipatov - altarelli - parisi ( dglap ) equation @xcite which can be written as _
t d_i^h=_jp_ij(z)d_j^h(x / z , t ) , [ dglap - eq ] where @xmath34 is the scale , @xmath35 denotes the convolution @xmath36 , and @xmath37 is the splitting function which describes the emission of parton @xmath24 by parton @xmath21 .
apart from the experimentally rather well determined quark fragmentation function @xmath38 , also the gluon fragmentation function @xmath39 is needed for the evolution of eq .
( [ dglap - eq ] ) .
the gluon ff can be taken either from mc simulations or from fits to experimental data , in particular to the longitudinal polarized @xmath18 annihilation cross - section and three - jet events .
the first application of the dglap method for the calculation of hadron spectra from decaying superheavy particles has been made in refs .
the most detailed calculations have been performed by barbot and drees @xcite , where more than 30 different particles were allowed to be cascading and the mass spectrum of the susy particles was taken into account .
although at @xmath40 , which is normally the initial scale in the dglap method , the fragmentation functions for supersymmetric partons are identically zero , they can be calculated at larger scales @xmath41 : susy partons are produced above their mass threshold , when their splitting functions are included in eq .
( [ dglap - eq ] ) . in @xcite
we proved that this method is correct .
also , the lsp spectrum can be computed within the dglap approach @xcite . in this paper
we shall study the agreement of the two methods : mc and dglap equations for the calculation of spectra produced in the decay or annihilation of superheavy particles .
we shall also compare the results obtained by different groups comparing the calculated spectra with recent ones measured by uhecr experiments .
the paper is organized as follows : in section 2 we will compare the two computation schemes described , referring the reader to our paper @xcite for more details .
photon , neutrino and proton spectra , needed for uhecr calculations , are computed in section 3 and compared with the spectra obtained in @xcite . in section 4
we will discus the consequences of our results for models of shdm and td in the explanation of the uhecr spectrum .
we will conclude in section 5 .
in this section we shall compare the hadron spectra computed by the two methods discussed above , mc and dglap . in figure [ fig1 ] we plot the ffs @xmath30 calculated by susy mc and susy dglap methods for @xmath42 gev and @xmath43 . in the dglap method the susy ffs have been evolved from the ones obtained with the susy mc at the scale @xmath44 tev .
one can see the good agreement between dglap ( solid curve ) and mc ( dotted curve ) .
this good agreement holds also for other ( lower ) scales @xmath3 and for other initial partons , e.g. gluon , squark or gluino .
one can see that the mc and dglap spectra slightly differ at very low @xmath45 and have a more pronounced disagreement at large values of @xmath45 .
the discrepancy at low @xmath45 is due to coherent branching , that is included in the mc scheme while it can not be embedded in the dglap one . at large @xmath45 ,
the calculations by both methods suffer from uncertainties , particularly the mc simulation . in this region
the results are sensitive to the details of the hadronization scheme ( see , e.g. , the problem of herwig @xcite with the overproduction of protons at large @xmath45 )
. 1.0 cm when one does not have the initial susy ffs from a mc simulation , the question arises how to proceed . as was first suggested by rubin @xcite ,
the initial ffs can be taken as the ones for ordinary qcd at the low scale @xmath46 , while the production of susy partons is included in the splitting functions assuming threshold behavior at @xmath47 tev .
we can check this assumption computing the susy ff in both ways . in figure [ fig1 ]
we present the susy ffs @xmath30 for @xmath43 and @xmath48 gev , evolved from the initial scale @xmath46 ( dashed curve ) .
the good agreement between the two dglap curves proves the validity of the assumption made .
the spectra of photons , neutrinos and nucleons produced by the decay of superheavy particles are of practical interest in high energy astrophysics . these spectra @xmath49 with @xmath50
can be also considered as ffs .
because the dependence on the type @xmath21 of the primary parton is weak , we shall omit the index @xmath21 from now on , keeping @xmath51 as subscript .
till now we concentrated our discussion on the total number of hadrons ( @xmath52 ) described by the ff @xmath53 , but in fact we have performed similar calculations separately for charged pions and protons+antiprotons .
the procedure of the calculations is identical to that already described for the dglap and mc computation schemes . for charged pions and protons+antiprotons we used experimental data from refs .
below we shall present results of our susy mc simulations in terms of ffs for all pions @xmath54 , all nucleons @xmath55 and all hadrons @xmath56 .
we introduce the ratios @xmath57 and @xmath58 as : @xmath59 and @xmath60 .
1.0 cm the spectra of pions and nucleons at large @xmath3 have approximately the same shape as the hadron spectra , and one can use in this case @xmath61 and @xmath62 @xcite , taking into account the errors in the experimental data @xcite . we can calculate now the spectra of photons and neutrinos produced by the decays of pions neglecting the small contribution ( @xmath63 ) of @xmath64 , @xmath65 , @xmath66 and other particles . including these particles
affects stronger neutrinos than nucleons and photons , which are the main topic of this section .
the normalized photon and neutrino spectrum from the decay of one @xmath17 particle at rest can be computed using the pion and nucleon ff following the recipe given in @xcite .
we shall compare our photon spectra with those calculated by the dglap method in @xcite .
the photon spectrum is most interesting to compare , because it is straightforwardly related to the hadron spectrum which is the basic physical quantity .
moreover , the photon spectrum is the dominant component of radiation produced by superheavy particles .
to be precise , we compare the ff @xmath67 at @xmath42 gev .
figure [ fig2 ] , that is referred to nucleons , demonstrates good agreement between our spectrum and those from @xcite at @xmath68 . as it was mentioned above
, the disagreement at large @xmath45 is not surprising .
apart from @xmath38 taken directly from the experiments , both calculations use the much more uncertain @xmath69 . in our case , @xmath69 is taken from our mc simulation @xcite , in the case of barbot & drees in @xcite from the fit performed in ref . @xcite . in both cases ,
rather large uncertainties exist at large @xmath45 @xcite .
the agreement between the three curves as presented in figure [ fig2 ] is good .
the same conclusion holds for the comparison of photons and neutrino spectra @xcite .
as follows from the previous section , the accuracy of spectrum calculations has reached such a level that one can consider the spectral shape as a signature of the model .
the predicted spectrum is approximately @xmath70 in the region of @xmath45 at interest .
another interesting feature of these new calculations is a decrease of the ratio of photons to nucleons , @xmath71 , in the generation spectrum . at @xmath72
this ratio is characterized by a value of 2 3 only @xcite .
the decrease of the @xmath71 ratio is caused by a decrease of the number of pions in the new calculations .
this result has an important impact for shdm and topological defect models because the fraction of nucleons in the primary radiation increases .
however , in both models photons dominate ( i.e. their fraction becomes @xmath73 ) at @xmath74 ev .
1.0 cm in this section we shall consider : uhecr from superheavy dark matter ( shdm ) @xcite and topological defects ( td ) @xcite .
production of shdm particles naturally occurs in the time - varying gravitational field of the expanding universe at the post - inflationary stage .
the relic density of these particles is mainly determined ( at fixed reheating temperature and inflaton mass ) by their mass @xmath3 .
the range of practical interest is @xmath75 gev , at larger masses the shdm is a subdominant component of the dm .
shdm is accumulated in the galactic halo with the overdensity @xmath76 , where @xmath77 gev/@xmath78 , @xmath79 g/@xmath78 and @xmath80 @xcite . with these numbers , @xmath81 .
because of this large overdensity , uhecrs from shdm have no gzk cutoff .
clumpiness of shdm in the halo can provide the observed small - angle clustering .
the ratio @xmath82 of relic abundance @xmath83 and lifetime @xmath84 of the @xmath17 particle is fixed by the observed uhecr flux as @xmath85 .
the numerical value of @xmath86 is theoretically calculable as soon as a specific particle physics and cosmological model is fixed . in the most interesting case of gravitational production of @xmath17 particles , their present abundance is determined by their mass @xmath3 and the reheating temperature @xmath87 . choosing a specific particle physics model one can fix also the life - time of the @xmath17 particle . there
exist many models in which sh particles can be quasi - stable with lifetime @xmath88 yr .
the measurement of the uhecr flux , and thereby of @xmath86 , selects from the three - dimensional parameter space @xmath89 a two - dimensional subspace compatible with the shdm hypothesis . in figure [ fig3 ]
we have performed a fit to the agasa data using the photon flux from the shdm model and the proton flux from uniformly distributed astrophysical sources .
for the latter we have used the non - evolutionary model .
the photon flux is normalized to provide the best fit to the agasa data at @xmath90 ev .
the fits are shown in figure [ fig3 ] with @xmath91/d.o.f . indicated there .
_ one can see from the fits in figure [ fig3 ] , that the shdm model with the new spectra can explain only the excess of agasa events at @xmath92 ev : depending on the shdm spectrum normalization and the details of the calculations for the extragalactic protons , the flux from shdm decays becomes dominant only above @xmath93 ev . _ _ topological defects _ ( for review see @xcite and reference therein ) can naturally produce uhe particles .
the following td have been discussed as potential sources of uhe particles : superconducting strings , ordinary strings , monopolonium ( bound monopole - antimonopole pair ) , monopolonia ( monopole - antimonopole pairs connected by a string ) , networks of monopoles connected by strings , vortons and necklaces ( see ref .
@xcite for a review and references ) .
monopolonia and vortons are clustering in the galactic halo and their observational signatures for uhecr are identical to shdm . however
the friction of monopolonia in cosmic plasma results in monopolonium lifetime much shorter than the age of the universe . of all other td which are not clustering in the galactic halo , the most favorable for uhecr
are _ necklaces_.
their main phenomenological advantage is a small separation which ensures the arrival of highest energy particles to our galaxy .
we shall calculate here the flux of uhecr from necklaces .
necklaces are hybrid td produced in the symmetry breaking pattern @xmath94 . at the first symmetry breaking monopoles
are produced , at the second one each ( anti- ) monopole get attached to two strings .
this system resembles ordinary cosmic strings with monopoles playing the role of beads .
necklaces exist as the long strings and loops .
the symmetry breaking scales of the two phase transitions , @xmath95 and @xmath96 , are the main parameters of the necklaces .
they determine the monopole mass , @xmath97 , and the mass of the string per unit length @xmath98 .
the evolution of necklaces is governed by the ratio @xmath99 , where @xmath100 is the average separation of a monopole and antimonopole along the string . as it is argued in ref .
@xcite , necklaces evolve towards configuration with @xmath101 .
monopoles and antimonopoles trapped in the necklaces inevitably annihilate in the end , producing heavy higgs and gauge bosons ( @xmath17 particles ) and then hadrons .
the rate of @xmath17 particles production in the universe can be estimated as @xcite @xmath102 , where @xmath41 is the cosmological time .
the photons and electrons from pion decays initiate e - m cascades and the cascade energy density can be calculated as @xmath103 , where @xmath104 is the redshift and @xmath105 is the fraction of the total energy release transferred to the cascade .
the parameters of the necklace model for uhecr are restricted by the egret observations @xcite of the diffuse gamma - ray flux .
this flux is produced by uhe electrons and photons from necklaces due to e - m cascades initiated in collisions with cmb photons . in the range of the egret observations , @xmath106 mev , the predicted spectrum is @xmath107 with @xmath108 @xcite .
the egret observations determined the spectral index as @xmath109 and the energy density of radiation as @xmath110 ev/@xmath78 .
the cascade limit consists in the bound @xmath111 .
according to the recent calculations , the galactic contribution of gamma rays to the egret observations is larger than estimated earlier , and the extragalactic gamma - ray spectrum is not described by a power - law with @xmath112 . in this case , the limit on the cascade radiation with @xmath108 is more restrictive and is given by @xmath113 ; we shall use this limit in further estimates . using @xmath114 with @xmath115 and @xmath116 gyr @xcite we obtain from the limit on the cascade radiation @xmath117 gev@xmath118 .
the important and unique feature of this td is the small separation @xmath65 between necklaces .
it is given by @xmath119 @xcite .
since @xmath120 is limited by e - m cascade radiation we can obtain a lower limit on the separation between necklaces as @xmath121 , this small distance is a unique property of necklaces allowing the unabsorbed arrival of particles with the highest energies .
the fluxes of uhecr from necklaces are shown in figure [ fig4 ] .
we used in the calculations @xmath122
gev@xmath118 which corresponds to @xmath123 ev/@xmath78 , i.e. twice less than allowed by the bound on @xmath114 .
the mass of the @xmath17 particles produced by monopole - antimonopole annihilations is taken as @xmath124 gev .
> from figure [ fig4 ] one can see that the necklace model for uhecr can explain only the highest energy part of the spectrum , with the agasa excess somewhat above the prediction .
this is the direct consequence of the new spectrum of particles in @xmath17 decays obtained in this work .
thus uhe particles from necklaces can serve only as an additional component in the observed uhecr flux .
in this paper we have compared the mc and dglap methods for the calculation of hadron spectra produced by the decay ( or annihilation ) of superheavy @xmath17 particles with masses up to @xmath125 gev .
we found an excellent agreement of these two methods .
the calculations have been performed both for ordinary qcd and susy qcd .
the inclusion of susy partons in the development of the cascade results only in small corrections , and it justifies our computation scheme with a single mass scale @xmath126 @xcite . in comparison to the dglap method , the mc simulation has the advantage of including coherent branching .
it allows reliable calculations at very small @xmath45 .
the gaussian peak , the signature of the qcd spectrum , can not be obtained using the dglap equations .
we have calculated the all - hadron spectra , as well as spectra of charged pions and nucleons , using the dglap equations .
our nucleon spectrum agrees well with that of refs .
we compared also our spectrum of photons with the calculations of ref .
the comparison of the photon spectra is interesting , because of physical reasons ( photons can be observable particles ) , and because the photon spectra are connected directly with the hadron spectra .
the spectra are in good agreement .
we conclude that all calculations are in a good agreement especially in the most interesting low @xmath45 regime and the predicted shape of the generation spectrum ( @xmath70 ) can be considered as a signature of models with decaying ( annihilating ) superheavy particles .
the predicted spectrum of shdm model can not fit the observed uhecr spectrum at @xmath127 ev .
only events at @xmath128 ev , and most notably the agasa excess at these energies , can be explained in this model .
the robust prediction of this model is photon dominance . in present calculations
this excess diminishes to @xmath129 @xcite .
according to the recent calculations the muon content of photon induced eas at @xmath130 ev is high , but lower by a factor 5 10 than in hadronic showers .
the muon content of eas at @xmath130 ev has been recently measured in agasa @xcite .
the measured value is the muon density at the distance 1000 m from the shower core , @xmath131 . from 11 events at @xmath130 ev the muon density
was measured in 6 . in two of them with energies about @xmath132 ev
, @xmath133 is almost twice higher than predicted for gamma - induced eas .
taking into account the contribution of extragalactic protons at this energy , the ratio @xmath134 predicted by the shdm model is 1.2 1.4 .
it is lower than the upper limit @xmath135 obtained by agasa at @xmath136 ev on the basis of a much larger statistics .
the muon content of the remaining 4 eas marginally agrees with that predicted for gamma - induced showers .
the contribution of extragalactic protons for these events is negligible , and the fraction of protons in the total flux can be estimated as @xmath137 .
this fraction gives a considerable contribution to the probability of observing 4 showers with slightly increased muon content .
not excluding the shdm model , the agasa events give no evidence in favor of it . the simultaneous observation of uhecr events in fluorescent light and with water cherenkov detectors has a great potential to distinguish between photon and proton induced eas .
an anisotropy towards the direction of the galactic center is another signature of the shdm model
. both kinds of informations from auger @xcite will be crucial for the shdm model and other top - down scenarios .
topological defect models are another case when short - lived superheavy particle decays can produce uhecr . in figure [ fig4 ] the spectra from necklaces are presented .
one can see that at @xmath92 ev photons dominate , and the discussion in the previous paragraph applies here too .
in contrast to previous calculations , the agreement with observations is worse : necklaces can explain only the highest energy part of the spectrum in figure [ fig4 ] , with the agasa excess somewhat above the prediction .
in the other energy ranges , uhe particles from necklaces can provide only a subdominant component .
other tds suffer even more problems .
m. takeda _
[ agasa collaboration ] , astro - ph/0209422 .
n. hayashida _
[ agasa collaboration ] , phys .
lett . * 73 * , 3491 ( 1994 ) .
k. shinozaki _
[ agasa collaboration ] , astrophys .
j. * 571 * , l 117 ( 2002 ) . t. abu - zayyad _ et al . _ [ hires collaboration ] , astro - ph/0208243 .
d.j . bird _
[ fly s eye collaboration ] , ap.j .
* 424 * , 491 ( 1994 ) .
j. blmer _
et al . _ [ auger collaboration ] , j. phys .
* g29 * 867 ( 2003 ) . n.a .
rubin , thesis , cavendish laboratory , university of cambridge ( 1999 ) . s. sarkar and r. toldr , nucl . phys .
* b621 * , 495 ( 2002 ) . z. fodor and s.d .
katz , phys .
lett . * 86 * , 3224 ( 2001 ) .
c. barbot and m. drees , astropart . phys .
* 20 * , 5 ( 2003 ) .
[ topaz collaboration ] , phys .
b345 * , 335 ( 1995 ) .
d. buskulic _
[ aleph collaboration ] , z. phys .
* c66 * , 355 ( 1995 ) ; p. abreu _ et al . _
[ delphi collaboration ] , _ ibid _ * c73 * , 11 ( 1996 ) ; g. alexander _
[ opal collaboration ] , _ ibid _ * 72 * , 191 ( 1996 ) . | in this work we present a detailed computation of the spectra of uhecr in the top - down scenario .
we compare the spectra of hadrons obtained by two different methods in qcd and supersymmetric ( susy ) qcd with large primary energies @xmath0 up to @xmath1 gev .
the two methods discussed are a monte carlo ( mc ) simulation and the evolution of the hadron fragmentation functions as described by the dokshitzer - gribov - lipatov - altarelli - parisi ( dglap ) equations .
the hadron spectra obtained by the two methods agree fairly well in the interesting energy range @xmath2 ( @xmath3 is the energy scale of the process @xmath4 gev ) .
we have also computed the spectra of photons , neutrinos and nucleons obtaining a good agreement with other published results .
the consistency of the spectra computed by different methods allows us to consider the spectral shape as a signature of the production model for uhecr , such as the decay of super heavy relic particles or topological defects . |
the neutrinos at the main injector ( numi ) beamline at the fermi national accelerator laboratory @xcite will generate an intense @xmath2 beam from the decays of mesons produced in the collisions of 120 gev protons on a graphite target .
the mesons are focused into a 675 m long evacuated volume where they decay to neutrinos .
the meson decays @xmath3 produce an energetic muon for every neutrino , allowing monitoring and validation of the neutrino beam focusing to be accomplished by monitoring of the muon flux . as in several previous experiments ,
the muon flux and remnant hadron flux at the end of the decay volume will be measured by arrays of ionization chambers @xcite .
in addition to charged particle fluxes , the ionization chambers are exposed to large fluences of neutrons , primarily produced in the aluminum / steel beam absorber upstream of the first muon monitor ( see figure [ fig : alcove_layout ] ) .
the neutron fluxes , as estimated by a geant@xcite simulation of the beamline , are as great as 13@xmath410@xmath5/@xmath6/spill at the hadron monitor and 9@xmath410@xmath7/@xmath6/spill in the first muon station .
the neutron fluences are a factor of 10 larger than the charged particle levels the chambers should measure .
most of the neutrons are low energy , with a mean energy of 30 mev .
the response of the ion chambers to a charged particle flux has been studied with electron and proton beams @xcite . in this article , we study the response of ion chambers to neutrons .
previous neutrino beamlines have also utilized ionization chamber arrays to measure the muon beam profile and rate .
because of background particles produced in the beam dump , previous experiments have reported having to contend with higher - than - anticipated fluxes @xcite , or even to disregard the rate information from the muon detectors altogether @xcite . unlike charged particles , neutrons do not ionize the gas directly . instead
, fast neutrons interact with nuclei and create a charged recoil nucleus .
this nucleus very quickly deposits its energy through ionization .
the recoil particle can be produced either in the chamber gas or in the plates of the chamber electrodes .
we studied the response of four parallel plate ion chambers to neutron fluxes at the university of texas nuclear engineering teaching laboratory .
the neutrons for this test are provided by pube radionuclide sources , which produce 1 - 10 mev neutrons via the capture on be of alpha particles emitted by the pu @xcite .
the total source activity is approximately 55 ci , leading to approximately 1.5@xmath410@xmath1 neutrons / sec .
the objective of this test is to determine the ionization due to neutron flux , an effective measurement of the cross section for neutron scattering times the ionization left by the recoiling particles .
the four parallel plate ion chambers are housed in two separate aluminum gas vessels ( see figure [ fig : setup ] ) .
one vessel contains two chambers with 2 mm gas electrode separation , the other has two chambers with 5 mm electrode separation .
electrical feedthroughs to the ion chambers are made using peek insulating plastic and aluminum compression fittings .
a grounded aluminum mesh inside the vessel separates the two chambers .
the parallel plate ion chambers are fabricated from 1 mm thick 100@xmath4100 mm@xmath8 ceramic plates with electrodes made of pt - ag alloy , 10@xmath9 m in thickness .
the sense plane has a central 80@xmath480 mm@xmath8 collection electrode with a 10 mm wide guard ring .
the chamber materials are identical to that to be used in the numi beam monitoring chambers .
one ion chamber in each vessel has a 1 @xmath9ci am@xmath10 @xmath11 source mounted such that the @xmath11 particles enter its gas volume .
the alpha sources permit calibration of temporal effects such as barometric pressure variations .
each vessel also has a signal feedthrough with no internal connection to an ion chamber .
these `` dummy '' feedthroughs serve to check for spurious signals .
signals and high voltage are delivered to the setup through a shielding wall via coaxial cable with kapton insulator .
the exterior of all feedthroughs is potted with epoxy , then wrapped in aluminum foil to provide a complete grounded shield .
this exterior shielding reduces stray electric fields from the high voltage feedthroughs to the signal feedthroughs which can drift ionized charges in the surrounding air to the signal lines .
the chambers are tested under helium or argon gas flow at 1000 mbar .
the gas used in this test is rated as 99.998% pure and is passed through an oxygen getter before going to the chambers which are connected in series to the gas delivery .
the exhaust gas is monitored with an oxygen analyzer , showing 0.5 - 1.5 ppm of o@xmath12 impurity level during the tests .
the ionization current of each chamber is read by a picoammeter .
the nominal sensitivity of the digital display of a picoammeter is 1 pa , however we found that by using the voltage output of the current - to - voltage amplifier we could achieve a sensitivity of better than 0.1 pa for steady currents .
the reason for the 1 pa rating of the picoammeter appears to be drift in the internal amplifier circuit which occurred sporadically during the tests . with frequent monitoring of the signals
, we could discard data runs in which non - physical drifts are observed .
the picoammeter output is digitized by an adc ( least significant bit is equal to 0.06 pa ) and read into a computer .
the neutrons for this test are obtained from six individual pube sources .
four of the sources are @xmath13pu totalling 8.99 curies .
the two others provide 46 ci of @xmath14pu . the neutron yield from the @xmath13pu
is measured to be 57 per 10@xmath15 alpha decays , but calculations suggest 65 per 10@xmath15 @xmath11 decays @xcite . given that yields may vary for specific alloy samples @xcite , we take @xmath16 neutrons/@xmath11 . the theoretical neutron yield for @xmath14pu is @xmath17 @xcite .
we estimate all sources combined to emit @xmath18 neutrons / sec . during the test
the neutron sources are moved to several locations around the gas vessels .
the neutron flux through each chamber in these runs is estimated using the source activities and the solid angle subtended by the chambers . in performing the solid angle integral over the ion chamber
faces , each pube source is assumed to be point - like .
the neutron flux is quoted in units of pa , where 1pa=6.2@xmath410@xmath15 neutrons / sec .
the fluxes through each chamber for each run are tabulated in table [ tab : runs ] . in table
[ tab : runs ] is calculated as well a cosine - weighted flux .
a neutron traversing the chamber at normal incidence receives a weight 1.0 , while a neutron entering at 45@xmath19 to the normal receives an extra weight of @xmath20 .
this cosine factor weights for the pathlength of the neutron passing through the ion chamber gas gap at non - normal incidence , which increases its probability of interaction .
the error in the fluxes due to the estimation method is about 10% for the smaller fluxes and
20% for the largest , which are very sensitive to the source positions .
the pube source also emits gamma rays , which form a background for the measurements of this paper . in approximately 57% of @xmath5be@xmath21c
captures which lead to an emitted neutron from the pube source , the @xmath22c@xmath23 excited state emits a 4.4 mev gamma @xcite .
these @xmath24 s may compton scatter in our chamber materials , yielding additional ionization from the compton electron .
this background is discussed further in section [ neutint ] .
in each of the runs , the bias voltage for each chamber is ramped from zero volts until gas amplification is observed .
the plateau ionization currents from runs 0 and 10 , in which no neutron sources are present , are subtracted from the plateau ionization currents observed in subsequent runs to determine the neutron - induced ionization in the gas .
as expected , only the chambers with an alpha source show an appreciable signal during runs 0 and 10 .
all chambers show an additional signal when significant neutron fluxes are present . because of voltage offsets in the the current - to - voltage amplifier in the picoammeters , the definition of zero ionization current is determined by measuring the chamber plateau curve with both positive and negative bias voltages and assuming the plateau values to be symmetrical about zero signal .
this definition also calibrates away the effect of any contact potentials in the chambers , which cause ionization current to be collected even with zero applied bias voltage .
figure [ fig : platsar ] , for example , shows the voltage plateau curves for the four chambers during different runs taken with argon gas flow . while the 2 mm chambers are fairly symmetric about zero ionization current , the 5 mm chambers show either some effect of a contact potential or an offset in the picoammeters .
figures [ fig : platsar ] and [ fig : plats ] show the plateau curves for the runs taken with argon and helium gas flow , respectively . in the helium data ,
the negative chamber bias data is not always available .
therefore , only the positive voltage portion of the he runs is plotted , where the voltage plateaus have been scaled by the ratio of positive and negative ionization currents observed in other runs .
the inavailability of the positive and negative polarities in some helium runs leads to somewhat larger uncertainties in the ionization currents on plateau . to deduce the ionization current due to the pube source
, we subtract from the ionization current of a given run the ionization current observed in run 0 ( he data ) or run 10 ( ar data ) .
the chambers without alpha sources are therefore more sensitive to small pube signals .
the ionization current attributed to pube - induced interactions is plotted as a function of neutron flux in figures [ fig : scatter4 ] and [ fig : scatter2 ] for the argon and helium data , respectively .
we have fit the data in figures [ fig : scatter4 ] and [ fig : scatter2 ] to find the average pube source - induced ionization per neutron .
the fits constrained to go through the origin gives slope values indicated in table [ tab : results ] .
allowing the vertical intercept to float in the fit changes the slope in the helium fits , for example , by only 0.03 ionizations / neutron , and the fitted vertical intercepts are consistent with zero ( [email protected] and [email protected] ionizations for the 5 mm and 2 mm chambers , respectively ) .
as stated above , some of the observed ionization per neutron in figures [ fig : scatter4 ] and [ fig : scatter2 ] and in table [ tab : results ] is due to the accompanying 4.4 mev @xmath24 s from the pube source . in this section
we estimate the ionization rate induced by the pube source , accounting separately for ( 1 ) the ionization from recoil gas molecules scattered elastically by the pube neutrons ; and ( 2 ) the ionization from compton electrons ejected by the 4.4 mev @xmath24 s .
`` wall effects , '' whereby @xmath24 rays are produced by inelastic interactions of neutrons in the chamber wall materials @xcite , are not considered .
similar estimates of neutron scattering in gaseous chambers have been used to estimate rates in drift chambers downstream of a beam dump @xcite .
the results of the calculations are summarized in table [ tab : expectations ] .
the estimate of the ionization expected from neutron recoils in the gas first requires the measured total cross section for neutron scattering on helium or argon in the energy range 2 - 40 mev @xcite , which are repeated in table [ tab : recoil ] .
these cross sections indicate an interaction length @xmath26 of 1 - 2@xmath27 cm in the chamber gas .
the scattering is predominantly elastic below 10 mev @xcite .
the mean energy of an elastically - scattered nucleus is @xmath28 , where @xmath29 is the recoil nucleus energy , @xmath30 the incident neutron energy , and @xmath31 the target nucleus mass .
these recoil energies , along with the estimated ranges of these recoils in the chamber materials , are listed in table [ tab : recoil ] .
the ranges are extrapolated from measured ranges of alpha particles @xcite using the scaling relation @xcite @xmath32 , where @xmath33 and @xmath34 are the mass and charge of the recoiling nucleus , @xmath35 is the range of an alpha at the same velocity as the nuclear recoil .
the ranges of recoil al , o , pt , or ag nuclei from the walls are insufficient to traverse the chamber wall materials and reach the chamber gas volume , with the exception of pt ions liberated very near the electrode surface
. even the pt ions will contribute insignificant ionization in the chamber gas , since the pt ions energies are quite low .
in fact , the utility of pt electrodes for reducing `` wall - effects '' in neutron interactions has been noted previously @xcite .
the ionization of the recoils created in the gas is then calculated .
because of the short ( 1.2 mm ) range of recoil ar ions in ar gas , we estimate their ionization as the recoil energy divided by the energy @xmath36 ev to create an ionization in ar gas @xcite , yielding an average of 8900 ionizations . for helium recoils , we take @xmath37 kev / cm @xcite for @xmath381.6 mev alpha s , divided by @xmath39 ev to create an ion pair @xcite , yielding 11,900 ionizations for a chamber with 1 cm electrode spacing . to estimate the average ionization per neutron , we multiply the probability for neutron scattering in the gas by the ionization deposited per recoil .
the results are given in table [ tab : expectations ] . to estimate the ionization caused by @xmath24 rays emanating from the pube source
, we performed a parametric monte carlo calculation of @xmath24 s compton scattering in the chamber wall materials , followed by energy loss and multiple scattering by the compton - scattered electrons .
the ionization induced in the chamber gas volume is calculated for those electrons which arrive there .
compton scattering of the 4.4 mev @xmath24 s occurs with approximately 1 barn cross section @xcite .
this yields a probability of scattering in the aluminum chamber walls of 25% and a probability of scattering in the ceramic plates of 11% .
including the fact that only 57% of neutrons are accompanied by a 4.4 mev @xmath24 , this yields 19% of neutron events producing a compton scatter in the chamber walls .
after the compton scatter , the electron energy is calculated .
its energy loss is calculated using tabulated stopping power @xmath40 data @xcite .
multiple scattering of the electron in the aluminum wall and in the ceramic is estimated as in @xcite .
if the path through each material ( aluminum or ceramic ) is greater than the electron s range , the electron does not reach the chamber gas volume .
approximately 74% of the compton electrons are either ranged out in the chamber walls or are multiple scattered away from the gas volume .
the @xmath40 of the electrons arriving in the chamber gas volume is multiplied by the path length of the electrons through the gas volume .
the electron pathlength through the gas can be larger than the electrode separation due to their large multiple scattering .
the range of the electrons is @xmath41 cm in he @xcite , so no electrons range out in the gas . for the he gas , we find a mean of 3.7 ( 8.8 ) ion pairs created in the 2 mm ( 5 mm ) chamber .
the results after multiplying by the probability [email protected] to have a compton electron , are shown in table [ tab : expectations ] .
we have studied the signals induced in ionization chambers exposed to low energy ( 1 - 10 mev ) neutrons in the form of a pube source .
an ionization current from the source is observed whose magnitude is ( [email protected]@xmath250.1 ) ionizations / neutron / cm in the helium gas , and ( [email protected]@xmath251.0 ) ionizations / neutron / cm in the argon gas , based on our observations at several neutron fluences and two chamber gas gaps .
the first uncertainty is that due to the estimate of the neutron fluxes , while the last uncertainty is due to the pube source activity into neutrons .
our measurements of the pube source - induced ionization are limited in precision because of the large number of @xmath24 rays from the pube sources .
however , the neutron - specific signal inferred is consistent with that expected from pure elastic scattering of the neutrons in the chamber gas : 0.35 ion pairs per neutron per centimeter in he , and 0.75 ion pairs per neutron per centimeter in argon .
because of the problem of backgrounds from the 4.4 mev @xmath24 s from the pube source , a study of the individual pulse heights for each particle interaction in our chambers is desirable ; such will be the subject of further study .
our measurements indicate the numi beam monitors will have a background signal of order 20 - 30% from neutron interactions in the chambers . while the yield of ion pairs per neutron is only 2 - 3% for that of charged particles ( assuming elastic scattering ) ,
geant@xcite - based monte carlo estimates indicate that the hadron monitor and the first muon monitor will be exposed to the order of 10 times as many neutrons as signal particles ( protons in the case of the hadron monitor and muons in the case of the muon monitors ) .
the neutron background signal will be much less correllated spatially with beam misalignments than the signal , so its effect will be important to model .
we thank the staff of the university of texas nuclear engineering department for providing the neutron source facility , and especially sean okelly and bill kitchen , for their valuable assistance .
we thank evan fuller and james hall for assistance with the ionization chamber readout .
this work was supported by the u.s .
department of energy , de - ac02 - 76ch3000 de - fg03 - 93er40757 , and de - fg02 - 05er40896 , and the fondren family foundation .
h. wachsmuth , `` neutrino beam monitoring , '' proceedings of the symposium on beam intensity measurement , daresbury , england , dnpl / ri ( 1968 ) , pp.68 - 74 ; r. bertolotto _ et al .
6th intl . conf .
high energy accel .
, c.e.a . , 11 - 15 sept .
1967 ( ceal-2000 - 1967 ) .
a. a. belkov _ et al .
_ , `` calibration of muon detectors for ihep neutrino beam monitoring system , '' serpukhov ihep preprint ifve-82 - 99 , apr .
1982 , proceedings of the eigth international symposium on nuclear electronics , dubna , 1975 , pp 392 - 395 .
p. auchincloss , _ a measurement of the total cross section for neutrino nucleon interactions _ , phd thesis , columbia university , 1987 .
g. f. knoll , _ radiation detection and measurement _ , ( new york : john wiley & sons , inc . ) , 2000 .
.runs taken during the study of neutron interactions .
the gas in the chambers is noted , as well as the expected neutron flux on each of the chambers during each run .
the neutron fluxes are quoted in units of picoamperes , where 1 pa = 6.2@xmath410@xmath15 neutrons / sec .
both the unweighted and cosine - weighted fluxes ( in parentheses ) are listed .
[ tab : runs ] [ cols="^,^,^,^,^,^,^ " , ] | we present studies of helium- and argon - filled ionization chambers exposed to intense neutron fluxes from pube neutron sources ( @xmath0 mev ) .
the sources emit about 10@xmath1 neutrons per second . the number of ion pairs in the chamber gas volume per incident neutron is derived . while limited in precision because of a large gamma ray background from the pube sources , our results are consistent with the expectation that the neutrons interact purely elastically in the chamber gas .
5 cm 1.0 cm -.25 cm ionization chambers , neutron scattering , elastic scattering , electrical phenomena in gases 25.40 , 29.40 , 51.50 |
single - molecule electronics has shown significant progress in the recent years . a variety of interesting effects have been observed in the transport properties of single ( or a few ) conjugated organic molecules including rectification , negative differential resistance , and switching @xcite . in these quasi - one dimensional systems , which present well delocalised @xmath0-electrons
, the electronic current flowing through the quite flexible backbone of the molecule affects the ground state properties of both electronic and mechanical degrees of freedom of the molecule .
the importance of inelastic effects in the transport properties has been demonstrated in several ground - breaking experiments @xcite ; these effects are related to the interaction between electron and mechanical degrees of freedom of the molecule .
the interaction between an injected charge carrier ( electron or hole ) and the mechanical degrees of freedom ( phonon , vibron ) in the molecular junctions is important in order to understand energy transfer , heating and dissipation in the nanojunction @xcite .
the electron - vibron interaction is also at the heart of inelastic electron tunneling spectroscopy ( iets ) .
iets is a solid - state - based spectroscopy which gives information about the vibration modes ( vibrons ) of the molecules in the nanojunction @xcite .
it is now possible to measure such vibrational spectra for single molecules by using scanning tunneling microscopy ( stm ) @xcite to build iets maps @xcite , or by using other electromigrated junctions or mechanically controlled break junctions @xcite . at low applied bias ( typically 100 - 400 mev ) the iets show features ( peaks , dips , or peak - dip - like lineshape ) which have been attributed to selective excitation of specific vibration modes of the molecule .
the position in energy ( bias ) of the features correspond approximately to the frequency of the vibration , as given by other spectroscopic data ( ir , raman ) obtained on the same molecules in a different environment .
there have been many theoretical investigations focusing on the effects of electron - vibron coupling in molecular and atomic scale wires @xcite .
most of them have focused on the interpretation of the features in iets .
however , most of these studies have been performed by using the lowest - order expansion possible for treating the effects of the electron - vibron interaction ( i.e using the so - called self - consistent born approximation scba ) . in the language of many - body perturbation theory
, it corresponds to a self - consistent hartree - fock calculation for the electron - vibron coupling .
however , in analogy to what is obtained at the hartree - fock level for interacting electrons , there are good reasons to believe that this approximation is not enough to correctly describe the physics of the electron - vibron interacting system , especially beyond the weak electron - vibron coupling regime .
for example , the limits of scba have already been investigated in ref .
[ ] but without introducing remedies to go beyond scba . in this paper ,
we examine this using a true non - equilibrium greens - function ( negf ) technique @xcite which allows us to study all the different transport regimes in the presence of electron - vibron interaction . following the spirit of many - body perturbation theory and feynman diagrammatics , we go beyond the commonly - used scba approximation by introducing higher - order diagrams for the electron - vibron interaction
. we study the simplest possible model system which nonetheless contains the relevant physics of the transport properties of the molecular junction @xcite . furthermore , because of the uncertainty of the exact geometry of the single - molecule junction in the experiments , there is a corresponding uncertainty about how to model the coupling between the molecule and the electrodes and correspondly for the potential drops at each molecule - electrode contacts .
hence we take the quantitites characterizing the potential drops at the contacts as phenomenological parameters @xcite .
we concentrate in this paper on the electronic properties of the molecular junction in both equilibrium and non - equilibrium conditions as well as on the linear - response properties of the junction ( prior to considering the full non - equilibrium transport properties in a forthcoming paper ) .
such properties are given by the density of electronic states and represented by the spectral functions , which are at the very heart of all physical properties of the system , such as the charge density , the current density , the total energy , etc .
spectral functions are most closely related to photoemission and adsorption spectroscopies . to our knowledge
such experiments have not yet been performed on single - molecule junctions , though photoemission spectra have been measured on quasi one - dimensional supported atomic scale metallic wires ( see for example ref .
[ ] showing interesting results on one - dimensional collective electronic excitations ) .
the paper is structured as follows .
we start with a description of our model system in section [ sec : hamiltonian ] and a discussion of the relevant underlying theory of non - equilibrium green s functions in section [ sec : non - equil - greens ] .
our calculated spectral functions are presented in section [ sec : res : spec_fnc ] , where we consider first the equilibrium case ( section [ sec : at - equilibrium ] ) and then the non - equilibrium case ( section [ sec : noneq - specfns ] ) at the hartree - fock level .
we discuss especially the effects of including or not the hartree diagram in the calculation .
we also compare negf calculations with results obtained from inelastic scattering techniques @xcite for equivalent model systems in section [ sec : comp - with - inelastic - scattering - technique ] .
we show that it is indeed necessary to go beyond scba to obtain correct results for the relevant range of electron - vibron coupling .
we then present the effects of the second - order diagrams in the spectral functions in section [ sec : vert - corr - spectr ] .
the second - order diagrams correspond to two classes of process ; the first is related to vertex corrections of the scba calculation and the second to polarisation effects ( i.e. partial dressing of the vibron propagator by the electron - hole bubble diagram ) . in section
[ sec : linear_reponse ] we discuss the effects of different levels of approximation for the electron - vibron coupling on the linear conductance of the molecular junctions . throughout the paper
, we will use the term vibron to define a quantum of vibration of a mechanical degree of freedom .
our model is based on a system with an interacting central region connected to two non - interacting electrodes ( see figure [ fig : leadsmolecule ] ) .
is broadened by the coupling to the leads ; the electronic transport is resonant if the fermi levels @xmath1 are very close to this broadened level , and off - resonant ( as shown ) otherwise .
the system is shown in equilibrium , with @xmath2 , non - equilibrium effects are studied by moving the fermi levels .
, title="fig : " ] + the total hamiltonian for the system is given by @xmath3 where @xmath4 and @xmath5 represent the electronic , vibron , and electron - vibron coupling parts of the hamiltonian respectively .
the electronic part of the hamiltonian is broken into sections describing the left ( right ) electrode @xmath6 ( @xmath7 ) , the central interacting part @xmath8 and the potentials coupling the central region to the left and right electrodes @xmath9 respectively : @xmath10 the hamiltonians for the electrodes are given by @xmath11 where @xmath12 ( @xmath13 ) creates ( annihilates ) a _ non - interacting _ electron with energy @xmath14 on electrode @xmath15 .
the electronic hamiltonian for the central region and the coupling potentials are given by @xmath16 where the interacting electrons in the central region are created ( annihilated ) on electronic level @xmath17 by the operators @xmath18 ( @xmath19 ) . in principle , our hamiltonian for the central region may contain electron - electron interactions , and is built from a complete set of single - electron creation and annihilation operators . in our current work
we do not include any electron - electron interactions , and so the electronic part of the total hamiltonian for the central region @xmath20 becomes @xmath21 meanwhile , the vibron part of the hamiltonian is represented by @xmath22 where @xmath23 creates ( annihilates ) a vibron in vibron mode @xmath24 with frequency @xmath25 .
the electron - vibron coupling term is taken to be linear in the vibron displacement and its most general expression is then given by @xcite @xmath26 where @xmath27 is the coupling constant for exciting the vibron mode @xmath24 by electronic transition between the electronic levels @xmath17 and @xmath28 .
we concentrate on the simplest version of the hamiltonian of the central part : the single - site single - mode sssm model , in which one considers just one electron level coupled to one vibration mode . despite the simplicity of this model , remarkably it not only contains all the physics we require but also ensures that we isolate the properties we are interested in without the complication of the added electronic levels .
the reasons for this are as follows .
firstly , when the fermi energy of the leads is pinned around the midgap , then at low and intermediate biases one of the frontier orbitals ( either homo or lumo ) dominates the transport properties . secondly in conjugated organic molecules ( mostly used in single - molecule junction experiments )
, it is known that the optically relevant vibration modes are mostly coupled to either the homo or lumo levels @xcite .
we can therefore be confident that our model contains the relevant physics .
the total hamiltonian for the central region thus becomes @xmath29 where we now have just one electronic level @xmath30 and just one vibron mode @xmath31 , coupled via the electron - vibron coupling constant @xmath32 .
the expression for the lead - central - region coupling ( equation ( [ eq : couplingpot ] ) ) also simplifies to become @xmath33 where we have replaced the coupling potentials @xmath34 with hopping integrals @xmath35 .
non - equilibrium ( negf ) within the keldysh formalism @xcite represent an extremely useful tool for studying the non - equilibrium properties of many - particle systems .
the are functions of two space - time coordinates , and are obviously more complicated than the one - particle density which is the main ingredient of density - functional - based theories .
one of the great advantage of negf techniques is that one can improve the calculations in a systematic way by taking into account specific physical processes ( represented by feynman diagrams ) which is what we do in this paper for the electron - vibron interaction .
the provide us directly with all expectation values of one - body operators ( such as the density and the current ) , and also the total energy , the response functions , spectral functions , etc .
in appendix [ app : higherorderdiagrams ] , we provide more details about negf and how to obtained the electron - vibron self - energies from a feynman diagrammatic expansion of the electron - vibron interaction .
we now briefly describe how we apply the negf formalism to the sssm model .
the green s functions are calculated via dyson - like equations .
for the retarded and advanced green s functions @xmath36 these are @xmath37 where @xmath38 is the non - interacting green s function for the isolated central region . for the greater @xmath39 and lesser @xmath40 green s functions , we use a quantum kinetic equation of the form @xmath41 here @xmath42 is a total self - energy consisting of a sum of the self - energies from the constituent parts of the system : @xmath43 @xmath44 are the self - energies arising from the non - interacting leads @xmath45 and as such are simple to calculate : @xmath46 \ ( 1 - f_\alpha(\omega ) ) , \\
\sigma^<_\alpha = -2i\ \im m[\sigma^r_\alpha(\omega ) ] \ f_\alpha(\omega),\end{gathered}\ ] ] where @xmath47 is the fermi - dirac distribution for lead @xmath15 , with fermi level @xmath48 and temperature @xmath49 .
the fraction of potential drop at the left contact is @xmath50 and @xmath51 at the right contact @xcite , hence @xmath52 is indeed the applied bias , and @xmath53 $ ] .
the component of the retarded green s function for the isolated ( non - interacting ) lead @xmath15 corresponding to the sites ( or energy levels , depending on the representation used to the electrodes ) connected to the central region is given by @xmath54 . in this paper
, we have chosen a simple model , a semi - infinite tight - binding chain with on - site energy @xmath55 and nearest - neighbour hopping integral @xmath56 .
this model gives a semi - elliptic density of states of the terminal lead site connected to the central region , and so each lead s green s function becomes @xmath57 with @xmath58 .
we have chosen this model because it is one of the most simple , although in principle and in practice there are no limitations for taking any other more complicated or more realistic models for the lead , such as bethe lattices with @xmath59-coordination , or a nanotip supported by a semi - infinite surface as shown in ref ) since all their electronic properties are wrapped up in the lead self - energies @xmath60 . the self - energy for the interacting central region , @xmath61 is somewhat more complicated .
it consists of the sum of the self - energies due to interactions between the electrons and to interactions between the electrons and the quantum vibration modes ( vibrons ) . in this paper
we consider only the coupling between each electron and a single vibron of the central region ( the molecule ) , hence the electron - vibron self - energy @xmath62 .
\(a ) @xmath63 ( b ) @xmath64 \(a ) @xmath65 ( b ) @xmath66 for the current work , it is necessary to calculate several types of self energy .
firstly we have the fock - like self - energy @xmath67 , which is a function of energy , and is represented in diagrammatic form by figure [ fig : fockhartree](a ) .
we also have the hartree - like self - energy @xmath68 which is independent of energy , given by figure [ fig : fockhartree](b ) .
calculations using only the hartree and fock diagrams , and performed in a self - consistent way are usually referred to as the self - consistent born approximation ( scba ) @xcite . however , as explained in the introduction , we also want to go beyond the scba , and thus we will also calculate two further self - energies that include two - vibron processes .
the first of these is the double - exchange self - energy @xmath69 given by figure [ fig : dx - dph](a ) . in the many - body language , it is part of the vertex correction to the fock diagram .
the second is given by figure [ fig : dx - dph](b ) and corresponds to the dressed vibron ( or gw - like ) self - energy @xmath70 , which consists of the vibron propagator renormalized by a single electron - hole bubble ( the polarization ) .
this is why we refer to the effects of @xmath70 as polarization effects in the following .
the details of how we implement these self - energies are given in appendix [ app : higherorderdiagrams ] .
once we have calculated all the different green s functions , any of the physical properties of the system , such as the electron density , the electronic current denstiy , the total energy , the current noise , the heat transfer , etc . can be calculated .
for example , the electronic current passing through the @xmath15 contact is given by @xmath71 i.e the first term describes the transfer of an electron from the interacting region to electrode @xmath15 , while the second transfers an electron from electrode @xmath15 to the central region . we can then express this in terms of green s functions and derive an expression for the expectation value of the current@xcite : @xmath72 \right . \\
& \qquad\quad \left .
\gamma_\alpha(\omega)\ g^<(\omega ) \right \ } \ .
\end{split}\ ] ] all physical properties may be expressed in terms of the spectral function @xmath73 which is at the heart of this paper .
the spectral function is related to the imaginary part of the retarded or advanced electron green s functions , as @xmath74 = + \im m [ g^a(\omega ) ] \ .\ ] ] for non - interacting systems , it is simply proportional to the density of electronic states @xmath75 . for interacting systems , it gives information about the excitations ( electron or hole ) of the system .
furthermore , when the system is at equilibrium ( @xmath76 ) , there are some relationships between the lesser ( greater ) and the advanced and retarded green s functions : @xmath77 and @xmath78 these relationships are at the centre of the fluctuation - dissipation theorem for equilibrium , which also be recast as a relationship betweeen the greater and lesser green s functions : @xmath79 for statistical averages at finite temperature in the grand canonical ensemble .
this equation is related to the kubo - martin - schwinger boundary conditions @xcite . for non - equilibrium conditions , there is no unique fermi level at finite bias ( or no unique temperature if @xmath80 ) in the whole system , and the relationships given by eqs .
( [ eq : glesser_equi]-[eq : glessgreat_kms ] ) no longer hold .
this is an important feature of the non - equilibrium formalism for which conventional equilibrium statistics need to be reformulated .
the calculations start by constructing the non - interacting of the entire system @xmath81 with @xmath82 being any three of the possible @xmath83 .
the other are obtained by using the relationships between them as shown in appendix [ app:1st_order_diagrams ] .
for example , the retarded is given by @xmath84^{-1}$ ] .
one then calculates three `` keldysh components '' for the self - energies corresponding to any of the diagrams @xmath85 \{f , dx or dph } : @xmath86 , @xmath87 and @xmath88 .
the hartree diagram has only one component @xmath89 as shown above .
the advanced and retarded self - energies @xmath90 are then obtained by simple algebra using the relationships between the different self - energies as explained in appendix [ app:1st_order_diagrams ] .
the new are then calculated by using the dyson equations for @xmath36 @xmath91^{-1},\ ] ] and the quantum kinetic equations for @xmath92 @xmath93 where the total self - energies are given by @xmath94 the new self - energies are then recalculated and the process re - iterates until full self - consistency is achieved . actually at each iteration of the calculations , we use a simple mixing scheme of the self - energies obtained at the present iteration and at the previous iteration .
this mixing scheme permits us to achieve full self - consistency in a maybe slightly longer but more stable iterative process .
note finally that by using eq .
[ eq : totalgfkeldysh ] , instead of the more general formulation given by eq .
[ eq : greens - greater - lesser ] , we assume that , after switching on the interactions , there are no bound states in the system ( i.e. there are no interaction - induced electron states located outside the spectral supports of the left and right leads ) @xcite which is indeed the case .
in this section , we present results for the spectral functions @xmath73 for the different transport regimes and for different applied biases in the low vibron - temperature regime .
we divide the calculations into four types .
firstly , we calculate the spectral functions at equilibrium for two transport regimes .
the first of these is where @xmath95 or @xmath96 , known as the off - resonant regime as in order to create a current between the left and right leads one puts an electron in the empty ( for electron transport , @xmath97 ) or full ( for hole transport , @xmath98 ) electronic level @xmath99 .
the second transport regime is when @xmath100 , known as the resonant transport regime .
the linewidth is the width of the peak in the spectral function which arises from the electronic coupling of the central region to the left and right leads . in this case
the electronic level @xmath99 is , at equilibrium , half - filled by an electron ( and thus also half - filled with a hole ) .
for each of these transport regimes , we calculate the spectral function both at equilibrium ( applied bias @xmath101 ) and non - equilibrium at finite bias ( @xmath102 ) . for each of these four groups (
resonant / off - resonant transport regimes , at / out of equilibrium ) , a large amount of different negf calculations have been performed for different values of the relevant parameters and within different levels of approximation ( hartree fock , hartree fock+second - order , partially or fully self - consistent calculations ) . in the rest of this section ,
we present only a limited and selected number of results which we found the most relevant for each case , and we analyse and compare in detail the effects of the different diagrams on the spectral functions of the system at and out of equilibrium .
we also , in section [ sec : comp - with - inelastic - scattering - technique ] , compare perturbation expansion based calculations ( negf ) to a reference calculation which is exact in term of electron - vibron coupling but which however is only valid for a specific transport regime .
we first consider the spectral functions at equilibrium , with no applied bias .
figure [ fig : specdens_equi_offres ] shows @xmath73 for the off - resonant regime , and figure [ fig : specdens_equi_res ] shows the resonant transport regime . in the equilibrium many - body language ,
the features in the spectral functions obtained at positive energies ( @xmath103 , above the fermi level ) correspond to electron excitations , while the features at negative energies ( @xmath104 , below the fermi level ) correspond to hole excitations . for the off - resonant ( a ) electron and ( b ) hole transport regime .
calculations were performed with the fock - like electron - vibron diagram ( solid line ) and with both the hartree and fock - like diagrams ( symbols and dotted line ) . for the electron transport regime
the inclusion of the hartree self - energy has no effect , but for the hole transport regime it shifts the entire spectral function to lower energies .
the parameters are @xmath105 for electron ( hole ) transport @xmath106 and @xmath107 .
, title="fig:",width=302 ] + for the off - resonant ( a ) electron and ( b ) hole transport regime .
calculations were performed with the fock - like electron - vibron diagram ( solid line ) and with both the hartree and fock - like diagrams ( symbols and dotted line ) . for the electron transport regime
the inclusion of the hartree self - energy has no effect , but for the hole transport regime it shifts the entire spectral function to lower energies .
the parameters are @xmath105 for electron ( hole ) transport @xmath106 and @xmath107 .
, title="fig:",width=302 ] for the off - resonant electron transport regime ( figure [ fig : specdens_equi_offres](a ) ) all the features in the spectral function are above the fermi level @xmath108 and hence correspond to electron excitations .
the main peak corresponds to adding an electron in the single available level .
this peak is broadened by the coupling to the leads , and its position in energy @xmath109 is renormalised by the electron - vibron interaction , i.e. @xmath109 is close to the static the polaron shift @xmath110 .
the lesser peaks in the electron - transport spectral function are vibron side - band peaks arising from resonance with @xmath111 excitations in the vibration mode .
these peaks correspond to vibration excitation ( vibron emission ) only . at zero vibron temperature ,
these are the only available mechanisms for vibrational excitations .
we note , and discuss further in section [ sec : comp - with - inelastic - scattering - technique ] , that these side - band peaks should occur at integer multiples of @xmath112 away from the main peak , but that for both our fock - only and hartree - fock scba calculations the peak - peak separation is slightly wider than this . in this regime ,
the hartree self - energy is negligible , because most of the spectral weight is above the fermi level and @xmath113 .
this implies that the polaron shift is mainly due to the fock - like self - energy . for the off - resonant hole - transport regime ( figure [ fig : specdens_equi_offres](b ) ) , all the features in the spectral function occur at @xmath114 and therefore correspond to hole excitations .
the vibron side band peaks are at lower frequencies than the main peak because they correspond to the emission of vibrons by holes rather than electrons .
when we include just the fock - like self - energy , the hole spectral function is symmetric ( with respect to the equilibrium fermi level @xmath108 ) with the electron spectral function , as can be seen clearly in figure [ fig : specdens_equi_res ] .
adding the hartree self - energy , however , breaks this electron - hole symmetry .
as most of the spectral weight is below @xmath115 , the expression for @xmath116 given in equation eq.([eq : hartreeselfenergy ] ) reduces to a constant @xmath117 ( i.e. twice the polaron shift ) as @xmath118 . as a result of this the whole spectral function is shifted to the left by this amount . for the resonant transport regime .
calculations were performed with the fock - like ( solid line ) and with both the hartree and fock - like diagrams ( dashed line ) .
the parameters are identical to those used in figure [ fig : specdens_equi_offres ] , except for the value of the electron level which is @xmath119 for the resonant regime.,width=302 ] we now turn to the resonant transport regime , with the spectral function shown in figure [ fig : specdens_equi_res ] . here
our electronic level @xmath99 is broadened by the coupling to the leads and is partially filled with electrons .
we can see that calculations performed with only the fock - like self - energy preserve the electron - hole symmetry .
the spectral function presents peaks located both at positive and negative energies , which correspond to the emission of vibrons by electrons or holes respectively . as in the previous section
, the inclusion of the hartree self - energy ( equation ) breaks down the electron - hole symmetry and the features are shifted to lower energies with a corresponding modification of the spectral weights for each peak .
we note here that the inclusion of the hartree self - energy thus modifies so drastically the spectral function that it will also strongly affect the @xmath120-@xmath121 characteristics of the junction in comparison to calculations performed with only the fock self - energy . in this section ,
we present results for the spectral functions @xmath73 for the off - resonant transport regimes , for different applied biases , both with and without the hartree contribution .
we first present calculations for an asymmetric potential drop with @xmath122 , i.e. @xmath123 and @xmath124 ( @xmath125 ) .
figure [ fig : specdens_ne_offres_f](a ) shows the spectral function calculated with just the fock component at different applied biases .
we then increase the bias @xmath121 by increasing the chemical potential of the left contact while keeping that of the right contact at zero .
for low values of @xmath121 ( @xmath126 ) there is little change in the spectral function .
however , once the value of @xmath121 exceeds that of @xmath112 , the spectral function becomes increasingly modified , especially when there is significant spectral weight inside the bias window @xmath127 . in particular , the lineshapes of both the main peak and the vibron side - band peaks become deformed , with a noticeable asymmetry of the main peak for biases where the main peak , but not the right - hand vibron side - band peak , lies within the bias window .
there is a saturation regime once the vibron side band peaks , and thus nearly all of the spectral weight , is within the bias window ( curves for @xmath128 ) .
here the main peak becomes pinned around @xmath129 ( effectively midway between @xmath130 and @xmath131 ) , and a symmetric lineshape is restored .
we postulate that this is owing to the bias being large enough to achieve simultaneous electron and hole transport .
.,title="fig:",width=302 ] + .,title="fig:",width=302 ] figure [ fig : specdens_ne_offres_f](b ) shows the spectral function with both the hartree and fock diagrams included .
as we have already noted , at zero bias there is no change from the fock - only spectral function . for small biases ( @xmath132 )
there is little difference , but for larger biases the effect of the fock diagram becomes increasingly evident .
this is because once the bias exceeds the vibron frequency @xmath112 , the non - equilibrium electron density becomes strongly perturbed .
the main peak is more stable in position than for the fock - only spectral function , although it shifts slightly towards lower energies with increasing bias before stabilising .
the righthand vibron side - band peak becomes strongly deformed , and moves to lower , rather than higher , energies as the bias is increased .
the left - hand vibron side - band peak appears at a lower bias than for the fock - only spectral function , at a frequency just above zero , then tends towards zero as the bias increases .
as for the fock - only case however , once both vibron side - band peaks are within the bias window , saturation is reached and the peak positions stabilize .
for both the fock - only and hartree - fock spectral functions , the separation between both side - band peaks and the main peak is @xmath133 for the parameters we used .
.,width=302 ] figure [ fig : hartreepot_vs_v_offres ] shows the hartree potential ( i.e. the real part of @xmath89 ) plotted against applied bias for different values of the electronic level @xmath99 .
the solid line corresponds to the value of @xmath99 used to calculate the spectral functions of the lower panel in figure [ fig : specdens_ne_offres_f ] .
the hartree potential is small and almost constant in the low bias region .
hence , in the quasi - equilibrium regime , such a potential could be neglected in the calculations , because it simply corresponds to an energy reference shift . however when the bias window @xmath134 $ ] starts to encompass features in the spectral functions ( either the main peak or vibron side - band peaks ) , the corresponding non - equilibrium electron density starts to vary substantially with the applied bias .
thus the hartree potential shows a strong dependence on the value of the bias ( as also shown in ref . [ ] ) , until the saturation regime is reached and the hartree potential is once again constant for very large biases .
the asymptotic saturation value of the hartree potential is dependent on the value of the electronic level @xmath99 as one would expect .
the first drop in the value of the hartree potential happens at @xmath109 , with the subsequent steps , which become progressively smaller and broader , at @xmath135 and then @xmath136 .
this behaviour indicates that one can not in principle neglect the hartree diagram contribution in the calculations , unless one is interested in calculating only the properties of the system for a bias range for which the electron density is ( almost ) constant .
it is worth mentioning that it is not straightforward to relate the modification of the peak positions in the two panels in figure [ fig : specdens_ne_offres_f ] to the hartree potential alone . in self - consistent calculations
, a highly non - linear system needs to be solved , since the hartree potential is obtained from one element of the which are themselves dependent on the values of the hartree potential .
we now consider what happens when we apply a bias in the resonant transport regime .
we apply a symmetric potential drop ( i.e. the potential of the left electrode is raised by an amount @xmath137 while that of the right electrode drops by the same amount ) .
this allows us to keep the electron - hole symmetry and see under which circumstances the electron - hole symmetry is broken . for the resonant transport regime for different applied biases for ( a )
the fock diagram only , and ( b ) the fock and hartree diagrams .
the curves are shifted vertically ( + 10 in @xmath138-axis ) for clarity , with the values of the left and right chemical potentials given by the triangular symbols , here we have a symmetric potential drop . adding the hartree diagram breaks the electron - hole symmetry .
the other parameters are the same as for figure [ fig : specdens_ne_offres_f ] except for @xmath139 and @xmath140.,title="fig:",width=302 ] + for the resonant transport regime for different applied biases for ( a ) the fock diagram only , and ( b ) the fock and hartree diagrams .
the curves are shifted vertically ( + 10 in @xmath138-axis ) for clarity , with the values of the left and right chemical potentials given by the triangular symbols , here we have a symmetric potential drop . adding the hartree diagram breaks the electron - hole symmetry .
the other parameters are the same as for figure [ fig : specdens_ne_offres_f ] except for @xmath139 and @xmath140.,title="fig:",width=302 ] the spectral functions for the fock diagram are shown in figure [ fig : specdens_ne_res_f](a ) . as we have no hartree term , the spectral function is electron - hole symmetric at equilibrium . moreover , as we have chosen a symmetric potential drop ( @xmath141 ) the spectral functions stay symmetric for all applied biases .
constraining the symmetry in this way allows us to concentrate on the effects of increasing the bias , and we can see that the most important modifications of the spectral function are in the width of the central and satellite peaks .
the vibron side - band peaks show only a very small change with increasing applied bias .
figure [ fig : specdens_ne_res_f](b ) shows the spectral function when we add in the hartree term while keeping all other variables unchanged .
we note immediately that the electron - hole symmetry is broken , even at equilibrium .
in addition the main peak is shifted towards negative energies with respect to the fock - only calculation , and the right - hand vibron side - band peak is completely suppressed the whole spectral function takes on the qualitative appearance of a spectral function in the hole - dominated regime ( compare with figure [ fig : specdens_equi_offres ] ) . on increasing the bias , the width and shape of the left - hand vibron side - band peaks
is varied , and for higher biases new vibron side - band peaks appear above the main peak , until at high biases the spectral function is very similar in shape to that for the off - resonant spectral function ( figure [ fig : specdens_ne_offres_f ] ) . in this section , we will check the validity of the self - consistent born approximation ( scba ) , i.e. self - consistent calculations using only the hartree and fock diagrams , versus another method which gives more exact results as far as electron - vibron coupling is concerned . since negf is a many - body perturbation expansion theory , by definition it does not contain all diagrams even though self - consistency allows us to achieve a partial resummation of a subclass of diagrams . as an alternative to negf
, one can calculate the transport properties using an extension of conventional scattering theory to include the interaction of incoming single - particle states with some bosonic degrees of freedom inside the central region of interest @xcite .
this technique , termed the multi - channel inelastic scattering technique ( mcist ) @xcite , has the advantage of being an exactly solvable problem , even in the presence of many non - interacting electronic states coupled to many vibration modes in the central region @xcite .
mcist is based on many - body perturbation theory for polaron and it is exactly solvable in the sense that mcist is treating the electron - vibron coupling to all orders . in the language of polaron theory , mcist contains all the diagrams corresponding to the electron - vibron interaction ( in the corresponding transport regime ) . to be more precise
, mcist contains all orders of crossing and non - crossing diagrams in terms of the bare vibron propagator @xcite . in conventional polaron theory for one electron , there are no diagrams with electron - hole loops in them @xcite .
however , mcist is a single - particle scattering technique and treats the statistics of the fermi seas of the left and right leads only in an approximate manner . in the language of negf , this means that the results given by mcist are only valid for a specific transport regime ( as we will see below ) . in the case of the sssm model ,
the retarded of the central region has the usual form @xmath142^{-1 } \ .\ ] ] within mcist , the retarded electron - vibron self - energy containing all orders of the electron - vibron coupling is expressed as a continued - fraction as shown analytically in refs . [ ] and [ ] : @xmath143 where we recall that @xmath144 is the retarded gf of the central region connected to the leads , i.e. @xmath145^{-1}$ ] , without electron - vibron coupling .
as show in ref .
[ ] , the lowest born approximation , fully consistent with scba ( hartree - fock calculations as shown above ) , is given in mcist by an electron - vibron self - energy equivalent to eq .
( [ eq : ser_mcist ] ) but where the integer @xmath17 factors at each level of the continued fraction are all replaced by the integer @xmath146 .
obviously , this approximate substitution is good enough in the ( very ) weak electron - vibron coupling for which only the first level of the continued fraction contributes the most @xcite .
below we compare the spectral functions obtained by exact mcist calculations , by mcist approximated to the scba ( hartree - fock ) level , and to negf - scba calculations . ) , while the lower panel shows results for strong electron - vibron coupling ( @xmath147 ) .
mcist calculations give exact results in this case , and can also be performed at the same level of approximation as scba ( see main text for details ) .
mcist - scba and negf - scba are virtually identical , but show large discrepancies for the peak positions ( main peak and more especially for the vibron side - band peaks ) in comparison to the exact results .
the parameters are @xmath148 ( electron transport ) , @xmath149 ( upper panel ) and @xmath150 ( lower panel ) , @xmath151.,title="fig:",width=302 ] + ) , while the lower panel shows results for strong electron - vibron coupling ( @xmath147 ) .
mcist calculations give exact results in this case , and can also be performed at the same level of approximation as scba ( see main text for details ) .
mcist - scba and negf - scba are virtually identical , but show large discrepancies for the peak positions ( main peak and more especially for the vibron side - band peaks ) in comparison to the exact results .
the parameters are @xmath148 ( electron transport ) , @xmath149 ( upper panel ) and @xmath150 ( lower panel ) , @xmath151.,title="fig:",width=302 ] the spectral functions obtained from mcist are shown in figure [ fig : specdens_mcistversusscba ] for the weak / intermediate electron - vibron coupling regime ( @xmath152 ) and for the strong electron - vibron coupling regime ( @xmath153 ) .
the overall lineshapes correspond to a main peak with vibron side - band peaks located only above the main peak .
these are typical results fully consistent with an off - resonant transport regime situation .
furthermore , the spectral functions obtained with ba - based approximation ( mcist - scba and negf - scba ) are virtually identical , especially in the intermediate ( @xmath152 ) to very weak ( not shown here ) electron - vibron coupling regime . for strong electron - vibron coupling ,
one obtains the same peak positions , however the amplitude of the peaks ( especially the main peak ) is slightly different .
since mcist calculations will always give similar lineshapes independent of the value of @xmath99 , one can conclude that mcist calculations are only valid for the off - resonant transport regime at and near equilibrium .
mcist is not able to reproduce the spectral features of the resonant transport regime ( i.e. vibron side - band peaks on both sides of the main peak ) .
this is essentially due to the fact that in mcist , one does not take properly into account the statistics of the fermi seas of the left and right leads .
see for example eqs .
( [ eq : gr_mcist ] ) and ( [ eq : ser_mcist ] ) , there are no leads fermi distributions in the retarded component of the leads self - energies @xmath154 .
now , comparing ba - based calculations with the exact mcist calculations , one can see from figure [ fig : specdens_mcistversusscba ] that the ba - based calculations give the wrong polaron shift , i.e. the normalized position of the main peak @xmath99 , especially in the strong electron - vibron coupling regime .
furthermore ba - based calculations also give the wrong energy separation between the main peak and the first vibron side - band peak .
this energy difference should be equal to the vibron energy @xmath112 , as it is given by exact mcist calculations .
note that the limits of ba - based calculations were also been studied by lee _
_ in a somewhat different context in ref .
[ ] . in conclusion
, this means that hartree - fock ( or ba ) based calculations for electron - vibron interaction are only valid for weak coupling , as can be expected from a perturbation - expansion based theory .
hence one needs to include higher - order diagrams in the electron - vibron self - energies to go beyond the commonly - used self - consistent born approximation ( hartree - fock ) in order to obtain correct results for a wide range of parameters .
the effects of the higher - order diagrams ( here second - order - dx and dph diagrams ) are explored in detail in the following sections . additionally ,
although mcist calculations are only valid in the off - resonant transport regime at and near equilibrium , they include all possible higher - order diagrams ( with bare vibron propagator ) and hence can be used as a reference for any perturbation - expansion - based negf calculations performed at equilibrium or in the quasi - equilibrium regime . in this section ,
we present results for the spectral functions when the second - order diagrams ( see figure [ fig : dx - dph ] ) are included in the calculations of the green s functions .
the reader can find more information about the mathematical expressions for the self - energies corresponding to the second - order diagrams in appendix [ app : higherorderdiagrams ] .
these diagrams fall into two types the double - exchange dx diagram , corresponding to vertex corrections , and the dressed vibron diagram , which includes a single electron - hole bubble , renormalizing the vibron propagator and hence giving rise to polarization effects .
we have used three different levels of approximation to calculate these green s functions : firstly , calculations with no self - consistency the green s functions are simply calculated using the diagrams in figure [ fig : fockhartree ] and figure [ fig : dx - dph ] using the bare propagator @xmath155 as the electron green s function . in our model , @xmath155 is the green s function of the central region connected to the leads with no electron - vibron interactions .
this is a first - order perturbation expansion for which @xmath156 $ ] .
we use the abbreviations ba ( born approximation for non self - consistent hartree and fock diagrams ) and ba+dx ( dx for double exchange ) and ba+dx+dph ( dph for dressed vibron , for the @xmath157-like diagram ) in the following . secondly , we perform partly self - consistent calculations , where the green s functions are calculated with the first loop of self - consistent calculations with the hartree and fock diagrams .
we use these green s functions as a starting point to calculate new , corrected , green s functions including the second - order diagrams @xmath158 and/or @xmath159 $ ] in which the green s functions are the corresponding scba green s functions .
finally , we perform fully self - consistent calculations , in which the green s functions are calculated in a self - consistent manner with the first and second - order diagrams included within each iteration of the self - consistency loop .
our rationale for choosing to do calculations in this manner is as follows .
we are able to test the different levels of approximation and see very precisely the effects of vertex corrections and polarization are on both the bare green s functions @xmath155 and the scba - level green s functions .
we will also show in section [ sec:2ndorder_off_res_equi ] that by using the scba green s functions rather than @xmath155 as a starting point , one achieve a better convergence in the calculations .
in particular we will show later on in section [ sec:2ndorder_off_res_equi ] that in some cases , for example the off - resonant transport regime when @xmath160 ) , the use of the bare @xmath155 green s functions as a starting point for a fully self - consistent calculation with second - order diagrams actually gives unphysical results . additionally , fully self - consistent calculations are extremely computationally intensive , and hence it is both interesting and useful to explore the range of parameters for which the second - order - diagram corrections to the scba calculations give a sufficiently accurate description of the green s functions in comparison to the fully self - consistent calculations .
figure [ fig : specdens_eq_ofres_hfdxdph ] shows the spectral functions of the off - resonant transport regime at equilibrium , calculated for weak / intermediate electron - vibron coupling ( @xmath161 ) and for different diagrams and levels of self - consistency .
as already mentioned in sections [ sec : at - equilibrium ] and [ sec : noneq - specfns ] , the energy separation between the main peak and the first vibron side - band peak ( which should be equal to the vibron energy , here @xmath162 ) is not well reproduced by ba - based ( hartree - fock based ) calculations and is much larger than @xmath112 . the self - consistency introduced in the calculations give a marginally smaller energy separation .
the effects of the second - order dx diagram are firstly to bring the vibron side - band peak closer to the main peak , hence giving an energy separation closer to the exact value ; and secondly , one observes a strong narrowing and a larger amplitude of the vibron side - band peak in both self - consistent and non - self - consistent calculations .
both these effects thus qualitatively modify the spectral functions towards better agreement with the exact results as shown in section [ sec : comp - with - inelastic - scattering - technique ] .
it is , however , worth mentioning that including the dx diagram in the calculation does not greatly affect the position of the main peak , a result which may be understood from the fact that the dynamical polaron shift is a quantity difficult to obtain exactly from a perturbation expansion theory beyond weak coupling @xcite .
, @xmath163 .
self - consistent calculations give better spectral functions , with vibron side - band peak closer to the main peak .
the second - order dx diagram narrows the side - band peaks and bring them even closer to the main peak as expected .
the second - order dph diagram does not contribute much in the off - resonant regime .
in this regime and for small electron - vibron coupling , second - order corrections to scba calculations are a good approximation to corresponding full self - consistent calculations.,width=302 ] the effects of the second - order dph diagram in the spectral function are virtually nil for the case of weak / intermediate coupling and the off - resonant regime .
this might not be that surprising since the dph diagram corresponds to a fock - like diagram with a renormalised vibron propagator .
the renormalization of the vibron is due to a single electron - hole bubble .
however in the off - resonant transport regime , the spectral function is almost empty ( in the case of electron transport ) or almost full ( in the case of hole transport ) which implies that there are not many electron - hole excitations available in this transport regime . hence the polarization ( i.e. the contribution of the electron - hole bubble ) is very small , subsequently giving very small values for the dph self - energy . as an example , we have checked our numerical values in the case shown in figure [ fig : specdens_eq_ofres_hfdxdph ] and we have found , as expected , that the maximum values of @xmath164 are 30 to 50 times smaller than the maximum values of @xmath165 .
furthermore , in the off - resonant regime and for weak - ish electron - vibron coupling , it seems that the fully self - consistent ( full sc ( ba+dx ) curve in figure [ fig : specdens_eq_ofres_hfdxdph ] ) results @xmath166 are well approximated by the results given by a second - order correction to a self - consistent hartree - fock ( scba ) calculations ( scba+dx curve in figure [ fig : specdens_eq_ofres_hfdxdph ] ) @xmath167\ ] ^{-1}$ ] .
this is an interesting result as it implies that the physical properties of the system , at least in these conditions , could be well described by a second - order correction of the lowest - order sc calculations , without the need to perform a fully self - consistent calculation up to the second - order .
now we turn to the analysis of results obtained for stronger electron - vibron coupling .
examples of such calculations are given in figure [ fig : specdens_eq_ofres__strg_hfdxdph ] for @xmath168 .
in the strong electron - vibron coupling regime , one obtains qualitatively the same contributions of the second - order diagrams as explained above for the weak coupling regime .
the dph diagram does not have a large role in the off - resonant transport regime , although slightly affecting the width of the main peak . and the dx diagram shifts the vibron side - band peak towards the main peak ( and hence towards the exact results ) as well as narrowing the peak width and increasing the peak amplitude .
these effects are amplified in figure [ fig : specdens_eq_ofres__strg_hfdxdph ] because the electron - vibron coupling constant @xmath32 is bigger than in figure [ fig : specdens_eq_ofres_hfdxdph ] .
furthermore , for strong electron - vibron coupling , the dx diagram modifies the energy position of the main peak and seems to give a slightly better polaron shift .
, @xmath169 .
the calculations show that no self - consistency gives poor results in comparison to exact calculations given in figure [ fig : specdens_mcistversusscba ] .
a closer spectral function to the exact result is obtained from sc(ba+dx ) calculation . the second - order dph diagram
do not contribute much in the off - resonant regime .
see main text for a detailed analysis of the spectral functions.,title="fig:",width=302 ] + , @xmath169 .
the calculations show that no self - consistency gives poor results in comparison to exact calculations given in figure [ fig : specdens_mcistversusscba ] .
a closer spectral function to the exact result is obtained from sc(ba+dx ) calculation . the second - order dph diagram
do not contribute much in the off - resonant regime .
see main text for a detailed analysis of the spectral functions.,title="fig:",width=302 ] there is however another interesting effect observed from such a set of calculations , which can be seen in the upper panel of figure [ fig : specdens_eq_ofres__strg_hfdxdph ] .
the first iteration of a fully self - consistent calculation ( including the hartree , fock , and dx diagrams ) starting with the non - interacting gfs @xmath155 gives non physical results , i.e. negative values of the spectral function ( see curve ba+dx in figure [ fig : specdens_eq_ofres__strg_hfdxdph ] ) , and such an unphysical behaviour does not self - correct in the following iterations .
this a well known problem , which has already been encountered in the past by several authors in the context of electron - vibron interaction ( for example in ref [ ] ) and also in the context of electron - electron interaction when considering topologically equivalent diagrams @xcite . in ref .
[ ] , a somewhat different approach than ours was used .
it is based on a linked cluster expansion for the non - equilibrium steady - state regime , and negative densities of states were obtained when including second - order diagrams , and in some cases even higher - order cluster approximations did not seem to give a convergent solution at intermediate electron - vibron strength and in the presence of the fermi seas .
however , our calculations reveal that it is possible to solve such a problem by starting the fully self - consistent calculations ( up to second order ) from a different starting point , namely by starting from the @xmath170 green s functions ( i.e. the gfs obtained from a fully sc calculation including only the lowest - order diagrams ) .
this is shown by the curve ( scba+dx ) in the upper panel of figure [ fig : specdens_eq_ofres__strg_hfdxdph ] and by the lower panel in which all the fully converged self - consistent results are shown . for the moment
, we do not have a full physical explanation of the reason why starting from a scba calculations is better to achieve full self - consistency with higher - order diagram than hartree - fock , apart from the simple fact that an hartree - fock calculation is probably closer to the true interacting solution than the non - interacting solution . to conclude this section
, we can say that in the off - resonant regime at equilibrium , the second - order dx diagram dominates over the second order dph diagram . in this section ,
we present calculations for the resonant regime at equilibrium .
though we have shown , in sections [ sec : spectralfunctionsresonant ] and [ noneq - specfns - resonant ] , the importance of the hartree diagram , we will consider below results obtained without the hartree diagram .
the calculations were performed with the fock and second - order dx and/or dph diagrams which conserve ( at numerical accuracy ) the electron - hole symmetry of the system . the reasons why we have choosen to perform this model calculation are twofold :
first , even for a single electronic level , we expect to have the maximum possible electron - hole excitations available when the spectral functions are electron - hole symmetric .
hence we expect the polarization effects be to more pronounced in a system with electron - hole symmetry .
second , the electron - hole symmetric model permits us to emphasize the competitive effects between the dx and dph diagrams as will be shown below .
( resonant transport ) , @xmath171 .
the top inset shows the whole spectral functions , while the left ( a ) and right ( b ) parts are zooms , for positive @xmath172 , of the central peak and first vibron side - band peak respectively .
the second - order diagrams have the following effects on the scba spectral functions : dph broadens the central and side - band peaks as well as lowers their height , in opposition to dx which narrows the peaks and brings the side - band peaks slightly closer to the center , with a strong and unphysical increase of the amplitude of the central peak .
calculations with both dx and dph give in appearance a broadening of the central peak , which however recovers the correct height necessary to conserve the fermi - liquid property of the electron - hole symmetric system ( see section [ sec : linear_reponse ] for more detail ) .
, width=302 ] figure [ fig : specdens_eq_res__intermed_hfdxdph ] shows the equilibrium spectral functions of the resonant regime ( at weak / intermediate electron - vibron coupling ) obtained from different self - consistent calculations including first and second - order diagrams .
on one hand , the second - order dph diagram corresponds to a partial dressing of the vibron propagator by one electron - hole bubble .
it gives an extra lifetime in the retarded / advanced electron green s functions in comparison to hartree - fock ( scba ) calculations .
hence the main effect of the dph diagram is to introduce an extra broadening of the peaks in the spectral functions .
since the dph diagram is one of the so - called conserving approximations @xcite , the broadening of the peaks leads to a reduction of their height to keep globally the same total spectral weight .
these effects can be seen on the sc(ba+dph ) curve in figure [ fig : specdens_eq_res__intermed_hfdxdph ] . on the other hand , the second - order dx diagram , which is a conserving approximation ,
has an opposite effect : a strong narrowing of the peaks ( especially of the central peak ) accompanied with an increase of their height , as can seen on the sc(ba+dx ) curve in figure [ fig : specdens_eq_res__intermed_hfdxdph ] .
now it is interesting to see what happens when the calculations are performed with both second - order dx and dph diagrams .
this is shown on the sc(ba+dx+dph ) curve in figure [ fig : specdens_eq_res__intermed_hfdxdph ] : we obtain a hybrid behaviour , in the sense that in appearance the central peak is broadened in comparison to scba calculations .
however , the height of the peak at the fermi level is conserved ( up to numerical accuracy ) .
this is a very important result which proves that for the model calculation of an electron - hole symmetric system , one has to include both the dx and dph second - order diagrams in order to conserve the expected fermi - liquid properties of the system . in this regime
both the dx and dph second - order diagrams play an equally important role which determines the linear response properties of the system as shown below in section [ sec : linear_reponse ] . by comparison , in a diagrammatic treatment of the _ electron - electron _ interaction on the electron propagator , the situation is often different : the electron - hole bubble diagram that appears here in the dph contribution to the propagator is large , especially in highly polarizable metallic and open - shell systems where electron - hole pairs may be created with low energy cost , because the coulomb interaction operates at all energy scales . in that case , summing the bubble diagrams to infinite order as is done in hedin s @xmath157 approximation @xcite is much more important than including the second - order exchange diagram .
the key difference for the electron - phonon interaction is that the vibron frequency @xmath112 imposes a restricted energy scale on the interaction , reducing the importance of the bubble diagrams , and correspondingly increasing the importance of the second - order exchange diagram .
figure [ fig : specdens_ne_ofres_hfdxdph ] shows the spectral functions of the off - resonant transport regime at finite bias .
the calculations have been performed for weak / intermediate electron - vibron coupling ( @xmath161 ) and for different diagrams and levels of self - consistency .
we have considered the case for which real excitations of vibrons are possible @xmath173 .
the interpretation of the results is not as straightforward as in the equilibrium case , because non equilibrium effects are sometimes counter intuitive .
however , the overall shapes of spectral functions are quite similar to that obtained within hartree - fock based calculations ( see section [ noneq - specfns - offres ] ) in the sense that they present a central peak with vibron side - band peaks on both sides . similarly to the resonant regime at equilibrium ,
the main effects of the second - order dx diagram is to narrow the width of all peaks , and to shift slightly the side - band peaks towards the main peak . as mentioned in the previous section ,
the main effect of the second - order dph diagram is to broaden the peaks , in opposition to the effects of the dx diagram .
however here , the broadening in the non - equilibrium condition appears less important than in the case of the resonant regime at equilibrium .
hence full self - consistent calculations performed with both dx and dph second - order diagrams give a narrowing of the peaks with a corresponding increase of their amplitude in comparison to scba calculations . finally , out of equilibrium
, it can be seen that the results given by a full self - consistent calculation ( curve sc(ba+dx ) in figure [ fig : specdens_ne_ofres_hfdxdph ] ) are strongly different from a second - order correction to an hartree - fock calculation ( scba+dx curve in [ fig : specdens_ne_ofres_hfdxdph ] ) which gives an excessively narrowed central peak .
this means that in the weak / intermediate electron - vibron coupling , second - order corrections to a scba calculations are only good enough at equilibrium , however at non - equilibrium full self - consistency needs to be performed with all diagrams of the same order . ) for the off - resonant transport regime for the weak / intermediate electron - vibron coupling .
calculations were performed self - consistently with first - order hartree and fock - like diagrams ( scba curve ) , and with second - order dx and/or dph diagrams ( sc(ba+dx,+dph,+dx+dph ) curves ) .
second - order dx correction to scba is also shown ( scba+dx curve ) .
the top - left inset shows a zoom on the central peak around @xmath174 , and the top - right insets show a zoom on the vibron side - band peaks around @xmath175 and @xmath176 .
the other parameters are @xmath148 , @xmath177 .
the second - order dx diagram narrows the peaks with a slight shift of the side - band peaks towards the central peak .
the narrowing is too strong in the case of partially self - consistent calculations ( scba+dx ) .
the second - order dph diagram broadens the peak , but not as much as in the resonant case .
full self - consistent calculations including both second - order diagrams result in an intermediate behaviour for the modifications of the spectral functions.,width=302 ] as we have already shown in the first - order electron - vibron diagrammatic calculations , the spectral functions for the off - resonant and resonant regime at non - equilibrium are qualitatively similar , in the sense that they present a central peak with vibron side - band peaks on both sides .
one can compare for example the spectral functions obtained for hartree - fock - like calculations at non - equilibrium shown in figure [ fig : specdens_ne_offres_f ] and figure [ fig : specdens_ne_res_f ] .
hence , and again on a qualitative level , the effects of the second - order dx and dph diagrams on the resonant case at finite bias are similar to what has been obtained for the off - resonant non - equilibrium case described in the previous section .
the effects of these higher - order diagrams on the full non - equilibrium transport properties for the different transport regimes will be presented in a forthcoming paper .
however before turning the discussion to the linear - response properties of the system at and near equilibrium , we would like to comment on a specific aspect of the effects of higher - order diagrams . the narrowing of the vibron side - band peaks and of the main central peak due to higher - order ( dx ) diagrams was also obtained by other authors ( see for example ref . [ ] ) . in this paper ,
a different non - equilibrium approach was used .
it consists of starting with an electron dressed by a vibron ( a polaron ) in the isolated central region , then using perturbation expansion theory ( with partial resummation ) in terms of the coupling of the central region to the non - equilibrium left and right leads .
however , the results for the spectral functions in ref .
[ ] were only given for the resonant transport regime at equilibrium and all calculations were performed without the hartree - like diagram . though we have already shown that such a diagram plays a crucial role in the spectral properties of the system in the resonant regime at and out of equilibrium .
we now briefly discuss the effects of electron - vibron interaction taken at different levels of the diagrammatic expansion on the linear - response transport properties of the single - molecule nanojunction . before doing so
, we explain how to derive the linear conductance from the value of the spectral functions calculated at equilibrium . in our model , there is a direct proportionality between the left and right leads self - energies @xmath178 because , the central region is coupled to the leads via the single hopping matrix element and we have chosen identical leads . then the current ( eq . ( [ eq : expectation_value_current ] ) ) can be recast as follows ( for the details of the derivation , see for example refs . [ ] ) : @xmath179,\ ] ] where @xmath180 and @xmath181 $ ] .
the linear conductance @xmath182 is obtained from @xmath183 = 2\gamma(\mu^{\rm eq } ) \
a(\mu^{\rm eq})\ ] ] where @xmath184 is the quantum of conductance .
table [ tab : linear_conduc ] shows the different values for the linear conductance @xmath185 obtained from the equilibrium spectral functions shown in figures [ fig : specdens_eq_ofres_hfdxdph ] , [ fig : specdens_eq_ofres__strg_hfdxdph ] and [ fig : specdens_eq_res__intermed_hfdxdph ] .
[ tab : linear_conduc ] .linear conductance @xmath186 obtained from the value of the equilibrium spectral density at @xmath187 .
@xmath185 is given in units of the quantum of conductance @xmath184 .
the equilibrium spectral functions are shown in figs .
[ fig : specdens_eq_ofres_hfdxdph ] , [ fig : specdens_eq_ofres__strg_hfdxdph ] and [ fig : specdens_eq_res__intermed_hfdxdph ] ) and were done for different different levels of approximation as explained in the corresponding figure captions . [ cols="<,^,^,^ " , ] for the off - resonant regime at the weak electron - vibron coupling , there are not many differences in the linear conductance @xmath185 values calculated for all the different diagrams , as one might expect from the spectral function behaviour shown in figure [ fig : specdens_eq_ofres_hfdxdph ] .
this is especially true when the main peak in the spectral function is located well away from the fermi level , i.e. @xmath188 linewidth of the peaks . for strong electron - vibron coupling ,
the differences between different levels of approximations are more pronounced .
all @xmath185 values are smaller than the exact mcist result .
this is essentially due to the fact that perturbation expansion gives only an approximate value of the polaron shift .
the differences from the exact result are more important for scba - based calculations ( difference of @xmath189 with exact result ) and of course calculations performed up to second order give a better polaron shift and hence better values for the linear conductance ( @xmath190 difference only ) . in the quasi - resonant case
when @xmath191 ( 1 to 2 ) linewidth , the effects described above will be more pronounced , because the linear conductance is no longer given by the extreme tail of the main peak crossing the fermi level . for the resonant case in the absence of the hartree potential , the scba - based calculations should give a perfect linear conductance .
the reason why @xmath192 is because the calculations were performed with a tiny but finite value of @xmath193 as explained in detail in appendix [ app : numericalbits ] .
hence one may say that , for the chosen set of parameters , the value @xmath194 represents an upper bound for the linear conductance ( the corresponding numerical perfect conductance ) . and
as explained in appendix [ app : numericalbits ] , linear conductance values can only be compared between calculations performed with the same value of @xmath193 , which is what we have done . the linear conductance is strongly renormalized ( decreased and increased ) when including only one of the two second - order diagrams ( dph and dx respectively ) .
in contrast , it becomes close again to the expected quantum of conductance when the calculations are performed with both second - order diagrams .
this dependence of the conductance is well understood from the behaviour of the spectral functions shown in figure [ fig : specdens_eq_res__intermed_hfdxdph ] .
sc(ba+dph ) calculations introduce an extra broadening and a corresponding decrease of amplitude of the peaks in the spectral functions , hence a decrease of @xmath185 .
however , the strong narrowing , with increased amplitude , of the peaks due to sc(ba+dx ) calculations leads to an increase of @xmath185 .
the sc(ba+dx ) calculations give a linear conductance larger than that obtained from scba , i.e. @xmath195 , which is an unphysical result for our non - degenerate electronic level in the central part of the system .
this implies a strong constraint on the validity of the values of @xmath185 obtained from a many - body perturbation expansion : in order to conserve the fermi - liquid properties ( see appendix [ app : numericalbits ] ) in the electron - hole symmetric resonant regime , one has to perform the calculations by including all diagrams order by order .
results obtained from partial resummation of a subset of diagrams will probably give an incorrect linear conductance .
hence calculations performed by fully renormalizing the vibron propagators @xcite or all crossing diagrams @xcite , while containing diagrams higher than second order , will probably break important physical properties of the system at @xmath196 , because they lack important electronic processes , to second - order the dx vertex correction - like diagram or dph diagram respectivelly . in the resonant ( or quasi - resonant ) regime ,
polarisation effects and vertex corrections play an equally important role in the electronic structure and transport properties of electron - vibron interacting nanojunctions which electron - hole symmetry .
in this paper , we have presented a method based on negf to calculate the equilibrium and non - equilibrium electronic structure of electron - vibron interacting single - molecule junctions .
we have applied the method to a model system which consists of a single electronic level coupled to a single vibration mode in the central region , the latter in contact with two non - equilibrium electron reservoirs . in comparison to previous studies performed within a similar approach @xcite , the novelty of our method lies in the fact that it goes beyond the conventionally - used self - consistent born approximation ( scba ) .
higher - order diagrams for the electron - vibron interacting have been implemented in our calculations . in this paper
we have considered the second - order diagrams which contain two - vibron processes : the so - called double exchange dx diagram which is part of the vertex corrections to scba , and the partially dressed dph diagram in which the vibron propagator is renormalized by one electron - hole bubble .
we have studied the effects of the first and second - order diagrams on the spectral functions for a large set of parameters and for different transport regimes ( resonant and off - resonant cases ) at equilibrium and in the presence of a finite applied bias driving the system out of equilibrium .
we have shown the important role played by the hartree diagram in calculations based only on first - order diagrams .
such a diagram should not be neglected unless one works within an applied bias range for which the correspond hartree potential is constant . for calculations including both first and second - order diagram we have found that the effects of the individual second - order diagram are as follows :
the dx diagram reduces the width of the peaks in the spectral functions , with an increase of their height , while the dph has the opposite effect : it increases the width and decreases the height .
furthermore , the dx diagram moves the vibron side - band peak position towards the main peak , and hence gives a better peak separation as should be obtained from exact calculations .
calculations performed with both dx and dph diagram give intermediate results for the spectral functions , which are not simply an average / superposition of the individual effects because of the strong non - linearity involved in solving the problem self - consistently .
furthermore , the effects of the second - order diagrams also depend on the transport regime : at and near equilibrium , the dx diagram dominates over the dph diagram in the off - resonant transport regime ( essentially because there are not many electron - hole excitations available in this regime ) . in the resonant case , however , both dx and dph play an equally important role . for large non - equilibrium conditions
( @xmath197 ) , both second - order diagrams play an important role , since the corresponding spectral functions look qualitatively similar to those obtained the resonant regime .
we have thus shown that it is indeed necessary to go beyond scba to obtain the correct results for a wide range of parameters .
this has also been confirmed by comparing our negf results to an exact calculation in terms of electron - vibron interacting ( the multi - channel inelastic scattering technique mcist @xcite ) , though it should be noted that comparison between negf and mcist calculations are only valid for the off - resonant transport regime .
we have also studied in detail the effects of self - consistency on the calculations , especially for the cases including the second - order diagrams ( see section [ sec:2ndorder_off_res_equi ] and figure [ fig : specdens_eq_ofres__strg_hfdxdph ] ) .
we have found a solution to an old problem , well known in many - body perturbation theory : in order to avoid negative spectral densities when including higher order diagrams in the calculations , it seems more appropriate to start the self - consistent loop with scba ( hartree - fock like ) electron .
finally , we have studied the linear response ( linear conductance @xmath185 ) of the nanojunctions , and have found that in the off - resonant regime , the value of @xmath185 is governed by the behaviour of the tail of the main peak at the fermi level , hence it depends on both the position and the width of this peak . in the off - resonant regime , the dph diagram contribution is negligible and the dx diagram gives a better polaron shift ( better position of the main peak ) and hence the second - order ( mostly dx ) calculation gives a better agreement for @xmath185 with the exact result . for the near - resonant regime , the contribution of dph will become more important and both dx and dph will play an important role in determining the value of @xmath185 by changing both the position and the width of the main peak . for the resonant regime with electron - hole symmetry , it is necessary to perform the calculations by including both dx and dph diagrams in order to conserve the fermi - liquid properties of the system .
we anticipate that this will be also true for higher - order diagrams and calculations need to be performed by including all diagrams of the same order .
calculations performed with partial resummation of a subset of diagrams will break the expected fermi - liquid properties of the system and will probably lead to an incorrect value of the linear conductance .
finally , we expect that the effects of the second - order diagrams , shown in this paper for the spectral functions only , are also important in the full non - equilibrium transport properties of the nanojunctions .
the study of these effects on the non - linear conductance are currently undertaken and will be considered in a forthcoming paper .
in this appendix , we show how to derive the expression for the self - energies corresponding to one- and two - vibron process diagrams in terms of electron - vibron coupling . starting from the sssm hamiltonian
, the general definition of the one - particle gf of the central region is obtained from the time - loop contour - ordered product of the creation and annihilation operators in the central region @xmath198 the time - loop contour @xmath199 contains two branches , the upper @xmath200 and the lower @xmath201 branch . on the upper branch ,
time starts in the infinitely remote past and evolves forwards , then at the turning point , which can be placed at any arbitrary time , one passes onto the lower branch where the system evolves backwards in time back to the initially non - interacting starting point at @xmath202 .
then any expectation value of products of operator reduces to @xmath203 where @xmath204 is the average over the non - interacting ground state .
the operators are then given in the interaction picture , and @xmath205 is the generalization of the time evolution operator on the keldysh contour @xmath206 where @xmath207 is the time - ordering operator on the contour @xmath199 and @xmath208 implies integration over @xmath199 .
and @xmath209 is the `` perturbation '' to the reference hamiltonian , which in our case would be the interaction electron - vibron as well as the coupling of the central region to the leads ) . expanding @xmath205 as a series in terms of the electron - vibron coupling hamiltonian @xmath210
, one can derive the electron - vibron self - energies to any order of the electron - vibron coupling by calculating any time ordered products in the series using the usual rules of many - body perturbation theory , such as feynmann diagrammatic expansion or wick s theorem@xcite .
the one - vibron process self - energies , corresponding to the hartree and fock - like diagrams ( see figure [ fig : fockhartree ] ) are given by @xmath211 and @xmath212 where @xmath213 are times on the time - loop contour .
the projections onto the real ( physical ) times are given by , for example , @xmath214 for the fock self - energy .
the index @xmath215 labels the branch of the time - loop contour corresponding to the forward ( @xmath216)/ backward ( @xmath217 ) time evolution respectively .
there are several useful relationships between the different projections ( or keldysh components ) of the green s functions @xmath218 : @xcite .
they are @xmath219 or equivalently in terms of time - ordered ( @xmath220 ) , anti time - ordered ( @xmath221 ) , greater ( @xmath222 ) and lesser ( @xmath223 ) green s functions : @xmath224 and @xmath225 ( or equivalently @xmath226 ) , and hence @xmath227 .
similar relationships exist for the self - energies @xmath228 ( @xmath229 ) . using these relationships and taking the steady state limit , i.e. @xmath230 , and after fourier transformation into an energy representation @xmath231
, we obtain the usual expressions @xcite for the hartree @xmath232 and fock @xmath233 electron - vibron self - energies ( @xmath234 ) .
for example , the hartree and fock self - energies are @xmath235 where @xmath138 represents one of the tree keldysh compoentns @xmath236 . using the relationship @xmath237
, we find @xmath238 with the usual definitions for the bare vibron green s functions @xmath239 : @xmath240 , \end{split } \\
\begin{split } d_0^>(\omega ) & = -2\pi i \left [ { \langle n_{\text{ph}}\rangle } \delta(\omega + \omega_0)\right . \\ & \qquad \left . + ( { \langle n_{\text{ph}}\rangle } + 1 ) \delta(\omega-\omega_0 ) \right ] , \end{split } \\ \begin{split } d_0^r(\omega ) & = \frac{1}{\omega - \omega_0 + i\eta } \\ & \qquad - \frac{1}{\omega + \omega_0 + i\eta } , \quad \eta\rightarrow 0^+ \end{split } \\
d_0^a(\omega ) = [ d_0^r(\omega)]^*,\end{gathered}\ ] ] where @xmath241 is the averaged number of excitations in the vibration mode of frequency @xmath112 given by the bose - einstein distribution at temperature @xmath242 .
one can see that @xmath243 because @xmath244 unless @xmath245 which would be an odd case of study . and
since @xmath246 , one has @xmath247 furthermore the lesser and greater fock self - energies can be expressed in a more compact form : @xmath248 .
\end{split}\ ] ] the two - vibron process self - energies correspond to the two diagrams shown in figure [ fig : dx - dph ] , i.e. the so - called double exchange ( dx ) diagram and gw - like ( dressed phonon / vibron dph ) diagram in which the vibron propagator is renormalized by a single electron - hole bubble polarization .
the expressions for the self - energies for these diagrams are : @xmath249 and @xmath250 the factor @xmath251 comes from the series expansion of the exponential in the time evolution operator @xmath205 and the fact that one obtains 8 equivalent diagrams of the @xmath252 order . taking the steady state limit and after fourier transformation
, the different keldysh components of the double - exchange ( dx ) self - energy are given by @xmath253 the expression for the self - energy @xmath254 can actually be recast in a fock - like diagram with a renormalised vibron propagator @xmath255 @xcite : @xmath256 where the dressed vibron @xmath257 is given by @xmath258 and the polarization @xmath259 is given by the electron - hole bubble diagram : @xmath260 so , in principle , we already have all the ingredients to perform calculations using the fully dressed vibron propagator @xmath261 @xcite whose advanced and retarded components are given by @xmath262^{-1 } - \gamma_0 ^ 2\ \pi^{\{r , a\}}(\omega ) \right]^{-1 } \ .\ ] ] the vibron spectrum is renormalized by the polarization .
ther is a shift of the vibron energy related to @xmath263 and a finite linewidth of the peaks related to @xmath264 , instead of dirac @xmath265 peaks as in eq.([eq : d0_pvanddelta ] ) .
we have not performed the full vibron renormalization in the present work , because we want to study the effects of different diagrams of the same order , and order by order . @xmath261
correspond to a partial resummation of all the bubble diagrams and contains all ( even ) orders of the electron - vibron coupling .
the effect of this gw - like electron - vibron interaction will be presented in a forthcoming paper .
in this section we provide more details about how calculations for the self - energies are performed . for the imaginary part of the bare vibron gf and also by using a tiny , but finite value for @xmath193 ( 0.001 and 0.025 ) .
see main text for details .
the other parameters are @xmath148 ( electron transport ) , @xmath266 ( upper panel ) , @xmath267 .
for a small enough value of @xmath193 , one recovers the exact mcist as well as the @xmath268 negf results obtained from hilbert transform ( ht ) .
larger values of @xmath193 widen the peaks ; however the spectroscopic information remains .
for such values of @xmath193 , one can work with fewer grid points and makes second - order self - energy calculations tractable.,width=302 ] first , we start by using the following decomposition of the retarded and advanced vibron propagator @xmath269 : @xmath270 and the definition of the hilbert transformation ( h.t . ) : @xmath271 = \frac{1}{\pi } \mathrm{p.v . }
\int_{-\infty}^\infty dx \frac{f(x)}{x - y}.\ ] ] these expressions are in a form convenient for computation , since , as for @xmath274 , they involve the direct evaluation of @xmath275 on a energy grid shifted by @xmath276 in addition with the hilbert transform of @xmath40 and @xmath36 on a shifted energy grid .
the hilbert transformation is actually a conventional convolution product of the trial function with an appropriate @xmath277-like kernel .
calculation of convolution products can be made faster , instead of scaling as the square of the number of grid points , by the use of fft routines .
however since fft routines introduce artifical periodic boundary conditions , and one has to make sure that the functions , which decay slowly as @xmath278 ( the kernel of the hilbert transform , and real part of @xmath36 ) , have sufficiently small values at the boundaries of the energy grid .
otherwise , the corresponding discontinuities introduce spurious oscillations in the fft .
this means that one has to work with a wide energy grid and with a correspondingly large number of grid points to keep a good enough resolution .
working with a large number of grid points @xmath279 is not really a problem when calculating convolution products with fft , since the number of operations reduces from @xmath280 to @xmath281 , or even when without using fft routines .
however , it is not possible to reduce the calculation of the second - order dx self - energy to some sort of convolution product , and the number of operations then scales as @xmath282 .
such a scaling starts to make calculations seriously impractical when working with a large number of grid points .
hence , we have adopted another stategy which consists of introducing , in the definition of @xmath269 , a tiny but finite imaginary part @xmath283 instead of taking the limit @xmath268 .
this is just a natural step towards working with a fully dressed ( renormalized ) vibron propagator as already discussed in section [ app : dressing_d0 ] . by using this tiny but finite imaginary part @xmath283
, we avoid having to deal numerically with the dirac @xmath265-function and consequently with the hilbert transform and all the problems associated with the slowly decaying kernel .
we can then work with a less wide energy grid and fewer points while keeping a good energy resolution .
hence the calculations of the second order dx self - energy become more tractable .
of course , there is a price to pay for that : there is a lowest bound for the possible values of @xmath193 , and this lowest bound is strongly linked to the value of the grid spacing . by trial and error ,
we have found that @xmath193 must be at least equal to 2 or 3 times the grid spacing .
now we have to check and compare the results for the spectral functions obtained from these two different methods of calculation .
figure [ fig : app : specdens_mcistvsscbavsfullsc ] show the spectral functions for the off - resonant regime at equilibrium and for weak / intermediate electron - vibron coupling .
calculations have been performed with mcist at the hartree - fock level ( mcist - scba ) which serves as a reference calculation .
the negf - scba calculations performed within the limit @xmath268 by using hilbert transform and fft for the fock self - energy ( in the present case with a 131072 grid points ranging from -30 to + 30 ) are identical to the mcist - scba results as expected .
the negf - scba calculations performed with a tiny but finite imaginary part @xmath283 have been done for different sets of parameters .
only two are shown in figure [ fig : app : specdens_mcistvsscbavsfullsc ] : @xmath284 corresponding to a energy grid ranging from -10 to + 10 and with 32769 points , and @xmath285 corresponding to a energy grid ranging from -10 to + 10 and with 2049 points .
the spectral function obtained with @xmath284 is virtually identical to the corresponding `` exact '' calculations , while the spectral function obtained with @xmath285 has slightly broadened peaks with a reduced amplitude , even though such a calculation has been performed with only 2049 grid points which is in the range of good values of grid points to make calculation of the second - order dx self - energy numerically tractable . in any case ,
the most important point is that the spectral information ( peak positions ) are not dependent on the finite value of @xmath193 .
now we also have to verify the influence of @xmath193 on another transport regime for which we do not have any exact reference calculations , but for which we may expect to recover conventional properties of an interacting fermi liquid at very low temperatures .
figure [ fig : app : specdens_negfresscba_vs_etad0+nbpts ] shows the equilibrium spectral functions for the resonant regime .
calculations are performed with only the fock self - energy and for different values of @xmath193 ( 0.03 , 0.01875 , 0.00938 , 0.001875 , 0.000938 ) and for different number of energy grid points ( 2049 , 32769 , 65537 , 131073 ) .
the energy grid is ranges from -10 to + 10 .
as expected , the main central peak and the vibron side - band peak narrow with increasing amplitude when the value of @xmath193 decreases , until reaching an asymptotic behaviour for @xmath286 .
( 0.03 , 0.01875 , 0.00938 , 0.001875 , 0.000938 ) and for different number of energy grid points ( 2049 , 32769 , 65537 , 131073 ) .
the energy grid is ranging from -10 to + 10 .
the inset shows a zoom around @xmath287 of the rescaled spectral functions , giving then the linear conductance @xmath185 at @xmath287 .
the other parameters are @xmath288 , @xmath289 , @xmath290 .
peaks in the spectral functions narrow with decreasing values of @xmath193 as expected .
and one recovers a perfect linear conductance @xmath291 at @xmath196.,width=302 ] the most interesting result is given by the behaviour of spectral function at the equilibrium fermi level @xmath292 .
such a behaviour is exemplified in the inset of figure [ fig : app : specdens_negfresscba_vs_etad0+nbpts ] .
the inset shows the spectral function rescaled by @xmath293 so that the value at @xmath287 corresponds to the linear conductance @xmath294 ( see section [ sec : linear_reponse ] ) .
one can see that @xmath185 goes towards the unit of conductance as @xmath295 .
such a limit corresponds to a perfect conductance , as obtained for the corresponding non - interacting system .
this result shows that even in the presence of electron - vibron interactions , the linear conductance is perfect , as one also obtains for the non - interacting case . the electron - vibron interaction ( as least at the fock level )
does not renormalize the linear conductance . for the fully electron - hole symmetric system ( no hartree self - energy )
, we have checked that the value of @xmath296 is zero at @xmath287 and at @xmath196 ( at equilibrium ) .
the real part of the interacting self - energy @xmath297 at @xmath287 by definition because it is an odd function of @xmath172 .
our electron - vibron interacting system conserves its fermi - liquid properties at zero temperature and equilibrium , as also obtained for interacting electron systems ( see for example refs . [ ] ) .
we expect that such a behaviour also hold for other electron - vibron diagrams than the fock diagram . out of equilibrium
, the system may lose its conventional fermi - liquid properties , as we have already shown in the main part of the paper . in conclusion , we want to stress that by introducing a finite , but tiny , value for the imaginary part @xmath193 , one might not get the exact linear conductance value .
however this is a numerical artifact which can be controlled by increasing the number of energy grip points and reducing the value of @xmath193 .
nonetheless we expect that the relative effects of different electron - vibron diagram are correctly obtained from our calculations , whatever the small values of @xmath193 are .
and of course , in order to perform a correct analysis of such effects , one should only compare results for either spectral functions or linear conductance obtained numerically with the same value of @xmath193 .
we would like to thank martin stankovski for useful discussions of the second - order diagrams and ulf von barth for discussion of the negative dos problem .
this work was funded in part by the european community s seventh framework programme ( fp7/2007 - 2013 ) under grant agreement no 211956 ( etsf e - i3 grant ) .
rwg was also supported in part by the national science foundation under grant no.phy05-51164 . | we consider the interaction between electrons and molecular vibrations in the context of electronic transport in nanoscale devices .
we present a method based on non - equilibrium to calculate both equilibrium and non - equilibrium electronic properties of a single - molecule junction in the presence of electron - vibron interactions .
we apply our method to a model system consisting of a single electronic level coupled to a single vibration mode in the molecule , which is in contact with two electron reservoirs .
higher - order diagrams beyond the usual self - consistent born approximation ( scba ) are included in the calculations . in this paper
we consider the effects of the double - exchange diagram and the diagram in which the vibron propagator is renormalized by one electron - hole bubble .
we study in detail the effects of the first- and second - order diagrams on the spectral functions for a large set of parameters and for different transport regimes ( resonant and off - resonant cases ) , both at equilibrium and in the presence of a finite applied bias .
we also study the linear response ( linear conductance ) of the nanojunction for all the different regimes .
we find that it is indeed necessary to go beyond the scba in order to obtain correct results for a wide range of parameters . |
the huge effort on the experimental side at present b facilities ( babar , belle and cdf ) to increase the precision on data measurements force us to revise the strategies on the theory side to produce more accurate predictions .
non leptonic b decays offer different strategies to determine the unitarity triangle , to search for new physics ( np)@xcite but also to rule out models @xcite .
while a lot of attention has been devoted to the @xmath7@xcite decay modes , here we will focus on @xmath8 decays that has been observed at cdf @xcite ( @xmath3 ) and at babar @xcite and belle @xcite ( @xmath1 ) .
there are two main approaches in the literature to describe @xmath9 decays : flavour symmetries and @xmath10expansion methods ( qcd factorisation @xcite/ soft collinear effective theories @xcite or pqcd @xcite ) .
each of those methods has pros and cons , that we will discuss in turn .
flavour symmetries , like u - spin symmetry that relates @xmath11 with @xmath12 @xcite , provide a model independent analysis and extract most of the needed hadronic parameters from data . however , this method has the disadvantage that it relies strongly on the accuracy of data , and , at present , there is still not full agreement between babar and belle data on the cp asymmetries of the @xmath13 mode . as a consequence ,
error bars are still quite large , see for instance , the prediction @xmath14 @xcite or , more recently , @xmath15 @xcite .
also when relating @xmath13 with @xmath11 some of the needed u - spin parameters can only be roughly estimated and they are usually taken to be of the order of @xmath16 .
concerning @xmath17-expansion methods , here we will focus on qcd factorisation ( qcdf ) @xcite .
the main idea is to exploit the existence of a large scale @xmath18 together with colour transparency , that applies when the outgoing meson that does not contain the spectator quark is very energetic . at leading power in @xmath19
all long distance contributions can be parametrized in terms of form factors and light cone distribution amplitudes , while the contribution from energetic gluons comes in a perturbative series in @xmath20 and it is incorporated into the hard scattering kernels .
qcdf predicts some of the hadronic parameters reducing the error bars , however in the computation one has to face chirally enhanced ir divergences .
they are formally suppressed by a power of @xmath17 , but can be numerically significant .
they are modelled and induce an important uncertainty in the predictions .
however , there is a third possibility and it is the proposal presented in @xcite that combines qcdf and flavour symmetries giving rise to rather accurate predictions for the branching ratios of the above mentioned @xmath0 decays in sm and in supersymmetry .
moreover , the method predicts some of the su(3 ) breaking parameters which can be useful for other flavour approaches and , at the same time , deals with the problem of the chirally enhanced ir divergences coming at order @xmath19 ( see also @xcite ) .
since ir divergences play a central role in this discussion , it is worth to mention the two sources of ir divergences in qcdf : * hard spectator - scattering : hard gluons exchange between spectator quark and the outgoing energetic meson gives rise to integrals of the following type ( see @xcite for definitions ) : @xmath21 , \nonumber\ ] ] where the second term ( formally of order @xmath19 ) diverges when @xmath22 . *
weak annihilation : these type of diagrams also exhibit endpoint ir divergences as it is explicit in the corresponding integrals : @xmath23 + r_\chi^{m_1 } r_\chi^{m_2}\,\phi_{m_2}(x)\,\phi_{m_1}(y)\ , \frac{2}{\bar x y } \right\}. \end{aligned}\ ] ] both divergences are modelled in the same way @xcite : @xmath24 \!\equiv\ ! \phi_{m_1}(1)\,x_{h , a}^{m_1 } + { r , } \nonumber\ ] ] where @xmath25 is a finite piece and the divergent piece is cut - off by a physical scale of order @xmath26 with an arbitrary complex coefficient to take into account possible multiple soft scattering : @xmath27 .
the sm amplitude for a @xmath29-decay into two mesons can be split into tree and penguin contributions @xcite and penguin " the piece proportional to @xmath30 , even if applied to decays with no actual tree diagram . ] : @xmath31 with @xmath32 denoting the charge of the decay products , and the products of ckm factors @xmath33 .
tree and penguin contributions in @xmath34 in qcdf are : @xmath35 where @xmath36 , @xmath37 , the super - scripts identify the channel , the normalisation is @xmath38 ( see @xcite for the corresponding expression of the @xmath0 channels ) . following the observation in @xcite that the structure of the ir divergences is the same , independently of the charm or up quark running in the loop , we identified an ir - safe quantity at nlo in qcdf that we called @xmath39 .
all chirally enhanced ir divergences cancel exactly in this quantity at this order .
its explicit expression in terms of the coefficients in eq.([eq4 ] ) is : @xmath40,\ ] ] where electroweak contributions are neglected .
this quantity can be safely evaluated in qcdf and the result found in @xcite was : @xmath41 .
the largest uncertainty entering @xmath42 comes from the ratio @xmath43 and the scale dependence .
interestingly , this quantity can be expressed in terms of observables , providing a relation between the direct induced cp - asymmetry ( @xmath44 ) , the mixing induced cp - asymmetry ( @xmath45 ) and the branching ratio ( @xmath46 ) of @xmath34 : @xmath47 a_{mix}^{d0 } -[x_2 \cos\phi_d + x_3 \sin\phi_d ] a_{\delta}^{d0 } \}\ , , \nonumber\ ] ] where @xmath48 , @xmath49 is the weak mixing angle for the @xmath50 system , @xmath51 and @xmath52 are functions of @xmath53 and ckm elements .
all sm inputs are taken as in @xcite following @xcite .
moreover , it was found in @xcite that this sum - rule encodes also a very interesting information .
it provides a new way of measuring @xmath54 : @xmath55 ( see @xcite for a recent review on the extraction of @xmath28 ) . in a similar way , it was found in @xcite a corresponding ir safe quantity @xmath56 and sum - rule for the decay @xmath57 : @xmath58 a_{mix}^{s0 } -[y_2 \cos\phi_s + y_3 \sin\phi_s ] a_{\delta}^{s0 } \}\ , , \nonumber\ ] ] which provides a completely new way to determine the weak mixing angle @xmath5 that we present here ( see @xcite for definitions ) : @xmath59 this implies that a measurement of the branching ratio , direct cp asymmetry and mixing induced cp asymmetry of the decay @xmath60 automatically translates into a value for @xmath61 .
finally , given the relation @xmath62 , where @xmath63 a new relation between @xmath54 and @xmath64 immediately emerges .
the sm amplitude of this @xmath66 penguin decay is given by : @xmath67 its dynamics is described in terms of three parameters : @xmath68 , @xmath69 and the relative strong phase @xmath70 ( remember that @xmath71 stands for the piece proportional to @xmath72 but it is not due to an actual tree diagram in this case ) .
its hadronic parameters can be related via u - spin with the hadronic parameters ( @xmath73 , @xmath74 and @xmath75 ) of the also penguin governed mode @xmath76 .
this has several advantages : first , we can expect similar final state interactions ( although not equal ) , second , the sources of u - spin breaking can be better controlled using qcdf .
these sources are : i ) the factorisable ratio @xmath77 ( extrapolated from the lattice ) ii ) u - spin breaking @xmath17 suppressed terms @xmath78 and @xmath79 : sensitive to the difference of @xmath50 and @xmath6 distribution amplitudes and spectator quark dependent contributions coming from a gluon emitted from the @xmath80 or @xmath81 quark .
this leads to the relations @xmath82 and @xmath83 .
the next step is to determine the hadronic parameters ( @xmath84 , @xmath85 ) of the decay @xmath86 .
this is done using as inputs the @xmath87 and @xmath42 ( from qcdf ) .
the direct cp asymmetry @xmath88(denoted by @xmath44 ) will be taken as a free parameter .
the combination of those constraints gives rise to a set of non - linear equations ( see definitions in @xcite ) : @xmath89-[{\sin\gamma x_{\delta}&=&y_{\delta}x_p - { \rho_0 ^ 2 a_{dir}^{d0}}/({2|\lambda_u^{(d)}\lambda_c^{(d)}|\sin\gamma})\ , , \end{aligned}\ ] ] that determines @xmath90 , then using @xmath42 one gets @xmath84 .
two remarks are important here : first , there is a twofold ambiguity in the sign of @xmath91 ( solved in the next subsection ) .
second , current data together with our knowledge on @xmath42 limits the @xmath44 asymmetry ( by means of eqs.([circ ] ) ) within a restricted range between @xmath92 .
finally , our sm predictions for the branching ratio and cp asymmetries of @xmath93 are obtained using the relations between @xmath94 and @xmath95 mentioned above and including all u - spin breaking sources together with the qcdf uncertainties in @xmath42 .
the resulting predictions are @xmath96 , @xmath97 and @xmath98 .
the sign of these asymmetries can be fixed once @xmath44 will be measured with enough accuracy .
the analysis of the decay mode @xmath11 follows similar steps , but with some important differences : i ) we will use u - spin and isospin to connect @xmath1 with @xmath11 , ii ) @xmath3 contains a tree ( denoted by @xmath100 in qcdf ) with no counterpart in @xmath101 , iii ) @xmath102 has been measured with excellent precision at cdf@xcite and iv ) we will use the information only on the _ sign predictions _ ( not the absolute value ) for the cp - asymmetries of @xmath3 from the strategy that uses u - spin to relate @xmath3 with @xmath12 .
the hadronic parameters describing the amplitude : @xmath103 are obtained from the relations containing the sub - leading @xmath17 u - spin breaking : @xmath104 , @xmath105 .
those errors , estimated within qcdf , are stretched roughly by a factor two to be conservative .
the two - fold ambiguity on the sign of @xmath106 is lifted here using u - spin arguments based on the @xmath107 strategy . while the signs of @xmath91 and @xmath108 are correlated ( being both positive or negative ) , the prediction based on the @xmath107 strategy points towards a positive sign for @xmath108@xcite , discarding then the solution with @xmath109 .
our result in the sm for the branching ratio of @xmath3 averaging over all values of @xmath44 is @xmath110 , where the last error in parenthesis stands for a rough estimate of finite , non - enhanced @xmath19 corrections . finally , confronting our predictions for @xmath3 with the data on @xmath13 @xcite : @xmath111 and @xmath112 , @xmath113 , provides a double information .
first , we can give predictions for the u - spin breaking parameters : @xmath114 and @xmath115 connecting @xmath3 with @xmath13 .
these parameters can be compared with the qcd sum rules predictions in ref .
notice that while qcd sum rules gives only the factorizable part , our predictions include , in principle , the full contribution .
second , a comparison between the two relative strong phases @xmath116 and @xmath117 selects @xmath118 . then , if we restricts only to positive values of @xmath44 according to the previous arguments , our sm predictions turn out to be @xcite : @xmath119 another argument in favour of @xmath118 comes from the preference of @xmath120 of the u - spin based @xmath13 strategy @xcite ( see the anti - correlation between @xmath44 and @xmath121 in table 1 of @xcite ) .
the accuracy on these cp - asymmetries will be substantially improved once a precise measurement on @xmath44 will be available or the error of @xmath122 and the qcd uncertainties on @xmath42 ( mainly @xmath43 and scale dependence ) will be reduced . the leading gluino - squark box and penguin contributions @xcite to @xmath60 and @xmath3
were evaluated first in @xcite using the u - spin flavour strategy with @xmath123 and , afterwards , in @xcite using the new method combining flavour and qcdf @xcite .
the relative size of this contribution compared to the sm penguin is @xmath124 .
the amplitude of these decays in presence of np contains an extra contribution : @xmath125 , @xmath126 .
these np amplitudes @xmath127 and @xmath128 in terms of wilson coefficients are : @xmath129a\nn\end{aligned}\ ] ] with @xmath130 , @xmath131 and @xmath132 ( see @xcite for definitions ) .
these wilson coefficients are sensitive to the @xmath133 mass splitting . after including the constraints coming from @xmath134 , @xmath7 and @xmath135 we found that a large isospin violation controlled by the mass splitting @xmath136-@xmath137 is possible between the np amplitudes @xmath138 and @xmath139 . for the region of parameters considered in this supersymmetric scenario , @xmath138 can be up to a factor three larger than @xmath139 .
the specific results in supersymmetry for each decay mode are @xcite : * @xmath140 : the branching ratio is very little affected by susy . at most , the sm prediction can be increased by 15% for @xmath141 , increasing a bit the already good agreement with the new cdf data @xcite .
the direct cp asymmetry within susy falls inside the range @xmath142 for @xmath143 .
the deviation depends on the relative size of the competing sm tree versus the np amplitude .
@xmath144 can take any value from [ -1,1 ] .
* @xmath145 : the impact of susy on @xmath146 is even smaller , reflecting the reduced allowed region for @xmath147 as compared to @xmath138 . the situation is very different for the cp asymmetries , that are particularly promising , due to the very small size of their sm prediction @xcite . the direct cp asymmetry in susy can be 10 times larger than the sm one .
@xmath148 covers the entire range , and so this asymmetry can be large in the presence of susy , contrary to the sm prediction .
the method discussed here is being applied to other non - leptonic @xmath29-decays .
_ acknowledgements _ : research partly supported by eu contract euridice ( hprn - ct2002 - 00311 ) , pnl2005 - 41 , fpa2002 - 00748 and the ramon y cajal program .
y. nir , hep - ph/0109090 .
m. ciuchini , e. franco , a. masiero and l. silvestrini , j. korean phys .
* 45 * ( 2004 ) s223 ; a. j. buras , hep - ph/0505175 ; r. fleischer , hep - ph/0505018 ; hep - ph/0608010 . t. hurth , aip conf .
proc . * 806 * , 164 ( 2006 ) ; w. s. hou and m. nagashima , hep - ph/0602124 .
for instance , the lrsm was ruled out in : p. ball , j. m. frere and j. matias , nucl .
b * 572 * , 3 ( 2000 ) .
for a @xmath66 penguin analysis of this model see also g. barenboim et al .
d * 60 * ( 1999 ) 016003 .
y. grossman , m. neubert , a. l. kagan , jhep * 9910 * ( 1999 ) 029 .
j. matias , phys .
b * 520 * ( 2001 ) 131 .
a.j.buras , r.fleischer , s.recksiegel , f.schwab , eur .
j. c * 45 * , 701 ( 2006 ) ; acta phys .
b * 36 * , 2015 ( 2005 ) .
s. baek , p. hamel , d. london , a. datta and d. a. suprun , phys .
d * 71 * ( 2005 ) 057502 .
m. morello , hep - ex/0612018 ; d. tonelli , fermilab - thesis-2006 - 23 . update of a previous paper : a. abulencia
_ et al . _
[ cdf collaboration ] , phys .
* 97 * ( 2006 ) 211802 and d. tonelli [ cdf collaboration ] , pos * hep2005 * ( 2006 ) 258 .
b. aubert _ et al . _ [ babar ] , phys . rev .
lett . * 97 * , 171805 ( 2006 ) . updated version of b. aubert _ et al .
_ [ babar ] , phys .
* 95 * ( 2005 ) 221801 .
k. abe _ et al . _ [ belle ] , hep - ex/0608049 .
updated version of h. c. huang [ belle ] , hep - ex/0205062 .
m.beneke , g.buchalla , m.neubert and c.t.sachrajda , nucl .
b * 591 * ( 2000 ) 313 ; nucl .
b * 606 * ( 2001 ) 245 .
m. beneke and m. neubert , nucl.phys.b * 675 * ( 2003 ) 333 .
et al . _ , hep - ph/0510241 ; a. r. williamson and j. zupan , phys .
d * 74 * ( 2006 ) 014003 [ erratum - ibid .
d * 74 * ( 2006 ) 03901 ] .
r. fleischer , phys .
b * 459 * ( 1999 ) 306 ; r. fleischer and j. matias , phys .
d * 61 * ( 2000 ) 074004 ; phys .
d * 66 * ( 2002 ) 054009 ; m. gronau and j. l. rosner , phys .
d * 65 * ( 2002 ) 113008 ; j. matias , eur .
j. c * 33 * ( 2004 ) s364 ; a. s. safir , jhep * 0409 * , 053 ( 2004 ) ; r. fleischer , j. phys .
g * 32 * ( 2006 ) r71 c. w. chiang and y. f. zhou , jhep * 0612 * ( 2006 ) 027 . | the interplay between flavour symmetries connecting @xmath0 decays with the recently measured @xmath1 decay and qcd factorisation opens new strategies to describe the decays @xmath2 and @xmath3 in the sm and in supersymmetry .
a new relation , emerging from the sum - rule for the @xmath4 decay mode , is presented offering a new way to determine the weak mixing angle @xmath5 of the @xmath6 system . |
exotic superconductivity with high-@xmath6 has been often discovered in doped layered compounds , such as cuprates@xcite , @xmath10-@xmath11ncl ( @xmath11=zr , hf)@xcite , and iron pnictides@xcite .
recently , materials with bis@xmath0 layers have attracted much attention as a new family of layered superconductors .
superconductivity was first reported in bi@xmath12s@xmath12o@xmath13@xcite ( @xmath14 k ) , followed by @xmath9o@xmath1f@xmath2bis@xmath0 ( @xmath9=la , ce , pr , nd and yb)@xcite , where the maximum @xmath6 reaches @xmath1510 k around @xmath7=0.5 for @xmath9=la .
a characteristic crystal structure common to these materials is the bis@xmath0 layers consisting of the bi and s ions positioned alternately on the square lattice , which are separated by insulating blocking layers .
the superconductivity has been believed to reside in the bis@xmath0 layer , when electron - type carriers are doped . for the @xmath9o@xmath1f@xmath2bis@xmath0 compounds , for instance ,
the partial chemical substitution of o ions with f ions in the blocking layer allows the tuning of electron concentration in the bis@xmath0 layer and hence the critical temperature .
such a stacking structure with mixed anions is reminiscent of so - called 1111-type fe - based superconductors.@xcite the variety of the blocking layer has been rapidly increasing , which includes sr@xmath16la@xmath16fbis@xmath0@xcite , la@xmath17obis@xmath0 ( @xmath18=ti , zr , hf , and th)@xcite , and bi@xmath13o@xmath0s@xmath13@xcite .
theoretically , the undoped bis@xmath0-based material was predicted to be a band insulator .
first - principles calculation suggested that the conduction band primarily consists of bi 6@xmath19 and 6@xmath20 orbitals arising from the bis@xmath0 layer , hybridized with s 3@xmath4 orbitals.@xcite assuming electron doping within a rigid - band scheme , the two - orbital model was suggested as a minimal model that considers only the bi 6@xmath4 orbitals as the relevant bands crossing the fermi surface.@xcite it was predicted that the topology of the fermi surface varies depending on the carrier concentration . in particular , hole pockets appear around the @xmath21 ( 0,0 ) and @xmath18 ( @xmath22,@xmath22 ) points , when the doping level is increased up to @xmath150.5 .
there , the fermi surfaces exhibit nesting and the associated spin fluctuation was proposed as a candidate of the superconducting pairing.@xcite on the other hand , some theoretical works indicated the electron - phonon coupling as another pairing mechanism@xcite , which is enhanced in the vicinity of the ferroelectric and/or charge - density - wave transitions@xcite .
however , this has not been supported by experiments yet.@xcite the superconducting phase diagram was experimentally revealed for @xmath9o@xmath1f@xmath2bis@xmath0 ( @xmath9=la , ce , and nd).@xcite except for the difference in @xmath6 , the overall features of variation in @xmath6 are similar for all the compounds ; a broad dome - shaped superconducting phase insensitive to the carrier density manifests itself around @xmath7=0.5 .
however , the behavior of @xmath6 around the phase boundary to a non - superconducting phase has been still ambiguous
. the carrier concentration above which the value of @xmath6 is non - zero appears to significantly differ depending on the materials ; it is approximately @xmath7=0.1 for @xmath9=la , whereas @xmath7=0.25 for @xmath9=ce .
there have been also several complicated transport properties in addition to the emergence of the superconductivity . for lao@xmath1f@xmath2bis@xmath0 , while the resistivity slightly decreases with increasing @xmath7 up to 0.2 , it starts to steeply increase with further increasing @xmath7 and exhibits semiconducting behavior above @xmath7=0.5.@xcite for ceo@xmath1f@xmath2bis@xmath0 , on the other hand , the pristine compound is a bad metal and the resistivity monotonically increases with increasing @xmath7.@xcite such a puzzling behavior may be attributed to the change in band structure via the partial substitution of f ions for o ions . it was recently suggested that the band gap is substantially enlarged by fluorine doping , since the top of the valence band mainly originates from o 2@xmath4 orbitals , which tends to sink as fluorine is doped.@xcite to reveal the intrinsic dependence of @xmath6 on the carrier density , it is essential to afford the carriers to the bis@xmath0 layer without modifying the original band structure .
for this purpose , we have here focused on the recently - synthesized srfbis@xmath0 as a pristine compound.@xcite importantly , a first - principles calculation predicted the top of the valence band consists mainly of the s 3@xmath4 orbitals , where the density of states of the sr and f orbitals in the blocking layer is negligibly small .
this would provide an ideal arena to realize the rigid - band carrier doping into the bis@xmath0 layer by substituting sr ions with la ones .
very recently , the superconductivity was actually discovered for sr@xmath16la@xmath16fbis@xmath0@xcite , which has put a strong emphasis upon the importance of the extensive study about the dependence on carrier concentration . in this study
, we aim to elucidate the genuine phase diagram associated with the superconductivity residing in the bis@xmath0 layer .
we have systematically synthesized polycrystalline sr@xmath1la@xmath2fbis@xmath0 ( @xmath3 ) samples and revealed the detailed electronic properties of both the superconducting and normal states by performing the x - ray diffraction , x - ray absorption , transport and thermodynamic measurements .
the obtained phase diagram is characterized by an insulator - superconductor transition with a steep phase boundary at the critical concentration around @xmath7=0.4 .
this is markedly different from that for @xmath9o@xmath1f@xmath2bis@xmath0 , indicating the strong impact of the blocking layer on the superconductivity .
unusual enhancement of the superconducting coupling strength as well as @xmath6 has been suggested as the carrier concentration is reduced toward the critical point .
we discuss these superconducting properties in comparison with the li - intercalated zrncl compound .
sr@xmath1la@xmath2fbis@xmath0 polycrystals ( @xmath23 ) were synthesized by the following steps of solid state reaction and high - pressure annealing .
first , the stoichiometric mixtures of srf@xmath0 , srs , laf@xmath13 , las@xmath13 , bi and s powders were pressed into pellets and heated in an evacuated quartz tube at 600@xmath24c for 10 h. for @xmath7=0 , the starting materials were the stoichiometric srf@xmath0 , srs , and the prereacted bi@xmath0s@xmath13 powders .
the pellets for @xmath25 were ground in an ar - filled globe box and sealed in a gold capsule .
a small amount of titanium powder was added at the both ends of the sample to avoid oxidization .
then , the capsule was treated at @xmath152.0 gpa and @xmath15600@xmath24c for 1h , using a conventional cubic anvil - type high - pressure apparatus .
the obtained compounds were characterized at room temperature by 2@xmath26-@xmath26 powder x - ray diffraction ( xrd ) using cu - k@xmath27 radiation .
the x - ray absorption spectroscopy measurements at the bi @xmath28 edge were carried out at room temperature by synchrotron radiation at the beamline bl01b1 of spring-8 , japan .
the magnetization was measured with a superconducting quantum interference device ( quantum design ) .
the four - probe resistivity ( @xmath29 ) , hall coefficients ( @xmath30 ) , and specific heat ( @xmath31 ) were measured using physical property measurement system ( quantum design ) . for the @xmath31 measurements at 0 t ( superconducting state ) and 9 t ( normal state ) ,
the thermometers on the calorimeter puck were calibrated at both fields .
the addenda signal was determined before mounting the sample .
figure [ fig : xrd](a ) displays the xrd profiles for sr@xmath1la@xmath2fbis@xmath0 ( @xmath32 ) at room temperature .
all the main peaks are assigned to bragg reflections calculated for the tetragonal space group @xmath33 , which is consistent with the previous reports.@xcite while minute impurity peaks arising from bi@xmath0s@xmath13 are discernible for all @xmath7 , those from bi metal impurity show up for @xmath34 .
note that the amount of both impurity phases largely increases for @xmath7=0.6 , indicating the la solubility limit may be almost reached . in fig .
[ fig : xrd](b ) , we show the @xmath7 dependence of lattice constants @xmath35 ( in - plane ) and @xmath36 ( out - of - plane ) estimated by le bail analysis .
the @xmath35 value is almost unchanged as a function of @xmath7 , whereas the @xmath36 value systematically decreases with increasing @xmath7 , which indicates the formation of solid solution . for @xmath34
, however , the reduction in @xmath36 with an increase in @xmath7 starts to saturate .
this again implies the solubility limit close to @xmath37 in the present compounds .
( color online ) ( a ) x - ray diffraction profile at room temperature for sr@xmath1la@xmath2fbis@xmath0 ( [email protected] ) . @xmath39 and @xmath40 denote the impurity peaks of bi@xmath0s@xmath13 and bi metal , respectively .
the numbers are miller indices based on the space group @xmath33 .
( b ) lattice constants @xmath35 and @xmath36 as a function of @xmath7 ( la concentration ) .
the dashed lines are guides to eyes . ] to clarify the variation in valence state of bi ion with la substitution , we have carried out the x - ray absorption spectroscopy ( xas ) .
figure [ fig : xas](a ) shows the xas spectra for @xmath7=0 , 0.25 , 0.45 and 0.6 around the bi @xmath28 ( @xmath41 ) edge , which should be sensitive to the change in the 6@xmath4 states caused by electron doping into bi@xmath42 ions .
the data for pure bi are also displayed as a reference corresponding to the zero - valence state .
each bi @xmath28 xas spectrum consists of a clear single peak . as shown in the inset of fig .
[ fig : xas](a ) , the intensity of the peak gradually decreases with increasing @xmath7 , reflecting the reduction in empty @xmath43 states . for quantitative comparison , we plotted in fig .
[ fig : xas](b ) the integrated intensity of the bi @xmath28 edge for each @xmath7 versus the nominal bi valence , i.e. , @xmath44 . we here assume that la@xmath42 substitution for sr@xmath45 affords one electron to the bis@xmath0 layer .
the integrated intensity deceases almost linearly with increasing ( decreasing ) @xmath7 ( nominal bi valence ) and the linear extrapolation coincides with the data for pure bi .
this can be a firm evidence that electrons are effectively transferred into the bis@xmath0 conducting layer by la substitution in the srf blocking layer .
( color online ) ( a ) bi @xmath28 ( @xmath41 ) edge x - ray absorption spectra for sr@xmath1la@xmath2fbis@xmath0 ( @xmath7=0 , 0.25 , 0.45 and 0.6 ) and pure bi .
the inset shows the enlarged view around the maximum of the peak structure .
( b ) the integrated intensities of the @xmath28 edge spectra as a function of nominal bi valence ( @xmath44 ) .
the data for bi metal is plotted as the zero - valence state .
the error bars denote the uncertainty in integration range . ]
figure [ fig : rt ] presents the temperature profile of resistivity at 0 t for sr@xmath1la@xmath2fbis@xmath0 ( @xmath32 ) .
the undoped ( @xmath7=0 ) compound shows insulating behavior in the entire temperature region . above 150 k
, the resistivity exhibits the thermal - activation type temperature dependence with the activation energy of @xmath1525 mev . with increasing la concentration ( @xmath7 ) , the resistivity systematically decreases . for @xmath7=0.4 and 0.42 ,
the resistivity value at 300 k is significantly reduced ( by nearly two orders of magnitude ) , but it still increases with decreasing temperature . at low temperatures below 25 k , in particular , the conductivity obeys the @xmath46 dependence , which signals the weak anderson localization effect.@xcite the sample with @xmath7=0.45 exhibits metallic behavior above 200 k , while it robustly shows a weak upturn at low temperatures probably due to the localization effect . for @xmath7=0.5 and 0.6 ,
the overall feature of resistivity is similar to that for @xmath7=0.45 , although the change in absolute value is not systematic ; the resistivity value for @xmath7=0.6 is slightly larger than those for @xmath7=0.45 and 0.5 , which may be owing to the larger volume fraction of impurity phases for @xmath7=0.6 [ fig .
[ fig : xrd](a ) ] .
a clear superconducting transition is discernible at around 3 k for @xmath47 ( see fig . [
fig : rtmt ] for details ) .
note here that this value of @xmath7 nicely corresponds to the insulator - metal transition of the normal state .
in fact , the conductivity of the normal state at zero temperature is non - zero for @xmath5 ( and nearly zero for @xmath7=0.42 ) , when the temperature profile above @xmath6 is fitted to the @xmath46 dependence [ see also fig . [
fig : phase](b ) ] .
thus , an insulator - to - metal transition takes place upon electron doping in the bis@xmath0 layer and the superconducting transition occurs in the doping - induced metallic phase at low temperatures .
( color online ) temperature dependence of resistivity for sr@xmath1la@xmath2fbis@xmath0 ( [email protected] ) at 0 t. ] to reveal the details of the superconducting state for the present system , we show in figs .
[ fig : rtmt](a ) and ( b ) the temperature dependence of dc magnetic susceptibility and resistivity below 4 k , respectively .
the magnetization was measured at 0.1 mt after zero - field cooling .
all the superconducting samples except for @xmath7=0.42 exhibit substantial diamagnetic responses at the lowest temperature .
the apparent shielding fractions for @xmath5 exceed 60% at 1.8 k without demagnetizing - field correction , which ensures that the superconductivity is a bulk property . as shown in the inset to fig .
[ fig : rtmt](a ) , the temperature where the diamagnetic signal starts to steeply evolve well coincides with the zero - resistivity temperature for each @xmath7 ( denoted as a dashed vertical line ) .
the diamagnetic signal for @xmath7=0.42 would be significant below 1.8 k , since the resistivity is still on the way of decreasing to zero at 1.8 k. the @xmath7 dependence of @xmath6 is noticeable in the resistivity data shown in fig .
[ fig : rtmt](b ) .
the highest @xmath6 was observed for @xmath7=0.45 , where the zero - resistivity temperature is @xmath153.0 k. with decreasing @xmath7 from 0.45 , @xmath6 rapidly drops and a clear superconducting transition disappears for @xmath7=0.4 . when @xmath7 is increased from @xmath7=0.45 , on the other hand , the superconducting state robustly remains up to @xmath7=0.6 with the zero - resistivity temperature gradually decreasing down to @xmath152.8 k. we here note that the resistivity for @xmath5 shows barely metallic behavior above @xmath6 . below @xmath156 k , in fact , the temperature profile of resistivity deviates from the @xmath48 dependence characteristic of weak localization , which might be caused by the superconducting fluctuation and/or disorder effects inherent in polycrystalline specimens .
( color online ) temperature dependence of ( a ) dc magnetic susceptibility ( at 0.1 mt after zero field cooling ) and resistivity ( at 0 t ) for sr@xmath1la@xmath2fbis@xmath0 ( [email protected] ) below 4 k. the resistivity is normalized by the value at 6 k for each compound .
inset in ( a ) is the enlarged view close to @xmath6 .
the vertical dashed lines correspond to the zero - resistivity temperature for each superconducting compound . ] in fig .
[ fig : phase](a ) , we summarize the electronic phase diagram for the sr@xmath1la@xmath2fbis@xmath0 system as functions of @xmath7 and temperature .
the values of @xmath6 corresponding to the zero - resistivity and the mid - point of the transition are defined as the temperatures where the resistivity is 10% and 50% of the normal - state resistivity , respectively .
the superconducting phase manifests itself above @xmath7=0.4 , where @xmath6 shows unusual @xmath7 dependence .
as @xmath7 is reduced from 0.6 to 0.45 , the zero - resistivity @xmath6 gradually increases from 2.8 k to 3.0 k. with further decreasing @xmath7 toward the phase boundary to the insulator , @xmath6 steeply drops and the superconducting transition disappears at @xmath7=0.4 .
the resultant phase diagram is characterized by an insulator - superconductor transition with a sharp phase boundary at approximately @xmath7=0.4 , which significantly differs from the broad superconducting dome observed for @xmath9o@xmath1f@xmath2bis@xmath0 ( @xmath9=la , ce , and nd).@xcite the critical concentration for the present compound ( @xmath8 ) is also much higher than those for @xmath9o@xmath1f@xmath2bis@xmath0 ( @xmath49 - 0.25 ) .
the difference in blocking layer thus strongly influences the emergence of the superconducting phase .
as mentioned in the introductory part , this may be caused by the modification of the band structure upon the chemical substitution in the blocking layer .
since the electronic states arising from the srf blocks are situated away from the fermi energy@xcite , the present result is regarded as intrinsic phase diagram of the electron - doped bis@xmath0-layer compound with a rigid band .
another important feature is the doping - induced insulator - metal transition in the normal state , as mentioned above ( fig .
[ fig : rt ] ) . to make it clear , in fig .
[ fig : phase](b ) , we show the conductivity for the normal state at 4 k versus @xmath7 . for @xmath50 ,
the conductivity is virtually zero , whereas it is non - zero and continuously increases with increasing @xmath7 above @xmath7=0.4 . a sudden decrease in conductivity for
@xmath7=0.6 may reflect the enhanced volume fraction of the impurity phases . in the normal state , thus , the insulating phase is replaced above @xmath7=0.4 by the metallic one that condenses into the superconducting state at low temperatures .
it is worth noting that similar phase diagram was reported for the li@xmath2zrncl system.@xcite this compound shows an insulator - metal transition upon li intercalation , although the critical carrier concentration is much lower ( @xmath150.06 ) than the present system . in the superconducting phase , furthermore , @xmath6 increases with reducing carrier density on the verge of insulator - superconductor transition , similarly to the present system .
as an origin of such an increase in @xmath6 , it has recently been proposed that the pairing interaction strength is enhanced with approaching a band insulator.@xcite interestingly , such a tendency is also likely in the sr@xmath1la@xmath2fbis@xmath0 compounds , as has been indicated by the thermodynamic measurements ( @xmath51 @xmath52 ) .
( color online ) ( a ) electronic phase diagram and variation of @xmath6 as a function of @xmath7 ( la concentration ) .
filled circles ( zero ) and open triangle ( mid ) denote @xmath6 determined by the temperature where the resistivity is 10% and 50% of that for the normal state , respectively .
the solid curve is a guide to the eyes .
( b ) the normal - state conductivity at 4 k ( @xmath53 ) as a function of @xmath7 .
the dashed curve is a guide to the eyes . ]
to further study the superconducting properties for the sr@xmath1la@xmath2fbis@xmath0 system , we have systematically measured the specific heat .
figure [ fig : ctx045 ] displays the temperature dependence of total specific heat @xmath31 for @xmath7=0.45 , plotted as @xmath54 versus @xmath55 . the data at 0 t and 9 t correspond to the superconducting and normal states , respectively .
an anomaly relevant to the superconducting transition is discernible at @xmath55=9 k@xmath56 in the zero - field curve , reflecting the bulk superconductivity in this compound .
the data for the normal state , on the other hand , are well fitted to the conventional relation : @xmath57 , as shown by a dashed line .
the best - fitted result gives the normal - state specific - heat coefficient of @xmath58 mj / mol k@xmath56 , which is roughly in accordance with the value previously reported for sr@xmath16la@xmath16fbis@xmath0.@xcite after subtracting the phononic contribution ( @xmath59 ) , we have presented the electronic specific heat @xmath60 in the inset of fig .
[ fig : ctx045 ] . at 0 t ,
a clear jump manifests itself at the superconducting transition . taking the entropy balance into consideration ,
we have obtained @xmath61 k and the specific - heat jump @xmath62 .
although the size of the observed specific - heat jump is less than that of the bcs weak - coupling limit ( = 1.43 ) , it can be comparable to the bcs value when considering the apparent shielding fraction of @xmath1575% [ fig .
[ fig : rtmt](a ) ] , . temperature dependence of specific heat @xmath31 at 0 t ( superconducting state ) and 9 t ( normal state ) for sr@xmath1la@xmath2fbis@xmath0 ( @xmath7=0.45 ) , plotted in the form of @xmath54 versus @xmath55 .
the closed triangle denotes the anomaly arising from the superconducting transition .
the dashed line is the linear fit ( @xmath63 ) to the low - temperature data at 9 t. inset shows the electronic part of the specific heat at 0 t and 9 t , obtained after subtracting the phononic contribution determined by the data at 9 t ( @xmath64 ) . ]
figure [ fig : ctall ] displays the temperature profiles of @xmath65/t$ ] at @xmath66=0 and 9 t for @xmath67 .
the value of @xmath68 at 0 t is the difference in electronic specific heat between the superconducting and normal states .
we here subtracted the 9 t data by fitting them to a polynomial . for @xmath7=0.4 and 0.42 , no distinct anomaly in specific heat was observed in the measured temperature range .
a weak upturn observed for @xmath7=0.42 near the lowest temperature is reminiscent of its relatively broad superconducting transition . for @xmath5 ,
on the other hand , a clear jump is discernible at the superconducting transition .
noteworthy is that the temperature profile of @xmath68 at 0 t varies depending on @xmath7 ; the gradient of @xmath68 below @xmath6 gradually decreases with increasing @xmath7 , as highlighted by the dashed lines .
in fact , the gradient of @xmath68 for @xmath7=0.5 decreases by approximately 35% compared to @xmath7=0.45 , in spite of the similar broadness of the transition . the gradient seems to be further reduced for @xmath7=0.6 , although the superconducting transition is broader than those for @xmath7=0.45 and 0.5 .
concomitantly , the size of specific - heat jump at @xmath6 also gradually decreases with increasing @xmath7 .
these facts suggest the change in either superconducting coupling strength or electronic specific heat with @xmath7 , i.e. , the carrier concentration , when considering the empirical theory , the so - called @xmath27 model @xcite , that has a parameter of coupling strength @xmath69 , where @xmath70 is the superconducting gap size at 0 k. we here roughly estimate the values of @xmath71 , where @xmath72 is the residual electronic specific - heat coefficient from the non - superconducting part of the sample . based on the values of @xmath73 and apparent volume fraction , we have obtained @xmath74=1.2 and 1.3 mj / mol k@xmath56 for @xmath7=0.45 and 0.5 , respectively , indicating the difference in electronic specific heat may be less than 10% between these two compounds.@xcite in this case , the observed increase in gradient of @xmath68 below @xmath6 with decreasing @xmath7 would essentially reflect the increase in coupling strength @xmath27 . in the present system ,
the superconducting coupling strength is thus likely to be promoted toward the phase boundary to the insulator , in parallel with @xmath6 . as mentioned above ,
the similar reinforcement of the pairing interaction was first suggested for li@xmath2zrncl.@xcite there , the possible pairing mediated by the spin fluctuation was proposed even in the doped band insulators , since the spin susceptibility as well as the pairing interaction was found to be enhanced with approaching a band insulator phase . in the sr@xmath1la@xmath2fbis@xmath0 compound , the pairing mechanism is still unclear .
the systematic measurements of the phonon spectra would be quite important to elucidate whether the conventional electron - phonon coupling alone can explain the observed @xmath6 variation , or an exotic paring mechanism related to electron correlation is applicable to the present case as well .
( color online ) temperature profiles of @xmath75/t$ ] ( the difference in electronic specific heat from the normal - state value ) for @xmath76 at @xmath66=0 and 9 t. the curves are shifted for clarity as denoted by horizontal solid lines . ] to study the normal state in detail , we have carried out the hall effect measurements on the samples that show a superconducting transition at low temperatures .
figure [ fig : hall ] presents the temperature dependence of the hall coefficient ( @xmath30 ) for @xmath7=0.45 .
@xmath30 was estimated by fitting linearly the hall resistivity ( @xmath77 ) data up to @xmath78 t , although the @xmath66 profile of @xmath77 is slightly curved as shown in the inset of fig .
[ fig : hall ] .
surprisingly , the value of @xmath30 is positive throughout the entire temperature range and shows a weak temperature dependence .
this clearly indicates that hole - type carriers are dominant for this compound , in spite of electron doping .
moreover , the @xmath7=0.4 and 0.42 specimens also have the positive @xmath30 value , although the @xmath77 data are scattered due to their too high resistivity .
in contrast , for @xmath7=0.6 , the @xmath77 data exhibit a steep negative gradient , resulting in the large negative @xmath30 of @xmath79 @xmath80/c . considering that the bi impurity phase was detected for @xmath34 [ fig . [
fig : xrd](a ) ] , it is reasonable to ascribe the observed negative @xmath30 to just extrinsic hall signals from the bi metal impurities , because a bi metal has been reported to show huge negative @xmath30 at low temperatures [ for instance , @xmath81 @xmath80/c at 4.2 k for polycrystalline bi@xcite ] .
actually , the temperature dependence of @xmath30 for @xmath7=0.6 shows strong enhancement toward low temperatures ( not shown here ) , which well coincides with the behavior of pure bi metal
. it would be thus impossible to determine the sign of the intrinsic hall coefficient for @xmath34 , due to the likely contribution from the bi impurities.@xcite note here that another impurity phase bi@xmath0s@xmath13 has negative @xmath30 as well@xcite , and hence the positive hall signal should arise from the main phase of sr@xmath1la@xmath2fbis@xmath0 .
the existence of the hole - like fermi surfaces in the normal state is consistent with the theoretical predictions.@xcite the calculations based on a two - orbital model ( bi 6@xmath19 and 6@xmath20 ) have revealed the topology of the fermi surface dramatically changes , when approximately 0.5 electrons per bi site are doped , i.e. , at around @xmath7=0.5 in the present notation . for the low - doping regime , electron pockets locate around the @xmath82 [ ( @xmath83,0 ) or ( 0,@xmath83 ) ] point .
when @xmath7 is increased up to @xmath150.5 , however , these pockets connect with each other and evolve into two hole pockets around the @xmath21 ( 0,0 ) and @xmath18 ( @xmath22,@xmath22 ) points .
thus , the observed positive @xmath30 is considered to originate from the hole pockets formed after the topological transition of the fermi surfaces . on the other hand ,
the thermopower for @xmath7=0.45 exhibits a negative value , for instance , @xmath15 - 15 @xmath84v / k at room temperature ( not shown here ) , which signals the coexistence of electron - type carriers with hole - type ones .
the observed small positive @xmath30 may result from the cancellation of the contributions from the carriers with different signs , rather than large hole - like fermi surfaces . in theory ,
it was actually pointed out that new electron pockets show up again around the @xmath82 point with further increasing the carrier concentration above 0.5.@xcite thus , our observations should put a strong constraint to the fermi surface topology and relevant pairing mechanism for the bis@xmath0-based superconductor .
( color online ) temperature profile of hall coefficient @xmath30 for @xmath7=0.45 .
the inset shows the hall resistivity @xmath77 versus field at 10 k for @xmath76 . ]
we have investigated the detailed electronic phase diagram and the superconducting properties for the electron - doped sr@xmath1la@xmath2fbis@xmath0 layered compounds by synthesizing the polycrystalline samples with la concentration ( @xmath7 ) systematically changed from 0 to 0.6 .
characterization using the powder x - ray diffraction and x - ray absorption spectroscopy has revealed that la substitution is successful up to @xmath7=0.6 and it effectively affords an electron into the bis@xmath0 layer .
the pristine compound is a band insulator and the resistivity systematically decreases with increasing @xmath7 . for @xmath5 ,
the system exhibits metallic behavior except for the anderson localization effect , where the superconducting transition manifests itself at low temperatures .
the bulk superconductivity was confirmed by the large diamagnetic response and clear jump in specific heat .
the variation in @xmath6 is distinctive ; it gradually increases with decreasing the carrier concentration toward the insulating phase and suddenly disappears at the critical concentration ( @xmath8 ) .
the obtained electronic phase diagram is characterized by the doping - induced insulator - superconductor transition with a steep phase boundary around @xmath7=0.4 .
this markedly contrasts to the broad superconducting dome reported for the @xmath9o@xmath1f@xmath2bis@xmath0 compounds , which indicates the blocking layer strongly affects the superconducting transition .
considering that all the bands arising from the sr and f orbitals are away from the fermi energy , the present phase diagram would purely reflect the effects of rigid - band - like electron doping into the bis@xmath0 layer .
the thermodynamic measurements have furthermore suggested that the superconducting coupling strength is enhanced , as the phase boundary to the insulator is approached .
the observed superconducting features resemble those for li - intercalated zrncl superconductors , where the transition to the superconducting phase takes place by as small as 6% doping . on the other hand ,
the small positive hall coefficient as well as small negative thermopower was observed for the normal state of the present compound , which indicates that the topology of the fermi surface significantly changes from that in the low - carrier density regime .
therefore , the sr@xmath1la@xmath2fbis@xmath0 system can be regarded as a layered superconductor derived from a doped band insulator , characterized by hole - like fermi surfaces coexistent with electron - type ones .
this study was partly supported by nippon sheet glass foundation for materials science and engineering , the murata science foundation , the first program by the japan society for the promotion of science , and jsps grant - in - aid for scientific research ( no .
23685014 , no . 25620040 , and no .
24224009 ) .
the measurement in spring-8 was performed under proposal number 2013a1902 .
99 j. g. bednorz and k. a. mller , z. phys .
* 64 * ( 1986 ) 189 .
s. yamanaka , k. hotehama , and h. kawaji , nature * 410 * ( 1998 ) 580 .
y. kamihara , t. watanabe , m. hirano , and h. hosono , j. am .
* 130 * ( 2008 ) 3296 .
y. mizuguchi , h. fujihisa , y. gotoh , k. suzuki , h. usui , k. kuroki , s. demura , y. takano , h. izawa , and o miura , phys .
b * 86 * ( 2012 ) 220510(r ) .
y. mizuguchi , s. demura , k. deguchi , y. takano , h. fujihisa , y. gotoh , h. izawa , and o. miura , j. phys .
* 81 * ( 2012 ) 114725 .
j. xing , s. li , x. ding , h. yang , and h. h wen , phys .
b * 86 * ( 2012 ) 214518 .
d. yazici , k. huang , b. d. white , a. h. chang , a. j. friedman , and m. b. maple , philos . mag . * 93 * ( 2012 ) 673 .
k. ishida , y. nakai , and h. hosono , j. phys .
* 78 * ( 2009 ) 062001 .
x. lin , x. ni , b. chen , x. xu , x. yang , j. dai , y. li , x. yang , y. luo , q. tao , g. cao , and z. xu , phys .
b * 87 * ( 2013 ) 020504(r ) .
d. yazici , k. huang , b. d. white , i. jeon , v. w. burnett , a. j. friedman , i. k. lum , m. nallaiyan , s. spagna , and m. b. maple , phys . rev .
b 87 ( 2013 ) 174512
. w. a. phelan , d. c. wallace , k. e. arpino , j. r. neilson , k. j. livi , c. r. seabourne , a. j. scott , and t. m. mcqueen , j. am .
* 135 * ( 2013 ) 5372 .
h. usui , k. suzuki , and k. kuroki , phys .
b * 86 * ( 2012 ) 220501(r ) . g.b .
martins , a. moreo , and e. dagotto , phys .
b * 87 * ( 2013 ) 081102(r ) .
b. li , z. w. xing , and g. q. huang , epl * 101 * ( 2013 ) 47002 .
t. yildirim , phys .
b * 87 * ( 2013 ) 020506(r ) .
j. lee , m. b. stone , a. huq , t. yildirim , g. ehlers , y. mizuguchi , o. miura , y. takano , k. deguchi , s. demura , and s. -h .
lee , phys .
b * 87 * ( 2013 ) 205134 .
k. deguchi , y. mizuguchi , s. demura , h. hara , t. watanabe , s. j. denholme , m. fujioka , h. okazaki , t. ozaki , h. takeya , t. yamaguchi , o. miura , and y. takano , epl * 101 * ( 2013 )
17004 . s. demura , y. mizuguchi , k. deguchi , h. okazaki , h. hara , t. watanabe , s. j. denholme , m. fujioka , t. ozaki , h. fujihisa , y. gotoh , o. miura , t. yamaguchi , h. takeya , and y. takano , j. phys .
* 82 * ( 2013 ) 033708 .
k. suzuki , h. usui , and k. kuroki , physics procedia * 45 * ( 2013 ) 21 .
h. lei , k. wang , m. abeykoon , e. s. bozin , and c. petrovic , inorg
* 52 * ( 2013 ) 10685 .
p. a. lee and t. v. ramakrishnan , rev .
* 57 * ( 1985 ) 287 .
y. taguchi , a. kitora , and y. iwasa , phys .
* 97 * ( 2006 ) 107001 .
y. kasahara , t. kishiume , t. takano , k. kobayashi , e. matsuoka , h. onodera , k. kuroki , y. taguchi , and y. iwasa , phys .
* 103 * ( 2009 ) 077004 .
h. padamsee , j. e. neighbor , and c. a. shiffman , j. low temp . phys .
* 12 * ( 1973 ) 387 .
the values of @xmath73 for @xmath7=0.45 , 0.5 , and 0.6 are estimated from the specific heat data at 9 t to be 1.6 , 1.3 , and 1.4 mj / mol k@xmath56 , respectively .
the apparent volume fraction was extracted from the data in fig .
[ fig : rtmt](a ) .
the @xmath74 value for @xmath7=0.6 is approximately 0.9 , which is reduced by approximately 30% compared to that for @xmath7=0.5 .
this reduction roughly corresponds in size to that in the slope of @xmath68 ( @xmath1520% ) , which implies the coupling strength does nt change significantly above @xmath7=0.5 .
f. brochin , b. lenoir , c. bellouard , h. scherrer , i. vurgaftman and j. r. meyer , phys .
b * 63 * ( 2001 ) 073106 .
the hall resistivity for @xmath7=0.5 is strongly curved as a function of @xmath66 .
this may be because the intrinsic positive signal from the main phase would be comparable to the negative signal from the bi impurities .
b. chen , c. uher , l. iordanidis , and m. kanatzidis , chem .
* 9 * ( 1997 ) 1655 .
y. li , x. lin , l. li , n. zhou , x. xu , c. cao , j. dai , l. zhang , y. luo , w. jiao , q. tao , g. cao , and z. xu , arxiv:1310.1695 ( 2013 ) . | effects of electron doping on the bis@xmath0-based superconductors sr@xmath1la@xmath2fbis@xmath0 ( @xmath3 ) have been investigated using the systematically synthesized polycrystals by means of x - ray diffraction , x - ray absorption spectroscopy , transport and thermodynamic measurements .
the pristine compound is a band insulator with the bis@xmath0 layer , which accommodates electron carriers through the la substitution for sr , as evidenced by the change in x - ray absorption spectra reflecting the occupancy of bi 6@xmath4 orbitals . with increasing the carrier density ,
the resistivity progressively decreases and a bad metallic state appears for @xmath5 , where bulk superconductivity manifests itself below approximately 3 k. the value of @xmath6 gradually increases with decreasing @xmath7 from 0.6 to 0.45 and immediately decreases down to zero at the critical concentration of @xmath8 , resulting in an insulator - superconductor transition highly sensitive to the carrier density .
thermodynamic measurements furthermore have revealed the possible enhancement of the superconducting coupling strength as the insulating phase is approached .
the obtained superconducting phase diagram is markedly different from the broad dome - shaped superconducting phase previously reported for @xmath9o@xmath1f@xmath2bis@xmath0 ( @xmath9 : rare - earth ion ) , suggesting a strong influence of the blocking layer on the superconductivity .
instead all these features are similar to those observed in li - intercalated zrncl superconductor , except for the critical electron concentration of as low as 6% in the latter compound . for the present superconductor ,
notably , the existence of hole - type carriers has been indicated in the normal state from the hall effect measurements .
the sr@xmath1la@xmath2fbis@xmath0 system providing the phase diagram for the rigid - band doping in the bis@xmath0 layer would be another prototypical example of superconductivity derived from a doped layered band insulator , hosting both hole - like and electron - like fermi surfaces . |
bipartite and multipartite entanglement is one of the features that give rise to many of the developments of quantum computation and information , like quantum teleportation @xcite and quantum cryptography @xcite , among others @xcite .
the evaluation of the entanglement of a composite state is thus a main task to be done . for this purpose ,
the schmidt decomposition @xcite has proven to be a valuable tool , for systems with just two components .
in this paper we consider the case of continuous variables entanglement . for us , these variables may be @xmath0 which commute as phase space variables do .
we also refer to continuous variable entanglement in systems described by momentum and/or energy observables .
precisely , the entanglement of continuous variables stems from the original epr article @xcite .
however , the treatment of systems with continuous variables is far from straightforward . until now ,
schmidt decomposition in the continuous case required solving the corresponding integral equations @xcite .
they had to be discretized , losing the continuous dependence of the initial state . here
we propose a method to perform the schmidt decomposition for this case , to the accuracy desired , keeping the continuous character of the variables .
this method consists of two steps : \1 ) we decompose the bipartite system wave function , @xmath1 , by using two discrete and complete sets of orthogonal functions , @xmath2 , @xmath3 , of @xmath4 , in the form : @xmath5 the purpose of this step is to transform the continuous problem into a discrete one ( a necessary step for the numerical computation ) , while preserving the continuous dependence of @xmath1 .
\2 ) then we apply the ( finite dimensional ) schmidt procedure to ( [ eq11 ] ) in order to write the wave function @xmath1 as diagonal sum of biorthogonal terms : @xmath6 the orthogonal functions @xmath7 , @xmath8 -the modes- will be some particular linear combinations of @xmath9 , @xmath10 , respectively . notice that we are using the schmidt procedure for discrete systems to obtain the decomposition for the continuous case .
this is much more tractable , as it implies diagonalizing matrices instead of solving integral equations .
the rationale for this procedure is the expectation that only a few @xmath11 will suffice : a handful of appropriate orthogonal functions will approximate @xmath1 to the desired accuracy .
we finish by pointing out some properties of this method , namely \a ) we obtain complete analytic characterization of the modes @xmath7 , @xmath8 to the desired precision .
our method surpasses the standard numerical procedures in that keeps the continuous features present in @xmath1 .
\b ) we remark the portability of the attained modes @xmath7 , @xmath8 that are ready for later uses .
\c ) for the physical system analyzed in this paper ( biphoton ) , we found that with @xmath12 @xmath13 matrices the error was of around @xmath14 , that is , the number of @xmath11 functions required is small .
for other systems studied that we do not include here , the convergence was even better : for the case of two electrons which interact electrostatically , the obtained error with @xmath15 matrices was of @xmath16 ( schmidt number @xmath17 ) . in this paper
we begin with a detailed exposition of our method in sect .
then , in sect
. [ s4 ] we apply it to a relevant case : the entanglement of two photons created by parametric down - conversion .
we compare our results ( leading to well known , continuous functions ) with those computed by numerical methods @xcite that produce sets of points : discrete functions .
both methods agree remarkably well . finally , in sect .
[ delta ] we decompose the dirac delta , a case of physical and mathematical interest .
we consider a bipartite quantum system formed by two subsystems @xmath18 and @xmath19 .
some examples are two photons entangled by parametric down - conversion , a photon emitted by an excited atom and as a result entangled with it or two charged particles which interact electrically .
this system is described by the vector state @xmath20 where @xmath21 , @xmath22 are the creation operators of a particle associated to the subsystems @xmath18 and @xmath19 .
@xmath23 and @xmath24 are continuous variables associated to @xmath18 and @xmath19 respectively , which can represent momenta , energies , frequencies , or the like . in general
, the analysis is made in an ad hoc kinematical situation in which @xmath23 and @xmath24 turn out to be one - dimensional variables , @xmath25 , @xmath26 . in the following we assume this is the case .
in addition , there can be discrete variables ( like the spin ) to be treated with the schmidt method , that we do not include here to avoid unwieldy notation .
our method works as follows : we consider two denumerable , complete sets of orthogonal @xmath4 functions @xmath2 , @xmath3 @xmath27 , each one associated to each particular subsystem @xmath28 ( @xmath29 ) .
these functions obey @xmath30 \1 ) our first step is to expand the wave function @xmath1 as a linear combination of the @xmath31 , translating the continuous problem into a discretized one .
thus we work with the discrete coefficients of the linear combination , though the continuous character of the state is preserved in the @xmath32 dependence of the @xmath31 functions .
the expansion reads : @xmath33 where the coefficients @xmath13 are given by @xmath34 \2 ) our second step is to apply the schmidt decomposition to the discretized bipartite state ( [ eq24 ] ) , as is usually done for finite dimension hilbert spaces ( diagonalizing matrices , instead of solving integral equations ) . in order to do this ,
it is necessary to truncate the expansion ( [ eq24 ] ) , something that is possible to a certain accuracy due to the fact that latexmath:[$\int dp dq and the expansion is in orthogonal functions , so the coefficients @xmath13 go to 0 with increasing @xmath36 ( see below ) .
we truncate the series ( [ eq24 ] ) at @xmath37 , @xmath38 , with @xmath39 , without loss of generality .
the schmidt procedure leads to ( [ eq12 ] ) , where @xmath40 here the matrix @xmath41 is the ( transposed ) eigenvectors matrix of @xmath42 : @xmath43 and @xmath44 are the eigenvalues of @xmath45 .
there are two sources of error in this procedure : \a ) truncation error : this is the largest source of error in our method .
inescapably , the series ( [ eq24 ] ) must end at some finite @xmath46 , @xmath47 when attempting to obtain some specific result .
this step is possible to a certain accuracy because the function @xmath1 is square - integrable and we are expanding it into orthogonal functions , so @xmath48 and thus @xmath49 when @xmath50 .
the particular choice of the orthogonal functions @xmath51 will affect how fast the @xmath13 go to zero .
hence , the election of these functions for a particular physical problem will be a delicate task . to reach the same accuracy with different sets @xmath52 it will be necessary in general to consider a different pair of cut - offs @xmath53 for each of the sets .
b ) numerical error : this is a better controlled source .
it includes the error in calculating the coefficients @xmath13 via ( [ eq25 ] ) and the one produced when diagonalizing the matrix @xmath54 .
the suitable quantity to control the convergence for a particular @xmath1 and a specific set @xmath52 is the well known ( squared ) distance @xmath55 between the function @xmath1 and the schmidt decomposition obtained with cut - offs @xmath53 ( mean square error ) : @xmath56 this expression gives the truncation error .
it will go to zero with increasing cut - offs according to the specific @xmath52 chosen .
another easily computable , less precise way of controlling the convergence is given by the fact that ( with no cut - offs ) @xmath57 and thus @xmath58 is other measure of the truncation error , where here @xmath59 is calculated with cut - offs @xmath53 .
would we compute the @xmath60 exactly , then @xmath61 . in practice this can not be done because our @xmath60 are the eigenvalues of the @xmath62 matrix @xmath63 , that depend slightly on @xmath64 .
both distances behave in a very similar way , as we show in fig .
[ fig1 ] and fig .
[ fig1bis ] , though @xmath65 is more easily computable than @xmath66 .
the election of the two sets of orthogonal functions for a particular physical problem , @xmath67 can be approached from two different points of view , according to the feature one desires to emphasize : fundamental or practical .
the election of the orthogonal functions in a particular problem can be done according to the specific intervals in which the variables @xmath68 take values for that case .
typical examples of discrete sets of orthogonal functions are the orthogonal polynomials , defined in a variety of intervals . for example , a possible choice to describe one dimensional momenta @xmath69 , are the hermite polynomials , @xmath70 .
the equivalence sign indicates here that the polynomial must be accompanied by the square root of the weight function in order to be correctly orthonormalized , and normalization factors must be included .
if , on the other hand , the variable of interest in a specific problem is bounded from below , like the energy of a free massless particle @xmath71 , then the election could be laguerre polynomials , @xmath72 .
the criterion for choosing the orthogonal functions @xmath51 according to the intervals in which @xmath23 , @xmath24 are defined has a fundamental character .
for example , the localizability in configuration space of the fourier transforms of the modes ( [ eq26 ] ) , ( [ eq27 ] ) , depends critically on the intervals in which these modes are defined @xcite .
only if we choose the functions @xmath51 to be defined exactly in the same intervals as the amplitude @xmath1 , may the fourier transforms of the modes have the right localization properties . in spite of that , this point of view may not be the most suitable one , as it may give slower convergence than the point of view presented below . in this case ,
the election is approached with the goal of improving the convergence .
the @xmath51 are chosen here according to the functional form of @xmath1 .
the closer the lowest modes are to @xmath73 the lesser the number of them necessary to obtain the required accuracy .
what we are looking for here are @xmath51 that maximize @xmath74 for low @xmath46 , @xmath47 . in some cases
this practical point of view will be more useful than the fundamental one .
for example , suppose the amplitude for a particular problem is of the form @xmath75 , with @xmath76 a slowly varying function of @xmath23 , @xmath24 .
in this particular case it is reasonable to choose the functions @xmath51 as hermite polynomials , because their weight functions are indeed gaussians .
this leads to @xmath77 , and similarly for @xmath10 .
we call this approach practical because the convergence is reached faster
. there is a price : the information relevant to localization might be lost .
in this section we consider a realistic case of biphotons already studied in the literature @xcite : two photons entangled in frequency through parametric down - conversion .
we apply our method to this physical system in order to obtain the schmidt decomposition and the structure of modes without losing the analytic character within the target accuracy .
the system under study is a biphoton state generated by parametric down - conversion ( pdc ) of an ultrashort pump pulse with type - ii phase matching .
the amplitude in this particular case takes the form @xcite
@xmath78\nonumber\\\times \mathrm{sinc}\{l[(\omega_o-\bar{\omega } ) ( k'_o-\bar{k})+(\omega_e-\bar{\omega})(k'_e-\bar{k})]/2\}\label{eq41}\end{aligned}\ ] ] where @xmath79 are the frequencies associated to the ordinary and extraordinary fields respectively , @xmath80 and @xmath81 are the inverse of group velocities at the frequency @xmath82 , @xmath83 is the inverse group velocity at the pump frequency @xmath84 , @xmath85 is the pdc crystal length and @xmath86 is the width of the initial pulse .
typical values for these parameters are @xmath87 , @xmath88 , @xmath89 , @xmath90 and @xmath91 .
we perform now the following change of variables @xmath92 and thus obtain @xmath93\label{eq46}\ ] ] we have applied our method to the function ( [ eq46 ] ) ( once normalized ) according to section [ s2 ] , in the following way : we choose as orthogonal functions hermite polynomials , because their weights are gaussians and a gaussian appears in ( [ eq46 ] ) .
these polynomials were used in @xcite for pdc in some particular cases which are exactly solvable .
the orthonormal sets we chose , looking for maximizing the @xmath13 ( [ eq25 ] ) for the lowest @xmath46 , @xmath47 , were @xmath94 this choice of polynomials is suitable for the practical approach ( subsection [ ss12 ] ) , taking into account that @xmath95 and thus the interval of definition of @xmath96 can be restricted to a region centered in @xmath82 of width @xmath97 in @xmath98 .
we did a careful analysis of this , that for brevity we do not show here . notice that our conclusions would not apply in the case @xmath99 .
we have considered cut - offs @xmath100 taking values @xmath101 and followed the steps of sect .
we have computed the eigenvalues @xmath60 of the schmidt decomposition ( [ eq12 ] ) for each pair @xmath53 .
we have also computed the modes ( [ eq26 ] ) and ( [ eq27 ] ) . as a function of the cut - offs @xmath53.[fig1 ] ] as a function of the cut - offs @xmath53.[fig1bis ] ] in fig .
[ fig1 ] we plot the distance @xmath55 ( [ eq29 ] ) as a function of @xmath100 , to show how fast the convergence is . with @xmath102
the truncation error is of @xmath14 .
we also plot in fig .
[ fig1bis ] the distance @xmath103 ( [ eq29bis ] ) , which serves as another measure of the convergence , as a function of @xmath100 .
we obtained @xmath104 .
regarding now the most precise case considered , @xmath102 , we plot in fig .
[ fig2 ] the eigenvalues @xmath60 for different values of @xmath47 , observing good agreement with the results existing in the literature @xcite . for this case
we also plot in fig .
[ fig3 ] the modes ( [ eq26 ] ) and ( [ eq27 ] ) for @xmath105 , confirming the validity of the method when comparing with @xcite . versus index @xmath47.[fig2 ] ] , @xmath8 as a function of @xmath106 and @xmath107 , for @xmath108.[fig3 ] ] the modes are given explicitly by : @xmath109 where the values of the coefficients @xmath110 are obtained through ( [ eq26 ] ) and ( [ eq27 ] ) . the actual properties of the modes ( [ eq49 ] ) depend on these values .
in fact , the parity and number of nodes is determined by them , taking into account that @xmath111 is a polynomial of degree @xmath47 , parity @xmath112 and having @xmath47 nodes .
a good approximation to the @xmath113 obtained with our procedure is @xmath114 this expression has a deviation ( squared distance ) of @xmath115 from the whole mode obtained including terms until @xmath116 , which is the greatest power appearing for @xmath102 . on the other hand , @xmath117 . from ( [ eq410 ] )
it can be seen that in this mode the even components are greater than the odd ones ( these are negligible ) , so it is an even state , as shown in fig .
[ fig3 ] .
another example is the approximation to @xmath118 @xmath119 this has a deviation ( squared distance ) of @xmath120 from the whole mode obtained including terms until @xmath121 . on the other hand , @xmath122 .
more terms are needed in ( [ eq411 ] ) , because they go to zero more slowly with increasing powers of @xmath24 . here
the most important components are the odd ones ( the even ones are negligible ) , leading to an odd parity state , as shown in fig .
[ fig3 ] . to show how the convergence of the method depends on the specific family pairs of orthogonal functions @xmath2 , @xmath3 chosen , we consider the cases of hermite orthogonal functions depending on a parameter @xmath123 related to the width of the gaussian , fixed for each family pair : @xmath124 we applied our method to the amplitude ( [ eq46 ] ) with these sets of orthogonal functions , for @xmath125 , and cut - offs @xmath126 . we show in table [ tab1 ] the values of @xmath103 for these specific parameters .
@xmath103 + .@xmath103 for @xmath125 and @xmath126 .
[ cols="^,^,^,^",options="header " , ] clearly , the convergence is better for the case @xmath127 , which we used in the preceding calculations . in case
we chose another type of orthogonal function for ( [ eq46 ] ) ( laguerre , legendre , ... ) , the convergence would have been much worse because of the specific shape of that amplitude .
another interesting case is the dirac delta .
here we have @xmath128 and we take the same interval @xmath129 for @xmath23 and @xmath24 .
we consider complete sets of orthonormal functions satisfying @xmath130 .
a particular case is when they are real functions , as for example the typical orthogonal polynomials ( legendre , hermite , laguerre , chebyshev , ... ) are .
we must take into account that the dirac delta is not a function but a distribution , and indeed is outside @xmath4 .
however , we can calculate the @xmath13 and study how much entanglement does this state have .
we obtain straightforwardly @xmath131 .
this gives @xmath132 which is just the resolution of the identity as given in ( [ eq23 ] ) .
the schmidt decomposition of the dirac delta is not unique , because all the weights @xmath133 are equal to one ( they are degenerate ) .
in fact , the decomposition can be done with any complete , denumerable set of orthonormal functions , in the form ( [ eq23 ] ) .
this expression can be seen as an infinite entanglement case , in the sense explained below .
the fact that all the weights are equal to one , makes sense only because we are considering a distribution , not an @xmath4 state .
the sum of the squares of the weights , which must be equal to the square of the norm of the function @xmath1 , diverges because the dirac delta is not square - integrable . a possible measure of the entanglement of a state @xmath1 in its schmidt decomposition ( [ eq12 ] ) is given by the von neumann entropy @xcite @xmath134 this is usually called the entropy of entanglement . the state of @xmath4 closer to ( [ eq33 ] ) is the case of an entangled state with @xmath135 diagonal terms with equal @xmath60 , when @xmath135 goes to infinity .
to be correctly normalized it verifies @xmath136 , @xmath137 and @xmath138 this is the maximum entanglement case .
this provides an estimate of the entropy of the dirac delta ( were it in @xmath4 ) .
the origin of our interest in the @xmath139 comes from the fact that @xmath1 may evolve in time towards a dirac delta , as it happens in time dependent perturbation theory of quantum mechanics .
the entanglement in these cases would increase with time towards its maximum , corresponding to the dirac delta .
in this paper we have introduced a method for computing the schmidt decomposition of a bipartite state with continuous degrees of freedom . in the existing literature
@xcite the decomposition produced sets of points as approximation to the modes .
our method gives linear combinations of the well known orthogonal functions as approximation to them .
when these functions are chosen properly , a handful of them is enough to reach the desired accuracy .
we introduce some criteria for the control of convergence and truncation error .
the result of our method for the decomposition of a biphoton state produced by parametric down - conversion agrees with the numeric results in the literature @xcite .
we also touch on the last stage of evolution of entanglement for determined systems , analyzing the dirac delta case .
we thank i. bialynicki - birula for useful comments and correspondence .
this work was partially supported by the spanish ministerio de educacin y ciencia under project bmf 2002 - 00834 .
the work of l. l. was supported by the fpu grant ap2003 - 0014 . | we propose a method for obtaining the schmidt decomposition of bipartite systems with continuous variables .
it approximates the modes to the prescribed accuracy by well known orthogonal functions .
we give some criteria for the control of errors .
we illustrate the method comparing its results with the already published analysis for entanglement of biphotons .
the agreement is excellent . |
the deuteron , as the simplest nuclear system , serves the dual role of an `` effective free neutron target '' @xcite and as a testing ground for sophisticated models of nucleon - nucleon interactions and scattering mechanisms @xcite .
electron scattering off the deuteron has been used as a means to extract information on its nuclear structure , including the @xmath10-wave ( @xmath11 ) contribution to the ground state wave function @xcite . on the other hand , experiments that look for modifications of nucleon structure due to nuclear binding
have also used the deuteron as a testbed @xcite . in all of these cases ,
a thorough and detailed understanding of the scattering mechanism is necessary . in particular , quasi - elastic scattering off the deuteron has been widely studied @xcite as an ideal reaction to disentangle various contributions to the reaction mechanism , such as relativistic effects , non - nucleonic components of the deuteron wave function , meson - exchange ( mec ) and isobar ( ic ) currents , and final state interactions ( fsi ) between the outgoing nucleons .
recent experiments @xcite have focused on higher momentum transfers , where one - nucleon currents are expected to dominate the cross section . because of the continuing ( and growing ) importance of the deuteron as an effective neutron target @xcite
, a particularly important question is whether there is a kinematic region where the simple picture of the plane wave impulse approximation ( pwia ) works reasonably well , in which the virtual photon is absorbed by only one nucleon inside the deuteron while the other is an unperturbed spectator to the reaction .
alternatively , one wants to test state of the art models of fsi to ascertain if they can yield a reliable description of the reaction mechanism . in this quest ,
polarization degrees of freedom are particularly interesting , yet few experiments exist . from a practical point of view
, quasi - elastic scattering off a polarized deuteron target ( with or without detection of a final state proton ) is often used as a direct measure of the product of beam and target polarization for spin structure function experiments @xcite .
this requires that the theoretical asymmetry for this process is well - known , an assumption that should be tested experimentally . in the following
, we first give a brief overview of the theoretical background for the reaction @xmath4h@xmath12 , followed by an overview of existing data .
section [ section : experiment ] describes the experimental setup , followed by details of the data analysis and our results .
the final section summarizes our findings .
the deuteron is the simplest stable nucleus , consisting of a proton and neutron bound by only 2.2 mev ( see , for instance , the review by garon and van orden @xcite ) .
its structure is amenable to detailed and sophisticated microscopic calculations that range from nonrelativistic approaches @xcite , based on the schrdinger equation , to fully relativistic treatments @xcite .
comparison of these calculations with experiment allows us to constrain properties of nucleons and of the nucleon - nucleon potential . in turn , given a model for the nucleon - nucleon interaction , form factors and momentum or spatial distributions of the nucleons in deuterium can be obtained @xcite .
most modern models of the deuteron wave function agree in the basic features of this momentum distribution : at low momenta , it is dominated by the s - wave ( @xmath13 ) where the proton and neutron spin are parallel to the overall deuteron spin , while at momenta beyond 250 - 300 mev/@xmath8 , the contribution from the much smaller d - wave ( @xmath14 ) component ( which overall accounts for about 4 - 6% of the deuteron ground state ) becomes more important . in that kinematic region ,
the expectation value for the nucleon spins is actually opposite to that of the deuteron as a whole .
experimentally , a large body of data on the quasi - elastic deuteron break - up reaction , @xmath15 , has been collected to access information on the nucleon momentum distribution in deuterium ( recent examples can be found in @xcite ) . a very important question in this context
is how the measured ( missing ) momentum distributions can be connected to the deuteron wave function @xcite , given their potential distortion by final state interactions ( fsi ) @xcite . on the other hand , deuteron targets are often used to extract information on the neutron , due to the absence of sufficiently dense free neutron targets . for example , both unpolarized ( see @xcite and references therein ) and polarized @xcite structure functions of the neutron are often extracted from measurements on the deuteron . in particular in the latter case , a clear understanding of the spin - dependent momentum distribution of nucleons in deuterium is of great importance , not only for a reliable extraction of neutron spin structure functions but also because the product of target and beam polarization ( which enters the measured asymmetries as a constant factor ) is often extracted using the polarized quasi - elastic reaction @xmath16 @xcite .
similarly , the @xmath17 reaction ( with and without polarization information ) is often used to access the neutron form factors @xcite .
furthermore , the novel technique of spectator tagging " , where a backward - moving proton spectator is detected in coincidence with an inelastically scattered electron , is being used both to access the free neutron structure ( at small spectator momenta @xcite ) and to study possible modifications of nucleons that are part of a high - momentum correlation @xcite . in all these cases , it is imperative to understand both the underlying spin - momentum structure of the deuteron as well as the reaction mechanism for electron scattering , including fsi effects .
the present paper focuses on the reaction @xmath16 with a deuteron target polarized along the direction of the incoming electron beam .
the differential cross section for this reaction , @xmath18 is a function of the ( negative of the ) squared four - momentum transferred by the scattered electron , @xmath19 with @xmath20 and @xmath21 being the four - momenta of the incoming and scattered electron ( in the ultra relativistic limit ) , respectively . here
, @xmath22 is the energy of the incoming electron and @xmath23 is the energy of the scattered electron , while the scattered electron direction is given by the polar angle @xmath24 and the azimuthal angle @xmath25 with respect to the incoming electron beam .
the cross section also depends on the missing momentum @xmath26 of the unobserved ( but inferred ) final state neutron ; we will parametrize this momentum by its magnitude , @xmath27 and its angle @xmath6 relative to the direction of the three - momentum transfer @xmath28 . for polarized beam and target
, this cross section can be expressed as @xmath29 , \label{poldeut : eq}\ ] ] where @xmath30 is the unpolarized cross section , @xmath31 \end{aligned}\ ] ] is the vector polarization and @xmath32 \end{aligned}\ ] ] is the tensor polarization of the target ( with @xmath33 the occupation numbers for the three magnetic quantum numbers @xmath34 ) , and @xmath35 is the helicity of the electrons .
we adopt here the notation of @xcite , where the vector ( @xmath36 , @xmath37 ) and tensor ( @xmath38 , @xmath39 ) asymmetries are normalized as components of spherical tensors of rank 1 and 2 , respectively .
integration over all azimuthal directions @xmath40 of the final state neutron ( around @xmath41 ) leaves only the asymmetries @xmath37 and @xmath38 ( because of parity conservation in the electromagnetic interaction ) .
both asymmetries are functions of the beam energy @xmath22 and @xmath42 as well as @xmath27 and @xmath43 . forming the difference between opposite - sign and equal sign pairs of helicity @xmath35 and target polarization @xmath44 and dividing by the sum
, we arrive at the double - spin asymmetry @xmath45 where @xmath46 is the magnitude of the electron beam polarization and @xmath47 is the average magnitude of the target vector polarization , both along the beam direction .
the target used in the present experiment was vector - polarized up to @xmath48 using dynamical nuclear polarization ( dnp)@xcite , which yields a tensor polarization @xmath49 , according to equal spin temperature ( est ) theory@xcite .
the simplest model for quasi - elastic deuteron breakup , the plane wave impulse approximation ( pwia ) , assumes that the virtual photon is absorbed by a single ( on - shell ) nucleon ( impulse approximation ) and the struck nucleon leaves the nucleus without further interaction ( i.e. , as a plane wave ) . in this model , the measured asymmetry is proportional to the initial polarization of the struck nucleon ( ignoring the small contribution from the tensor asymmetry for the moment ) : @xmath50 here , @xmath51 is the ratio of electric to magnetic sachs form factors of the struck nucleon , @xmath52 and @xmath53 are its polarization components along and transverse to the momentum transfer vector @xmath41 ( in the electron scattering plane ) , @xmath54 is the nucleon mass , and @xmath55\tan^2\frac{\theta_e}{2}\right)^{-1 } .\ ] ] is the virtual photon polarization ratio . within this pwia picture ,
measurements of @xmath0 can be used to extract information on the spin and momentum dependence of the nuclear wave function .
one goal of the present experiment is to determine the kinematic region where pwia is a reasonably good approximation .
a more realistic description requires a model that includes the interaction between the spectator and knocked - out nucleon ( fsi ) .
sabine jeschonnek and j.w .
van orden have developed a comprehensive theoretical model @xcite for this purpose .
the authors model the @xmath16 reaction in various approximations ( including pwia ) as well as with a complete treatment of fsi to learn more about its sensitivity to the initial nuclear state and the reaction mechanism .
they use a relativistic deuteron wave function by solving the gross equation @xcite .
a current said parameterization @xcite of the nucleon - nucleon scattering amplitude is used to calculate the interaction of the two nucleons in the final state , up to kinetic energies of about 1.3 gev in the lab frame .
this nucleon - nucleon amplitude includes central , spin - orbit , and double spin - flip terms . within the kinematic range of applicability
, we compare this model directly to our data , including the effects for both the vector and the tensor asymmetries in eq .
( [ eq : apara ] ) .
from nikhef . theoretical curves from arenhvel are also shown .
short - dashed curves are results for plane wave born approximation ( pwba ) , long - dashed curves include fsis , and solid curves represent the full calculation @xcite.,title="fig:",scaledwidth=35.0% ] from nikhef . theoretical curves from arenhvel are also shown .
short - dashed curves are results for plane wave born approximation ( pwba ) , long - dashed curves include fsis , and solid curves represent the full calculation @xcite.,title="fig:",scaledwidth=35.0% ] although the @xmath4h@xmath56 reaction has been studied in detail , there exist only a few measurements of the beam - vector asymmetry @xmath57 and tensor asymmetry @xmath58 .
these asymmetries are directly related to the double spin asymmetry @xmath0 as seen in eq .
( [ eq : apara ] ) .
the existing data are at a relatively low @xmath42 and were compared to a model formulated by arenhvel _
et al . _
this section will summarize the results of these experiments .
in contrast , the new data reported in the following sections cover a much wider range in kinematics ( beam energy and @xmath42 ) , and can therefore test models of fsi and deuteron structure in a region where different reaction mechanisms dominate . as a function of missing momentum at @xmath42 = 0.21 ( gev/@xmath59 @xcite.,scaledwidth=40.0% ]
the first measurements of the tensor asymmetry @xmath58 were performed at the dutch national institute for nuclear physics and high energy physics ( nikhef ) .
the experiment at nikhef used a polarized gas target with a 565 mev electron beam @xcite .
the tensor - polarized deuterium gas was altered between a polarization of @[email protected] and @[email protected] every 10 seconds . scattered electrons and protons
were detected by the bigbite magnetic spectrometer .
the tensor asymmetry was extracted as a function of the angle @xmath63 between the polarization axis and the missing momentum , as well as a function of the magnitude of the missing momentum .
the range of missing momentum was limited to below 150 mev / c .
the results of this measurement can be seen in fig .
[ fig : nikehfat ] .
additionally , the first measurements of @xmath57 were performed at nikhef several years later @xcite .
a longitudinally polarized beam of electrons of 720 mev was scattered off a vector - polarized deuterium target .
the scattered electron was measured at a fixed angle @xmath64 , with a solid angle coverage of 96 millisteradians ( msr ) and knocked - out protons were measured at a central angle of @xmath65 with a solid angle coverage of 250 msr .
the missing momentum range was increased up to 350 mev/@xmath8 at a @xmath42 of 0.21 ( gev/@xmath59 .
figure [ aved : fig ] shows that at momenta higher than 200 mev/@xmath8 , the vector asymmetry , @xmath66 , becomes sensitive to the @xmath10-state of the deuteron wave function .
( top ) and a@xmath67 ( bottom ) for [email protected] ( gev / c)@xmath4 , for both parallel ( left ) and perpendicular ( right ) kinematics .
theoretical curves have been calculated including meson exchange currents ( mec ) , isobar currents ( ic ) , and relativistic correction ( rc ) @xcite.,title="fig:",scaledwidth=48.0% ] ( top ) and a@xmath67 ( bottom ) for [email protected] ( gev / c)@xmath4 , for both parallel ( left ) and perpendicular ( right ) kinematics .
theoretical curves have been calculated including meson exchange currents ( mec ) , isobar currents ( ic ) , and relativistic correction ( rc ) @xcite.,title="fig:",scaledwidth=48.0% ] the bates large acceptance spectrometer toroid ( blast ) experiment used a polarized electron beam at 850 mev . ]
incident upon an internal polarized deuterium target @xcite at the mit - bates accelerator .
an atomic beam source was used for the polarized deuterium target , providing considerable freedom in the choice of vector and tensor polarization states .
two sets of deuteron data were taken with nominal spin angles of 32@xmath69 and 47@xmath69 to provide perpendicular ( @xmath70 ) and parallel ( @xmath71 ) kinematics ) describe the angle of the target polarization quantization axis relative to the momentum transfer vector , @xmath41 . ] .
there are two analyses of the blast data ; the latest work by a. degrush @xcite re - evaluated the work of a. maschinot @xcite to extract @xmath57 and @xmath58 .
these data were taken for a @xmath42 range of [email protected] ( gev / c)@xmath4 and ten missing momentum bins from 0.0 to 0.5 gev / c .
the results of this measurement can be seen in fig .
[ fig : blast ] . as measured in hall c , vs. scattered electron energy ( @xmath23 ) and vs. the angle between the neutron and the @xmath41 vector .
the theoretical curves are predicted asymmetries using different scaled values of @xmath72 @xcite.,title="fig:",scaledwidth=32.0% ] as measured in hall c , vs. scattered electron energy ( @xmath23 ) and vs. the angle between the neutron and the @xmath41 vector .
the theoretical curves are predicted asymmetries using different scaled values of @xmath72 @xcite.,title="fig:",scaledwidth=32.0% ] within arenhvel s model , all ingredients including isobar currents and relativistic corrections are needed to describe most of the blast data . in particular , a reasonable description of @xmath73 in parallel kinematics ( top right panel of fig .
[ fig : blast ] ) requires the inclusion of fsi effects at larger @xmath74 .
this is in qualitative agreement with our findings ( see below ) .
figure [ fig : blast ] also shows that @xmath57 is described rather well by the simpler pwba out to significantly higher missing momentum than is the case for @xmath73 .
the e93 - 026 experiment in hall c at the thomas jefferson national accelerator facility measured the neutron electric form factor @xmath75 in @xmath16 quasi - elastic scattering . in this measurement @xcite the neutron was detected instead of the proton and @xmath75 was extracted by comparing the measured @xmath57 to theoretical predictions by arenhvel using variations of the parameterization of @xmath75 by galster _
et al . _
@xcite .
the target used in this experiment was a @xmath76nd@xmath77 target similar to the target used in the present study .
the experiment was limited to @xmath42 = 0.5 ( gev/@xmath59 and the missing momentum was less than 180 mev .
the results are shown in fig .
[ avedhallc : fig ] .
this analysis is based on data from the eg1b group of experiments that took place at the thomas jefferson national accelerator facility ( jefferson lab ) located in newport news , virginia .
the continuous electron beam accelerator facility ( cebaf ) at jefferson lab provided polarized electron beams from 1.6 to 5.8 gev energy to the cebaf large acceptance spectrometer ( clas ) in jefferson lab s hall b. the beam polarization was periodically measured with a mller polarimeter .
the experimental run was conducted in 20002001 for a period of seven months .
clas is divided by six superconducting coils into six symmetric sectors with several layers of particle detectors .
the coils produce a mostly azimuthal magnetic field .
there are three layers of drift chambers ( dc ) for tracking in this field , followed by a layer of scintillator counters ( tof ) for time - of - flight measurements .
cherenkov counters and electromagnetic calorimeters in the forward regions are used to identify the scattered electrons .
a faraday cup is used to measure the total accumulated beam charge .
the clas data acquisition ( daq ) system collected data at a 3 - 4 khz event rate , triggered by a coincidence of the signals above threshold from the electromagnetic calorimeters and cherenkov counters .
a detailed description of clas and its systems can be found in @xcite . for the eg1b run , longitudinally
polarized electrons were scattered from several different targets placed alternatively on the center line of clas and immersed in a liquid helium bath at 1 k. these included longitudinally polarized proton ( @xmath76nh@xmath77 ) and deuterium ( @xmath76nd@xmath77 ) targets , as well as auxiliary @xmath78c and liquid @xmath79he ( `` empty '' ) targets .
beam and target polarizations were either parallel or anti - parallel with respect to each other and the beam direction .
the two polarized targets , @xmath76nh@xmath77 and @xmath76nd@xmath77 , were polarized by the dnp method . the deuterium target maintained roughly 20%-40% polarization during data collection .
the polarization was measured in real time using a nuclear magnetic resonance ( nmr ) system ; however , the final determination was based on measured double spin asymmetries , as explained below . to minimize depolarization of the targets due to heating and radiation damage , the electron beam was rastered over the surface of the targets in a spiral pattern during the experimental run .
the targets were periodically annealed to remove extra paramagnetic radicals and restore polarization .
further information on the polarized target can be found in @xcite .
the eg1b group of experiments collected data using several different experimental configurations .
the polarized electron beam had energies of 1.606 , 1.723 , 2.286 , 2.561 , 4.238 , 5.615 , 5.725 , and 5.743 gev with current from 0.3 na to 10 na .
the current of the main toroidal magnet was set at 2250 a or 1500 a and was switched from positive to negative polarity at times .
positive current polarity resulted in electrons being bent towards the beam axis ( inbending ) , while negative polarity led to outbending electrons and extended the accepted kinematics to lower scattering angles .
all eg1b run sets with usable electron beam data are labeled by beam energy and torus polarity ( _ e.g. _ , a 4.2 gev electron beam run with a positive torus current is labeled as 4.2 + ) and are listed in table [ table : run ] .
these set labels are used throughout this paper .
the 2.3 gev data had too few events for the present analysis to yield statistically significant results and was therefore not included .
the remaining beam energies were combined into four groups , with average ( nominal ) energies of 1.6 , 2.5 , 4.2 , and 5.7 gev .
cccc set label & e@xmath80(gev ) & i@xmath81(a ) + 1.6 + & 1.606 & + 1500 + 1.6@xmath1 & 1.606 & @xmath11500 + 1.7 + & 1.724 & + 1500 + 1.7@xmath1 & 1.724 & @xmath11500 + + 2.5 + & 2.562 & + 1500 + 2.5@xmath1 & 2.562 & @xmath11500 + 4.2 + & 4.239 & + 2250 + 4.2@xmath1 & 4.239 & @xmath12250 + 5.6 + & 5.627 & + 2250 + 5.6@xmath1 & 5.627 & @xmath12250 + 5.7 + & 5.735 & + 2250 + 5.7@xmath1 & 5.735 & @xmath12250 + 5.7@xmath1 & 5.764 & @xmath12250 + [ table : run ] additional experimental information can be found in the archival publications @xcite for eg1b .
the analysis presented here builds on the previously published standard analysis of the complete eg1b data set @xcite , including all calibrations , corrections , basic cuts and quality checks . from that analysis , we selected reconstructed events containing an electron and either only one proton or one proton and one neutral particle ( which could be the recoiling neutron ) .
electrons were identified through cuts on the signals in the cherenkov counters and electromagnetic calorimeters , while protons were selected based on their time of flight ( measured with the tof ) and their momentum .
fiducial cuts on both electrons ( to exclude regions of rapidly varying detection efficiency ) and protons ( to avoid both the clas torus and polarized target coil enclosures ) were applied .
all further details can be found in @xcite .
information on the neutron kinematics was gained through use of the conservation of energy and momentum .
the missing energy ( @xmath82 ) was calculated under the assumption that the reaction took place on a deuteron at rest , as @xmath83 where @xmath84 is the mass of the deuteron , @xmath85 is the energy of the virtual photon , and @xmath86 is the measured energy of the proton .
a cut @xmath87 gev was applied to reduce the size of the data sample , to include only events of interest . the missing momentum ( @xmath27 )
was calculated as @xmath88 where the 3-momentum of the virtual photon , @xmath41 , is calculated from the measured electron kinematics and @xmath89 is the momentum of the detected proton . because the nuclear background overwhelms the signal from the deuteron at high missing momenta , we only analyzed events with @xmath90 gev@xmath3 . finally , the missing mass ( @xmath91 ) was calculated as @xmath92 examples of missing mass distributions for different kinematic bins can be found in section [ sec : inelas ] .
the mass of the neutron is known to be 0.94 gev / c@xmath4 and a missing mass cut was implemented to remove multi - particle final states : @xmath93 the combination of these cuts can be seen in fig .
[ fig : mmvsem ] .
the curved lines identify the missing mass cut selecting exclusive @xmath94 final states from a deuteron .
the asymmetries presented in this paper were calculated as a function of three kinematic variables : the squared four - momentum transfer , @xmath2 , the cosine of the angle between the virtual photon and the neutron momenta , @xmath43 , and the missing momentum , @xmath27 .
we integrated over the angle @xmath95 between the leptonic and hadronic plane .
for compatibility with the main eg1b analysis , we combined several of the standard eg1b @xmath2 bins @xcite into four larger bins in the range @xmath96 ( gev/@xmath59 ( smaller bins would have yielded too limited statistics in the quasi - elastic region ) . table [
q2:table ] shows these four @xmath2 bins .
52ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty link:\doibase 10.1103/physrevc.50.r1749 [ * * , ( ) ] link:\doibase 10.1103/physrevd.58.112003 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0370-2693(99)00940-5 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.physletb.2006.12.076 [ * * , ( ) ] link:\doibase 10.1103/physrevd.75.012007 [ * * , ( ) ] link:\doibase 10.1103/physrevc.90.025212 [ * * , ( ) ] link:\doibase 10.1103/physrevc.89.045206 [ * * , ( ) ] link:\doibase 10.1103/physrevc.92.055201 [ * * , ( ) ] link:\doibase
10.1140/epja / i2004 - 10061 - 5 [ * * , ( ) ] link:\doibase 10.1103/physrevc.45.2094 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/0375-9474(83)90516-x [ * * , ( ) ] link:\doibase 10.1103/physrevc.54.646 [ * * , ( ) ] link:\doibase 10.1103/physrevc.73.035212 [ * * , ( ) ] link:\doibase 10.1103/physrevc.92.015211 [ * * , ( ) ] , @noop `` , '' link:\doibase 10.1103/physrevlett.89.062301 [ * * , ( ) ] link:\doibase 10.1146/annurev - nucl-100809 - 131956 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.98.262502 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.107.262501 [ * * , ( ) ] @noop `` , '' @noop `` , '' link:\doibase 10.1016/j.physletb.2006.08.011 [ * * , ( ) ] link:\doibase 10.1007/0 - 306 - 47915-x_4 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.70.743 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.88.102302 [ * * , ( ) ] @noop ( ) , link:\doibase 10.1103/physrevc.90.064006 [ * * , ( ) ] link:\doibase 10.1016/0370 - 2693(94)01436-g [ * * , ( ) ] link:\doibase 10.1007/jhep05(2011)126 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.87.081801 [ * * , ( ) ] link:\doibase 10.1103/physrevc.73.025205 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.192001 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.108.142001 [ * * , ( ) ] link:\doibase 10.1103/physrevc.89.045206 [ * * , ( ) ] link:\doibase 10.1103/physrevc.80.054001 [ * * , ( ) ] \doibase http://dx.doi.org/10.1146/annurev.nucl.47.1.67 [ * * , ( ) ] http://stacks.iop.org/0034-4885/41/i=3/a=002 [ * * , ( ) ] link:\doibase 10.1103/physrevc.76.025209 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.82.687 [ * * , ( ) ] link:\doibase 10.1103/physrevc.52.1232 [ *
* , ( ) ] _ _ , @noop ph.d .
thesis , ( ) link:\doibase 10.1146/annurev - nucl-100809 - 131956 [ * * , ( ) ] _ _ , @noop ph.d .
thesis , ( ) _ _ , @noop ph.d .
thesis , link:\doibase 10.1016/0550 - 3213(71)90068-x [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0168-9002(03)01001-5 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0168-9002(03)00429-7 [ * * , ( ) ] @noop ( ) , link:\doibase 10.1103/physrevc.67.055204 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0370-2693(99)00940-5 [ * * , ( ) ] link:\doibase 10.1103/physrevc.51.38 [ * * , ( ) ] link:\doibase 10.1103/revmodphys.41.205 [ * * , ( ) ] | background : : the deuteron plays a pivotal role in nuclear and hadronic physics , as both the simplest bound multi - nucleon system and as an `` effective neutron target ''
. quasi - elastic electron scattering on the deuteron is a benchmark reaction to test our understanding of deuteron structure and the properties and interactions of the two nucleons bound in the deuteron .
purpose : : the experimental data presented here can be used to test state - of - the - art models of the deuteron and the two - nucleon interaction in the final state after two - body breakup of the deuteron . focusing on polarization degrees of freedom , we gain information on spin - momentum correlations in the deuteron ground state ( due to the d - state admixture ) and on the limits of the impulse approximation ( ia ) picture as it applies to measurements of spin - dependent observables like spin structure functions for bound nucleons .
information on this reaction can also be used to reduce systematic uncertainties on the determination of neutron form factors or deuteron polarization through quasi - elastic polarized electron scattering .
method : : we measured the beam - target double spin asymmetry ( @xmath0 ) for quasi - elastic electron scattering off the deuteron at several beam energies ( [email protected] gev , 2.5 gev , 4.2 gev and [email protected] gev ) , using the cebaf large acceptance spectrometer ( clas ) at the thomas jefferson national accelerator facility . the deuterons were polarized along ( or opposite to ) the beam direction . the double spin asymmetries were measured as a function of photon virtuality @xmath2 ( [email protected] ( gev@xmath3)@xmath4 ) , missing momentum ( @xmath5 gev@xmath3 ) , and the angle between the ( inferred ) `` spectator '' neutron and the momentum transfer direction ( @xmath6 ) . results : : the results are compared with a recent model that includes final state interactions ( fsi ) using a complete parameterization of nucleon - nucleon scattering , as well as a simplified model using the plane wave impulse approximation ( pwia ) . we find overall good agreement with both the pwia and fsi expectations at low to medium missing momenta ( @xmath7 gev/@xmath8 ) , including the change of the asymmetry due to the contribution of the deuteron d - state at higher momenta . at the highest missing momenta
, our data clearly agree better with the calculations including fsi .
conclusions : : final state interactions seem to play a lesser role for polarization observables in deuteron two - body electro - disintegration than for absolute cross sections . our data , while limited in statistical power , indicate that pwia models work reasonably well to understand the asymmetries at lower missing momenta . in turn
, this information can be used to extract the product of beam and target polarization ( @xmath9 ) from quasi - elastic electron - deuteron scattering , which is useful for measurements of spin observables in electron - neutron inelastic scattering . however , at the highest missing ( neutron ) momenta , fsi effects become important and must be accounted for . |
it is known that quantum chromodynamics ( qcd ) , has two very important properties , namely asymptotic freedom and confinement .
the former implies that at high momentum transfers , the quarks behave as quasi free particles , i.e. , the interaction between two quarks due to gluon interchange can be treated using perturbation theory . for this momentum range
, the dispersion processes can be then be calculated with a very good precision . by contrast ,
at low momentum exchange among the quarks ( @xmath0 gev ) qcd is highly nonlinear and leads to quark confinement .
several approximate methods have been developed to study the physical processes among quarks in the low momentum range of qcd . among them
is lattice qcd , which tries to solve the qcd equations of motion numerically on a discretized space - time grid@xcite .
this method , however , presents problems if extended to finite chemical potentials@xcite .
another option is to use effective qcd models , like the nambu jona - lasinio model ( njl ) , which is described by a lagrangian which accounts for the main features of qcd at low energies .
the advantage of njl models is that they can be extended to finite chemical potentials easily .
local as well as non - local extensions of the model have been studied in the literature ( see , for instance , refs . , and references therein ) .
non - local extensions of the njl model at zero temperature have been used to study hybrid stars with and/or without a quark - hadron mixed phases in their inner cores ( see ref . , and references therein ) . here
, we present an extension of such studies to finite temperatures in order to explore the role of a quark - hadron phase transition for proto - neutron stars or core collapse supernovae .
the euclidean effective action of the model presented in this paper is for the non - local 3-flavor polyakov njl model , including the vector interactions among quarks .
this action is given by @xmath1 \psi(x ) - \frac{g_s}{2 } \left [ j_a^s(x ) \
j_a^s(x ) + j_a^p(x ) \
j_a^p(x ) \right ] \right .
\nonumber \\ & & \qquad \qquad \left . - \frac{h}{4 } \
t_{abc } \left [ j_a^s(x ) j_b^s(x ) j_c^s(x ) - 3\ j_a^s(x ) j_b^p(x ) j_c^p(x ) \right]\right .
\nonumber \\ & & \qquad \qquad \left.- \frac{g_{v}}{2 } \left[j_{v}^\mu(x ) j_{v}^\mu(x)\right ] + \ { \cal u } \,[{\mathcal a}(x ) ] \right\}\ ; , \label{eq : s_e}\end{aligned}\ ] ] where @xmath2 , @xmath3 and @xmath4 are the scalar , pseudoscalar and vector currents , respectively , and @xmath5 $ ] is the effective potential related with the polyakov loop . to extend the model to finite temperatures we use the matsubara imaginary time formalism@xcite . after the bosonization of eq .
( [ eq : s_e ] ) we obtain the grand canonical potential for the mean field approximation , from which the phase diagram ( fig .
( [ f1 ] ) ) and thermodynamic quantities such as the specific heat , @xmath6 , and specific entropy , @xmath7 , ( fig .
( [ f2 ] ) ) can be calculated .
, as a function of chemical potential , @xmath8 .
right panel : specific heat , @xmath6 , as a function of @xmath8 .
( cep stand for critical end point . )
the impact of the polyakov loop and vector interactions on @xmath9 is very small and therefore not shown separately .
the locations of irst - order phase transitions are shown by solid lines .
[ f1],width=453 ] the quark current masses and coupling constants in eq .
( [ eq : s_e ] ) can be chosen so as to reproduce the phenomenological values of the pion decay constant , @xmath10 , and the meson masses @xmath11 , @xmath12 , and @xmath13 , as described in ref . . in this work
we choose the strange quark mass to the updated value of @xmath14 mev , with the values of the other parameters given by @xmath15 mev , @xmath16 , @xmath17 and @xmath18 mev .
the vector interaction coupling constant , @xmath19 , has been assigned a canonical value of @xmath20 . finally , we mention that the parameter @xmath21 determines the range of the non - locality in momentum space . because of their short mean - free paths ,
the neutrinos produced in the core of a proto - neutron star are prevented from leaving the star on a dynamical time scale .
the number of lepton - to - baryon ratio of such matter is around @xmath22 , but the exact value depends on , width=453 ] the efficiency of electron capture reactions during the gravitational collapse of the supernova .
the number of muons per baryon is @xmath23 , because no muons are present in the stellar matter prior to the trapping . on a timescale of 10 to 20 seconds ,
the neutrinos diffuse from the star , but leave behind much of their energy which causes significant heating of the ambient matter@xcite .
the temperatures generally achieved in the inner 50% of the stellar core at the peak of the heating are in the range of 30 to 50 mev . following the heating
, the star cools by radiating neutrino anti - neutrino pairs , and consequently the temperature drops off to @xmath24 mev within minutes . a model for the equation of state of such matter , which describes the inner cores of proto - neutron stars ,
is shown in fig.([f2 ] ) .
we have constructed models for the equation of state of hot quark matter in the framework of the n3pnjl model .
knowledge of the equation of state , a possible critical end point ( cep ) , and the nature of the phase transition in the qcd phase diagram itself are critical for a better understanding of the strongly interacting matter at high densities and/or temperatures .
we have found that quark confinement as described by the polyakov loop modifies the phase diagram by increasing the transitions temperatures .
the inclusion of vector interactions among quarks , on other hand , changes the location of the cep .
also , vector interactions stiffen the equation of state , while the inclusion of neutrinos has the opposite effect ( softens the equation of state ) .
one of the main motivations of this study concerns the possible occurrence of quark matter inside of proto - neutron stars .
one of the mechanism for the formation of quark matter in the inner cores of such stars could be the nucleation of quark matter drops due to the high pressure environment ( high compression ) that exists in the centers of proto - neutron star@xcite .
such an investigation , as well as the effect of neutrino trapping in hot beta - stable quark - hybrid matter is currently begin carried out .
g.m . , g.a.c . and m.o .
thank conicet and unlp for financial support . m.o .
acknowledges financial support from american physical society s international research travel award program .
is supported by the national science foundation ( usa ) under grant phy-1411708 . | we study the qcd phase diagram using a non - local su(3 ) njl model with vector interactions among quarks .
we analyze several thermodynamic quantities such as entropy and specific heat , and study the influence of vector interactions on the thermodynamic properties of quark matter . upon imposing electric charge neutrality and baryon number conservation on the field equations ,
we compute models for the equation of state of the inner cores of proto - neutron stars and providing a non - local treatment of quark matter for astrophysics . |
there is a discrepancy between what is experimentally observed in heavy ion collisions @xcite , and theoretical computations of photon emission from a thermalized quark gluon plasma ( tqgp ) @xcite .
computations using realistic equations of state and state of the art computations of photon emission from the tqgp give results whose slope in @xmath0 is too shallow ( falls too slowly in @xmath0 ) and is about a factor of 4 smaller than the observed rate @xcite . even more significant
is the difference between theoretically computed and experimentally observed photon elliptic flow .
photon emission is dominated by early times when temperatures are high and flow is not yet built up .
this early time emission in theoretical calculations leads to a significant under - estimation of elliptic flow compared to experimental data .
this has led to suggestions that the photon rate may be due to late time hadronic processes where the spectrum of photons is enhanced by radial flow effects @xcite .
this is perhaps possible but difficult , as it involves rates for photon production that are quite large , and for @xmath1 photons requires substantial lorentz boosts of fundamental emission processes . on the other hand
, a recent analysis shows that the photon spectrum measured at rhic and lhc for p+p , d+au , au+au collisions at rhic energy and pb+pb collisions at lhc is well described by the geometric scaling hypothesis @xcite .
this follows from theories of gluon saturation , and should be preserved when the expansion is scale invariant , suggesting that the photons are emitted early in time when particle masses are not important . however , one faces the problem of photon flow , an effect which takes a time of the order the transverse size of the colliding region to set in .
in this paper we suggest that a description of the photon spectrum , in particular the photon flow , should be based on the picture that the early dynamics of the collision are described by a glasma , not a thermalized quark gluon plasma .
photon emission from the glasma was considered in a previous work @xcite .
following recent developments @xcite , we will assume that the glasma is described by hydrodynamic equations with perhaps some asymmetry between longitudinal and transverse pressure @xcite .
we fully understand that the assumption that the glasma expands hydrodynamically with near perfect fluid behavior is contentious , and as well not much is known from first principles concerning the distributions of gluons in the glasma @xcite
. we will nevertheless proceed assuming that such near perfect fluid behavior is true , and use the simplest possible model , developed in this paper , to describe the quarks and gluon distributions in the glasma .
hopefully with deeper understanding the semi - quantitative conclusions found here survive a much more sophisticated treatment .
hadron distributions are generated at late times during decoupling , and are affected little by the glasma , except as a remnant of hydrodynamic expansion .
photon emission is however different .
in addition to knowledge of the bulk dynamics , one needs to know the quark and gluon distribution functions as a function of the energy density of the system at early and intermediate times in order to compute the distribution of emitted photons .
we argue that these distribution functions and their time evolution can be quite different between the tqgp and the glasma . a qualitative effect we find is related to the two scales that characterize gluon distributions in the glasma .
there is an ultraviolet scale @xmath2 above which distributions become dilute and an infrared scale @xmath3 at which the distributions are highly coherent . for a thermal system ,
the uv scale is the temperature , and the ir scale is the magnetic has @xmath4 . in the glasma at the initial time , both the infrared and uv scale are equal and are proportional to the saturation momentum .
( within our analysis , we determine the numerical relation between these initial scales and the saturation momentum . )
as time evolves these scales split apart .
thermalization can begin when these scales reach @xmath5 @xcite .
( in the later analysis we will make this criterion more precise . ) for a thermal system , the temperature scale decreases approximately as @xmath6 .
however , for the glasma , we will show below that the infrared scale decreases more rapidly than this but the uv scale decreases more slowly . such evolution
is slowed down even more if there is an asymmetry between longitudinal and transverse pressure .
therefore , if we start both the glasma and the tqgp from the same initial ultraviolet energy scale , it will take longer in the glasma to reach the same ultraviolet energy scale at a lower value .
we will show that this has several consequences : the shape of the photon distribution is changed and the typical time for photon emission increases .
the story for us is a bit more complicated however , as when one properly models both the glasma and the tqgp , the initial temperature for the tqgp must be different from that of the glasma , in order that they can match together at thermalization .
in addition the photons are emitted both from the glasma at earlier times and the tqgp at later times .
we find nevertheless , that the presence of an early time glasma slows the emission process for photons , and steepens the spectrum of produced photons . for small asymmetry between the longitudinal and transverse pressure ,
we find significant steepening of the photon spectrum and significantly longer times of photon emission .
this longer emission time will allow for elliptic flow effects to become important , since flow develops on time scales of the order of the size of the system , which is parametrically large compared to the time scale set by the inverse saturation momentum .
these effects point in the correct direction but they may not be sufficient to fix the problem when these simple computations presented here are realized in realistic hydrodynamical or glasma simulations
. the simple considerations in this paper may nevertheless be useful in guiding such more detailed computations , where the answer to the photon puzzle can be ultimately resolved .
of course our simple results automatically satisfy geometric scaling , and to what degree this is maintained in more detailed computations remains to be seen .
such scaling is however very general and should be true in ideal gas hydrodynamic models where the initial time is specified as @xmath7 , where @xmath8 is the initial temperature .
in such a circumstance , the initial temperatures is determined by the multiplicity and is proportional to the saturation momentum .
the scaling follows because no dimensional scale is introduced in the hydrodynamic expansion .
we begin with a simple model of the glasma evolution . we will assume the distribution function for gluons in the glasma to be of the form @xmath9 in this equation , the constant @xmath10 is of order 1 .
this distribution becomes a thermal gluon distribution function when the the infrared scale satisfies @xmath11 this is the separation of scales in a thermal system for the relation between the magnetic mass scale and the temperature . at the magnetic mass scale configurations in the quark gluon
plasma become non - perturbative . for @xmath12
this distribution is a classical thermal distribution , @xmath13 , where @xmath14 .
it is cut off at the scale @xmath2 . before reaching the thermalization scale
, this distribution represents an over - occupation of gluonic states . for the time being , we will assume that the quarks and gluons are massless , and that the distribution functions are isotropic .
the distribution @xmath15 can be shown to follow from a statistical matrix @xmath16 , which is a generalization of the partition function .
@xmath17 where @xmath18 the gluon distribution then follows via @xmath19 for the quarks , we take @xmath20 there is no enhancement for the quark degrees of freedom because they are fermions , as there can be no multiple occupation of fermion states . the entropy associated with these distributions is @xmath21 using distributions ( [ eq : gluons ] ) and ( [ eq : quarks ] ) to express the energy density , pressures , and number densities in terms of the quantities @xmath2 and @xmath3 we find @xmath22 for gluons and @xmath23 for quarks , where @xmath24 is the number of light quark flavors .
the numbers of quarks and gluons are @xmath25 and @xmath26 it is amusing that the entropy for the gluons is not of order @xmath27 .
the disappearance of this factor is because the leading order term which is associated with a classical field cancels .
the classical term comes when @xmath15 is large , and in this limit the first and second terms in our expression for the entropy are both of order @xmath28 and they cancel .
one picks up the non leading term in the entropy density @xmath29 , so that the entropy is of order @xmath30 .
this means that the entropy of the initial state is low compared to the gluon number density , @xmath31 . at thermalization , when @xmath32 , the number density and entropy density are of the same order of magnitude .
conformal invariance of the presented expressions guarantees that @xmath33 for isotropic distributions , and this is satisfied for our expressions
.. we will also be interested in the case where there is a momentum anisotropy .
we can introduce such an anisotropy into the momentum distributions of the particles , using that the longitudinal pressure is taken to be @xmath34 with @xmath35 .
this will modify the expansion dynamics through the evolution equation which follows from @xmath36 dimensional hydrodynamics , @xmath37 this will have the energy density fall as @xmath38
. the corresponding modification of the momentum distributions can be parametrized by @xmath39 , where @xmath40 is the direction of the anisotropy and @xmath41 characterizes its strength @xcite .
following @xcite we find the relation @xmath42 note that @xmath43 and @xmath44 as it should be . for @xmath45
this leads to a particle number reduced by a factor of @xmath46 @xcite . in our computations
, we will consider a fixed momentum anisotropy .
this is slightly different from what is done in standard hydrodynamic computations , where the viscosity to entropy density ratio , @xmath47 is taken to be a constant . in our case , we interpret this ratio as scattering time divided by a thermal wavelength , @xmath48 .
we will argue in the next section that the scattering time for the out of equilibrium glasma is proportional to the time , @xmath49 . in this case
, the expression derived from hydrodynamics becomes @xmath50
the parameter @xmath51 can be determined by knowing the ratio of the initial number of gluons to quarks in the hadron wavefunction . at the initial time
, we have @xmath52 .
the ratio of the total number of gluons to quarks and antiquarks of all flavors is therefore @xmath53 where @xmath24 is the number of quark flavor degrees of freedom important at the saturation momentum scale .
we will take @xmath54 ignoring the effects of charm quarks . in the mstw parton distribution functions
@xcite , there is about an order of magnitude difference between the number of up quarks and the number of gluons , so that @xmath55 taking @xmath56 , this suggests that @xmath57 . the way that the time evolution is computed is from identifying the time with the collision time in the transport equations @xcite .
dimensional arguments give @xmath58 we further use that the energy density scales according to the hydrodynamic equations as @xmath59 .
the factor of @xmath60 allows for an asymmetry between longitudinal and transverse pressure as discussed above .
recent numerical results suggest that @xmath60 might be close to @xmath61 , corresponding to equal longitudinal and transverse pressures @xcite .
we will allow for some anisotropy in our analysis , but in fact find a good description of the photon spectrum and significant increase in emission times with @xmath62 . using the expression for the energy density in terms of @xmath2 and @xmath3 , eqs.([eq : edengluons ] ) and ( [ eq : edenquarks ] ) , we can determine the evolution of these energy scales .
note that this can also be done directly from results for @xmath63 obtained from hydrodynamic simulations .
we do not need to require a power law in time . nevertheless , the simplified structure we get when we assume a power law in time for the evolution of the energy density makes it useful to consider .
let us first assume that in the glasma phase the evolution is dominated by the gluons and that the coupling constant is fixed .
then we find @xmath64 and @xmath65 note that the ultraviolet scale which corresponds roughly to a temperature falls less rapidly than is the case for a thermal system . for a thermally equilibrated system @xmath66 , and @xmath67 .
let us consider how flow would affect a thermal system .
let us ask what flow ensues if we are at some fixed temperature t. for a thermal system , the initial temperature is given in terms of the particle multiplicity as @xmath68 then @xmath69 where in this equation @xmath70 is the center of mass energy per nucleon . from this
we learn that processes at a fixed temperature will happen at a time which scales like the radius of the system and the fourth root of the beam energy . on the other hand flow effects will become important at a time of order @xmath71 .
so given a fixed temperature scale , whether or not flow is important depends on how large is @xmath72 .
this is independent of the size of the system .
flow effects at some temperature scale are important or not depending upon how high the beam energy is and what value of @xmath73 we choose . in the glasma , the evolution of the ultraviolet scale is slower than the evolution of the temperature for the tqgp . in the gluon dominated case
the uv scale , for @xmath74 , cools off as @xmath75 , with the exponent @xmath76 , which is always less than a third .
however , when @xmath77 , we reach the thermalization time @xmath78 for the range of @xmath60 at hand , @xmath79 , and using @xmath80 and @xmath81 , we find that thermalization may occur at a time of 1 - 2 orders of magnitude larger than that of the initial time scale @xmath82 .
this can correspond to quite large times of order @xmath83 .
for the case of @xmath84 , the energy density of the tqgp and that of the glasma decrease at the same rate .
so when the system thermalizes , it thermalizes at the same temperature .
however , the time it has taken to get to this scale is longer for the glasma because it took a longer time to evolve to this scale .
this is consistent because the initial temperature of the tqgp phase was higher than that of the glasma . on the other hand for @xmath85 ,
the situation is a bit more complicated .
energy densities at the thermalization time must be equal , but because the glasma is anisotropic , the average temperature will be lower in the isotropic thermal system .
also , the number of particles will not be conserved .
for the cases we consider these effects are small but they must be taken into account .
including gluons and quarks , for general @xmath60 in the glasma case we need to solve @xmath86 numerically for @xmath87 .
@xmath88 follows from @xmath89 .
the initial values @xmath90 and @xmath91 are chosen to fulfill the requirement of energy conservation when thermalizing at the time @xmath92 , which is the time when @xmath93 ( see eq.([eq : eta ] ) ) is equal to @xmath94 .
the energy density in the thermal case is given by @xmath95 while in the glasma we have @xmath96 where @xmath97 $ ] , which results from having an anisotropic momentum distribution in the case of general @xmath60 .
@xmath41 can be obtained from the relation ( [ eq : deltaxi ] ) for a given @xmath60 .
given an initial temperature @xmath8 , the requirement @xmath98 can be fulfilled by iteratively solving for the initial @xmath99 and @xmath100 .
the particle number in the thermal case is constant in time and given by @xmath101 in the glasma the particle number depends on time and we need to evaluate the expression @xmath102 at time @xmath103 . for @xmath104 expressions ( [ eq : dndytherm ] ) and ( [ eq : dndyglasma ] ) agree at time @xmath92 . for general @xmath60
there will be a difference due to the anisotropic momentum distributions in the glasma expressions . for collisions at lhc energies
we take @xmath105 , @xmath106 , @xmath107 , @xmath108 .
figs.[fig : energyscales - delta1over3 ] and [ fig : energyscales - delta1over4 ] show the resulting time evolution of @xmath73 , @xmath2 , and @xmath3 for @xmath104 and @xmath109 , respectively .
for @xmath104 the thermalization time is @xmath110 . in this case
the temperature and multiplicity agree exactly with the thermal system at the time of thermalization . for @xmath109
the thermalization time is @xmath111 .
as one can see in fig.[fig : energyscales - delta1over4 ] , the ultraviolet scales are not exactly the same at this time , which is caused by the anisotropic momentum distribution in the glasma case . in a similar way the particle number in the glasma at thermalization differs from that in the thermal system by approximately @xmath112 . in a thermal system and
the ultraviolet scale @xmath2 and infrared scale @xmath3 in the glasma for @xmath104 .
[ fig : energyscales - delta1over3],width=453 ] in a thermal system and the ultraviolet scale @xmath2 and infrared scale @xmath3 in the glasma for @xmath109 .
[ fig : energyscales - delta1over4],width=453 ]
the leading order photon production rates from compton and annihilation processes in a thermalized qgp are given by @xcite @xmath113 in the glasma we use the generalized expression @xmath114 eq.([eq : glasmarate ] ) is an approximation ignoring the anisotropy in the momentum space distribution of the quarks and gluons when @xmath115 .
the generalized rates taking this anisotropy into account have been computed in @xcite . because here we are focusing on the effect of the energy scales and not on the rapidity dependence of photon production this approximation is appropriate .
we further use the fact that the gluon distribution in the initial or final state is large compared to 1 .
this leads to the simple form ( [ eq : glasmarate ] ) where we get an additional factor @xmath116 from the gluon distribution .
we chose the factor under the logarithm in the glasma rate as to make the rates equal at the time of thermalization when @xmath117 .
in figs.[fig : photonyield-1over3 ] and [ fig : photonyield ] we show the time integrated photon yields for the thermal case @xmath118 and the case of an initial glasma stage @xmath119 assuming @xmath120 and using a @xmath121-factor of 7 ( for @xmath104 ) or 5 ( for @xmath109 ) , which accounts for logarithmic uncertainties in the rate , neglected collinear enhancement of bremsstrahlung , neglected non - collinear processes induced by jets @xcite , next - to - leading order ( nlo ) effects and photons from hadronic sources . in the regarded momentum range , these processes should not significantly modify the shape of the spectrum .
we also include a curve assuming glasma emission for the entire lifetime of the system .
the photon distribution from the glasma is clearly steeper than it is in the thermal case . because we are using the thermal rates for times greater than the thermalization time in the combined evolution , low momentum photons that are emitted at later times ,
closely follow the thermal distribution .
higher momentum photons ( @xmath122 for @xmath104 and @xmath123 for @xmath109 ) on the other hand dominantly emerge from the glasma . , @xmath104 ) ( dotted ) and early time glasma combined with later time thermal evolution ( solid ) compared to experimental data from the alice collaboration @xcite .
[ fig : photonyield-1over3],width=453 ] , @xmath109 ) ( dotted ) and early time glasma combined with later time thermal evolution ( solid ) compared to experimental data from the alice collaboration @xcite .
[ fig : photonyield],width=453 ] .mean photon emission time in a thermal system compared to the glasma case with isotropic pressures ( @xmath104 ) and a pressure anisotropy ( @xmath109 ) . in the glasma case
we switch to thermal emission after the thermalization time @xmath92.[tab : meantimes ] [ cols="^,^,^,^,^",options="header " , ] for a more quantitative analysis we compute typical emission times for photons of different energies according to @xmath124 and @xmath125 where @xmath126 .
results for three different photon energies are shown in table [ tab : meantimes ] .
we find that photon emission is shifted to significantly later times in the glasma compared to the thermal system . for @xmath109 ,
the thermalization time is @xmath127 , for @xmath104 we find @xmath128 .
mean emission times that are significantly larger ( as is the case for e.g. @xmath129 photons ) indicate that a large fraction of photons are produced in a thermalized system , as we expected from the photon distribution shown in fig.[fig : photonyield ] . the slower expansion for @xmath109 causes emission times to increase , except at the lowest photon energies , where @xmath130 is smaller than for @xmath104 .
this is because for @xmath109 thermalization is reached at a slightly earlier time than for @xmath104 .
furthermore , because of the anisotropy , the rates make a small jump down at the thermalization time . at higher photon energies ,
the mean emission times are significantly smaller than the thermalization time and are not affected by this .
the later photon emission times for a system that undergoes evolution through a glasma phase can have an important effect on the photon elliptic flow .
the later the photons are produced the more flow the system will have built up by the time of emission .
this effect could help explain the large experimentally observed photon elliptic flow .
we have shown in a simple model how photons might be emitted during the glasma phase of ultra - relativistic heady ion collisions .
our computations explicitly build in the geometric scaling of photon distributions produced in such collisions .
our computations give a reasonable agreement with the shape of the photon spectrum , a feature which can not be extracted by simple scaling arguments .
the time for emission of these photons in a combined glasma+tqgp computation is significantly longer than is the case for a pure tqgp computation . whether or not these increased times are sufficient to produce the observed flow seen in experiment can be determined by realistic 3 + 1 dimensional computations .
this paper outlines some of the ingredients which may be needed in such a computation .
the research of l.m . and b.p.s .
is supported under doe contract no .
de - ac02 - 98ch10886 .
larry mclerran would like to thank klaus reygers and johanna stachel who organized the emmi rapid reaction task force meeting `` direct photon flow problem '' .
he gratefully acknowledges the hospitality of emmi .
this work was initiated during this meeting , and was motivated by conversations specifically with charles gale and ulrich heinz , and more generally as a result of dialog with participants in the meeting .
00 a. adare _ et al . _
[ phenix collaboration ] , phys .
lett . * 109 * , 122302 ( 2012 ) [ arxiv:1105.4126 [ nucl - ex ] ] .
d. lohner [ alice collaboration ] , j. phys .
ser . * 446 * , 012028 ( 2013 ) [ arxiv:1212.3995 [ hep - ex ] ] . c. shen , u. w. heinz , j. -f .
paquet , i. kozlov and c. gale , arxiv:1308.2111 [ nucl - th ] . c. shen , u. w. heinz , j. -f .
paquet and c. gale , arxiv:1308.2440 [ nucl - th ] .
h. van hees , c. gale and r. rapp , phys .
c * 84 * , 054906 ( 2011 ) [ arxiv:1108.2131 [ hep - ph ] ] .
e. l. bratkovskaya , s. m. kiselev and g. b. sharkov , phys .
c * 78 * , 034905 ( 2008 ) [ arxiv:0806.3465 [ nucl - th ] ] .
o. linnyk , v. p. konchakovski , w. cassing and e. l. bratkovskaya , phys .
c * 88 * , 034904 ( 2013 ) [ arxiv:1304.7030 [ nucl - th ] ] .
c. klein - bsing and l. mclerran , arxiv:1403.1174 [ nucl - th ] .
m. chiu , t. k. hemmick , v. khachatryan , a. leonidov , j. liao and l. mclerran , nucl .
phys . a * 900 * , 16 ( 2013 ) [ arxiv:1202.3679 [ nucl - th ] ] .
t. epelbaum and f. gelis , phys .
lett . * 111 * , 232301 ( 2013 ) [ arxiv:1307.2214 [ hep - ph ] , arxiv:1307.2214 [ hep - ph ] ] . j. -p .
blaizot , f. gelis , j. -f .
liao , l. mclerran and r. venugopalan , nucl .
a * 873 * , 68 ( 2012 ) [ arxiv:1107.5296 [ hep - ph ] ] .
a. kurkela and g. d. moore , jhep * 1112 * ( 2011 ) 044 [ arxiv:1107.5050 [ hep - ph ] ] .
j. berges , k. boguslavski , s. schlichting and r. venugopalan , arxiv:1303.5650 [ hep - ph ] .
j. berges , k. boguslavski , s. schlichting and r. venugopalan , arxiv:1311.3005 [ hep - ph ] .
j. berges , k. boguslavski , s. schlichting and r. venugopalan , arxiv:1312.5216 [ hep - ph ] .
p. romatschke and m. strickland , phys .
d * 68 * , 036004 ( 2003 ) [ hep - ph/0304092 ] .
m. martinez and m. strickland , phys .
c * 79 * , 044903 ( 2009 ) [ arxiv:0902.3834 [ hep - ph ] ] . m. martinez and m. strickland , nucl .
phys . a * 830 * , 615c ( 2009 ) [ arxiv:0907.3893 [ hep - ph ] ] .
m. martinez and m. strickland , nucl .
a * 848 * , 183 ( 2010 ) [ arxiv:1007.0889 [ nucl - th ] ] .
p. romatschke and m. strickland , phys .
d * 70 * , 116006 ( 2004 ) [ hep - ph/0406188 ] .
j. i. kapusta , p. lichard and d. seibert , phys .
d * 44 * , 2774 ( 1991 ) [ erratum - ibid .
d * 47 * , 4171 ( 1993 ) ] .
b. schenke and m. strickland , phys . rev .
d * 76 * , 025023 ( 2007 ) [ hep - ph/0611332 ] .
s. turbide , c. gale , e. frodermann and u. heinz , phys .
c * 77 * , 024909 ( 2008 ) [ arxiv:0712.0732 [ hep - ph ] ] .
m. wilde [ alice collaboration ] , nucl .
phys . a * 904 - 905 * , 573c ( 2013 ) [ arxiv:1210.5958 [ hep - ex ] ] . | we introduce distribution functions for quarks and gluons in the glasma and discuss how they satisfy various relationships of statistical physics .
we use these distributions to compute photon production in the early stages of heavy ion collisions .
photon rates satisfy geometric scaling , that is , the emission rate per unit area scales as a function of the saturation momenta divided by the transverse momentum of the photon .
photon distributions from the glasma are steeper than those computed in the thermalized quark gluon plasma ( tqgp ) .
both the delayed equilibration of the glasma and a possible anisotropy in the pressure lead to slower expansion and mean times of photon emission of fixed energy are increased .
this delayed emission might allow for larger photon elliptic flow . |
the wmap results ( spergel et al . 2003 ) have given rather strict limits on the matter and energy content of the universe , in agreement with the @xmath0cdm model .
typically , for @xmath1@xmath2 = 1 , @xmath1@xmath3 = 0.73 , and @xmath1@xmath4 = 0.27 , with @xmath1@xmath5 = 0.044 , and @xmath1@xmath6 = 0.23 .
some of these results depend on data from the 2df galaxy redshift survey .
initial results from the sdss redshift survey are not much different ( tegmark et al .
the ratio of baryonic to non - baryonic matter is thus 1 to @xmath7 5 - 6 .
the baryonic matter at z = 0 is mostly not in stars ( @xmath1@xmath8 = 0.005 ) , and the matter budget in e.g. the local group is still not well known : there could be hot gas in substantial amounts ( nicastro et al .
2003 ) .
these results further support the @xmath0cdm model for structure formation in the universe , and underscore the need to understand galaxies in the framework of this theory .
however , at the scale of galaxies , the `` predictions '' of the @xmath0cdm model depend on numerical simulations , which , despite their sophistication , suffer from inadequate resolution , and may miss some of the physics .
this meeting aims to see how well specific predictions from current cosmological simulations of dark matter halos fare on issues such as their central density profiles , the relation between dark to baryonic matter there , and their shape and extent .
over the last 30 years observations , in particular rotation curves derived from hi data , have clearly established that the `` visible '' parts of a spiral galaxy , i.e. the stellar bulge and disk , and the hi gas layer , can not account for the observed high rotation speeds in the outer parts of spirals , if the stellar mass - to - light ratio is constant with radius .
this discrepancy is attributed to unseen matter , often taken to be a ( spherical ) dark halo .
several issues complicate this picture : + _ assumptions going into the derivation of the rotation curve_. a large fraction of galaxies have a warped hi layer in the outer parts , which is modelled with a `` tilted ring '' model in which the orientation of the rings is gradually changed to allow for the peculiarities of the velocity field .
non - axisymmetric disturbances , such as bars or oval distortions , or peculiar velocities due to spiral structure , are in first instance ignored , and an azimuthally averaged rotation curve derived .
asymmetries , if small enough , are likewise ignored .
all these assumptions may introduce systematic errors in the derived rotation curves at the 5 - 15% level .
+ _ values of the mass - to - light ratio are not well determined_. this leads to a degeneracy in the mass models . at one extreme , the maximum disk approach tries to fit as much of the mass into the disk as possible , without overshooting the rotation curve . for optical curves , most galaxies
can then be fitted without a dark halo , due to the limited extent of the curve . in figure 1 i plot , for a sample of galaxies studied by palunas & williams ( 2000 ) ,
the m / l@xmath9-ratio they determined from maximum disk fits against the extinction corrected ( b - i ) colour i collected from the leda database . for large galaxies
a trend exists : on average redder disks have higher m / l ratios , as expected e.g. in models from bell & de jong ( 2001 ) .
yet there is a fair amount of scatter in each class of galaxies selected according to their maximum rotation velocity v@xmath10 , in particular for smaller ( bluer ) galaxies .
the required m / l values are too high compared to those predicted by bell & de jong s `` preferred '' initial mass function ( imf ) , the `` scaled imf '' = 0.7 * salpeter curve , based on maximum disk + halo fits for ursa major galaxies ( verheijen 1997 ) .
the extreme maximum disk models of the galaxies in the palunas & williams sample require a bottom heavier imf .
this problem is aggravated for low surface brightness ( lsb ) disk galaxies , as is shown by the few points taken from models of such galaxies by de blok & bosma ( 2002 ) .
this clearly illustrates that the dark matter problem in these galaxies is much worse , and has lead to the assumption that even a `` minimum disk '' ( m / l = 0 ) is a reasonable approximation to the real mass distribution .
+ _ scaling the hi can produce a good fit instead of a halo .
_ bosma ( 1978 ) noticed that the surface density ratio of hi gas mass to total mass becomes constant with radius in the outer parts .
reversely , one can thus scale the curve for the hi gas with a certain factor to fit the rotation curve , instead of with a halo .
hoekstra et al .
( 2001 ) did this for a number of spirals , and swaters ( 1999 ) for a number of dwarf galaxies .
the scaling factor is not constant , and varies with the maximum velocity of a galaxy ( cf .
figure 2 ) . for a number of galaxies
the fits may well break down at larger radii if more sensitive hi data become available .
cosmological numerical simulations invariably produce cuspy dark halos , but the precise value of the inner slope @xmath11 of the radial density profile @xmath12 is debated .
moore et al .
( 1999 ) and fukushige & makino ( 2001 ) advocate @xmath11 = -1.5 , and navarro , frenk & white ( 1996 , 1997 ) an inner slope of -1.0 . both white and navarro ( this volume ) report on work by hayashi et al .
( 2003 ) : there is no real convergence towards a unique value of the inner slope , the resolution prohibits predictions to be made inside a radius of @xmath7 1 kpc , but the profile remains cuspy , and there is some cosmic scatter ( see also fukushige , kawai & makino 2003 ) . this situation differs from an until recently widespread perception that the nfw profile , with inner slope
-1.0 , should be considered `` universal '' , and the yardstick against which the observations should be discussed .
i will use the slope -1.0 here as a fiducial mark to see where the observations stand , and which is to be improved as the cosmological numerical simulations become more realistic ( e.g. by having adequate resolution on the dwarf galaxy scale , and perhaps incorporating more relevant physics ) , and better understood theoretically ( e.g. by demonstrating which physical process , or combination of processes , sets up the value of the inner slope in dark matter dominated galaxies ) .
most observers conclude that @xmath11 in lsb galaxies is closer to 0.0 than to -1.0 , and almost all find that the decomposition mass models work better if the dark halo is modelled with a ( pseudo- ) isothermal sphere , or a burkert profile , rather than with the nfw profile .
however , some authors maintain that the problem of determining the inner slope is fraught with systematic effects , which all magically contribute towards shallower slopes .
the issue debated is not only about the best value for the inner slope , but also whether a slope of -1.0 can be reasonably excluded , so as to force a modification of the @xmath0cdm picture at galaxy scales .
see also the contributions by de blok and by swaters ( this volume ) . on the selection side ,
the debate is about the quality and relevance of the data .
as more galaxies are observed , stricter selection criteria can be used to retain only galaxies with sufficiently `` good '' data .
thus one excludes 1 ) galaxies with poor angular resolution , thereby requiring in most cases supplementary h@xmath11 data in addition to a hi rotation curve , so that the inner 1 kpc of a galaxy is well probed .
this excludes far away galaxies or galaxies for which there is only low resolution hi data , and thus minimizes slit width effects ; 2 ) edge - on galaxies , which are apparently too difficult to understand for some workers in the field , despite demonstrations by bosma et al .
( 1992 ) and matthews & wood ( 2001 ) that small galaxies seen edge - on are transparent and have rotation curves which are slowly rising in the inner parts .
the argument that somehow there is no emission at the tangent point leads to very peculiar h@xmath11 distributions in more face - on galaxies which are yet to be seen ; 3 ) galaxies with large asymmetries , faint emission , etc . , which have low quality rotation curves .
de blok , bosma & mcgaugh ( 2003 ) applied these criteria to data of mcgaugh et al .
( 2001 ) and de blok & bosma ( 2002 ) , and find for a restricted sample that the values of the inner slope are between 0.2 and -1.0 ( see de blok , this volume )
. swaters et al .
( 2003a ) , in a similar study , come to a similar conclusion .
de blok et al .
( 2003 ) also report about deliberate slit offsets and misalignments for one galaxy , ugc 4325 , and conclude that these effects could not magically bring the slope from about -0.2 to -1.0 , although of course scatter is introduced if the major axis is not observed correctly . finally , for several galaxies
there are more than one observation , done with different telescopes , and by different observers . even though the error bars are sometimes unsatisfactory large , there is good agreement in general
the situation for ugc 4325 , debated at the meeting , is shown in figure 3 .
the hi position - velocity diagram prompts me to think that the peak velocity is indeed around @xmath7 110 km / s as it is in the optical , after which the rotation curve declines .
the determination of the kinematical center , and the influence of non - circular motions on the rotation curves are also debated . the rather irregular nature of some of the dwarf or lsb galaxies studied , and the faintness of particularly the near - infrared images makes the determination of the center of such galaxies difficult .
interestingly , hst data on central star clusters in late type spirals is available for a number of the galaxies used in the core / cusp studies , and comparison of these positions and the centers of the inner disks show agreement to within 1 - 2 arcsec ( bker et al .
2002 ) . for the galaxy
ic 2574 , i reanalyzed the hi data obtained by walter & brinks ( 1999 ) , and deliberately varied the position of the center .
in some cases i also carved up the galaxy into annuli , and allowed the center of each annulus to vary with respect to the next one .
despite the variation of the center positions with an r.m.s .
scatter of about 250 pc , the determined slopes still come out to be -0.15 @xmath13 0.38 , well away from the fiducial nfw slope of -1.0 . bars could contribute to non - circular motions in the inner parts , and thus lead to a poor quality rotation curve when long slit data are used , or when a two - dimensional velocity field is not properly analyzed .
swaters et al .
( 2003a ) assert that bars in their long slit data sample predominantly have shallow slopes , but inspection shows that ugc 2259 , ugc 4499 , and ugc 5721 , as well as f568 - 3 are also barred , in addition to the galaxies they identify as such . for the remaining unbarred galaxies
, there is a wide variety of slopes .
swaters et al .
( 2003b ) observed the galaxy ddo 39 with an integral field spectrograph yielding a twodimensional velocity field , which shows that the inner parts of this galaxy are affected by non - circular motion caused by a bar - like distortion .
several other studies of twodimensional h@xmath11 velocity fields are underway .
garrido et al . (
2002 ) report on data obtained with a fabry - prot instrument , and bolatto ( this meeting , see simon et al .
2003 ) report on data obtained with an integral field spectrograph , combined with co data obtained with bima .
the results sofar for ngc 4605 and ngc 2976 are that the inner slopes are rather shallow , and that the non - circular motions are mainly in the inner parts .
one can look at bars in lsb galaxies in two ways : 1 ) they could be similar to bars in hsb galaxies , or 2 ) they could be just the response to the expected triaxiality of the dark halo .
since the reported non - circular motions are stronger in the inner parts for ddo 39 and ngc 2976 , it seems likely that those galaxies behave like hsb s , where the disk is dynamically important in the inner parts .
yet it can not be excluded that whole disks are elongated due to the triaxial halo .
one way to study this further is to look at results of a fourier analysis of velocity fields of a number of galaxies , performed by schoenmakers ( 1999 ) . from his work ,
i have collected the results on disk elongation in figure 4a , which shows that the scatter in @xmath14@xmath15 sin @xmath16 is larger for galaxies with smaller maximum rotation velocity .
thus disks in small galaxies could be globally elongated due to the forcing of a triaxial halo .
interestingly , the small galaxy with no elongation , ngc 247 , was declared in van den bosch et al .
( 2000 ) `` the only galaxy found [ sofar ] capable of limiting the slope of the dark matter profile '' .
true enough , its slope is -1.02 , close the the nfw slope , but perusal of schoenmakers s results and a near infrared image show the presence of a bar , and a strong lopsidedness . in future , the effort will thus shift towards evaluating the importance of the disc component in lsb galaxies , using near - infrared surface photometry , and twodimensional velocity fields at high spatial resolution . meanwhile
, some evaluation of the importance of non - circular motions can be had by studying bars in hsb galaxies .
disk elongation can be constrained using the tully - fisher relation ( franx & de zeeuw 1992 )
. early work on the barred spiral ngc 5383 ( athanassoula 1984 ) shows that the viewing angle is important : some angles are more favourable than others . neither the mean rotation curve , nor the position velocity cut along the major axis can be trusted as a good representation of the true rotation curve , but the deviations scale with the strength of the bar . for the weak bars in lsb galaxies , it is not obvious how strong these effects are .
this so - called `` maximum disk '' problem is related to the core / cusp problem , since in the current @xmath0cdm picture the dark halo dominates the potential in the central parts also in high surface brightness spirals .
several ways have been explored to break the mass model degeneracy using other dynamical considerations concerning the importance of the disk in the inner parts of spiral galaxies .
athanassoula et al .
( 1987 , abp ) use swing amplifier criteria , which depend on the rotation curve shape and on a characteristic x parameter dependent on the epicyclic frequency @xmath17 , the number of arms m , and the active surface mass density of the disk . by requiring that the swing amplification of the m = 2 perturbations is possible , the range of mass - to - light ratios is limited to a factor of 2 : a lower limit set by requiring that the disk is massive enough to just allow amplification of the m = 2 perturbations , and an upper limit set by requiring that amplification of the m = 1 perturbations is just prohibited .
usually the latter condition holds for a model with maximum disk and a non - hollow halo .
peculiar motions due to the spiral arms were clearly seen in the early m81 21-cm line data , and modeled with a spiral density wave response calculation by visser ( 1980 ) , who did not include a dark halo in his models .
the presence of `` wiggles '' in position - velocity curves from long slit data are thus associated with the spiral arms .
kranz et al . ( 2001 ) use such data for ngc 4254 and try to reproduce the observed velocity perturbations with a stationary gas flow model using the k - band image of this galaxy as input to the evalution of the disk part of the galactic potential .
they find that a maximum disk model produces too large velocity perturbations , and put an upper limit on the disk mass fraction ( the mass ratio between a given disk model and the maximum disk model ) of 0.8 .
however , this galaxy is lopsided in the hi , the spiral may be evolving , the small bar in the center of the galaxy might have a different pattern speed than the main spiral pattern , the inclination may be higher than the authors take it , and the adopted method may favour lower disk mass fractions ( slyz et al
kranz et al . ( 2003 ) report on a similar analysis for four more cases , and find a trend that the brightest spirals ( those with the highest rotational velocities ) , seem to have maximum disks , but that towards lower luminosity spirals the relative influence of the dark matter in the inner parts increases . comparison with data from abp shows good agreement with this trend ( cf .
figure 4b ) .
weiner et al . (
2001 ) model the stationary gas flow in the barred spiral ngc 4123 , using a potential derived from an optical image , and find that the best fit to the velocity data requires a maximum disk model for the mass distribution .
lindblad , lindblad & athanassoula ( 1996 ) find likewise a relatively good fit for the bright barred spiral ngc 1365 with a maximum disk model .
figure 4 suggests that faster rotators have rounder disks , which are more self - gravitating . a view
not necessarily in contradiction is voiced by courteau et al .
( 2003 ) , who contend that _ on average _ , disks with v@xmath10 @xmath18 200 km / s are sub - maximal .
they argue this on the basis of velocity dispersion data from bottema ( 1997 ) which i deem having too large error bars ,
work on disk stability by fuchs
yes , but see fuchs ( 2002 ) , the absence of the expected correlated scatter in the tully - fisher relation ( courteau & rix 1999 ) but disk maximality seems to depend on v@xmath10 , and the result on the lens 2237 + 0305 . nevertheless , they find that tully - fisher relations for barred and un - barred galaxies are similar , in agreement with previous work , so barredness does not affect maximality .
trott & webster ( 2002 ) combine for 2237 + 0305 their lens model with hi rotation data further out .
there is little need for a dark halo in the central parts , which are dominated by a bulge - bar system .
their statement that the disk is not maximal is partly influenced by their inclusion of the bar into the bulge , even though bars are thought to originate in the disk . for our own galaxy ,
data on the microlensing towards the bulge - bar system likewise suggest that dark matter does not dominate in the central parts ( bissantz & gerhard 2002 ) .
the @xmath0cdm simulations predict the presence of far more satellite galaxies than is observed around the milky way , and the concept of failed dwarfs has been advanced to rescue the `` theory '' . in this picture , studies of high velocity clouds are debated , but their individual distances are quite uncertain , and they do nt indeed seem to have stars ( e.g. simon & blitz 2002 ) .
the dwarf spheroidal companions of the milky way are very important in another way : the smallest ones are dark matter dominated .
new data on the draco system ( kleyna et al .
2002 ) show that this galaxy is more extended than previously thought ; stoehr et al .
( 2002 ) produce a good fit with @xmath0cdm .
but for ursa minor there is a subpopulation of stars with low velocity dispersion .
such a cold clump will survive easily in a cored halo potential , but breaks up rather rapidly in a cusped halo potential ( kleyna et al .
2003 ) .
finally , modelling the trail of the sagitarius dwarf galaxy ( e.g. ibata et al . 2001 ) suggests that the dark halo around the milky way is nearly spherical .
stellar kinematics of elliptical galaxies have turned up only 3 unambiguous cases where dark matter is needed to fit the data : ngc 2434 , ngc 7507 and ngc 7626
( kronawitter et al . 2000 ) , partly because the data is limited in radius to 1 - 2 r@xmath19 .
planetary nebulae can now be used as a tracer further out , and is becoming an industry .
interestingly , constant m / l models can explain the new data , though m
/ l values are rather high ( e.g. romanovsky et al .
2003 ) .
for field ellipticals , evidence for dark matter does come from hi studies , which show for several galaxies a flat rotation curve from @xmath7 0.3 r@xmath19 out to 5 - 6 r@xmath19 ( e.g. oosterloo et al . 2002 ) .
this implies a global ( m / l)@xmath20 ratio @xmath7 25 .
this value is in good apparent agreement with a similar value found based on x - ray studies . yet the new chandra and xmm data show a wealth of detail in the images of the x - ray gas of individual galaxies , so much so that one can question the validity of the hydrostatic equilibrium equation used to evaluate masses .
a further complication is the contribution to the x - ray flux of low mass x - ray binaries , occuring presumably in globular clusters . indeed , paolillo et al .
( 2003 ) argue that the core of the x - ray emission is associated with the stellar distribution ( gas from mass loss of evolving stars ) , and that the extended x - ray emission traces really the group or cluster potential rather than the potential of the elliptical galaxy itself .
this agrees with studies using globular clusters as a tracer , e.g. for ngc 4472 and m87 , which show that there is need for dark matter at larger radii there ( see kronawitter et al . 2000 and references therein ) .
a most interesting development is the use of lensing data , combined with more classical spectroscopy , to estimate the mass of ellipticals .
a very good example of this is the study of the lens 0047 - 281 by koopmans & treu ( 2003 ) , who find indications that 1 ) the total density follows a power law with slope -1.9 , 2 ) a constant m / l model can be eliminated using velocity dispersion data , and 3 ) the inner dark matter slope can be as shallow as -1.1 . for some lenses ,
the flux ratios can not be matched with a smooth potential , and therefore substructure needs to be added to the models to produce a fit ( e.g. bradac et al .
this is seen as consistent with the @xmath0cdm prediction .
athanassoula , e. 1984 , physics reports , 114 , 319 athanassoula , e. , bosma , a. , & papaioannou , s. 1987 , , 179 , 23 bell , e. , & de jong , r.s .
2001 , apj , 550 , 212 bissantz , n. , & gerhard , o. 2002 , mnras , 330 , 591 bker , t. , et al .
2002 , aj , 123 , 1389 bosma , a. 1978 , phd thesis , university of groningen bosma , a. , byun , y.i , freeman , k.c .
, & athanassoula , e. 1992 , apj , 400 , l21 bottema , r. 1997 , , 328 , 517 bradac , m. , schneider , p. , steinmetz , m. , lombardi , m. , king , l.j , & porcas , r. , 2002 , , 388 , 373 courteau , s. , andersen , d.r . , bershady , m.a . , macarthur , l.a . , & rix , h .- w .
2003 , apj , 594 , 208 courteau , s. , & rix , h .- w . 1999 , apj , 513 , 561 de blok , w.j.g . ,
& bosma , a. 2002 , , 285 , 816 de blok , w.j.g .
, bosma , a. , & mcgaugh , s.s .
2003 , mnras , 340 , 657 franx , m. , & de zeeuw , p.t .
1992 , apj , 392 , l47 fuchs , b. 2002 , astro - ph/0212485 fukushige , t. , & makino , j. 2001 , , 557 , 533 fukushige , t. , kawai , a. , makino , j. 2003 , astro - ph/0306203 garrido , o. , marcelin , m. , amram , p. , & boulesteix , j. 2002 , , 387 , 821 hayashi , e. et al .
2003 , astro - ph/0310576 hoekstra , h. , van albada , t.s .
, & sancisi , r. 2001 , mnras , 323 , 453 ibata , r. , lewis , g.f . , irwin , m. , totten , e. , & quinn , t. 2001 apj , 551 , 294 kleyna , j.t . , wilkinson , m.i . ,
evans , n.w . , gilmore , g. , & frayn , c. 2002 , mnras , 330 , 792 kleyna , j.t .
, wilkinson , m.i . , gilmore , g. , & evans , n.w .
2003 , apj , 588 , l21 koopmans , l.v.e . , & treu , t. 2003 , apj , 583 , 606 kranz , t. , slyz , a. , & rix , h .- w .
2001 , apj , 562 , 164 kranz , t. , slyz , a. , & rix , h .- w .
2003 , apj , 586 , 143 kronawitter , a. , saglia , r.p . ,
gerhard , o. , & bender , r. 2000 , , 144 , 53 lindblad , p.a.b . , lindblad , p.o . , & athanassoula , e. 1996 , , 313 , 65 marchesini , d. et al .
2002 , apj , 575 , 801 matthews , l.d . , & wood , k. 2001 , apj , 548 , 150 moore , b. , quinn , t. , governato , f. , stadel , j. , & lake , g. 1999 , , 310 , 1147 mcgaugh , s.s . ,
rubin , v.c .
, & de blok , w.j.g .
2001 , aj , 122 , 2381 navarro , j.f . , frenk , c.s .
, & white , s.d.m .
1996 , apj , 462 , 563 navarro , j.f . ,
frenk , c.s .
, & white , s.d.m .
1997 , apj , 490 , 493 nicastro , f. et al .
2003 , nature , 421 , 719 oosterloo , t.a .
morganti , r. , sadler , e.m . ,
vergani , d. , & caldwell , n. 2002 , aj , 123 , 729 palunas , p.a . , & williams , t.e.b .
2000 , aj , 120 , 2884 paolillo , m. , fabbiano , g. , peres , g. , & kim , d .- w .
2003 , apj , 586 , 850 romanowsky , a. et al .
2003 , science , 301 , 1696 schoenmakers , r.h.m , 1999 , phd thesis , university of groningen simon , j.d .
, & blitz , l. 2002 , apj 574 , 726 simon , j.d . ,
bolatto , a.d . , leroy , a. , & blitz , l. 2003 , apj , 596 , 957 slyz , a. , kranz , t. , rix , h .- w .
2003 , mnras ( in press ) , astro - ph/0309597 spergel , d. et al .
2003 , apjs , 148 , 175 stoehr , f. , white , s.d.m . , tormen , g. , & springel , v. 2002 , mnras 335 , l84 swaters , r.a . 1999 ,
phd thesis , university of groningen swaters , r.a .
, madore , b. , van den bosch , f.c . , & balcells , m. 2003a , apj , 583 , 732 swaters , r.a . ,
verheijen , m.a.w . ,
bershady , m.a . , & andersen , d.r .
2003b , apj , 587 , l19 tegmark , m. et al .
2003 , astro - ph/0310273 trott , c.m . , &
webster , r.a .
2002 , mnras , 334 , 621 van den bosch , f.c . , robertson , b.e . ,
dalcanton , j.j , & de blok , w.j.g .
2000 , aj , 119 , 1579 verheijen , m.a.w .
1997 , phd thesis , university of groningen visser , h.c.d .
1980 , , 88 , 159 walter , f. , & brinks , e. 1999 , aj , 118 , 273 weiner , b. , sellwood , j.a . , & williams , t.b .
2001 , apj , 546 , 931
write me a piece , albert .
i d really appreciate it , i wo nt have time to do it myself . "
i am sitting opposite seth shostak , editor of the magellanic times , whom i have known for a long time , and known of for even longer .
he and i owe a lot of intellectual indebtness and gratitude to a quiet man who patiently taught us the principles of 21-cm hi interferometry , david rogstad .
rogstad was one of the first to use a two element radio aperture synthesis interferometer to study the 21-cm line emission of nearby galaxies ( he was loc co - chair john whiteoak s first student . ) seth did his thesis on detailed synthesis observations of 5 late type galaxies with the owens valley radio observatory ( ovro ) twin element interferometer .
dave rogstad moved for a few years to the kapteyn laboratorium in groningen in the late sixties , where he guided a couple of students in various projects - henk olthof on warps , arnold rots on data of ngc 6946 and ic 342 , and me .
he himself worked on his ovro data of m101 . for me , it was my first research project .
i had to read reams of charts to mark off phase zero measurements of fringes from the just completed westerbork telescope .
rogstad taught me computing and offered valuable advice ( beware , if you start computing , you stop thinking . " ) after this , he gave me his 21-cm line data of m82 to reduce , along with all his programs - on punched cards .
i was enthusiastic about this , looked up the literature on this exploding galaxy " , and learned about the galaxy rotation work of burbidge , burbidge and prendergast .
the m82 hi data were interesting , and dave encouraged me to speak about them at a yerac ( young european radio astronomy conference ) in dwingeloo .
dave s work on m101 stimulated the young staff members in groningen , led by ron allen , to push for a temporary receiver to do 21-cm line work on nearby galaxies with the westerbork telescope - which was not explicitly constructed for this purpose . by the time this project was completed ,
i was fortunate enough to use it for my thesis work on a pilot hi study of spiral galaxies of various morphological types .
that was more or less [ considered ] a backwater project , the main attention for 21-cm line work being the study of spiral structure of m51 , m81 and m101 ( testing the density wave theory . " ) even so , i witnessed the debate about rotation curves in the outskirts of galaxies using hi data .
i vividly remember a seminar by mort roberts , arguing that m31 s rotation curve was flat , and all the young turks of the kapteyn lab arguing that he picked up signal of the main disk through the sidelobes of his single dish antenna ( ed salpeter was subjected to a similar treatment a few years later [ at the iau symposium 77 in bad mnstereiffel ] when he reported on his arecibo data , which had some after - effects if his annual reviews autobiography is any guide ) .
so 21-cm line interferometry was the creed .
my thesis work , with its basic result that the rotation curves of spiral galaxies of all morphological types stayed flat ( or were even still rising ) beyond the optical image , helped settle the debate of the presence of dark matter in the outer parts of spirals .
i also was fortunate to be able to go to palomar with piet van der kruit , to get h@xmath11 rotation data in the inner parts , and to obtain plates for surface photometry to determine the luminosity profile and the mass - to - light ratio as function of radius .
but the questions remain - how was the dark matter debate conducted , who said what when and why , and who had the correct " ( i.e. present day accepted ) picture first ?
this is not so easy as some people would have it . in historical development of modern cosmology
( asp conference series vol .
cs-252 ) , i read two articles - one by sidney van den bergh and one by jaan einasto - and there seems to be little overlap .
so there could be different stories , and different people were convinced at different times of the presence of the dark matter in spirals - just like the various events and characters which made up the french revolution . in any case , as far as the hi line rotation curves is concerned , i can only cite vera rubin et al.s ( 1978 ) paper where she writes that mort roberts and his collaborators deserve credit for first calling attention to flat rotation curves " , credit he got only sparingly .
what is clear now is that there is little need for a dark halo to fit the optical rotation curve data , a conclusion [ again ] reached in an apj paper as recently as 2000 .
mort roberts himself confided to me once that he thought that my thesis work vindicated his contention that rotation curves of giant spirals are flat .
ostriker and peebles , who pushed the idea for dark halos , first in a 1973 paper on disk stability ( a paper which nowadays can not withstand close scrutiny ) , and then , with yahil , in their classic 1974 paper on various indicators pointing to a linear increase of the mass of a giant galaxy with radius , cite for the rotation data the papers by roberts and rots ( 1973 ) and rogstad and shostak ( 1972 ) . which brings me back to seth shostak .
contrary to my thesis , which i had printed in 500 copies , and with its handy format became quite in demand ( my thesis advisor had his copy stolen from his office somewhere in the 1980s , he told me recently ) , shostak s thesis ( 1971 ) is not widely available .
i found a copy of it in the vla library , so i started reading it carefully , and yes , it contained the flat rotation curve of ngc 2403 , and its implication that the volume density of matter drops off as r@xmath21 ( mass rises linearly with radius , therefore ) .
so he had it !
okay , only for one case ,
but even so ; and it was a clean observation with an interferometer , so unlike roberts observation of m31 ( the primary piece of evidence in the roberts and rots ( 1973 ) paper ) .
oh well , it is unclear to me sometimes how scientific evidence is accepted in certain circles .
sometimes one has a result , correct after all , but at a time too far ahead of the pack , which is not yet ready to absorb it . in any case , the amusing part of seth s thesis is in the very end .
a brief acknowledgement to gordon stanley , then director of ovro .
a warm thanks to david rogstad for his patient advice .
and then this fantastic flight of fancy - this thesis is dedicated to ngc 2403 and its inhabitants , to whom copies can be furnished at cost " .
no wonder this man is now working on seti ! he just wants to know why nobody came along thus far to claim a copy .
if somebody from that galaxy had come along , seth would have been the most famous man on earth , offered an autographed copy of his thesis for free , and arranged for a public lecture at the current general assembly .
john whiteoak would have been happy to schedule that , and the harbourside auditorium would have been way too small ! | i review the observational side of the present state of the debate about the dark matter in galaxies , with emphasis on the core / cusp problem in low surface brightness galaxies , and the question of maximum / sub - maximum disks in spiral galaxies .
some remarks are made about the dwarf spheroidals around the milky way , and about elliptical galaxies .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
correlation and interference distinguish quantum from classical physics .
the former is manifest in the measurement of many - body coincidences predicted by a quantum joint probability density function ( pdf ) .
some observable correlations can not be realized classically @xcite .
quantum interference is most familiar as a one - body pdf for an outcome that can be achieved in at least two indistinguishable ways .
however , it can also be generated by superposing many - body states in indistinguishable ways @xcite .
a many - body interferometer is more difficult to treat since it is possible to measure correlations in the probability of finding each particle at different positions _ and _ times .
it differs from a one - body interferometer primarily in its beamsplitting mechanism . here ,
correlation and interference are combined in asynchronous correlation interferometry of a bipartite system , using measurement of one substate as the beamsplitter for the other .
experimental confirmation of quantum correlation has involved photons @xcite , atoms @xcite , and josephson phase qubits @xcite .
however , there is little experimental evidence of correlated interference between massive particles . analysis of correlated systems usually begins with an expression for the many - body state after it has been prepared by an interaction .
for bipartite systems this is routinely done with photon pairs from parametric downconversion .
the extremes of either a few or a large number of particles are most often treated . in the former case
the analysis typically involves a simultaneous correlation between the positions @xcite or angular momenta of two particles @xcite .
the formalism often deals only with identical particles . here
a mechanism is describe which generates interferometric correlations between distinguishable particles , which can be of disparate masses .
the issue of non - simultaneous correlations is explored here in the context of perhaps the simplest quantum correlation possible : a particle reflecting elastically from a mirror , both of which are distinguishable .
the mirror and particle have non - zero rest mass and motion is in free space along one dimension , with all states unbound .
measurements of particle reflection , but not associated with correlated interference , have involved mirrors that reflect atoms @xcite and bose - einstein condensates @xcite , atoms reflecting from a solid surface @xcite , neutrons @xcite and atoms @xcite reflecting from vibrating mirrors , and atoms reflecting from a switchable mirror @xcite .
consider first the particle and mirror in uncorrelated eigenstates of energy before reflection , referred to as the ` incident harmonic state ' .
the energy eigenstate after reflection , referred to as the ` reflected harmonic state ' , results in correlation between the particle and mirror via conservation of energy and momentum : an energy or momentum measurement of the reflected particle yields a correlated energy or momentum measurement of the mirror when given the incident harmonic state .
superposing such states yields the incident and reflected particle - mirror wavegroups .
these differ from the harmonic states in that a measurement of the energy ( or position ) of the particle does not uniquely constrain the energy ( or position ) of the mirror since the reflected particle - mirror wavegroup is not in an eigenstate of the energy ( or position ) operator .
such incident and reflected wavegroup particle - mirror states interfere when they overlap .
this is similar to the transient one - body interference of an electromagnetic wavegroup reflecting from a stationary mirror @xcite .
however , classically the mirror experiences only a continuous force due to radiation pressure .
quantum mechanically , interference occurs since the incident and reflected states are indistinguishable for a measurement of position ( but not for a momentum measurement ) .
interference is expected between the incident and reflected particle substates _ along with _ interference of the mirror substates which have and have not reflected the particle .
their correlation is perhaps not expected , being a consequence of the solution to the schrdinger equation from which a joint pdf is constructed .
the correlations in the two - body interference are manifest as coincidence rates , e.g. a correlation in the _ simultaneous _ measurement of particle and mirror positions .
this two - body wavefunction is then modified to incorporate predictions for measurement of the particle first and then later that of the mirror , using the copenhagen interpretation of measurement in quantum mechanics .
the resulting pdf is a function both of the different particle positions and different times at which each is measured .
an assumption used is that between the times of the two measurements , there is neither interaction between the particle and mirror nor with the environment .
the focus of the discussion is on asynchronous correlated interference . in this case , a measurement of only the particle in the correlated interference region splits the mirror substate into ones which have and have not reflected the particle .
later measurement of the mirror reveals this interference .
[ fig : overview ] is a schematic representation of this state splitting process .
actual two - body pdfs are shown in later sections .
the particle pdf before interaction is the gaussian at @xmath0 , moving to the right at speed @xmath1 , while the mirror is represented as the black rectangle , moving to the right at speed @xmath2 . at @xmath3 interference between the incident and reflected particle substates is shown as oscillations in its pdf while interference between the mirror substates which have and have not reflected the particle is represented by the rectangular checkerboard pattern rather than the solid rectangle .
this then is a region of correlated interference , although in this schematic the correlation can not be illustrated . for simplicity ,
speeds are not shown in this and the next snapshot .
measurement of the particle but not the mirror is represented by the ` photon ' coming from below , interacting with the particle , and then being measured by the detector above at time @xmath4 .
more details on this aspect of the process are given in section [ sec : measurementtheory ] .
the lower two schematics illustrate the states with and without this measurement , using the lower and upper halves of each schematic , respectively . without measurement ,
the particle wavegroup reflects from the mirror and continues to move to the right with reduced speed while the speed of the mirror increases . without measurement ,
the correlated interference disappears when the wavegroups no longer overlap as illustrated when @xmath5 .
in addition , only the mirror state which reflects the particle survives ( as is illustrated in the upper half of the @xmath5 schematic ) . this is a consequence of the complete particle wavegroup having reflected . a noise - free and non - destructive measurement of only the particle , however , collapses the particle substate as shown in the lower portion of the @xmath6 schematic , with @xmath7 .
yet the mirror remains in a superposition state . at time @xmath5
the two mirror states have separated a distance greater than the wavegroup size due to their different speeds and are then represented by solid rectangles .
although there is then no interference , the mirror remains in a superposition of having both reflected and not reflected the particle for all later times .
this splitting is similar to a beamsplitter producing photon states which traverse spatially separate and therefore distinguishable paths .
finally , it should be noted that the particle state after measurement spreads , as shown in the lower portion of the @xmath5 schematic . it is assumed that this spreading wavegroup does not interact with the split mirror states for times @xmath8 . to determine the consequences of such measurements , a two - body solution of the schrdinger equation for simultaneous measurement
is first obtained for the particle - mirror system by applying standard techniques to derive the energy eigenstates .
this is used to construct the asynchronous two - body wavefunction .
wavegroups are formed from a superposition of these solutions .
particular emphasis is given to correlated interference or the overlap region of the incident and reflected particle - mirror wavegroups .
a marginal pdf is illustrated , using an example of interference for the particle when the mirror is not measured , which yields an analog of the doppler effect for the interference of light in retro - reflection from a moving mirror .
the remainder of the paper deals predominately with issues which illustrate the unique properties of asynchronous correlation interferometry , which are most easily understood using the example of a microscopic particle reflecting from a mesoscopic / macroscopic mirror .
this example is not intended as a practical experimental proposal .
rather , observation of asynchronous correlations will more likely first occur with a microscopic particle reflecting from a `` microscopic '' mirror .
the particle - mirror interaction is modeled as a moving delta function potential where reflection is assumed to occur at the center of mass of the mirror with the schrdinger equation given by @xmath9+i\partial_{t})\psi=0 , \label{eq : schreqn}\end{aligned}\ ] ] where square brackets are used to indicate the argument of a function and @xmath10 and @xmath11 are the particle and mirror positions along the x - axis .
the mirror reflectivity , related to @xmath12 , goes to infinity for a lossless mirror . the standard separable solution to this equation results from a center of mass ( cm ) and relative ( rel ) transformation of the particle - mirror system ( not to be confused with the cm of the particle or mirror ) .
this does not change the total energy @xmath13 , where @xmath14 and @xmath15 are the wavevectors for the particle and mirror with @xmath16 , @xmath17 , masses @xmath18 , @xmath19 , and initial velocities @xmath1 and @xmath20 , respectively . the transformed schrdinger equation becomes @xmath21 \notag + i\partial_{t})\psi[x_{cm},x_{rel},t]=0 \label{eq : scheqncm}\end{aligned}\ ] ] where @xmath22 , @xmath23 , @xmath24 , and @xmath25 .
the next step in the derivation requires parsing the energy into @xmath26 and assuming the separable solution @xmath27=\psi_{cm}[x_{cm},t ] \psi_{rel}[x_{rel } , t ] \notag \\
= e^{-i e_{cm } t/\hbar } u[x_{cm } ] e^{-i e_{rel } t/\hbar}u[x_{rel } ] , \label{eq : schtot}\end{aligned}\ ] ] which reduces the schrdinger equation into two ordinary differential equations , @xmath28}{dx_{cm}^{2 } } = e_{cm } u[x_{cm } ] \label{eq : scheqode1}\ ] ] @xmath29}{dx_{rel}^{2}}+\beta \delta[x_{rel } ] = e_{rel } u[x_{rel } ] .
\label{eq : scheqode2}\ ] ] the solution to these equations is then obtained from the initial values and boundary conditions .
finally , this solution is transformed back to the particle - mirror system yielding @xmath30 $ ] .
an example of this procedure is found in the solution to the hydrogen atom where the schrdinger equation is first transformed from the laboratory to the cm - relative coordinates yielding _ uncorrelated _ substates .
transforming back to the electron - proton system yields _ correlated _ electron and proton substates @xcite . in subsection
[ sec : non - simultaneous ] , the infinite potential energy boundary condition at the mirror surface is used to obtain energy eigenstate solutions to eqns . [
eq : scheqode1 ] and [ eq : scheqode2 ] . the full solution is then transformed to the particle - mirror system . in subsection
[ sec : wavegroupsimultaneous ] , wavegroups are formed via a superposition of these states .
this solution , @xmath30 $ ] , is next used to construct the wavefunction which predicts the outcome of asynchronous measurements . for illustrative purposes ,
let the particle be measured first .
experimental realizations of such a measurement procedure are discussed in subsection [ sec : measurementtheory ] . to account for different temporal measurements of the particle and mirror , respective time parameters @xmath4 and @xmath31
are used instead of @xmath32 in @xmath30 $ ] .
coefficients of these time parameters in the phase are then the energies of the particle and mirror , respectively .
this is a parsing of the energy in the solution to the schrdinger equation for the particle - mirror system , similar to that applied after eqn .
[ eq : schreqn ] to transform it into the cm and rel system . since that transformation yielded a simultaneous pdf
there was no need for the two - time notation . to modify it for asynchronous predictions ,
relative and cm times are introduced in subsection [ sec : non - simultaneous ] , which then both account for different temporal measurements of the cm and relative positions and isolate the cm and rel energy terms .
however , transformation to the cm - rel system is simply a mathematical technique to facilitate a solution to the schrdinger equation .
for example , there exists no observable particle with a reduced mass as described in the cm - rel system .
once a solution , which satisfies the initial conditions and boundary values , is obtained in the cm - rel system the inverse transformation is applied to describe the system that can be measured .
these time parameters provide labels for the particle and mirror energies in the wavefunction s phase and also label the different times at which the particle and mirror are measured .
they assist in isolating the energy of the particle from that of the mirror , just as @xmath10 and @xmath11 isolate the effect of the particle and mirror wavevectors in the phase of the wavefunction .
this then allows all particle parameters to be fixed in the two - body wavefunction when the particle is measured .
such labeling , while simple in a separable system , is not trivial in the correlated particle - mirror system .
the different time labels do not indicate evolution of the subsystems via different hamiltonians as do the different time variables used by mcguire @xcite nor are they used to generate a manifestly lorentz invariant theory as done by petrat @xcite .
they neither tick at different rates nor are they out of phase , but rather act as only a label , just as @xmath10 and @xmath11 label the particle and mirror spatial coordinates along the @xmath33 axis .
such notation then results in the wavefunction derived in subsection [ sec : theoryparticlemirror ] being expressed as @xmath30 \rightarrow \psi[x_{1},t_{1},x_{2},t_{2}]$ ] .
now let the particle be measured first , at position @xmath34 at time @xmath35 .
this then collapses the particle substate into an eigenstate of that operator , @xmath36=\delta[x_{1}-x_{10}]$ ] .
the first measurement forces the two - body correlated wavefunction into an uncorrelated state which is , at the time that the particle is measured , a product of the particle substate with the mirror substate , given by @xmath36 \psi[x_{10},t_{10},x_{2},t_{10}]$ ] . the particle state after
measurement is irrelevant to the subsequent time evolution of the mirror since they are no longer correlated and no longer interact .
the mirror state is determined by fixing all the particle coordinates in the two - body wavefunction at the time of the measurement of the particle .
the two time formalism facilitates this procedure by fixing the particle parameters while allowing only those of the mirror to time evolve .
the coordinates of this measurement of the particle , @xmath37 , then determine the initial state of the mirror from the two - body wavefunction as @xmath38 $ ] .
this one - body wavefunction for the mirror then continues to time evolve with @xmath39 $ ] for @xmath40 .
for example , the phase of the mirror substate evolves in time only with the mirror s energy times the time @xmath31 .
the particle s energy times the time @xmath4 term in the phase should not influence the time evolution of the mirror substate after the particle has been measured .
it does not in this formalism since all of the energy terms associated with the particle are fixed at @xmath41 .
similarly , the particle s wavevector times the position @xmath10 is fixed by setting @xmath34 while the mirror s wavevector times its position @xmath11 is allowed to vary .
the probability of _ simultaneously _ measuring the particle and mirror within a small region @xmath42 around @xmath43 , over which the pdf is essentially constant , is @xmath44=\textmd{pdf}\delta x_{1 } \delta x_{2}$ ] with @xmath45 \psi^{*}[x_{10},x_{20},t]$ ] .
the probability of measuring the particle first and the mirror later is given by a product of two probabilities for one - body measurements .
the first , the probability of measuring the particle around @xmath46 at time @xmath35 without knowledge of the mirror s position , is determined from the average of the two - body pdf over the mirror coordinate @xmath11 , @xmath47=\textmd{pdf}_{1}\delta x_{1}$ ] with @xmath48 \psi^{*}[x_{10},t_{10},x_{2},t_{10}]dx_{2}$ ] .
the next , the probability of measuring the mirror at @xmath11 and @xmath31 after having measured the particle at @xmath49 within @xmath50 at @xmath35 , is given by @xmath51=\textmd{pdf}_{2}\delta x_{2}$ ] with @xmath52 \psi^{*}[x_{10},t_{10},x_{2},t_{2}]\delta x_{1}$ ] .
the probability of first measuring the particle at @xmath49 and time @xmath35 and then the mirror at @xmath11 and time @xmath31 is therefore @xmath53 .
the predictions for synchronous measurement presented below are plots of the two - body pdf as a function of @xmath10 and @xmath11 for snapshots at @xmath54 .
the probability of measuring both the particle and mirror at the same time in a small region of these plots is @xmath30 \psi^{*}[x_{1},x_{2},t ] \delta x_{1 } \delta x_{2}$ ] .
in the asynchronous predictions presented , it is assumed that the particle has been measured and therefore the plots are of the one - body pdf for the mirror as a function of both @xmath11 and the position @xmath49 and time @xmath35 at which the particle was measured .
these are snapshots at @xmath31 .
the probability of measuring the mirror in a small region at time @xmath31 , having already measured the particle at @xmath49 and @xmath35 , is then @xmath39 \psi^{*}[x_{10},t_{10},x_{2},t_{2 } ] \delta x_{2}$ ] . from these figures ,
only the probability of measuring the mirror after the particle has been measured ( not the probability of first measuring the particle _ and _ then the mirror ) can be determined .
conservation of probability for this one - body mirror wavefunction is addressed in the appendix . while the validity of this asynchronous model is limited by the assumption of no interaction after the first measurement
it also makes subtle assumptions about collapse of a two - body wavefunction : a measurement of the particle substate influences neither the mirror substate at @xmath35 nor the subsequent evolution of the mirror substate apart from fixing the parameters of the particle substate in the two - body wavefunction .
a more detailed discussion is found in subsection [ sec : measurementtheory ] . before reflection
an uncorrelated solution to the schrdinger equation is given by @xmath55 .
\label{eq : scheqnunentangled}\end{aligned}\ ] ] an incident wavegroup constructed from this then leads to uncorrelated predictions about the probability of finding the particle at @xmath56 and mirror at @xmath57 .
for the reflected wavefunction , the solution to eqn .
[ eq : scheqode2 ] must vanish at @xmath58 to satisfy the boundary condition at the mirror and not exist for @xmath59 ( or @xmath60 ) since the particle can not move through the mirror ( for the uncorrelated incident state , however , there is no interaction and the particle does move past the mirror ) . in this transformed system , a solution to eqn .
[ eq : scheqode2 ] can be constructed from the superposition of incident and `` reflected '' wavefunctions in the cm - rel system ( in much the same way as is the solution for a wave traveling along a string toward a rigidly clamped boundary is constructed from free string solutions traveling in opposite directions ) , @xmath61 , \label{eq : psiseparable}\ ] ] where @xmath62 $ ] is the unit step function .
the only difference between the arguments of the two exponentials is the sign of the relative wavevector @xmath63 corresponding to reflection in the relative coordinate .
that is , @xmath64 where the initial velocities must allow reflection to occur .
relative and center of mass times @xmath65 and @xmath66 are introduced and associated with the relative and center of mass energies .
these time variables satisfy the same properties as do @xmath4 and @xmath31 but in this case provide the notation needed in separating the energies associated with the relative and center of mass subsystems . the solution to eqn .
[ eq : scheqode1 ] for reflection is given by @xmath67 the complete solution for an eigenstate of energy in reflection is then @xmath68=\psi_{cm}\psi_{rel}$ ] .
the particle - mirror system has now been partitioned into separable center of mass and relative coordinate subsystems .
this separable solution , for the two uncorrelated substates , can be used to construct a wavegroup whose substates yield uncorrelated wavegroups associated with the cm and relative coordinates which satisfies initial and/or boundary conditions .
a measurement of the cm position affects neither the time evolution of the wavegroup associated with the relative motion nor introduces any correlation between the cm and relative positions .
while the energy and momentum of the center of mass subsystem are unaffected by reflection , that is not the case for the relative subsystem where the relative wavevector changes sign upon reflection .
therefore , in this separable system interference of incident and reflected wavefunctions , determined by the pdf @xmath69 , \label{eq : separableinterference}\end{aligned}\ ] ] is associated only with the relative coordinate subsystem .
the change from cm - rel to particle - mirror systems is accomplished by using the following relations in the separable solutions given by equations [ eq : schtot ] , [ eq : psiseparable ] , [ eq : scheqphase ] , and [ eq : psicm ] : @xmath70 , @xmath71 , @xmath25 , @xmath24 , @xmath72 , and @xmath73 .
these relations , however , do not address the two - time labeling issue in the particle - mirror system . to do so , note that the energy of the reflected particle , given by @xmath74 with @xmath75 , is associated with the temporal coordinate @xmath4 . similarly , the energy for the mirror is @xmath76 with @xmath77 and is associated with the temporal coordinate @xmath31 . both of these energies and momenta are consistent with those of a classical particle reflecting from a moving mirror .
these are manifest in the two - body wavefunctions , however , as a doppler shift .
application of these transformation relations then changes @xmath78 in equation [ eq : psiseparable ] into equation [ eq : scheqnunentangled ] .
the two - time expression for @xmath79 in the particle - mirror system , although simple to calculate using the procedure just outlined , is too large to present here .
this complexity is a consequence of the correlations generated in reflection .
correlated interference of these incident and reflected wavefunctions in the particle and mirror subsystems is then given by ( transforming eqn .
[ eq : separableinterference ] ) @xmath80 .
\label{eq : entangledinterference}\end{aligned}\ ] ] this is similar to gottfried s joint or simultaneous pdf for the interference obtained in the correlation between two particles produced in a momentum - conserving decay after each has traversed separate double slits @xcite when @xmath81 .
note also that interference of the mirror and particle are coupled in eqn .
[ eq : entangledinterference ] , illustrating how many - body systems interfere with ` themselves ' rather than only the particle and mirror each interfering with itself ( in which case there would be no correlation ) @xcite . to gain familiarity with this result ,
consider a simultaneous measurement .
for fixed @xmath10 , the approximation @xmath82 leads to interference for the mirror which varies from maximum to minimum through a distance @xmath83 similarly for fixed @xmath11 this approximation leads to interference for the particle which varies from maximum to minimum through a distance @xmath84 . for @xmath85 both the mirror and particle fringes are spaced at half the debroglie wavelength of the _ particle _ , which can be up to @xmath86 m for ultra cold atoms @xcite .
the time dependence in equation [ eq : entangledinterference ] is determined by the time components of the phase of the incident wavefunction , @xmath87=p_{in}^{2}t_{1}/2 m + p_{in}^{2 } t_{2}/2m$ ] , and the reflected wavefunction , @xmath88=p_{ref}^{2}t_{1}/2 m + p_{ref}^{2 } t_{2}/2m$ ] .
the temporal part of the joint pdf depends on @xmath87-\phi_{ref}[t_{1},t_{2}]$ ] . for simultaneous measurements ( @xmath54 )
this phase difference is zero since the time variable factors from all energy terms and the total energy before and after reflection does not change : @xmath89-\phi_{ref}[t]=(p_{in}^{2}/2 m + p_{in}^{2}/2m - p_{ref}^{2}/2 m - p_{ref}^{2}/2 m ) t=0 $ ] . that , however , is not the case for non - simultaneous measurements since the times of measurement of the particle and mirror differ and are therefore no longer a common factor of all energy terms in the phase .
this leads to the doppler effect which is described next . consider an ensemble of identically prepared particle - mirror systems .
let a measurement on every member of the ensemble be made at both fixed particle - mirror positions , @xmath10 , @xmath11 , and time @xmath31 while the particle is measured at different times @xmath4 for different members of the ensemble .
the time dependent interference pattern from eqn .
[ eq : entangledinterference ] emerges from these ensemble measurements as the expected beat frequency " @xmath90 @xcite associated with interference of the incident and reflected particle substates ( the superposition of states with different energies commensurate with the energy exchanged in reflection ) . if instead , @xmath10 , @xmath11 , and @xmath4 are fixed while the mirror position is measured at different times @xmath31 , the beat frequency " is that associated with superposing mirror substates differing in energy .
these particle and mirror beat frequencies are identical due to the same energy being exchanged between the particle and mirror in reflection and are given by @xmath91 to better understand the experimental consequences of these results , wavegroups are next formed from a superposition of the incident and reflected ` energy eigenstates ' ( given by eqn .
[ eq : psiseparable ] ) expressed in terms of the correlated particle and mirror substates rather than the cm and relative substates .
it is assumed that the initial particle and mirror gaussian substates are sufficiently spatially separated that any probability of the particle initially being on the `` wrong '' side of the mirror is negligible .
an analytic expression for such wavegroups can be obtained for a gaussian distribution in wavevector components @xmath14 and @xmath15 ( or velocities @xmath1 and @xmath20 ) . for the mirror
this is proportional to @xmath92/(2 \delta k^{2})$ ] where the peak of the distribution is at @xmath93 and @xmath94 is its width while for the particle this is proportional to @xmath95/(2 \delta k^{2})$ ] where the peak of the distribution is at @xmath96 and @xmath97 is its width . for a particle reflecting from a mirror .
the lower pdf waveform moves toward the diagonal white line , corresponding to @xmath58 , then reflects in the middle snapshot where the incident and reflected two - body wavefunctions ` overlap ' , and finally it moves away from the diagonal in the upper snapshot .
the correlated interference fringes are spaced by about half the debroglie wavelength of the _ particle_. the upper left inset is a schematic of the ` classical ' analog before reflection while the upper right inset is that after reflection with initial and final particle and mirror velocities @xmath1 , @xmath20 , @xmath98 , and @xmath99 respectively .
there is no classical analog for the middle snapshot . ]
consider first _
simultaneous _ measurement of the particle and mirror . in fig .
[ fig : interference ] snapshots of such a two - body joint pdf are shown at three times , @xmath100 , for @xmath101 , @xmath102 , and @xmath103 .
the incident wavegroup propagates in the @xmath104 plane along a line whose slope is determined by a ratio of the group velocities of each substate and spreads due to dispersion independently in each direction .
the particle and mirror initially move in the positive direction for the parameters chosen . during reflection the incident and reflected wavegroups overlap resulting in interference .
after reflection the speed of the mirror increases while the particle continues moving in the positive direction with decreased speed ( see the insets for the classical analog ) .
careful inspection of the lower and middle snapshots just to the left of the diagonal white line confirms the prediction that the spatial location of the interference maxima and minima do not depend on time as given in eqn .
[ eq : entangledinterference ] when @xmath81 .
wavegroup distortion shown in the upper reflected snapshot is discussed below . in the cm - rel system (
not shown ) the fringes " are aligned parallel to the cm coordinate illustrating the result given in eqn .
[ eq : separableinterference ] . a slice of fig .
[ fig : interference ] for @xmath105 along the @xmath11 coordinate is shown in fig .
[ fig : fringespacing ] ( the solid line ) along with a slice of this figure for @xmath106 along @xmath10 ( the dashed line ) for different bandwidth wavegroups .
this demonstrates essentially the same fringe spacing for the particle and mirror substates with narrow bandwidth wavegroups , as discussed for the approximation @xmath107 following eqn .
[ eq : fringespacing ] .
the minima of the interference shown in fig .
[ fig : interference ] correspond to positions where the particle and mirror can never simultaneously be found .
verification of this result requires simultaneous cm measurement of the particle and mirror with instruments which have a spatial resolution that is smaller than this fringe spacing along both coordinates . which show the fringe spacing along the @xmath11 axis for @xmath105 ( dashed lines ) and along the @xmath10 axis for @xmath106 ( solid lines ) .
the @xmath10 axis has been inverted to display both the dashed and solid lines together .
although each graph has @xmath102 the value of @xmath94 increases sequentially by a factor of @xmath108 from the front to the back of the figure . ]
next consider non - simultaneous measurements of the particle at position @xmath49 and time @xmath35 while allowing @xmath31 to vary until a measurement is made of the mirror s cm position .
this can be categorized into two regimes .
measurement of the particle either occurs in the region where the incident and reflected wavegroups do not overlap , regime a , or in the region where the interference similar to that shown in fig .
[ fig : interference ] occurs , regime b. to illustrate such asynchronous measurements , the particle is measured at a particular @xmath49 and @xmath35 while the mirror substate then evolves with @xmath39 $ ] .
this is shown as one - body pdf plots of snapshots at different times @xmath31 using a 3-d graph of @xmath11 vs fixed values of @xmath49 and @xmath35 .
all three joint pdf snapshot results for regime a shown in fig .
[ fig : regimea ] are for @xmath111 .
the one farthest to the right is similar to the upper pdf waveform snapshot in fig .
[ fig : interference ] since it occurs at @xmath112 .
the other two snapshots are for times , @xmath113 and @xmath114 with @xmath115 , @xmath102 , and @xmath116 .
note that this mirror wavegroup moves only along the @xmath11 axes and disperses .
it is useful to compare the physical interpretation of the joint pdf shown in fig .
[ fig : interference ] with that of the one - body pdf of fig .
[ fig : regimea ] . in the former case the probability of finding the particle and cm of the mirror in a region centered around @xmath117 and @xmath118 simultaneously at time @xmath32
is given by @xmath119 dx_{1 } dx_{2}$ ] . in fig .
[ fig : regimea ] the particle is measured at @xmath111 while the snapshots are given for different values of when and where the cm of the mirror is measured . for the leftmost pdf in this figure the probability of measuring just the mirror , once the particle has been measured ,
is given by @xmath120 dx_{2}$ ] .
when the particle has been measured at @xmath111 and @xmath49 . since @xmath112 for the rightmost pdf , it is similar to the upper snapshot in fig .
[ fig : interference ] .
the diagonal white line is the same as that in fig .
[ fig : interference ] . ]
all three joint pdf snapshot results for regime b , shown in fig . [ fig : regimeb ] , are for @xmath121 .
the one farthest to the right is similar to the middle pdf waveform snapshot in fig .
[ fig : interference ] since it occurs at @xmath122 . the middle snapshot of fig .
[ fig : regimeb ] occurs predominately between @xmath123 and @xmath124 for @xmath125 , while the leftmost snapshot consists of two bumps " encompassing the region between @xmath126 and @xmath127 for @xmath128 . as in the previous figure @xmath115 , @xmath102 , and @xmath116 . the complexity of the joint pdf in fig . [
fig : regimeb ] , in comparison with that of regime a , is a consequence of measuring the particle in the interference region .
that is , there are two indistinguishable ways that the particle could have reached the interference region .
it could have come from the incident _ or _ reflected particle wavegroup substates .
this lack of knowledge about the particle is manifest in the subsequent mirror wavegroup which then consists of a superposition in which the mirror has yet to reflect _ and _ has already reflected the particle .
these two mirror states have different speeds due to the mirror recoil in one but not the other . to see this one need only change the speed of the incident particle wavegroup
. the speed of the mirror state which has not reflected the particle does not change while the speed of the mirror state which reflected the particle increases .
overlap of these two mirror wavegroup states is shown in fig .
[ fig : regimeb ] from complete overlap ( right hand snapshot ) to partial overlap ( middle snapshot ) and finally to virtually complete separation ( left hand snapshot ) due to the differing speeds of the two mirror states . and
3 sequential times , @xmath129 . the rightmost pdf waveform is similar to the middle pdf waveform in fig .
[ fig : interference ] .
the diagonal white line is the same as that in fig .
[ fig : interference ] .
this is a pdf illustration of the splitting of the mirror states shown schematically in fig .
[ fig : overview ] ] although these results illustrate important issues in non - simultaneous measurement , they have explored only a limited set of parameters and interferometric geometries . rather than present a comprehensive treatment , we use three examples to further probe the consequences of asynchronous measurements .
first , the beamsplitting effect , that a measurement on the particle has on the mirror , is explored in more depth using a particle whose mass is equal to that of the mirror .
second , a manifestation of the doppler effect is illustrated , using a time scale shorter than that used in the figures above .
finally , a specific example of reflection of a microscopic particle from a macroscopic mirror is described .
after reflection , the spatial width of the mirror wavegroup substate is exchanged with that of the particle wavegroup substate when @xmath130 .
this is most easily seen by constructing a particle - mirror wavefunction with different bandwidths for the particle and mirror wavegroup substates and is shown in fig .
[ fig : coherencetransfer ] , which is a contour plot of joint pdfs similar to figure [ fig : interference ] , but without the snapshot in the interference region .
the solid and dashed contours correspond to @xmath131 and @xmath132 respectively with the spread in velocities given by @xmath133 . but without the interference or `` overlap '' region snapshot , illustrating coherence transfer for two different particle - mirror mass ratios .
the solid and dashed lines are for a spread in velocities given by @xmath133 with @xmath131 and @xmath132 , respectively .
note the exchange of wavegroup widths between particle and mirror substates for @xmath131 .
the diagonal white line is the same as that in fig .
[ fig : interference ] . ]
this result can be understood by comparing classical and quantum reflection . in a one - dimensional classical collision ,
conservation of energy and momentum result in the exchange of particle - mirror velocities independent of either velocity for @xmath134 .
this is manifest quantum mechanically in the exchange of commensurate substate parameters @xmath14 and @xmath15 between the incident and reflected two - body wavefunctions .
if an incident particle substate , consisting of only one harmonic component ( corresponding to speed @xmath1 ) reflects from a mirror substate with many velocity components , then each harmonic component of the mirror substate ( corresponding to different values of @xmath20 ) reflects the particle substate and therefore acquires velocity @xmath1 , while the reflected particle substate acquires different velocity values for each reflected component of the mirror wavegroup .
this results in the reduction of the mirror bandwidth and an increase in the particle bandwidth , which is manifest in fig . [
fig : coherencetransfer ] as the exchange of incident and reflected wavegroup shapes .
it also is responsible for the distortion of the reflected wavegroup shape in fig .
[ fig : interference ] . except that the particle was measured at @xmath49 and @xmath121 while the mirror is measured at @xmath11 with incremental increases in the time @xmath31 .
note the difference in scale along the axes .
the diagonal white line is the same as that in fig .
[ fig : coherencetransfer ] .
the distinct mirror substates in part ( d ) correspond to those which either have or have not reflected the particle .
this is another pdf illustration of the splitting of the mirror states shown schematically in fig .
[ fig : overview ] ] consider next non - simultaneous joint pdfs for equal particle and mirror masses .
however , rather than choosing measurement of the particle at @xmath135 , as shown in fig .
[ fig : coherencetransfer ] , the particle is measured at @xmath121 to illustrate asynchronous measurement in the interference region .
the results , shown in fig .
[ fig:2tequalmass ] , again illustrate splitting of the mirror .
however , these states differ dramatically in their shapes and in their splitting ratios . that is , depending on the position that the particle is measured at , a slice of the joint pdf along the @xmath11 axis can reveal one or two peaks of unequal height .
the long pdf waveform parallel to the @xmath49 axis is that of the mirror substate associated with no reflection , while the narrow peak corresponds to the mirror substate which reflected the particle .
the differing speeds of these two waveforms , due to mirror recoil in one but not the other , are evident in the figure .
the time order of measuring the particle and then the mirror can be reversed . in regime
b , the splitting then generates two states consisting of a superposition in which the particle has yet to reflect _ and _ has already reflected from the mirror .
variation of @xmath31 , on a temporal scale smaller than that shown in fig .
[ fig : regimeb ] , reveals the energy differences between the superposition of these two mirror substates , in regime b , as a `` beat frequency '' in the joint pdf .
this is related to the result for harmonic states given in eqn .
[ eq : beat ] and is illustrated for wavegroups in fig .
[ fig : doppler ] . a slice , taken from fig .
[ fig : interference ] , along the @xmath11 axis for @xmath136 and @xmath137 , is shown in fig .
[ fig : doppler ] as the leftmost plot .
plots to the right of this are shown at @xmath136 and @xmath121 for @xmath138 , and @xmath139 , respectively .
all other parameters are the same as used in fig .
[ fig : interference ] .
this illustrates the expected beat frequency from the `` doppler shift '' in reflection , although it is shrouded in two - body interference .
equation [ eq : entangledinterference ] predicts no such beat frequency for simultaneous measurements ( @xmath81 ) , which is discussed in more detail in subsection [ sec : asynchdoppler ] . and @xmath121 , starting at time @xmath140 ( which is a slice of fig .
[ fig : interference ] ) as the leftmost plot .
the remaining plots to the right progressively increase the time @xmath31 when the mirror is measured by @xmath141 . ]
one constraint for regime b is that the fringe visibility function must be non - zero and the incident and reflected particle - mirror wavegroups must ` overlap ' in the @xmath104 plane .
the interference fringes are then determined predominately by a superposition of ` energy eigenstates ' @xcite .
for example , the interference shown in fig .
[ fig : interference ] is determined predominately by eqn .
[ eq : entangledinterference ] , when the wavegroups ` overlap ' in the center snapshot , where the longitudinal coherence lengths for both the particle and mirror are greater than the fringe spacing . in the upper snapshot of fig .
[ fig : interference ] there is neither ` overlap ' nor such interference .
the fringe `` visibility function '' is non - zero if each wavegroup substate ` overlaps ' within approximately a coherence length @xcite , which is given by @xmath142 @xcite . for particle substates
, this can be @xmath143 for ultracold atoms @xcite or @xmath144 for slow neutrons @xcite .
this longitudinal mirror coherence length can be estimated from the uncertainty in the mirror velocity . if it is in thermal equilibrium with the environment then @xmath145 , yielding @xmath146 , which is consistent with results for atoms in a bose - einstein condensate @xcite .
[ fig : spreadvelocity ] illustrates how variation in the longitudinal coherence length of the mirror substate affects the particle - mirror interference , for fixed particle coherence length and _ simultaneous _ measurements .
part ( a ) shows a longer mirror coherence length than is used in fig . [
fig : interference ] , while parts ( b ) through ( d ) progressively reduce the coherence length of only the mirror substate . in fig .
[ fig : spreadvelocity ] ( d ) , the coherence length of the mirror is so small that overlap is prevented , over a range of @xmath10 values where it was present before .
nevertheless , a slice along the @xmath11 axis for measurement of the particle at a particular @xmath10 , indicates a splitting of the mirror substate into two states which do not overlap and are therefore distinguishable .
this is a consequence of two ways that the particle could have reached @xmath10 : it could have come from the incident _ or _ reflected particle wavegroup substates , due to the large particle coherence length . as the mirror
s coherence length increases , these two ways overlap and generate correlated interference as shown in fig . [ fig : spreadvelocity ] ( a ) . as the mirror
s coherence length decreases , the position of the mirror before reflection is distinguishable from that after reflection , resulting in no interference , as shown in part of fig .
[ fig : spreadvelocity ] ( d ) . , but with different spreads in mirror velocities , for a fixed spread in particle velocity while all other parameters are the same .
@xmath147 , and @xmath148 for @xmath149 in parts ( a ) , ( b ) , ( c ) , and ( d ) respectively . note the change of scales in graph ( d ) . ] one might expect that the small coherence length associated with a mesoscopic / macroscopic mirror mass would not allow the interference shown in fig .
[ fig : spreadvelocity ] ( a ) .
this is modeled in fig .
[ fig : largemass ] where snapshots of two - body pdfs are shown at three times for a rubidium atom with @xmath150 kg reflecting from a mirror with @xmath151 kg . the spread in wavevectors is determined by thermal equilibrium using @xmath152 k for the atom ( which is released from a bose - einstein condensate ) and @xmath153 k for the mirror .
the snapshots are for @xmath121 with @xmath154 and @xmath155 s. this mirror longitudinal coherence length , determined by thermal equilibrium at @xmath153 k , corresponds to the interference in reflection shown in fig . [
fig : spreadvelocity ] ( c ) , for simultaneous measurement . for non - simultaneous measurements ,
[ fig : largemass ] shows only snapshots over a small range of times @xmath31 , when the mirror is measured , while the particle has been measured at @xmath121 .
however , varying @xmath31 over a much larger range , for a mirror initially at rest , does not change the character of the joint pdf shown in this figure .
initial mirror and particle velocities of @xmath156 and @xmath157 m / s are used in fig .
[ fig : largemass ] , which results in a beat frequency of @xmath158 hz and changes the joint pdf in a manner similar to that shown in fig .
[ fig : doppler2 ] along the @xmath11 axis .
although these two interfering mirror states ( which either reflect or do not reflect the particle ) have different speeds , they do not completely separate , as do those shown in the leftmost plot of fig .
[ fig : regimeb ] .
this is a consequence of the small difference in mirror speeds due to the large @xmath159 ratio , resulting in a mirror offset which is small compared with the size of the wavegroup .
the longitudinal coherence length of the atom and mirror are determine by thermal equilibrium at @xmath152 k ( released from a bose - einstein condensate ) and @xmath153 k , respectively .
each snapshot progressively increases the time @xmath31 . ]
these results illustrate that a measurement of the particle at one of the minima of the interference pattern , in the joint pdf of fig .
[ fig : largemass ] , yields no possibility of measuring the cm of the mirror for any value of @xmath11 and subsequent time @xmath31 ( within a duration limited by the beat frequency ) .
this is to be contrasted with the result shown in fig .
[ fig : interference ] , where simultaneous measurement of the particle and mirror at an interference minimum _ is _ constrained temporally and spatially as follows .
first , the duration of the simultaneous interference is finite . for a static mirror ,
it is essentially determined by the coherence length of the particle divided by its speed , @xmath160 when @xmath161 .
second , the spatial extent of the interference along the @xmath11 axis is not infinite but severely limited by @xmath162 .
experimental confirmation of the interference in fig .
[ fig : interference ] therefore requires both a small spatial and a short temporal resolving power of the measuring instrument .
destructive interference , on the other hand , is not similarly constrained .
the interference minimum in fig .
[ fig : largemass ] occurs over a range of @xmath11 much larger than the mirror coherence length .
it is also robust over a wide range of mirror masses .
correlated measurements are more difficult to perform than only a measurement on the particle , while making none on the mirror . predictions of such effects are given by an average of the pdf ( or trace ) over the mirror coordinate , converting it into a marginal pdf , or for this two - body system a ` one - body ' pdf @xcite .
in the context of this work , the probability for detecting the particle without measuring the mirror for incident and reflected energy eigenstates states is then manifest as an integral of eqn .
[ eq : entangledinterference ] over the mirror coordinate ( or given by the reduced density matrix obtained as a trace over the mirror coordinates @xcite ) .
such a procedure , applied to eqn .
[ eq : entangledinterference ] , `` washes out '' the one - body interference .
gottfried carries out a similar averaging on the two - body pdf for simultaneous measurement in a momentum - conserving decay after each particle traverses separate double slits with a similar result @xcite .
however , a one - body standing wave interference pattern is certainly expected for the particle reflecting from a static mirror . for a moving mirror ,
the doppler shift introduces a time dependence in this pattern .
therefore , the prediction that there is no such interference between incident and reflected particle states , when the mirror is not measured , seems surprising .
this prediction , however , must be re - evaluated for wavegroups .
consider the averaging procedure when applied to the four pdfs shown in fig .
[ fig : spreadvelocity ] .
it is apparent that the one - body interference ` washes out ' in graph [ fig : spreadvelocity](a ) since many cycles in the mirror coordinate of the joint pdf are averaged over , while that is not the case for fig .
[ fig : spreadvelocity](c ) . as shown in fig .
[ fig : largemass ] , the mesoscopic / macroscopic mirror has a coherence length smaller than the fringe spacing determined by an ultracold atom reflecting from it and therefore most closely resembles fig .
[ fig : spreadvelocity](c ) .
it is also apparent from this figure that averaging over @xmath11 does _ not _ ` wash out ' the one - body interference .
however , this averaging of fig .
[ fig : spreadvelocity](c ) is only for simultaneous measurement and does not change with time even if the mirror moves ( see subsection [ sec : asynchdoppler ] for more details ) . for an asynchronous measurement of the particle at @xmath49 and @xmath35 ,
the joint pdf must be integrated over @xmath11 to obtain the marginal pdf for only the particle , if the mirror is not measured . yet at what time @xmath31 must this integration occur for asynchronous measurements ? this is addressed in fig .
[ fig : doppler2 ] , where four sequential plots are shown for different @xmath35 values , while within each plot snapshots at different times @xmath31 are shown .
the bandwidth parameters used in fig .
[ fig : doppler2 ] have been adjusted to model those of a mesoscopic / macroscopic mirror mass .
it is apparent from this figure that averaging over @xmath11 does not ` wash out ' the one - body interference for any value of @xmath31 .
that is , the result of averaging over @xmath11 is independent of @xmath31 .
yet variation of the one - body pdf with @xmath35 at fixed @xmath49 ( the doppler shift ) is evident in fig .
[ fig : doppler2 ] . as mentioned above , the initial mirror and particle velocities of @xmath156 and @xmath157 m / s that are used in fig .
[ fig : largemass ] result in a beat frequency for the particle of @xmath158 hz .
this effect of the doppler shift on the particle pdf ( shown along the @xmath10 axis ) differs from that in fig .
[ fig : doppler ] , which illustrates the effect of the doppler shift in the mirror s pdf ( shown along the @xmath11 axis ) .
there is no measurable doppler interference effect for the mirror in fig .
[ fig : doppler2 ] , due to its coherence length being much shorter than the fringe spacing . the choice of short coherence length mesoscopic / macroscopic mirror and a long coherence length particle substates yields the expected interference for a particle reflecting from a mirror when the mirror is not measured , but requires the use of the two - time formalism ( again see subsection [ sec : asynchdoppler ] for more details ) .
however , such a prediction is limited to a special case of a more general result .
for example , this averaging procedure applied to fig . [
fig : spreadvelocity](d ) , for simultaneous measurement , does not yield the expected particle interference in reflection when the mirror is not measured , since there is no interference over a large range of @xmath10 values where it exists in fig .
[ fig : spreadvelocity](c ) .
( d ) , with different @xmath35 values while within each plot @xmath31 is varied .
this illustrates how the result of averaging the pdf over @xmath11 is independent of @xmath31 , leading to the expected one - body interference ( derived from the two - body joint pdf ) when the mirror is _ not _ measured . ]
consider the opposite case of measurement of the mirror without a measurement of the particle , for the assumptions used in fig .
[ fig : doppler2 ] . in this case , the averaging is over @xmath10 , which washes out the fringes .
therefore , measurement of just the mirror results in no interference .
this interference for the particle , but not the mirror , is that expected for a classical wave reflecting from a mirror .
yet , it is a special case rather than a general result since it depends on the coherence length assumptions .
synchronous correlated interference in reflection is temporally and spatially limited by the particle - mirror coherence lengths as the incident and reflected two - body wavegroups separate .
we have presented a simple model illustrating the magnitude of these effects .
measurement of the particle first can prolong interference of the mirror , which is measured later .
the resulting two one - body mirror states travel in the same direction but with momenta and energies differing by that either exchanged or not exchanged with the particle
. this superposition state can either interfere or not depending on the coherence length of the mirror and the time between measurements . for an asynchronous measurement with a small mirror coherence length
, interference persists only for a small momentum exchange . since these states have different energies their interference results in a ` beat ' or time dependent fringe pattern for the mirror pdf .
for longer times between measurements , these two mirror wavegroups then separate beyond their coherence lengths , due to their different average momenta .
this then results in a spatially separated superposition state without interference .
a microscopic particle reflecting from a mesoscopic / macroscopic mirror at rest is well suited to prolong this one - particle interference of the two mirror states for two main reasons .
first the two mirror wavegroups maintain overlap longer since their difference in speeds is @xmath163 , while their coherence length is @xmath164 .
this results in an overlap time @xmath165 .
second their energy difference is negligible for a stationary mirror . that is , the `` beat '' frequency goes to zero in the limit of large mirror mass while the difference in momenta of these mirror states remains small but finite .
it is this difference in mirror momenta which leads to the phase difference @xmath166 when superposing these two mirror wavefunctions .
that is , the mirror fringe spacing is only determined by the change in particle momentum since conservation of momentum in reflection requires @xmath167 .
the fringe structure of the mesoscopic / macroscopic mirror is then commensurate with that of the microscopic particle .
double slit interference with a massive particle , on the other hand , superposes two one - body states with the same momentum whose difference in phase , @xmath168 , is due to the difference in path lengths @xmath169 from each slit to the measurement point times an extremely large wavevector ( rather than the small @xmath94 for the mirror reflecting a microscopic particle ) .
this results in an imperceptible fringe spacing on the mesoscopic / macroscopic object traversing the double slits . the interference of the mesoscopic / macroscopic mirror states does not require utilizing standard division of amplitude nor division of wavefront interferometric methods .
it does , however , use measurement acting as a beamsplitter to generate two mirror states which interfere for times greater than that for synchronous measurement .
in addition , path lengths need not be carefully matched for interference to be manifest .
these are distinct experimental advantages over division of amplitude or wavefront interferometers , particularly for a mesoscopic / macroscopic mirror mass .
asynchronous measurement contains some subtle issues .
foremost among them is what constitutes a measurement of the particle , after which no interaction with the mirror occurs .
an experimental realization of such a measurement might consist of the particle - mirror reflection occurring only along the x - axis while another particle called the `` probe '' moves along the y - axis , as illustrated in fig .
[ fig : overview ] .
a measurement then involves the probe absorbing , being absorbed by , being scattered by , or scattering from the particle .
the effect of this measurement on the particle ( e.g. back action ) has no effect on the mirror substate by the no interaction assumption .
the operational definition of such a measurement , when the particle is in the correlated interference region , is then that which causes splitting of the mirror state .
this could result in the observation of prolonged interference of the mirror states or observation of these states when they no longer overlap ( this is discussed in more detail in section [ sec : measuring ] ) .
note that the probe need not interact with the particle at any particular location apart from it occurring within this correlated interference region ( and not at a pdf minimum ) .
related issues involve modeling this sequential measurement . in the copenhagen interpretation
, a measurement of the particle s position collapses its substate into an eigenstate of that position operator .
the assumption of no interaction after the first measurement then requires that the particle does not reflect a second time from the mirror s substate .
the state of the particle after measurement is therefore irrelevant to the subsequent evolution of the mirror substate and is treated as such in the formalism presented above by fixing the particle coordinates in the two - body function at its measurement position and time .
a more realistic model would account for a measurement which occurs over a distance @xmath50 rather than at a point in space .
such a treatment is beyond the scope of this work .
this assumption of the lack of interaction can be relaxed to further study the implications of the collapse postulate . to illustrate this
let @xmath170 $ ] be the solution to the schrdinger equation for the incident uncorrelated plus the reflected correlated wavegroups .
also let the first measurement be of the particle in the correlated interference regime at @xmath34 and time @xmath35 .
the particle substate then collapses into an eigenstate of the position operator and therefore generates the two - body uncorrelated wavefunction @xmath171 \psi_{1}[x_{10},t_{10},x_{2},t_{10}]$ ] where @xmath36=\delta[x_{1}-x_{10}]$ ] . to determine the effects of allowing interaction ( a second reflection ) before a later measurement of the mirror ,
the solution to the schrdinger equation @xmath172 $ ] for @xmath173 and @xmath40 is needed .
this is determined from the initial condition @xmath174 along with the boundary condition at the mirror surface .
however , the particle must not be absorbed in the measurement process since it is now allowed to interact again with the mirror after its initial measurement . as a particular example
, consider measuring the particle at a position in the correlated interference region given by the middle snapshot in fig .
[ fig : interference ] . the particle - mirror state just after this measurement is a cut along the @xmath11 axis at @xmath34 of the particle - mirror wavefunction associated with this pdf , due to the delta function nature of this collapse .
this wavefunction in the @xmath104 plane is then the initial condition used in solving the two - body schrdinger equation for the subsequent particle - mirror state .
the reflection boundary condition at @xmath58 must also be satisfied .
the narrow particle substate along the @xmath10 axis will lead to its `` diffusion '' and subsequent `` re - reflection . '' note that if the particle is measured a second time , before a measurement of the mirror , then the procedure just mentioned must be repeated .
measurement - induced multiple reflections then occur and are , in themselves , of fundamental interest .
a time dependent interference pattern is expected classically when a harmonic wave reflects from a moving mirror .
for a stationary mirror the incident and reflected waves form a standing wave which is measured by placing a detector at a fixed position where the incident and reflected waves overlap @xcite .
motion of the mirror introduces a doppler shift in the reflected wave which generates time dependence in this interference .
the detector fixed in space then measures a time dependent intensity variation .
consider the analogous process for the particle - mirror system .
the incident and reflected two - body wavefunctions have the same total energy in reflection from a moving mirror .
interference in the synchronous formalism involves superposing these two wavefunctions , each with the same frequency .
the result , given by eqn .
[ eq : entangledinterference ] with @xmath81 , is then independent of time . to illustrate this ,
consider the middle snapshot of fig .
[ fig : interference ] . for different snapshots (
not shown ) , all of which still partially overlap spatially with this middle snapshot , the envelope of the pdf waveform moves through the @xmath104 plane , yet the maxima and minima of the fringes for these different snapshots remains in the same location .
the doppler shift for the particle - mirror system , analogous to a classical wave reflecting from a moving mirror , involves only measuring the particle flux at a fixed position with no measurement of the mirror .
not measuring the mirror involves averaging the two - body pdf ( trace ) over the mirror coordinates .
the mirror coherence length determines if this averaging `` washes - out '' the interference .
an example , for the synchronous case is shown in fig .
[ fig : spreadvelocity ] ( a ) .
however , a smaller mirror coherence length results in a pdf which , when averaged over @xmath11 , maintains the interference along @xmath10 , as is illustrated in fig . [ fig : spreadvelocity ] ( c ) .
the pdf for simultaneous correlation , with a coherence length associated with that shown in fig .
[ fig : spreadvelocity ] ( c ) , then generates only static interference and not the expected time dependent interference effect associated with the doppler shift
. however , as shown in subsections [ subsec : beat ] and [ subsec : trace ] , the two - time formalism predicts a doppler interference effect similar to that classically when measuring only the particle , but only as a special case ( for certain coherence lengths ) of a more general result .
this is a consequence of using the two - time formalism to parse the total energy and therefore the frequency of the wavefunction into terms associated with the particle and mirror rather than only using the expression for the total energy .
interference in a typical one - body interferometer ` washes - out ' as the fringe spacing becomes imperceptible with increasing mass .
another mechanism that eliminates interference is a measurement to determine along which path the object traveled through the interferometer .
the former issue is mitigated using two - body reflection of a microscopic particle from a mesoscopic / macroscopic object , as described above , while the latter is the first topic of this section .
a method to determine path information and therefore destroy interference utilizes a ` heisenberg microscope , ' where an additional particle scatters from the object traversing the interferometer .
if the wavelength of this particle , @xmath175 , is smaller than the spatial separation along the two paths , @xmath169 , then path information can be obtained in principle from the state of this scattered probe particle and interference is destroyed .
this was verified using a three grating mach - zehnder interferometer traversed by an atom and using a photon as the probe to determine path information @xcite .
the contrast of the interference , as a function of @xmath169 , was shown to drop dramatically for @xmath176 .
generating such decoherence for particle - mirror interference involves using a probe particle to determine the position of the mirror during interference ( after measurement of the particle ) .
there are two distances that the mirror could have moved : that associated with and that without reflection . if the probe particle can resolve the difference in these distances then the ` path ' associated with reflection is distinguishable from that without reflection . for a synchronous measurement of a near stationary mirror with @xmath82 and @xmath177 ,
the mean difference in these mirror distances is @xmath178 , which when equal to the wavelength of a probe particle of mass @xmath179 then requires that the probe have velocity @xmath180 . for asynchronous measurement @xmath181 requiring a probe velocity @xmath182 , where @xmath183 .
this calculation can also be used to approximate the effect of such decoherence on the mirror from the collisions with gas atoms forming the thermal bath which surrounds the mirror if these gas atoms are considered to be the probe which determines path information and they scatter from only one of the superposed mirror states .
the interference in reflection described above is then maintained if there is a small probability of these atoms having the velocity @xmath184 .
for a mesoscopic mirror , @xmath185 kg , and both microscopic particle and probe atoms of mass @xmath186 kg , @xmath187 m / s .
atoms with such thermal speeds are not likely to be found in a low temperature gas .
this decoherence mechanism can also be applied to the emission of three examples are used by the mirror during interference . if the photon s wavelength is small enough to allow for the location of the slab associated with one of these positions , interference is destroyed since it
can then be determined at which position the reflection occurred @xcite . for an ultra cold atom reflecting from a mirror of mass @xmath188 kg , @xmath181 , requiring a wavelength of @xmath189 m for @xmath190 s.
the probability of such a thermal photon emission is small .
however , for such a probe to destroy mirror interference it must interact with only one of the mirror states ( how this could be accomplished with closely spaced mirror states is not obvious ) . if not , then the probe ( with a coherence length longer than the separation of the mirror states ) itself is placed into a superposition state of having interacted with both mesoscopic / macroscopic mirror states .
this results in probe interference , which can be used as a method to determine that the mirror is in a superposition state , as discussed in section [ sec : measuring ] , these calculations indicate that decoherence associated with obtaining path information can be mitigated with a judicious choice of parameters .
this is due in part to the displacements associated with the interfering mirror substates being proportional to @xmath191 . to measure such small displacements
requires probe particles with proportionally smaller wavelengths which are typically not found in the environment .
another decoherence mechanism is through coupling with the many degrees of freedom of the surrounding environment .
zurek @xcite predicted exponential decay of the off - diagonal density matrix element terms for a mesoscopic / macroscopic object in a superposition state , @xmath192 $ ] , with time constant @xmath193 , where @xmath194 is the thermal relaxation time , @xmath169 is the difference in distance between these two states , and @xmath195 is the thermal de broglie wavelength of the mesoscopic / macroscopic object . the decay is into a mixed rather than a pure state . applying this decoherence model to the mesoscopic / macroscopic mirror states after measurement of the particle , as described above , yields @xmath196 .
reflection of an atom from a mirror of mass @xmath197 kg at @xmath153 k yields @xmath198 for @xmath190 s. however , @xmath194 can vary dramatically depending on environmental and material properties along with the dimensional values of the mirror .
macroscopic yet small masses can have short thermal relaxation times .
nevertheless , it appears that environmental decoherence of the mesoscopic / macroscopic mirror states generated from asynchronous measurement can be made either negligible or non - negligible in this model of decoherence .
this decoherence time was derived to show that mesoscopic / macroscopic bodies in superposition states quickly decohere .
therefore , any interference due to such superposition should not be observed . on the other hand ,
asynchronous measurement generates a superposition of macroscopic mirror states which mitigate this environmental decoherence , due to their small spatial separation .
it has been shown that correlated interference for a microscopic particle reflecting from a mesoscopic / macroscopic mirror , does not disappear with increasing mirror mass , is difficult to eliminate via a path measurement , and can be made insensitive to environmental decoherence .
it should also be pointed out that the robust character of interference for objects with many degrees of freedom is reinforced by measurements which demonstrate that even if the size of the object is larger than both the coherence length and debroglie wavelength , interference can still be observed @xcite .
a third particle , reflecting from the mirror after asynchronous measurement , can act as a probe to confirm that the mirror is in a superposition state .
such reflection is again described by a two - body wavefunction , if the measurement of the initial particle has eliminated this particle from interacting with both the split mirror states and the probe .
the result is then a sum of two - body solutions to the schrdinger equation , one for the probe reflecting from each mirror state .
wavegroups are then formed from these harmonic states as described above .
the interference in reflection of these two - body wavegroups is robust since overlap is maintained as they both travel in the same direction . measuring the probe , without any measurement of the mirror ,
does not destroy the probe interference for sufficiently small mirror coherence lengths , as described in subsection [ sec : asynchdoppler ] .
therefore , the more difficult correlation measurement of both the probe and mirror need not be performed . the coherence length of the probe need only be larger than the very small displacement of the mirror states for this probe interference to be exhibited , while the phase difference between these probe states is determined by the mass of the probe . with such interference
, the probe can not distinguish between the mirror states as discussed in section [ sec : path ] .
verification of interference from only the reflected probe is then a method to measure the mesoscopic / macroscopic mirror in a superposition state , without any ` direct ' measurement of the mirror .
details of such a calculation are beyond the scope of this foundational work .
asynchronous measurement in quantum mechanics has been treated for a particle reflecting from a mirror .
the results presented assume measurement of the particle first , collapse of the particle substate , and a later measurement of the mirror while no particle - mirror interaction occurs between the times of these measurements .
the two - body schrdinger equation , solved with standard techniques , is parsed to completely separate the variables associated with the particle from those of the mirror , so that when the particle is measured only its parameters are fixed while those of the mirror can continue to evolve in time until the mirror is measured .
the subsequent interference of the mirror states which have and have not reflected the particle is in many ways familiar from examples in quantum mechanics in which an outcome can be achieved in indistinguishable ways . yet this process of splitting differs by using measurement as its mechanism .
this is not possible in a one - body system . in addition , the example used here of a microscopic particle reflecting from a mesoscopic or macroscopic mirror illustrates why these interference effects do not vanish with increasing mirror mass . _ without _ such measurement , the two - body wavegroup time evolves into one in which the wavegroup has completely reflected and then no longer overlaps with the incident wavegroup .
correlated interference then disappears since there is then no uncertainty about the particle and mirror belonging to the incident or reflected states .
correlated interference in reflection is an example of one of the simplest interferometers , utilizing neither division of amplitude nor division of wavefront methods to generate interference .
in addition , path lengths need not be carefully matched for interference to be manifest .
the complexity of _ asynchronous _ correlated interferometry is revealed as the coherence lengths of the particle and mirror substates are varied .
examples are : ( 1 ) when both are small then essentially classical reflection results .
measurement of the particle either occurs after _ or _ before reflection , resulting in a later measurement of the mirror either having recoiled _ or _ not recoiled .
( 2 ) when both are large , compared with the particle s wavelength , then measurement of the particle in the correlated interference region acts as a beamsplitter generating mirror states which have _ and _ have not reflected the particle .
( 3 ) measurement of the particle only , for short mirror and long particle coherence lengths , yields the expected interference due to the doppler effect but only as a special case of a more general result .
other issues of asynchronous measurements presented include : transfer of coherence between the particle and mirror , the effects of substate coherence on asynchronous interference , distortion of the two - body wavegroups due to reflection , interference effects on mesoscopic / macroscopic mirrors , and the prolonging of interference of the mirror substate when the particle is measured first in the region of correlated interference .
finally , the superposition of mesoscopic / macroscopic mirror states was shown to be insensitive to environmental decoherence .
some consequences of asynchronous measurements not presented include : multiple sequential particle measurements before a measurement of the mirror ( two - body quantum zeno effect ) , treatment of asynchronous measurement in other interpretations of quantum mechanics , two - body effects for spatially varying potentials , reflection of a massless particle , dealing with an inelastic collision , and both synchronous and asynchronous two - body treatment of quantum tunneling .
perhaps the one modification which would result in a more familiar interferometer is to incorporate a partially transmitting , rather than totally reflecting , mirror .
one example is of a thin aluminum slab whose two surfaces partially retro - reflect a neutron .
this then leads to a two - body interferometer analogous to the one - body thin - film interference of light .
finally , the more likely domain in which these prediction will be tested is in the reflection of a microscopic particle from a microscopic `` mirror , '' the details of which have also not been discussed .
there is little experimental support for asynchronous correlation interferometry .
yet , it involves perhaps the most elementary of only a few exact solutions to the two - body schrdinger equation .
it is surprising that such theoretical simplicity is not reflected experimentally .
measurement in quantum mechanics plays a role different from that in any other physical theory and its interpretation remains controversial .
it is shown here that measurement can act as a beamsplitter , which , under the appropriate conditions , can prolong the transient interference associated with reflection in a two - body quantum system .
although far from being comprehensive , these results also indicate a direction , heretofore unexplored , for further research in understanding asynchronous quantum correlation , probing decoherence , and extending quantum measurements to larger masses .
the authors would like to thank professors c. durfee and j. scales for useful comments , along with referee guidance .
without collapse the two - body wavefunction , expressed in terms of the two times , evolves and conserves probability .
the probability of measuring the particle at @xmath56 and the mirror at @xmath57 is given by @xmath199 dx_{1 } dx_{2}$ ] with the joint pdf determined by the solution of equation [ eq : schreqn ] as @xmath200 but expressed in terms of the two time variables in the particle - mirror system .
using this equation , conservation of probability can then be expressed locally as , @xmath201 + \partial_{t_{2 } } pdf[x_{1},t_{1},x_{2},t_{2 } ] + \notag \\ \partial_{x_{1 } } j_{1}[x_{1},t_{1},x_{2},t_{2 } ] + \partial_{x_{2 } } j_{2}[x_{1},t_{1},x_{2},t_{2}]=0 , \label{eq : consprob}\end{aligned}\ ] ] where @xmath202=\hbar ( \psi^ { * } \partial_{x_{1 } } \psi-\psi \partial_{x_{1 } } \psi^{*})/(2 i m)$ ] and @xmath203=\hbar ( \psi^ { * } \partial_{x_{2 } } \psi-\psi \partial_{x_{2 } } \psi^{*})/(2 i m)$ ] . while the expressions for these current densities appear similar to that for one particle systems there are subtle but important differences for a two body system @xcite .
multiplying equation [ eq : consprob ] by @xmath204 , integrating over the segment from @xmath205 to @xmath206 along the x - axis ( @xmath207 and @xmath208 ) , and then rearranging terms yields a solution to equation [ eq : consprob ] if @xmath209 dx_{1 } + \notag \\
j_{1}[b , t_{1},x_{2},t_{2}]-j_{1}[a , t_{1},x_{2},t_{2}]=0 \notag \label{eq : consprob2a}\end{aligned}\ ] ] and @xmath210 dx_{2 } + \notag \\
j_{2}[x_{1},t_{1},b , t_{2}]-j_{2}[x_{1},t_{1},a , t_{2}]=0 \notag .
\label{eq : consprob2b}\end{aligned}\ ] ] these equations indicate that the time rate of change of probability within the @xmath211 segment on the x - axis is determined separately by a net change in particle and mirror probability fluxes in that region , which is similar to conservation of probability in a one - body system . now
, however , probability of the two - body system is conserved for the mirror even when @xmath10 and @xmath4 are fixed .
the asynchronous procedure described here therefore maintains conservation of probability for the mirror after the particle has been measured .
bell , physics * 1 * 195 ( 1964 ) .
k. gottfried , am .
* 68 * 143 ( 2000 ) .
d. salart , a. baas , j.a.w .
vanhouwelingen , n .
gisin , h. zbinden , phys .
lett . * 100 * 220404 ( 2008 ) .
s. grblacher , et .
nature , * 446 * 871 ( 2007 ) .
ma rowe , et .
nature , * 409 * 791 ( 2001 ) .
m. ansmann , et .
nature , * 461 * 504 ( 2009 ) .
a. aspect , p. grangier , and g. roger , phys .
* 49 * , 91 ( 1982 ) .
d. bohm and y. aharonov , phys .
* 108 * , 1070 ( 1957 ) .
d. kouznetsov and h. oberst , phys .
a * 72 * , 013617 - 1 ( 2005 ) .
t. a. pasquini , m. saba , g .- b .
jo , y. shin , w. ketterle , and d. e. pritchard , phys .
lett . * 97 * , 093201 - 1 ( 2006 ) .
f. shimizu , phys .
* 86 * , 987 ( 2001 ) .
t. hils et .
al . , phys .
a * 58 * , 4784 ( 1998 ) .
y. colombe , b. mercier , h. perrin , and v. lorent , phys .
a * 72 * , 061601 ( 2005 ) .
szriftgiser , p. , d. guery - odelin , m. arndt , and j. dalibard , phys .
lett . * 77 * , 4 ( 1996 ) .
o. wiener , ann .
d. physik * 40 * , 203 ( 1880 ) .
h. rauch , contemp .
* 2 * 345 ( 1986 ) .
j. p. gordon , phys .
* 14 ( 1973 ) .
s.m . barnett and r. loudon , phil .
r. soc.a * 368 * 927 ( 2010 ) .
mcguire and a.l .
godunov , phys .
a , * 67 * 042701 ( 2003 ) . s. petrat and r. tumulka , proc .
a , * 470 * 20130632 ( 2014 ) .
p. tommasini et.al . ,
* 66 * , 881 ( 1998 ) .
siverman , _ more than one mystery _ , p-96 ( springer verlag , new york , 1995 ) .
a. d. cronin , j. schmiedmayer , and d. e. pritchard , rev .
phys . * 81 * , 1051 ( 2009 ) .
c.u.serge and j.d .
sullivan , am .
* 44 * 729 ( 1976 ) .
a beat frequency is usually associated with an ac coupled voltage from a detector placed at the output port of an interferometer which combines waves of different frequencies . this voltage , proportional to the intensity , has the same trigonometric dependence as the joint pdf given in eqn .
[ eq : entangledinterference ]
. however since @xmath212 , the frequency of this ac coupled output voltage is @xmath213 .
hamilton , a.g .
klein , and g.i .
opat , phys .
a , * 28 * , 3149 ( 1983 )
. the coherence length does not change even though the size of the wavegroup does .
klein , g.i .
opat , and w.a .
hamilton , phys .
* 50 * , 563 ( 1983 ) .
m. ferrero and a. van der merwe , _ fundamental problems in quantum physics _ , p-125 ( kluwer academic publishing , dordrecht , 1995 ) .
pushin , m .
arif , m.g .
huber , d.g .
cory , phys .
100 * , 250404 ( 2008 ) .
chapman , et.al .
* 75 * , 3783 , ( 1995 ) .
w. h. zurek , rev .
* 75 * , 715 , ( 2003 ) .
in the particle - mirror system the incident and reflected wavevectors for the particle ( or mirror ) are not necessarily equal and opposite as they are in the cm - rel system .
the probability current density to the left of the mirror then exhibits spatial and temporal interference due to the effect this difference in wavevectors has on the incident and reflected wavefunctions .
this results in a non - additive current density , a consequence of the increasing space being generated as the mirror moves away from the particle .
the net flux must account for the build - up of probability within this `` new '' space . | a two - body quantum correlation is calculated for a particle reflecting from a moving mirror . correlated interference results when the incident and reflected particle substates and their associated mirror substates overlap .
using the copenhagen interpretation of measurement , an asynchronous joint probability density ( pdf ) , which is a function both of the different positions and different times at which the particle and mirror are measured , is derived assuming that no interaction occurs between each measurement .
measurement of the particle first , in the correlated interference region , results in a splitting of the mirror substate into ones which have and have not reflected the particle .
an analog of the interference from the doppler effect for only measurements of the particle ( a marginal pdf ) , in this two - body system , is shown to be a consequence of the asynchronous measurement .
the simplification obtained for a microscopic particle reflecting from a mesoscopic or macroscopic mirror is used to illustrate asynchronous correlation interferometry . in this case , the small displacement between these mirror states can yield negligible environmental decoherence times . in addition , interference of these mirror states does not vanish in the limit of large mirror mass due to the small momentum exchange in reflecting a microscopic particle . |
strong and rapid variability is a key observational feature of systems driven by the accretion of material onto a central compact object .
the power spectral density ( psd ) of emission from such objects is characterized by broadband noise across several decades in frequency .
this broadband component is generally well described by a broken - powerlaw form , however the physics that governs such breaks is poorly understood .
additionally , the psd can feature distinct and resolved peaks indicative of quasi - periodic oscillations ( qpos ) in the emission .
the broadband noise structure displays many interesting characteristics beyond the psd .
light curves show log - normality in their structure and a strong linear relationship between the root - mean square ( rms ) variability and the flux of the source ( uttley & mchardy 2001 ; negoro & mineshige 2002 ; gaskell 2004 ) . studies of frequency dependent coherence between energy bands in the time - series data indicates that the emission is coherent at low frequencies then becomes incoherent at high frequencies ( e.g. cygnus x-1 , see nowak et al .
1999 for details ) .
consequently , many sources also show frequency - dependent phase / time lag behavior between emissions at different wavebands .
this was first clearly demonstrated in accreting stellar mass black holes which often displayed a constant time lag at low frequencies and constant phase at high frequencies ( nowak et al .
qualitatively similar behavior is now seen in active galactic nuclei ( agn , vaughan , fabian & nandra 2003 ) and accreting non - magnetic white dwarfs ( scaringi et al . 2013 ) .
the broadband variability is often attributed to inward propagating fluctuations driven by stochasticity in the angular momentum transport mechanism ( lyubarski 1997 , hereafter l97 ) . within the framework of the standard shakura & sunyaev ( 1973 ) disk model in which the internal `` viscous '' shear stresses are proportional to the pressure ( @xmath0 ) , l97 attempted to explain this broadband flicker - noise by introducing a small stochastic perturbation into the @xmath1-parameter .
this perturbation generated broadband fluctuations in the accretion ( @xmath2 ) .
however , the analysis of l97 was ultimately of a linearized form of the disk equation .
in fact , many of the above features , most notably the rms - flux relationship , are not formally captured by this model since linearization means that all perturbations are applied to the base state and do not build upon each other multiplicatively . still , it was suggested by uttley & mchardy ( 2001 ) that the log - normality and rms - flux relation are a result of propagating fluctuations , with the intuitive but formally unproven statement that such features would arise from an analysis of the disk equations with non - linear fluctuations ( see also uttley et al .
2005 ) .
this problem has also been examined by kelly et al .
( 2011 , hereafter k11 ) using the tools of stochastic differential equations . specifically , they studied the use of the stochastic ornstein - uhlenbeck ( ou ) process to model the observed light curves and power spectra of systems with accretion - driven variability .
this was motivated by the fact that a solution of the stochastically perturbed linearized disk equation is a linear combination of ou processes which , in the case where the stochastic driving terms are brownian motion , are simply damped random walks ( also see wood et al . 2001 , titarchuk et al . 2007 )
. k11 found that a linear combination of only one or two such processes was sufficient to well describe the data , and that they could reproduce breaks in the observed power - spectra .
in this paper , we construct numerical realizations of _ non - linear _ fluctuations in a simple one - dimensional viscous disk .
we show that these disk models produce mass accretion rates and luminosities that feature a linear rms - flux relationship and approximate log - normality .
we also examine the fourier space relations between the mass accretion rates and dissipations at different radii , one possible proxy for light curves in different energy / wavelength bands .
we find coherence between radii for frequencies less than the viscous frequency .
however , we fail to cleanly reproduce the frequency dependence of the observed time - lags .
therefore , this toy non - linear model provides some theoretical underpinning for attempts to understand the rms - flux relationship but further developments of the model are required to capture the full behavior of real systems .
this work is complimentary to the analysis of the propagating fluctuation model performed by ingram & van der klis ( 2013 ) , however they begin from a statistical model for the mass accretion whereas we base our analysis on the non - linear disk equations .
this paper is structured as follows : section 2 outlines our modifications to the standard viscous disk model and numerical setup , section 3 presents the results of our simulations , and section 4 discusses the results of the previous section and provides suggestions for future work .
we consider accretion disks in the framework of a simple 1d viscous disk ( e.g. , see pringle 1981 ) .
we operate within the thin disk approximation which assumes that the material in the disk is narrowly confined in a common orbital plane .
this ensures that the vertical scale of the disk ( h ) is much less than the radial scale ( r , @xmath3 ) .
consequently , we must also assume that the disk is radiatively efficient . if this were not the case , viscous dissipation would cause the disk to heat to temperatures of order the virial temperature and hence puff up into a quasi - spherical configuration , violating our thin disk assumptions .
furthermore , also required to be consistent with the thin - disk assumption , we assume a fluid element s orbit is essentially keplerian in nature , with a very small radial velocity responsible for the actual accretion . while an obvious application of this model is to black hole accretion disks , we neglect relativistic effects and assume a newtonian gravitational potential ( @xmath4 ) .
the canonical equation for such a disk ( in units where @xmath5 ) is the non - linear diffusion equation given in pringle ( 1981 ) as @xmath6,\label{eq : basiceqn}\end{aligned}\ ] ] where @xmath7 is the distance from the central engine , @xmath8 is the surface density of the disk material , and @xmath9 is the effective shear viscosity of the disk . as we discuss in detail below , @xmath10 will be taken as a stochastic variable with some ( prescribed ) stationary base state @xmath11 to which we add spatially - uncorrelated but temporally - correlated stochastic fluctuations .
we now wish to make solutions to eqn .
( [ eq : basiceqn ] ) more computationally convenient with this eventual application in mind .
first , we introduce the change of variables @xmath12 .
equation ( [ eq : basiceqn ] ) then becomes @xmath13 multiplying both sides of equation [ eq : mixeq ] by the term @xmath14 ( which has no @xmath15 dependence ) while defining @xmath16 and @xmath17 , equation ( 2 ) can be written as @xmath18 which is the form of the disk evolution equation that serves as the basis for our numerical solution .
we are primarily concerned with the local mass accretion rate across , and viscous dissipation at , each radius .
the instantaneous mass accretion rate is @xmath19 which in our new variables translates to @xmath20 we can perform a similar manipulation of the dissipation , @xmath21 which gives @xmath22 the total disk luminosity is then given by @xmath23 ( 2 \pi r ) dr = \int_{\text{disk } } \frac{9 \psi}{x^4 } dx.\end{aligned}\ ] ] we now assume a stochastic perturbation in the viscous parameter of the form @xmath24 . here
, @xmath25 is the fluctuation relative to the baseline level @xmath26 , and physically represents the turbulent fluctuations in the maxwell stresses of a magnetohydrodynamic accretion disk . unlike previous treatments ( l97 and k11 ) , we do not assume that @xmath27 is small .
to proceed , we need to specify the nature of the random process @xmath27 .
we employ assumptions similar to those of l97 .
specifically , we suppose that the fluctuations are spatially uncorrelated ( likely to be valid for @xmath28 ) but have temporal correlations on some timescale @xmath29 .
thus , we impose the condition that @xmath30 when @xmath31 . the characteristic timescale imposes the condition that @xmath32 when @xmath33 .
a specific realization of these conditions can be generated by an ornstein - uhlenbeck ( ou ) process .
the ou process , models the evolution of a statistical variable of interest as it responds to an additive input noise with an exponential decay to its mean ( see kelly et al . 2009 , k11 ) .
this evolution is governed by a stochastic differential equation of the form @xmath34 where @xmath35 is the characteristic frequency , @xmath36 is the mean of the process , @xmath37 is the driving noise amplitude , and @xmath38 denotes a wiener process . in the typical case where @xmath38 is brownian motion ,
@xmath39 is gaussian white noise . equation .
[ eq : ou ] then simply represents a damped random walk .
equation [ eq : ou ] can be made physically relevant for our disk with the appropriate choice of parameters .
we assume the mean in the fluctuations is zero ( since the average component of the viscosity is already described via the @xmath40 term . ) .
next , we take guidance from l97 and , for most of the work presented here , assume that the relevant timescale for the fluctuations is the local viscous timescale such that @xmath41 , where @xmath42 .
we take @xmath39 to be gaussian white noise with zero mean and unity standard deviation . to ensure that the rms amplitude of the fluctuations @xmath27 are independent of radius we note that when @xmath39 has zero mean the variance is given by k11 as @xmath43 .
therefore , we must take @xmath37 as proportional to the square root of the local accretion time .
we normalize this factor to the accretion time at the inner edge of the disk yielding @xmath44 .
the final form of the ou equation is then : @xmath45 which satisfies all of the physically relevant requirements for the distribution of fluctuations .
we need one additional restriction on the viscous fluctuations in order to make our mathematical model well - posed .
if , due to a downward fluctuation , the overall viscosity @xmath10 becomes negative over an extended region of the disk , the basic diffusion equation is unstable and singularities will develop in the solution .
thus , we impose a positivity criterion whereever and whenever the solution of eqn .
[ eq : ou2 ] results in @xmath46 ( and hence @xmath47 ) , we peg @xmath48 .
lastly , when evolving @xmath49 , we scale the amplitude of @xmath27 by a fiducial value of 0.5
. this is done to control the effect of high sigma spikes on the stability of the solution .
we performed our simulations using an explicit numerical integration algorithm .
the simulation was run on a spatial grid with @xmath50 and @xmath51 with @xmath52 .
we chose a fiducial value for the baseline viscosity @xmath53 .
simulations were run with a time - step computed to safely satisfy the necessary cfl stability conditions , typically @xmath54 .
the initial condition on @xmath55 was chosen such that the disk began in a steady state with unity accretion rate , i.e. @xmath56 .
lastly , we imposed dirchlet boundary conditions @xmath57 and @xmath58 . in order to protect the simulation from stochastic effects at the outer boundary we established a buffer zone between @xmath59 and @xmath60 where @xmath61
the simulation was run for a duration of @xmath62 , and the state of the disk was recorded with a cadence of @xmath63 .
our primary goal is to examine whether the phenomenology of the broadband noise in real sources can be reproduced by our toy model .
consequently , we focus our analysis on the dissipation of the disk as this most closely corresponds to the observational emissions of real astrophysical disks .
we integrate the dissipation across the entire disk at each time to produce the total time - dependent luminosity of the disk .
this light curve can be seen in figure [ fig : lightcurve ] .
qualitatively , it resembles the light curve for an accreting source .
log - normality can be visually inferred by observing the decreased amplitude of fluctuations during dips compared with the flares .
the luminosity distribution is shown in figure [ fig : distribution ] .
we fit this distribution with a standard log - normal function of the form @xmath64,\end{aligned}\ ] ] where @xmath65 is the luminosity and the fitting parameters @xmath36 and @xmath66 are the mean and standard deviation respectively .
we found best fit parameters of @xmath67 and @xmath68 .
it must be noted that we also found a reasonable , albeit visually slightly worse , fit using a normal distribution with @xmath69 and @xmath70 .
the best - fit lines are plotted in figure [ fig : distribution ] .
we investigated the existence of an rms - flux relationship in our data using the integrated luminosity as a function of time .
the resulting light curve was divided into 600 equal segments of @xmath71 . for a black hole of 10@xmath72
this corresponds to @xmath73s per segment .
the mean luminosity @xmath74 was computed for each segment .
we then computed the rms fluctuation in each segment as @xmath75 where @xmath76 is the luminosity at the @xmath77 point in the segment .
the resulting rms - flux plot can be seen in figure [ fig : rmsflux ] . due to the nature of the stochastic noise in the simulation it is necessary to bin the resulting rms - flux data .
the data were linearly binned into 50 segments and then averaged over each bin .
the linear nature of the binned data is immediately apparent . following the prescription of uttley & mchardy ( 2001 ) we fit the binned data with a line of the form @xmath78 where @xmath79 and @xmath80 are fitting constants .
we obtained a fit to the data with @xmath81 and @xmath82 .
the fitted line is shown on fig .
[ fig : rmsflux ] to trace the data qualitatively well .
the fact that our simulation produces a linear result is an important confirmation of the log - normality of the light curve .
furthermore , this directly supports the phenomenological model that the multiplicative effect of numerous inward propagating fluctuations are the cause of the observed log - normality .
we next wish to understand how our light curves behave in frequency space .
we divided the total integrated disk luminosity into 36 segments , each of length 819200 m ( @xmath83 40 seconds at 10 @xmath72 ) . the fast fourier transform was taken for each segment .
we used these fourier segments , denoted @xmath84 , to construct the power spectrum density ( psd ) of the @xmath77 segment as per the standard prescription @xmath85 where the use of angle brackets denotes an averaging over segments .
the resulting power spectrum across the inner boundary is shown in figure [ fig : psd ] .
it can be seen that the psd is consistent with a system dominated by broadband noise .
there are no qpos , as expected given the simple nature of the model .
we attempted to fit the psd to a broken power - law of the form @xmath86 and @xmath87 , where we take the threshold frequency to be a free parameter of the model .
the best - fit returned @xmath88 and @xmath89 at a break frequency of @xmath90 .
the slope is likely not consistent with the @xmath91 expectation put forth by l98 because we are not dealing with a strict @xmath92 accretion disk .
however , while the power - law dependence is flatter than that indicated by observational data , this indicates a clear break in the power spectrum as expected .
the break frequency is comparable to the viscous frequency for the inner regions of the disk ( i.e. @xmath93 ) . given that the dissipation is dominated by the inner regions it is reasonable to expect the psd to obey these timescales .
a key aspect of the propagating fluctuation model is the behavior of the fluctuations as they propagate inward from the outermost radii .
therefore , it is important to analyze to what extent various disk radii are capable of communicating and at what frequencies this communication may occur .
this can be accomplished through the use of the coherence function . from nowak et al .
( 1999 ) we define the coherence function , @xmath94 , as follows : consider two time series @xmath95 and @xmath96 with fourier transforms @xmath97 and @xmath98 .
the coherence function between @xmath95 and @xmath15 is then @xmath99 where , again , the angle brackets indicate an averaging over segments of the data .
it then becomes clear that the coherence function is merely the modulus of the averaged cross - spectrum between two time series normalized by the averaged power spectra of the individual series .
consequently , the coherence function takes on values between zero and unity . in the case of our data
we take the dissipation at different radii as our time signals .
we can choose the inner radius as the first time series and compute its coherence function with all of the other radii in order to produce a coherence map , @xmath100 .
such an approach can be thought of as being analogous to the common practice of studying the coherence function between different wavebands . if we assume the emission is locally a blackbody then the natural temperature gradient in the disk will cause emission from two radii to occur at different wavebands . the resulting coherence map for the disk as shown in figure [ fig : coherencemap_visc ] .
there is clear structure in the map as radii become coherent on roughly the viscous frequency or less .
this provides some indication of the timescales on which radii communicate information about the perturbations .
we should note that coherence does not necessary arise because the radii are globally in perfect phase .
rather , it indicates that a linear transformation exists between the time series in the time domain , i.e. , @xmath101 where @xmath102 is the transfer function ( nowak et al .
our coherence map confirms the notion of l97 that , at a given radius , fluctuations are passed down to smaller radii only if they fall below the local viscous frequency .
if two radii of the disk have time - coherent dissipations , this implies the existence of well - defined ( frequency - dependent ) phase shifts between the dissipation time - series .
more precisely , if we write the cross - spectrum between two time series as a complex number then we have @xmath103 where @xmath104 is the frequency - dependent fourier phase shift .
the calculation of the phase shift produces values in the domain from @xmath105 $ ] with the assumption that @xmath106 as @xmath107 .
we can now scale the phase by the frequency yielding the frequency - dependent time lag : @xmath108 . given the nature of our model , the fourier time
lags represent the physical reprocessing time of the fluctuations as they are diffused in the disk . to illustrate the behavior of the phases and time lags , fig .
[ fig : phase_1_2 ] shows these quantities derived from the dissipation at @xmath109 and @xmath110 ( corresponding to @xmath111 and @xmath112 ) . from the coherence map ( fig .
[ fig : coherencemap_visc ] ) , we can see that these two radii are extremely coherent at low frequencies ( @xmath113 for @xmath114 ) , and then the coherence gradually falls to low values ( @xmath115 ) over a span of about 1.5 decades of frequency around the viscous frequency .
consistent with this , at low frequency there is little scatter in the phases from the individual segments .
the mean phase gradually increases magnitude as frequency increases in a manner consistent with approximately equal time - lag .
we can see that at frequencies around the local viscous frequency ( @xmath116 ) the scatter in the phase increases ( corresponding to the decreasing coherence ) . at frequencies above @xmath117 , the scatter in phase becomes large and phases get wrapping , becoming almost randomly scattered between @xmath118 and @xmath119 at @xmath120 .
the sharp decrease in the averaged phase shift above this frequency is entirely an artifact of averaging incoherence processes with a phase wrapped cross spectrum argument . in order to counteract some effects of phase wrapping we binned the data at the level of the cross - spectrum .
the phases and lags were then computed from these binned data .
looking at the time lag information across the same radii we see that the binned lag data shows a lag which is small for very low frequencies , increases in magnitude to @xmath121 at @xmath122 and then very slowly declines in magnitude to @xmath123 before coherence is lost .
given our convention , the sense of this lag is that the inner ( @xmath109 ) dissipation profile lags the dissipation further out , expected from inwardly propagating fluctuations . at frequencies above the local viscous frequency
the lag begins to weakly rise before the radii become incoherent . by comparison
, studies of the fourier lag in cygnus x-1 have shown lags with a strong frequency dependent decline ( @xmath124 and hence consistent with approximately constant phase shift ; nowak 1999 ) .
this is clearly not visible in the dissipation data .
it is interesting to note that the lag times are much less than the naively computed local viscous timescale , @xmath125 . when looking at the innermost radii it is important to check for the possible effects of the computational boundary on our analysis .
we repeated the fourier analysis , this time computing the coherence between the dissipation at radii @xmath110 and @xmath126 .
the viscous frequency between these two radii is @xmath127 .
we see in fig .
[ fig : phase_2_5 ] an interesting pattern of time lags , with slowly decreasing lags below the viscous frequency , and negligible lags ( approximately constant phase ) above the viscous frequency . coherence is completely lost ( resulting in essentially randomly distributed phases ) for @xmath128 .
this pattern of time lags has tantalizing similarities to the behavior seen in cygnus x-1 and other accreting sources .
the above analysis considers the behavior of the dissipation across the entire disk .
we also investigated the local behavior of key observables ( e.g. light curves , histograms , psds , and the rms - flux relationship ) at the specific radii @xmath129 , @xmath126 and @xmath130 .
we focused on the inner radii as that is the region of the disk where the dissipation is most dominant .
qualitatively , we observe the same behavior at each radii as the disk shows globally .
we observe dissipation light curves and histograms with indications of log - normal behavior , psds that show a clear break , and a linear rms - flux relationship . while we have considered the observational features of our simulated disk it is important to consider other physical processes .
given its dynamical importance , there is motivation to study the behavior of @xmath2 in our simulations . repeating the above analysis for @xmath2 we found many of the same behavioral properties .
there is clear log - normality in the light curve along with the expected linear rms - flux relationship .
the psd is composed entirely of broadband noise and shows no break at the viscous frequency at the inner radius .
the most striking differences between the two datasets are the coherence map and subsequent phase / lag information . the coherence map shows the same functional form as the as the dissipation , however @xmath2 appears to become coherent only at lower frequencies when compared to the dissipation ( fig .
[ fig : coherencemap_mdot ] ) .
we can understand this by examining the phases / lags at two illustrative radii ; again we choose @xmath109 and @xmath110 .
we see in fig .
[ fig : phase_mdot ] , for @xmath131 , there is a constant time lag .
however , this lag is substantially larger than that found for the dissipations , @xmath132 as opposed to the @xmath133 for the dissipations .
it is clear from fig .
[ fig : phase_mdot ] that the corresponding larger phase shift starts to phase wrap well before there is a loss of coherence .
thus , the coherence function @xmath134 is suppressed at intermediate frequencies by phase wrapping well before there is a genuine loss of communication between the radii . for completeness
, we examine the fourier relationship between the dissipation and mass accretion rate
. this could be astrophysically relevant if , for example , the hard x - ray emission from an accreting black hole was tied closely to mass accretion rate through the innermost circular stable orbit but the soft x - rays were thermal emission from a dissipative disk .
looking at the coherence map , we see that the regions of high coherence again trace the viscous frequency with regions becoming coherent at frequencies similar to those seen for @xmath2 ( figure [ fig : coherencemap_mdot_diss ] ) .
interestingly , there is evidence of convergence to a constant phase of @xmath135 ( which is also phase wrapped to @xmath118 ) in this phase plot ( figure [ fig : phase_mdot_diss ] ) . while the viscous timescale is one natural choice as the characteristic frequency for the driving perturbations it may be illuminating to consider other possibilities .
the only other relevant quantity in our toy model is the dynamical frequency , @xmath136 , assuming keplerian orbits .
we ran a second simulation with the characteristic frequency of the ou process set to the dynamical frequency .
analysis of the resulting data showed a disk that was strongly inconsistent with the aforementioned phenomenologies . while the simulation was able to produce a light curve with log - normality and a linear rms - flux relationship , the disk was not able to produce reasonable behavior in frequency space . the coherence map for both the dissipation and the mass accretion rate show large regions of little to no coherence , while the regions of high coherence are concentrated at low frequencies or along the inner boundary .
as an example , the coherence map for the dissipation is show in figure [ fig : coherencemap_dyn ] .
there is no hint of any meaningful structure in these plots . due to lack of general coherence
we do not expect to see any frequency - dependent phase behavior apart from at the lowest possible frequencies .
the phase and lag plots are shown in figure [ fig : phase_dyn ] .
we do note that there is a weak trend in the binned data above the viscous frequency likely due to a combination of noise and diffusion effects in the disk .
we have presented one of the first numerical attempts to understand the propagating fluctuation " model of lyubarskii ( 1997 ) in the context of non - linear accretion disk models . while the observational effects of this model have long been intuited from detailed studies of observational data , many key phenomenological features are not rigorously attributable to the linear model .
one of the most striking features is the linear relationship between the rms flux - variability and the average flux .
we have shown that our toy model , which starts with the basic disk equations , is capable of reproducing this relationship .
going beyond the rms - flux relationship , we can examine the behavior of our model disk in fourier space .
we find that the time - series of the dissipations at two different radii have temporal coherence up to frequencies comparable to , or slightly greater than , the viscous frequency relating the two radii . across most of this range of coherence frequencies
, the relationship between the dissipations appears to be one of constant time lag , with the inner radius lagging the outer radius by about a tenth of a viscous timescale . at higher frequencies
there is some hint , seen most clearly when comparing radii away from the inner computational boundary , for a constant phase regime ( with a time - lag that correspondingly declines with increasing frequency ) , but this only lasts over about a factor of @xmath137 in frequency before all coherence is lost .
this frequency behavior in real sources has been well studied observationally .
for example , detailed studies of cygnus x-1 with the _ rossi x - ray timing explorer ( rxte _ ; nowak et al .
1999 ; poutanen 2000 ; gilfanov et al . 2000 ; uttley et al . 2011 ) reveal coherence between the soft and hard x - ray bands at frequencies below the viscous frequency , possible regions of constant time - lag , and extended regions of approximately constant phase shift .
our model is able to reproduce the coherence behavior , but does not show extensive regions of constant phase .
nowak et al .
1999 also noted that black hole x - ray binaries show stepped " lag behavior , with the appearance of several sub - components each with their own constant lag .
it was thought that these steps in lag were due to the lorentzian sub - components used to model the psd .
it may then be the case that our observed constant lag behavior is the result of a single lorentzian component seen when comparing specific pairs of radii ( i.e. @xmath109 and @xmath110 ) , whereas in observations of an astrophysical disk we are observing the contributions of many radii . of course , given the simplicity of our model and treatment , the inability to explain the detailed phenomenology of real sources is not surprising .
observationally , coherence and time - lag studies must be performed on lightcurves extracted in particular bands . even in the case
where the observed emission is thermal , a given waveband picks out emission from a broad range of radii and the detailed lightcurve may be poorly approximated by the dissipation at one radius .
in addition , of course , it is possible for one ( or even both ! ) of the energy bands under study to be well above the thermal energy of the disk and , hence , actually originating from an x - ray corona the uncertain relationship between the underlying state of the disk ( @xmath2 or dissipation ) and the emission from the x - ray corona is another barrier in matching this simple model to the observations
. future developments of our approach will include more realistic models for relating the disk dynamics to the observables . ultimately , real astrophysical accretion disks are extremely complex systems .
even our 1d model may miss key disk behaviors by assuming azimuthal symmetry .
for example , the work of dexter & agol ( 2011 ) showed that 2d accretion disks with stochastic temperature fluctuations end up strongly inhomogeneous .
however , it is not clear if their stochastic behavior is driven by stochasticity in the accretion flow as captured by our model or is yet another complication in understanding the behavior of the accretion disk . clearly , the full range of interactions present require 3d magnetohydrodynamical models , often with general relativistic effects and radiative transfer , to fully understand .
however such models are computationally expensive , and for the foreseeable future it will not be feasible to run a grmhd model for many outer disk viscous timescales .
thus , simple models such as that presented in this paper can be viewed as a bridge between the full grmhd simulations and the observations , into which lessons from the simulations ( such as the statistics of the fluctuations or the physics of the emission processes ) can be imported .
future development of these models will follow such a direction .
we thank phil uttley for helpful discussion about this work .
we also thank our anonymous referee for their helpful comments about this work .
psc is grateful for support provided by the nsf through the graduate research fellowship program , grant dge1144152 .
csr thanks nasa for funding under the astrophysical theory program grant nnx10ae41 g .
dexter , j. , & agol , e. 2011 , , 727 , l24 gaskell , c.m . ,
apj , 2004 , 612 , l21 gilfanov , m. , churazov , e. , & revnivtsev , m. , 2000 , mnras , 316 , 4 ingram a. , & van der klis , m. , 2013 , mnras , 434 , 1476 - 1485 kelly , b. c. , bechtold , j. , & siemiginowska , a. 2009 , apj , 698 , 895 kelly , b. c. , sobolewska , m. , & siemiginowska , a. 2011 , apj , 730 , 52 lyubarskii , y. e. 1997 , mnras , 292 , 679 negoro , h. , & mineshige , s. , 2002 , pasj , 54 , l69 nowak , m. , et al . , 1999 ,
apj , 510 , 874 poutanen , j. , 2000 aas , head meeting # 5 , # 31.17 ; bulletin of the american astronomical society , vol .
32 , p.1235 pringle , j. , 1981 , ara&a , 19 , 137 scaringi s. et al . , 2013 , mnras , 431 , 2535 shakura , n. i. , & sunyaev , r. a. , 1973 , a&a , 24 , 337 titarchuk , l. , shaposhnikov , n. , & arefiev , v. 2007 , apj , 660 , 556 vaughan s. , fabian a.c . , nandra k. , 2003 , mnras , 339 , 1237 uttley p. , & mchardy i. m. , 2001 , mnras , 323 , l26 uttley , p. , mchardy , i. m. , & vaughan , s. 2005 , mnras , 359 , 345 uttley , p. , wilkinson , t. , cassatella , p. , et al . , 2011 , mnras , 414 , 1 wood , k. s. , titarchuk , l. , ray , p. s. , wolff , m. t. , lovellette , m. n. , & bandyopadhyay , r. m. 2001 , apj , 563 , 246 | we present a non - linear numerical model for a geometrically thin accretion disk with the addition of stochastic non - linear fluctuations in the viscous parameter .
these numerical realizations attempt to study the stochastic effects on the disk angular momentum transport .
we show that this simple model is capable of reproducing several observed phenomenologies of accretion driven systems .
the most notable of these is the observed linear rms - flux relationship in the disk luminosity .
this feature is not formally captured by the linearized disk equations used in previous work . a fourier analysis of the dissipation and mass accretion rates across disk radii show coherence for frequencies below the local viscous frequency .
this is consistent with the coherence behavior observed in astrophysical sources such as cygnus x-1 . |
to identify the true nature of the ways in which radio galaxies grow and evolve is not a trivial problem to tackle : we can not observe any radio galaxy at well - separated intervals throughout its life .
we can only make observations of a particular radio galaxy if it intercepts our light - cone . whether or not any radio galaxy is intercepted by our light - cone , and at what point in its life ,
is entirely random .
[ fig : rain1 ] schematically illustrates the randomness of the sampling of any radio galaxy by our light - cone .
unavoidably imposed on the random sampling by our light - cone is the effect of more selective sampling of the radio galaxy population determined by the _ finding - frequency _ of the survey .
different frequencies even within the radio regime , for example 151mhz and 1.4 ghz , find different populations with rather different selectivities : added complexities come from the fact that across different radio - frequency regimes different physical processes are responsible for the emission .
this is schematically illustrated in fig .
2 . at low frequency ( e.g.151mhz )
emission is largely isotropic while at ghz regimes doppler boosting preferentially selects objects with a small angle to the light - of - sight . in this way objects having low intrinsic jet - powers may be found by the survey .
with ghz surveys very young sources such as gpss / csos may be found which are synchrotron self - absorbed or free - free absorbed at lower frequencies .
unfortunately , the effect of a single finding - frequency varies as a function of the redshift of the sources being surveyed . for a given survey finding - frequency @xmath0
, the emission of a given radio galaxy at redshift @xmath1 is sampled at rest - frame @xmath2 , as illustrated in fig.[fig : rain3 ] . for objects characterised by steep spectra ,
this makes their detection at high redshift harder for a given observing frequency sensitivity .
-correction which is convolved with the finding - frequency of any survey : an object at redshift @xmath1 is sampled on the basis of its rest - frame emission at @xmath3 the finding - frequency .
the red colours here indicate the low - frequency at which the survey samples the emission .
the increasingly blue colours of the arrows at higher-@xmath1 represent the higher rest - frame frequencies at which they are sampled .
, width=377 ]
_ direct _ evidence for luminosity evolution in classical doubles , for example the decreasing luminosity of a radio galaxy with increasing size , could not be possible as fig.[fig : rain1 ] simply demonstrates : no radio galaxy has more than a _
brief encounter _ with our light cone .
we can not make well - spaced measurements along the time - axis of any given radio galaxy .
even if two properties such as luminosity and linear size might be closely related within an individual source , it does _ not _ follow that an anti - correlation will be manifested in the luminosities and linear sizes of a sample of many radio sources .
the finding of neeser et al .
( 1995 ) that the luminosities and linear sizes of classical doubles from two complete samples , 3c and 6c , are uncorrelated ( which is substantiated by our own study of three complete samples ( 3c , 6c and also 7c ) in blundell , rawlings & willott 1999 , see [ sec : pd ] therein ) should _ not _ be taken as the absence of evidence of a dependence of source luminosity on source size / age .
absence of evidence of such a dependence would not in any case constitute evidence of absence of such a dependence . in order to establish a full picture of the essential nature of how a radio galaxy evolves
as it gets older , it is essential to correctly understand the influence of the sampling functions of the surveys and to correctly measure the radio source observables ( luminosity , spectral index , redshift and linear size ) .
we now address these issues .
the consequences of a flux - limit , together with the rarity of classical doubles in the local universe , mean that for a single sample , such as 3c , there will be a tight correlation between the luminosities ( @xmath4 ) and redshifts ( @xmath1 ) of the members of that sample .
this is known as the @xmath4@xmath1 degeneracy .
this is illustrated in the left panel of fig .
[ fig : pz ] , which shows the distribution on the @xmath4@xmath1 plane of sources from the samples studied by leahy , muxlow & stephens ( 1989 ) and liu , pooley & riley ( 1992 ) .
though these are two different samples , they are both drawn from the 3c sample and subject to the @xmath4-@xmath1 degeneracy which afflicts the 3c or any other single flux - limited sample .
developing models of radio sources based on just such a set of sources runs the risk of modelling the selection effect of the @xmath4@xmath1 degeneracy . to circumvent such a problem ,
we have increased coverage of the @xmath4@xmath1 plane by using complete samples of radio sources which are selected at lower flux - limits ( rawlings et al .
while the flux - limit of the 3c sample is 12jy at 151mhz , the flux - limit of the faintest sample ( 7c ) is 0.5jy .
the substantial improvement in coverage of the @xmath4@xmath1 plane achieved in this way is illustrated by the right panel in fig.[fig : pz ] .
@xmath1 plane by the 3c sample is shaded in grey ; the filled circles represent sources from the sample studied by leahy , muxlow & stephens ( 1987 ) and the stars represent those sources studied by liu , pooley & riley ( 1992 ) _ right : _ our study is based on three complete samples selected at different flux - limits at 151mhz , a frequency sufficiently low that doppler biases are unimportant .
see discussion on the finite widths of these regions in blundell , rawlings & willott ( 1999 ) .
, width=510 ]
the integrated spectrum of a classical double radio source is usually either concave ( see fig . [
fig : spix ] ) or described by a simple power - law . in order to properly measure the spectral index at a common rest - frame frequency ,
it is necessary to fit the curvature of the spectrum and perform a @xmath5-correction to take account of the redshifted sampling of the spectrum .
we performed spectral fitting and @xmath5-corrections as described in blundell , rawlings & willott ( 1999 ) to all members of the 3c , 6c and 7c complete samples used in our study , after obtaining from the literature flux - density measurements for each object at 365 mhz , 408 mhz , 1.4 ghz and 4.86 ghz in addition to measurements from our own vla observations and from the survey measurements at 151mhz . hereafter in this paper we refer to the instantaneous slope of the @xmath6 flux - density
@xmath6 frequency plot evaluated at rest - frame 151mhz as the spectral index at 151mhz or @xmath7 , and evaluated at rest - frame 5ghz as the spectral index at 5ghz or @xmath8 .
flux - density _
v_.@xmath9 plot ) assumes that the spectrum follows a simple power - law .
a correct measurement of spectral index involves fitting for the curvature of the spectrum and correcting for the redshifting of the emission .
, width=377 ] luminosity is not as fundamental a property of a radio source as jet - power @xmath10 is .
this is because in addition to depending on the details of the environment into which the source is expanding , luminosity also depends on the frequency regime in which the luminosity is being evaluated and on the age of the source , even if the bulk kinetic jet - power remains constant . in the context of comparisons between radio - loud and radio - quiet quasars
we discuss how luminosity at ghz frequencies is a poor indicator of the underlying jet - power ( blundell & rawlings , in prep . ) ; this is especially but not exclusively true for those objects having significantly doppler boosted core emission . for the study
we describe here , all luminosities are evaluated at rest - frame 151mhz , following the spectral fitting procedure mentioned in
[ sec : spix ] .
many previous studies of samples of radio sources have required the use of redshift estimators for a significant fraction of the sample ( e.g. dunlop & peacock 1990 ) .
the most common estimator employs the @xmath11band ( 2.2 @xmath12 m ) magnitude .
use of this estimator is motivated by the small spread in the near - infrared hubble diagram ( @xmath11@xmath1 relation ) for radio galaxies in the 3c sample ( lilly & longair 1984 ; best , longair and rttgering 1998 ) .
it is now known that simple extrapolation of the 3c relation to higher redshifts and/or to fainter radio samples ( which contain fewer radio luminous sources at any given redshift ) leads to both large uncertainties in the estimated redshifts , and to systematic biases . for the sources in the 3c ,
6c and 7c samples used in our study , the redshifts are almost exclusively spectroscopic . in the very few cases where we have been unable to secure spectroscopic redshifts we have estimated redshifts by obtaining multi - colour ( optical and near - infrared ) information and fitting spectral energy distributions ( seds ) with template galaxy spectra ( e.g. dunlop & peacock 1993 ; willott et al . _ in prep _ ) .
angular sizes of radio sources were measured by taking the sum of the separation from the core to each hotspot or , in the absence of a core , the hotspot - to - hotspot separation .
angular sizes were converted into linear sizes in the usual way ( e.g. carroll , press & turner 1992 ) for three different cosmologies .
the low - frequency spectral index @xmath7 ( evaluated at 151mhz in the rest - frame ) is less subject to the effects of synchrotron and inverse compton losses off the cmb than higher - frequency spectral indices ( @xmath8 ) evaluated in the ghz regime . as such
, @xmath7 is a closer indicator of the energy distribution as initially injected than is @xmath8 .
we find that there is no dependence on redshift of @xmath7 , as shown in fig .
[ fig : za ] ( upper panel ) .
in contrast with the picture in 5.1 an increased dependence of ghz spectral index is seen with redshift in fig .
[ fig : za ] ( lower panel ) .
this points to the increasing importance of inverse compton losses off the cmb with increasing redshift , which was first pointed out by krolik & chen ( 1991 ) .
7 shows that the greater the linear size of a radio source the steeper is its low frequency spectral index .
this was the strongest correlation we found out of all of those we examined for the set of complete samples under study . we interpret this as being a consequence of adiabatic expansion losses and telling us that the magnetic fields decrease as the lobes expand , for the following reason .
( we consider and reject other possibilities in blundell , rawlings & willott 1999 . ) for a fixed observing frequency @xmath14 , the relationship between the lorentz factor ( @xmath15 ) of the particles giving rise to that emission and the ambient magnetic field is given by : @xmath16 where @xmath17 is the charge on an electron and @xmath18 is its rest - mass .
thus as the magnetic field decreases , higher-@xmath15 particles will be responsible for radiation at that fixed observing frequency .
since for lobe emission the spectrum of the energy distribution of particles is steep and frequently curved these higher-@xmath15 particles will be from a steeper part of the energy distribution , and hence a higher rest - frame 151mhz spectral index will be observed .
a correlation between the luminosities of classical doubles and their spectral indices was first suggested in 1960 by heeschen , and more recently by laing & peacock ( 1980 ) .
we find this correlation to be present for the combination of 3c , 6c and 7c complete samples in the sense that the more powerful the radio source , the steeper is its low - frequency spectral index .
one way of depicting the @xmath4@xmath19 correlation is to show for the @xmath4@xmath1 plane the effects of low - frequency spectral index on the flux - limits and how sources from the different complete samples cluster around the flux - limits ( see fig .
[ fig : pa ] ) .
all avoid the region between the @xmath20 flux - limit and the @xmath21 flux - limit .
there is no such avoidance seen for the lower - power 6c sources over the same redshift range .
, width=377 ] we have modelled this @xmath4@xmath19 correlation as coming from a dependence of the steepness of the energy distribution of particles injected into the lobe on jet - power .
we do not presume that the injection spectrum is universally given by the canonical value of @xmath20 .
we model the steepening of the energy distribution of particles as being controlled largely by classical synchrotron loss suffered in the high magnetic field of the hotspot .
we specifically invoke that the hotspot magnetic fields are higher in those sources with higher jet - powers ( as detailed in blundell , rawlings & willott 1999 ) .
this point is discussed further by de young ( these proceedings ) .
we choose to normalise this dependence of hotspot magnetic field on jet - power by assuming equipartition ( of the energy in the particles and the energy in the magnetic field ) in the hotspot ( for which there is good evidence in the case of cygnus a , harris et al .
we discuss further the role of the hotspot in
[ sec : model ] .
the ` linear - size evolution ' of radio sources has been known for many years ( kapahi et al .
1987 , barthel & miley 1988 ) .
we find for our combination of complete samples selected at low frequency that the decrease of linear size with increasing redshift is quite mild , in the sense that the parameterisation of linear size at constant luminosity and spectral index is given by @xmath22 .
we return to the physical mechanism responsible for linear - size evolution of classical double radio sources in [ sec : model ] .
the @xmath4-@xmath13 plane shows that there is no strong correlation between the luminosities and linear sizes of radio sources in the combination of complete samples .
given the sampling functions alluded to at the start of this paper which influence the nature of the objects which make it into our surveys , it should _ not _ be claimed that the absence of a correlation between @xmath4 and @xmath13 in any way might imply that for any given source as @xmath13 grows , @xmath4 will not evolve .
we have formulated a model for evolving radio galaxies which reproduces the above dependencies , and others , and this is detailed in blundell , rawlings & willott ( 1999 ) .
this model is the first model for radio galaxy evolution which incorporates the role played by the hotspot .
this role has two aspects .
_ first _ , the magnetic field of the hotspot causes significant radiative losses on synchrotron particles during their dwell - time in the hotspot , in the period of time before they are injected into the lobe .
we were led to this as discussed in
[ sec : pa ] by our observation that the radio sources in our complete samples seemed to strongly indicate that the more luminous a radio - source was , the steeper its _ low - frequency spectrum _ was ( for example , when evaluated at rest - frame 151mhz ) .
this rest - frame frequency regime is less affected by synchrotron and inverse compton losses than the ghz regime and thus more closely informs us on the energy distribution of particles _ as initially injected _ into the lobe of a radio - source .
modelling a steeper spectrum in a more powerful source came naturally out of invoking stronger magnetic fields ( hence stronger radiative losses ) in sources with higher jet - powers . by equating the jet - thrust and the pressure in the compact hotspot , and invoking equipartition in the hotspot
, we find that the magnetic energy density in the hotspot is proportional to the bulk kinetic power transported in the jet .
_ second _ , as a plasma element expands out of a compact hotspot into a lobe whose pressure is considerably lower ( and continues to become lower as the radio lobe expands and gets older ) it suffers enhanced adiabatic expansion losses compared to the expansion losses suffered by that element of plasma once it continues to dwell in the lobe .
this is consistent with observations of classical doubles which invariably show highly compact hotspots embedded towards the outermost edges of the smooth , low surface - brightness , extended emission which comprises the lobe albeit with a bewildering menagerie of shapes and structures ( leahy et al .
1997 , hardcastle et al .
1997 & black et al .
1992 ) .
the earliest ( scheuer 1974 ; baldwin 1982 ) to the most recent ( kaiser , dennett - thorpe & alexander 1997 ; blundell , rawlings & willott 1999 ) models of radio source evolution , with reasonable assumptions about the environments into which radio sources expand , predict that the luminosity of any individual radio source decreases as time or age increases .
this luminosity decline is enhanced when the role of the hotspot is included and illustrative ` tracks ' across the @xmath4@xmath13 plane of radio sources with given jet - powers and redshifts are shown in fig . [
fig : tracks ] . a combination of the dramatically declining luminosity - with - age of the high-@xmath10 sources
, their scarcity in the local universe , together with the harsh reality of survey flux - limits means that very powerful sources with large linear sizes are rarely seen .
this resolves the old problem first posed by baldwin in 1982 that there is a dearth of large _ and _ powerful classical doubles .
this problem had been a puzzle : intuitively [ and as later modelled by falle ( 1991 ) ] the more powerful radio - sources were expected to expand faster than those with lower jet - powers , so the laws of probability alone suggest it should be easy to find large , powerful sources .
but they are not : sources with @xmath23 and with @xmath24 are extremely rare in existing surveys ( see fig . [
fig : pd ] ) .
@xmath13 ' plane for the 3c sample in _ * a * _ are model tracks tracing out the evolution in luminosity and linear size of four example radio sources , with from top to bottom @xmath25 w at @xmath26 , @xmath27 w at @xmath28 , @xmath29 w at @xmath30 and @xmath31 w at @xmath32 .
small dots lie on all tracks to indicate the ( rest - frame ) ages 0.1 , 1 , 10 , 20 100 , 200 myr .
the dashed line indicates how the lower track luminosity reduces by @xmath33 half an order of magnitude if the ambient density becomes an order of magnitude lower . in _ * b * _
the horizontal lines represent the luminosities at which the 3c flux - limit of 12 jy takes its effect at the different redshifts indicated .
the same tracks as in _ * a * _ are plotted but here in _ * b * _ they are truncated when they reach the flux - limit .
source expansion in these plots is described by the model of falle ( 1991 ) . in his model
higher jet - power sources expand more quickly ; thus these objects fall through the flux - limit at a younger age .
the luminosity evolution is as described by the model of blundell , rawlings & willott 1999 , which incorporates the role of the hotspots .
, width=529 ]
in a recent letter to _ nature _ ( blundell & rawlings 1999a ) we demonstrated how declining luminosity tracks in combination with the consequences of applying a flux - limit at different redshifts would inevitably give rise to an effect which we termed the ` youth - redshift degeneracy ' : the fact that in a given flux - limited sample , those sources at higher redshift must be younger than their low redshift counterparts .
this effect applies very strongly to powerful sources when the role of the hotspot is incorporated into the model @xmath4@xmath13 tracks , as described in 6 .
the youth - redshift ( yz ) degeneracy means it is necessary to be careful about the correct interpretation of ` trends with redshift ' . in many cases
it is wrong to attribute such trends to changes in _ cosmic epoch _ when the effects arise because of _ source age_. we consider briefly five such examples of effects which occur because the high redshift sources are young sources , and hence are being observed not long after the jet - triggering event ( which is possibly a galaxy - galaxy merger , sanders et al 1988 ) .
* linear - size evolution and the yz - degeneracy * the linear size evolution observed in low - frequency flux - limited samples of classical double radio - sources ( kapahi et al .
1987 , barthel & miley 1988 , blundell , rawlings & willott 1999 ) arises because the high@xmath1 sources are younger , hence tend to be shorter .
it is the positive dependence on jet - power of the rate at which the lobe - lengths grow ( falle 1991 ) which contributes to the linear size evolution being as mild as it is ( see [ sec : dz ] and blundell , rawlings & willott 1999 ) . * structural distortion and the yz - degeneracy *
barthel and miley ( 1988 ) had suggested that higher redshift environments are denser and more inhomogeneous than at low redshift since they found increased distortion in the structures of their high-@xmath1 sample of steep - spectrum quasars compared with their low-@xmath1 sample .
sources which are younger may have the passage of their jets considerably more disrupted where there is a higher density and greater inhomogeneity in the ambient post - recent - merger environment .
a general trend of denser inter - galactic environments at high-@xmath1 can not be inferred from their result . * the alignment effect and the yz - degeneracy * where the alignment effect of optical light along radio - jet axes is caused by star - formation , it will be more easily triggered close in to the host galaxy or within the product of a recent merger [ assuming this is the jet - triggering event ( sanders et al 1988 ) ] than at distances further out sampled by the head of an expanding radio - source later in its lifetime . where the alignment effect is caused by dust - scattered quasar light , the certain youthfulness of distant radio - galaxies alleviates the near discrepancy ( de young 1998 ) between radio - source ages and the time - scale for which dust grains can survive in the presence of shocks caused by the advancing radio - jets . the youth
redshift degeneracy is consistent with the finding that the smallest sources in a sample of @xmath34 radio galaxies ( all with very similar luminosities ) are those which are most aligned with optical emission ( best et al 1996 ) .
best et al . remarked that the sequence of changing optical aligned structure with increasing radio size could be naturally interpreted by comparing it with different phases of the interaction of the radio jets with the interstellar and intergalactic media as the radio - sources age .
* faraday depolarisation and the yz degeneracy * garrington & conway ( 1991 ) have found a tendency for depolarisation to be higher in sources at higher @xmath1 .
objects with higher @xmath1 which are younger will be in much more recently merged environments with the consequence that inhomogeneities in density or magnetic field will more readily depolarise the synchrotron radiation from the lobes .
moreover , higher-@xmath1 sources being younger and tending to be somewhat shorter will be closer in to the centre of the potential well .
the higher density in this region will enhance the observed depolarisation .
* the yz degeneracy and dust emission * many of the highest-@xmath1 radio - galaxies have gas masses comparable to gas - rich spiral galaxies ( dunlop et al . 1994 ; hughes et al . 1998 ) and inferred star - formation rates which , in the local universe , are rivalled only by galaxy - galaxy mergers like arp 220 ( genzel et al 1998 ) . if high-@xmath1 objects are being viewed during a similar merging of sub - components the associated star formation could be responsible for a significant fraction of the stellar mass in the remnant galaxy . since the high-@xmath1 radio - galaxies those which hughes et al . detected with scuba are necessarily young ( @xmath35 years , see fig .
[ fig : tracks ] ) , and since the whole merger must take a few dynamical crossing times , or @xmath36 years , the implication is that the event which triggered the jet - producing central engine is synchronised with massive star formation in a gas - rich system , perhaps as material streams towards the minimum of the gravitational potential well of the merging system . the yz degeneracy may help explain why few lower-@xmath1 radio - galaxies show similarly large ( rest - frame ) far - infrared luminosities compared to the high-@xmath1 population : they are being observed significantly longer after the jet - triggering event .
dead or relic classical doubles appear to be very rare .
( giovannini et al
. 1988 , harris et al . 1995 ,
harris et al.1993 , cordey 1987 ) .
that there are so few implies that the turn - off timescale , both for the cessation of jet - activity and for the depletion of highly relativistic particles , must be very rapid .
once there is a cessation of jet - power , this implies that the radio source luminosity will decrease into oblivion .
what will govern this decline in luminosity ?
the radio source will expand causing adiabatic losses in the ways described in 5.3 ( see also scheuer & williams 1968 ) , but the expansion of the plasma lobe can only occur on timescales consistent with the sound speed so this can not be a rapid process . a rapid decay time does however come from considering the brevity of the radiative lifetimes of the synchrotron emitting particles . to radiate at rest - frame 5ghz in a magnetic field strength of 1nt , equation [ eq : lorentz ] tells us that the required lorentz factor of the electron is given by @xmath37 .
the radiative lifetimes of synchrotron emitting particles are often purported to be @xmath38 years , but such values overlook the fact that the lobe of a radio source is an evolving entity and the adiabatic expansion losses ( especially the decreasing magnetic field ) discussed earlier mean that particles with increasingly higher @xmath15 are required to produce the emission at the same chosen frequency as the source ages .
for example , in the situation above , a @xmath37 particle will emit most of its radiation at 5ghz if the magnetic field strength is 1nt .
if the field lowers adiabatically to 0.1nt then a particle whose lorentz factor is @xmath39 is required to produce the emission at the same ` observing frequency ' of 5 ghz . in a radio lobe , changes in the magnetic field go hand - in - hand with changes in the energies of the particles themselves since it is the expansion of the lobe which governs both .
thus the particle now radiating in the lower @xmath40-field of 0.1nt at 5ghz with @xmath41 would , prior to the decrease in magnetic field strength , have had @xmath42 .
the synchrotron cooling rate increases as @xmath43 , shortening the relevant radiative lifetimes further .
we explore quantitatively the radiative lifetimes of synchrotron emitting particles in blundell & rawlings ( 1999b ) , and find that they appear to be substantially shorter than source ages ( see figure 1 of that paper ) .
if we accept that the synchrotron radiative lifetimes are short then there is little difficulty explaining why synchrotron radiation in the ghz regime should cease on the same timescales . there would thus be no problem explaining the rarity of dead classical doubles .
but this picture leaves us with a problem in explaining how many large linear - size , low - powered radio sources have lobes with synchrotron emission often all the way back to the core .
these sources are almost certainly the oldest examples of classical doubles which we currently know of and at distances of many 100s kpc from the hotspot , synchrotron emission is seen from particles with a radiative lifetime substantially shorter than the age of the source .
reconciliation of these has traditionally come from assuming a fast bulk backflow speed within the lobe which transports the relativistic particles away from the hotspot rather faster than the lobe itself advances away from the core .
but this brings us to another problem with this standard picture : there must be a sink for all of this backflowing emission .
as jenkins & scheuer ( 1976 ) first pointed out , if synchrotron losses play a major part in limiting lobe lengths , then lobes lengths ( and by extension backflow reservoirs ) should be longer and more extensive at low frequencies than at high frequencies .
however , we do not find this to be the case .
images of bright 3c sources made using the new 74mhz receivers on the vla have remarkably similar silhouettes to fully sampled images at 1.4ghz .
some examples are shown in fig .
11 illustrating this , with images at ghz frequencies from the 3crr atlas web - site ( leahy , bridle & strom 1996 , http://www.jb.man.ac.uk/atlas ) shown for comparison .
these images show how the fractional contribution of the hotspots changes as a function of frequency and especially that there is no hint of any backflow emanating from the oldest parts of the lobes : in all of the 3c sources we have observed to date there is no evidence for any extended structure which is not already seen at ghz frequencies .
a full set of these images together with associated spectral index maps and analysis will be presented in kassim , perley & blundell ( in prep . ) .
a rapid diffusion of the high lorentz factor particles would reconcile the rapid disappearance of classical doubles ( note that converting classical doubles into fris would be too slow a process to solve this problem ) and the presence of synchrotron emitting material in lobes so far separated from the hotspot that their radiative lifetimes are shorter than the hotspot - travel - time from their location .
particle re - acceleration and inhomogeneous magnetic field strengths would only solve the latter , but not the former , problem . while it is clear that classical diffusion is too slow a process , other diffusion mechanisms which depend on the tangling scale and strength of magnetic fields within the lobes ( such as that discussed by duffy et al .
1995 ) merit further investigation .
thanks the royal commission for the exhibition of 1851 for a research fellowship . basic research in radio astronomy at the us naval research laboratory
is supported by the us office of naval research .
the vla is a facility of the national radio astronomy observatory operated by associated universities , inc .
, under co - operative agreement with the national science foundation .
hughes , d.h . , dunlop , j.s . ,
archibald , e.n . , rawlings , s. & eales , s.a . in _ cosmological evolution of the submillimetre luminosity of high - redshift radio galaxies in _ the birth of galaxies _ _ ( ed guiderdoni , b. ) rawlings , s. , blundell , k.m . , lacy , m. , willott , c.j . & eales , s.a . , in _ `
observational cosmology with the new radio surveys ' _ , eds m.n .
bremer , n. jackson and i.prez-fournon , kluwer academic publishers , p171 , ( 1998 ) .
* peter barthel : * another explanation for the @xmath4-@xmath19 correlation would be enhanced radiative efficiency in a hot confining medium ( x - ray halo ) such as is present around cygnus a and proposed to boost its radio luminosity ( barthel & arnaud , 1996 , mnras , 283 , l45 ) .
cosmologically evolving halos ( gopal - krishna & wiita , 1991 , apj ) plus the bias would explain all the cosmological effects you attribute to youth .
but , i do like your youth model guess several mechanisms are at work
* katherine blundell : * it is true that a more dense confining medium results in a higher radio luminosity ( see fig.[fig : tracks ] , which shows how an order of magnitude increase in density will increase the radio luminosity by half an order of magnitude ) .
it is not true that denser media will cause higher spectral indices to be observed , and especially not for low - frequency spectral indices .
it is worth reminding ourselves that we are trying to explain an observed correlation of luminosity with _ low - frequency _ spectral - index .
in fact the reverse effect will be important : for a more dense medium the expansion of the source is slower so the magnetic field does not fall off so quickly . in a _ higher _ magnetic field ,
lower lorentz factor particles are responsible for the radiation at a fixed observing frequency . given the usual shape of energy distributions in these objects ( i.e. steep curved spectra ) , lower lorentz factor particles will give rise to _ flatter _ not _ steeper _ spectra .
thus this mechanism can not be responsible for the @xmath4@xmath19 correlation .
* paul wiita : * some of the points you made concerning the youth - redshift degeneracy were made in a different context by gopal - krishna , myself and saripalli ( mnras , 1989 , 239 , 173 ) when we modelled the paucity of giant radio galaxies .
* katherine blundell : * our recent study which led to the ` youth - redshift degeneracy ' has illustrated that in fact the inverse compton losses are not by any means the most important factor in causing the deficit of large linear - size radio galaxies .
it is actually the drastic effects of a survey _ flux - limit _ together with luminosity tracks which decline because of adiabatic expansion losses which predominantly reduces the number of giant radio galaxies found in complete samples above @xmath44 .
( it should be noted that giant radio galaxies out to @xmath45 have been found by cotter , rawlings & saunders 1996 . )
inclusion of the role of the hotspot also enables us to explain not just the low - numbers of _ low_-powered large radio galaxies , but also the problem posed by baldwin ( 1982 ) that there is an _ absence _ of _ high_-power large size sources . | we describe the pitfalls encountered in deducing from classical double radio source observables ( luminosity , spectral index , redshift and linear size ) the essential nature of how these objects evolve .
we discuss the key role played by hotspots in governing the energy distribution of the lobes they feed , and subsequent spectral evolution .
we present images obtained using the new 74mhz receivers on the vla and discuss constraints which these enforce on models of the backflow and ages in classical doubles . , , , & |
on may 24 , 1997 ( day 3743 after core collapse ) , sonneborn _ et al . _
( 1998 ) observed supernova remnant ( snr ) 1987a with the space telescope imaging spectrograph ( stis ) .
they discovered very broad ( @xmath4 ) ly@xmath1 emission in the ultraviolet spectrum .
( 1998 , hereafter m98 ) interpreted this emission as the result of excitation of neutral hydrogen atoms in the supernova debris crossing a reverse shock .
the observed ly@xmath1 flux was close to the value predicted by borkowski _
( 1997 , hereafter b97 ) for such a model normalized to account for the soft x - ray flux observed by rosat .
m98 also pointed out that the broad ly@xmath1 emission should be accompanied by h@xmath1 emission that should be observable both from the ground and with stis .
the observations reported by sonneborn
_ et al . _
( 1998 ) were taken with a @xmath5 aperture that included the entire inner circumstellar ring of snr 1987a . with 24 arcsec@xmath6 dispersion ,
the resulting ly@xmath1 image seen on the stis detector is a convolution of the spatial ( @xmath7 ) and velocity ( @xmath8 ) structure of the emitting region .
m98 found that they could fit the observed image with a model in which the shock surface was a prolate ellipsoid slightly elongated in the direction normal to the ring plane , but they found evidence for an additional source of diffuse emission inside the shock surface .
this result was surprising , because the model in b97 , following chevalier & dwarkadas ( 1995 ) , implies that most of the emission should come from a ring shaped surface inside the inner circumstellar ring .
here we report a new set of stis observations of the fast shocks in snr 1987a , this time taken with slits covering only part of the emitting region and including h@xmath1 as well as ly@xmath1 .
these new observations enable us to distinguish the dynamics and morphology of the fast shocks more clearly .
we now see that the shock emission is indeed concentrated in the equatorial plane , contrary to our interpretation of the earlier stis observations .
we also see a diffuse component of emission that may indicate excitation and ionization by nonthermal particles .
we have not yet detected an unambiguous signal from high velocity nv @xmath9 emission .
the same set of observations also provides new information on the spectra of the supernova ejecta and the circumstellar ring , including a rapidly brightening `` hot spot '' that is apparently caused by the first encounter of the supernova blast wave with the ring .
we will interpret those data in subsequent papers .
the supernova intensive study ( sins ) collaboration obtained stis spectra of snr 1987a in ultraviolet and optical wavelengths in september and october 1997 respectively .
the ultraviolet spectrum , which covers the wavelength range 1130 1720 , was taken on jd 2,450,719.1 , or 3869.3 days since explosion , with the g140l low dispersion grating ( @xmath10 at ly@xmath1 ) .
the resolution along the spatial direction is @xmath11 .
five spectral images totaling 11200 sec of exposure were combined .
the optical spectrum , which covers the wavelength range 5260 10260 , was taken on jd 2,450,727.8 , or 3878.0 days since explosion , with the g750l low resolution grating ( @xmath12 at h@xmath1 ) .
the spatial resolution is @xmath13 .
two spectral images totaling 4800 sec of exposure were obtained .
the two images were dithered along the orientation of the slit so that both cosmic ray hits and hot pixels on the stis / ccd were removed simultaneously in the final combined data .
figure [ fig1]a illustrates the slit positions on the inner circumstellar ring for the two observations .
a @xmath14 slit was used for ly@xmath1 and a @xmath15 slit was used for h@xmath1 .
both slits were offset @xmath16 to the se of the centroid of the ring , with position angles of @xmath17 and @xmath18 for the ly@xmath1 and h@xmath1 observations , respectively .
= 0.5 figure [ fig2]a shows the ultraviolet spectrum in the vicinity of ly@xmath1 .
it displays the following components : ( 1 ) bright stationary geocoronal ly@xmath1 emission filling the slit ; ( 2 ) a nearly horizontal streak of intense blue - shifted ly@xmath1 emission below the centerline but inside the lower ( ne ) side of the ring extending to @xmath19 ; ( 3 ) a fainter horizontal streak of red - shifted emission above the centerline but inside the upper ( sw ) side of the ring extending to @xmath20 ; ( 4 ) a vertical band at @xmath21 due to diffuse , nv @xmath9 emission from nearly stationary nv ions in the vicinity of the ring ; and ( 5 ) a broad component of diffuse emission extending from @xmath22 to @xmath23 .
figure [ fig2]c shows the optical spectrum in the vicinity of h@xmath1 .
it displays the following components : ( 1 ) pairs of bright spots above and below the centerline due to emission by the nearly stationary inner ring at [ oi ] @xmath24 , [ nii ] @xmath25 , h@xmath1 , and [ sii ] @xmath26 ; ( 2 ) broad horizontal bands along the centerline due to emission in these same lines from atoms in the inner supernova debris excited by radioactivity ; and ( 3 ) a nearly horizontal streak of blue - shifted h@xmath1 emission below the centerline but inside the circumstellar ring extending to @xmath27 .
according to b97 and m98 , we expect to see ly@xmath1 and h@xmath1 emission from hi atoms in the supernova ejecta that are excited by electron and ion impacts as they cross a reverse shock .
( we do not expect to see ly@xmath1 and h@xmath1 emission from the blast wave because it is entering a region where the hydrogen atoms are already nearly fully ionized . )
since the hi atoms are freely streaming , with radial velocity @xmath28 , before they cross the shock surface , the doppler shifts of the ly@xmath1 and h@xmath1 are determined by the geometry of the surface .
m98 inferred from the earlier stis observations that the reverse shock surface was a slightly prolate ellipsoid .
for such a model , they predicted that a stis observation of ly@xmath1 through a @xmath14 slit would appear as an elliptical image ( their fig .
1d ) . the actual stis images of ly@xmath1 and h@xmath1 ( figs .
[ fig2]a , c ) do not agree well with the prolate ellipsoid model for the reverse shock . instead , they agree better with the model proposed by b97 , in which the reverse shock emission is confined primarily to an equatorial band .
we find also that the position and shape of the reverse shock surface vary from the far to the near side of the remnant , and probably on smaller scales as well .
we can fit both the blue - shifted ly@xmath1 and h@xmath1 observations with a model in which the shock on the near side has a radius which ranges from @xmath29 on the left side of the @xmath14 slit to @xmath30 on the right side of the slit and extends in latitude from @xmath31 to @xmath32 . to fit the red - shifted ly@xmath1 observations ,
the shock surface on the far side must have a radius @xmath33 and extend in latitude from @xmath34 to @xmath35 .
the red - shifted h@xmath1 emission was too faint to measure .
figure [ fig1]b illustrates the reverse shock surface and its orientation in the slits for the ly@xmath1 and h@xmath1 observations .
the surface outside of the slits is not constrained by the present data and is shown only as an extrapolation .
figures [ fig2]b , d show the resulting model ly@xmath1 and h@xmath1 stis images , respectively
. lines of constant latitude are shown on the simulated ly@xmath1 image assuming that all the emission comes from the center of the slit ( see 4 ) .
comparison of the model ly@xmath1 and h@xmath1 images demonstrates the advantage of a narrow slit in mapping the shock surface .
when we subtract the simulated images from the actual observations , the streaks vanish , leaving residual emission in the ly@xmath1 observation , which we discuss below in 4 .
the measured flux from the ly@xmath1 shock emission is @xmath36 on the blue side and @xmath37 on the red side .
the measured flux from the h@xmath1 shock emission is @xmath38 on the blue side and @xmath39 on the red side .
the blue - shifted ly@xmath1 flux from the shock is greater than the red - shifted flux by a factor of @xmath40 .
if we assume that the outer sn debris is spherically symmetric with a power law density profile given by @xmath41 ( eastman & kirshner 1989 ) , we can attribute this asymmetry to : the difference in shock radius ( which contributes a factor @xmath42 ) ; the fact that the reverse shock has a greater surface area on the near side ( a factor @xmath43 ) , and the @xmath44 year difference in light - travel time from the near side of the shock surface to the far side ( a factor @xmath45 ) .
this asymmetry should not come as a great surprise .
( 1997 ) found that the nonthermal radio emission seen inside the inner circumstellar ring was much brighter on the northeast side than on the southwest side .
the asymmetry in the reverse shock surface is not caused by an asymmetry in the supernova debris because then the red - shifted ly@xmath1 emission would be brighter due to the higher mass flux ( ram pressure ) required to drive the shock to a greater radius .
the asymmetry therefore represents an asymmetry in the circumstellar gas distribution which the ejecta is encountering ( the circumstellar gas on the far side must be more tenuous than the near side ) .
m98 calculated that , for the conditions behind the reverse shock , approximately 1 ly@xmath1 and 0.2 h@xmath1 photons will be emitted per hi atom entering the shock .
we can infer this emissivity ratio from the observations and our model , which allows us to correct for the fact that the ly@xmath1 observation is taken through a wider slit than the h@xmath1 observation . correcting for interstellar extinction factors of 0.15 at ly@xmath1 and 0.63 at h@xmath1 ( fitzpatrick _ et al .
_ 1990 ) we find , on the blue - shifted side , where the h@xmath1 streak is bright enough to measure , this ratio to be @xmath46 , consistent with our expectation .
since we know the number of ly@xmath1 photons emitted per hi atom crossing the shock surface , we can infer the normalization [ @xmath47 , the density at @xmath48 and @xmath49 years ] of the sn envelope s density profile from our model for the shock surface .
estimating that the reverse shock is propagating outward with velocity @xmath50 ( b97 ) and assuming a helium number density @xmath51 , we find @xmath52 , which may be compared with shigeyama & nomoto s ( 1990 ) model 14e1 ( @xmath53 ) and woosley s ( 1988 ) model 10h ( @xmath54 ) . if we had taken the value of @xmath55 inferred from radio observations of the remnant ( gaensler _ et al .
_ 1997 ) , @xmath47 would be decreased by @xmath56 .
when we subtract the model images from the actual images , a diffuse component of high velocity ly@xmath1 emission remains . this component was evident in the earlier slitless observation by sonneborn _
( 1998 ) , and led m98 to interpret those data with a model in which the reverse shock surface had the shape of a slightly prolate ellipsoid . now , with the present observations
, we can see clearly that the ly@xmath1and h@xmath1 emission from the reverse shock is confined to the equatorial plane , and that the residual ly@xmath1 emission comes from a volume interior to that surface . a corresponding diffuse component of h@xmath1presumably exists , but it is too faint to be seen with the present exposure . to analyze the distribution of the diffuse ly@xmath1 emission , we assume that the emitting gas is everywhere in free expansion .
if so , and if the slit were very narrow , there would be a unique mapping from the stis image to depth within the supernova debris .
from such a mapping , we could reconstruct a two - dimensional image of the supernova debris intercepted by the slit .
figures [ fig3]a - c are just such images , where the coordinates are radius and latitude measured from the equatorial plane of the inner circumstellar ring .
the curves of constant latitude are not radial lines because the slit is off - center .
= figure [ fig3]a represents the ly@xmath1 data of figure [ fig2]a in such a coordinate system , in which we assume that the slit is a line centered in the actual slit .
since the actual slit has width @xmath14 , the mapping is not unique ; this introduces a `` point - spread function '' for the mapping represented approximately by the error circles in figures [ fig3]a - c .
since we can not distinguish the contribution of red - shifted ly@xmath1 from that of nv @xmath9 with the present data , we can not interpret the signal on the lower right hand side of figure [ fig3]a .
figure [ fig3]b represents the model for ly@xmath1 emission from the reverse shock surface corresponding to figure [ fig1]b , including the broadening due to the @xmath14 slit . in this coordinate system , we can see clearly that the blue - shifted ly@xmath1 emission comes from a reverse shock surface that lies primarily above the equatorial plane .
figure [ fig3]c shows the residual flux on the blue side of ly@xmath1after subtracting the contribution from the reverse shock from the data .
this diffuse ly@xmath1 emission comes from supernova debris inside the reverse shock surface and appears to have a roughly uniform radial distribution . like the ly@xmath1 emission from the reverse shock surface ,
the diffuse ly@xmath1 emission is strongest within latitudes @xmath57 to @xmath58 ; but we see significant diffuse emission from @xmath59 to @xmath60 .
the flux of diffuse blue - shifted ly@xmath1 seen through the @xmath14 slit is @xmath61 , or @xmath62 times the flux of blue - shifted ly@xmath1 from the reverse shock .
each pixel in the spectrum observes a volume element in the debris so we are able to measure the volume emissivity of the diffuse interior emission . correcting for interstellar extinction and assuming a distance of @xmath63 to the remnant we find this emissivity to be @xmath64 . what causes this diffuse emission ?
since it is concentrated in the same latitude range as the reverse shock emission , one is tempted to look at the reverse shock itself as the source of the excitation .
we can rule out one obvious candidate soft x - rays emitted by the shock because the soft x - ray luminosity observed by rosat ( hasinger et al .
1996 ) is roughly an order of magnitude too faint to account for the diffuse ly@xmath1 .
likewise , the hot spot is still too faint to cause significant diffuse ly@xmath1 emission .
we suspect that the diffuse ly@xmath1 emission is caused by nonthermal particles that are accelerated at the reverse shock and diffuse into the freely expanding supernova envelope . we know that such particles are present from observations of nonthermal radio emission ( gaensler _ et al .
_ 1997 ) . the reverse shock may be a fairly efficient source of nonthermal particles ( jones & ellison 1991 ) .
moreover , those nonthermal particles that diffuse into the supernova envelope and deposit their energy may produce ly@xmath1 photons with efficiency approaching @xmath65 .
further observations with stis will give us a much clearer picture of the complex dynamics and kinetics of this extraordinary event .
it would be valuable to observe the system again in the same slit configurations to measure the rate of brightening in ly@xmath1 and h@xmath1emission from the shock surface .
we should map the high velocity emission from the entire system , especially near the east side where the radio lobes are brightest .
as m98 have emphasized , observations of the line profile of high velocity nv @xmath9 emission will provide a unique probe of the kinetics of a collisionless shock .
we believe that the emission seen on the red side of ly@xmath1 contains a significant contribution from nv @xmath9 , but we can not distinguish this emission from ly@xmath1 emission in the present observations . to do that , we must observe the high velocity emission from the same part of snr 1987a with stis in both ly@xmath1 and h@xmath1 .
the exposure should be deep enough to see the diffuse emission clearly in h@xmath1 as well as ly@xmath1 , and the slit should be narrow in order to separate the spatial structure from the velocity structure and to isolate the nv @xmath9 emission from stationary gas . then , since the ratio of ly@xmath1/h@xmath1 should be constant , we would be able to subtract the h@xmath1image from the ly@xmath1 image , leaving a remainder that is due to nv @xmath9 .
this research was supported by nasa through grant no .
go-2563.01 - 87a from the space telescope science institute , which is operated by the association of universities for research in astronomy , inc .
, under nasa contract nas 5 - 26555 , and by nasa grants nag 5 - 3313 to the university of colorado and 5 - 2844 to north carolina state university .
c. s. j. p. acknowledges funding by the stis idt through the national optical astronomy observatories , and by the goddard space flight center .
borkowski , k. , blondin , j. , & mccray , r. 1997 , apj , 476 , l31 ( b97 ) chevalier , r. a. , & dwarkadas , v. v. 1995 , apj , 452 , l45 eastman , r. g. , & kirshner , r. p. 1989 , apj , 347 , 771 fitzpatrick , e. l. , & walborn , n. r. 1990 , aj , 99 , 1483 gaensler , b. m. , manchester , r. n. , stavely - smith , l. , tzioumis , a. k. , reynolds , j. e. , & kesteven , m. j. 1997 , apj , 479 , 845 hasinger , g. , aschenbach , b. , & tr@xmath66mper , j. 1996 , a&a , 312 , l9 jones , f. c. , & ellison , d. c. , space sci .
rev . , 58 , 259 michael , e. , mccray , r. , borkowski , k. j. , pun , c. s. j. , sonneborn , g. 1998 , apj , 492 , l143 ( m98 ) shigeyama , t. , & nomoto , k. 1990 , apj , 360 , 242 sonneborn , g. , _ et al .
_ 1998 , apj , 492 , l139 woosley , s. e. 1988 , apj , 330 , 218 | we describe and model high velocity ( @xmath0 ) ly@xmath1 and h@xmath1 emission from supernova remnant 1987a seen in september and october 1997 with the space telescope imaging spectrograph .
part of this emission comes from a reverse shock located at @xmath2 of the radius of the inner boundary of the inner circumstellar ring and confined within @xmath3 of the equatorial plane .
departure from axisymmetry in the ly@xmath1 and h@xmath1 emission correlates with that seen in nonthermal radio emission and reveals an asymmetry in the circumstellar gas distribution .
we also see diffuse high velocity ly@xmath1 emission from supernova debris inside the reverse shock that may be due to excitation by nonthermal particles accelerated by the shock . |
recent studies have demonstrated that galaxies at @xmath0 can be efficiently selected by the simple observed near - infrared ( nir ) color criterion @xmath1 ( franx et al .
2003 ; van dokkum et al .
2003 ) . these distant red galaxies ( drgs )
are typically very faint in the rest - frame ultra - violet ( uv ) , and are complementary to the uv - selected lyman break galaxies ( lbgs ) .
studies of their broad - band spectral energy distributions ( seds ) ( frster schreiber et al .
2004a ) and their rest - frame optical emission lines ( van dokkum et al .
2004 ) indicate that at a given rest - frame optical luminosity drgs are dustier , more massive , and have higher ages than lbgs .
their star formation rates are quite uncertain , as they depend on the assumed star formation histories : median values for drgs with @xmath15 range from @xmath16 for constant star formation histories to @xmath17 for declining models ( frster schreiber et al.2004a ) .
in this paper we study the x - ray emission from drgs , using chandra observations of the field centered on the foreground cluster ms105403 at @xmath18 .
this field is one of the two faint infrared extragalactic survey ( fires ) fields and its population of red @xmath0 galaxies has been described by van dokkum et al .
( 2003 ; 2004 ) and frster schreiber et al .
( 2004a ) .
the x - ray properties provide information on the fraction of active galactic nuclei ( agn ) among drgs .
spectroscopy indicates that this fraction could be substantial ( van dokkum et al .
2003 , 2004 ) , but this interpretation is uncertain because current spectroscopic samples are very incomplete .
furthermore , the stacked x - ray flux of undetected sources provides an additional constraint on the average instantaneous star formation rate . where needed we assume @xmath19 , @xmath20 , and @xmath21kms@xmath22mpc@xmath22 .
galaxy identifications refer to the frster schreiber et al .
( 2004b ) fires catalog of the ms105403 field , which includes photometric redshifts for all sources .
the x - ray data were obtained from an archival 91ks exposure with the chandra acis - s3 detector .
the data were obtained by the acis gto team to study the foreground cluster ( jeltema et al .
the reduction of the data and an analysis of the point sources in the field are presented in johnson , best , & almaini ( 2003 ) .
after removal of periods with strong background flaring the effective exposure time is 74ks .
images of the full energy band ( 0.58 kev ) were made , as well as of the soft ( 0.52 kev ) and hard ( 28 kev ) bands .
the drgs were selected from the @xmath23 area imaged with the vlt infrared spectrograph and array camera in the @xmath24 , @xmath25 and @xmath26 bands as part of the fires project .
hst wfpc2 data is available in the @xmath27 and @xmath28 bands , as well as vlt fors1 data in the bessel @xmath29 , @xmath30 and @xmath31 bands .
the reduction , source detection , and photometry are described in frster schreiber et al .
( 2004b ) .
there are 45 sources with @xmath32 , @xmath33 , and photometric weights ( i.e. , the fraction of the maximal exposure time spent on - source ) in @xmath24 and @xmath26 greater than @xmath34 .
figure [ jk.plot ] shows the selection limits . from this group
, we removed 2 sources because they have photometric redshifts @xmath35 , and three sources within @xmath36of the center of the cluster , to minimize effects of the diffuse cluster light .
the final sample thus consists of 40 drgs .
johnson et al . ( 2003 ) selected point - sources in the ms105403 field using wavdetect with a rather conservative threshold ( giving @xmath37 spurious source over the entire field ) , and we first determined the overlap between drgs at @xmath3 and x - ray sources in the johnson et al .
catalog ( large solid circles in fig . [ jk.plot ] ) .
only one drg ( object 1100 ; @xmath38 ; @xmath39 ) was found to contain an x - ray source identified by johnson et al .
( object 30 in their catalog ) .
next we examined the prevalence of x - ray sources fainter than the conservative limits imposed by johnson et al .
the x - ray fluxes were determined using apertures of @xmath40 diameter , which contain @xmath41% of the flux for point - sources over our entire @xmath42 fires field . for each drg
a local background was determined from the mean flux in 30 randomly placed apertures near the location of the drg .
the locations of these apertures were constrained to be non - overlapping , at a distance of @xmath43 from the drg , and at least @xmath44 from point sources identified by johnson et al .
in addition to galaxy 1100 we find one other object ( 313 ; @xmath45 ; @xmath46 ) with a significant ( @xmath47 ) x - ray flux .
the seds of both objects are shown in fig .
[ seds.plot ] , along with the median sed of all other drgs ( in the observed frame ; for rest - frame seds see frster schreiber et al .
the sed of object 313 is very similar to the median , suggesting that the agn does not contribute significantly to the continuum flux .
however , galaxy 1100 has much bluer optical - nir colors than the majority of drgs indicating that its active nucleus may effect the sed , in particular in the rest - frame ultra - violet .
the @xmath48 limit of 5.3 counts ( after background subtraction ) corresponds to an x - ray flux of @xmath49 ( see below ) , and a k - corrected luminosity of @xmath50 in the rest - frame @xmath51kev band for the median @xmath52 of the drgs in our sample .
we conclude that the fraction of agn with @xmath53among drgs is @xmath4% ( assuming poisson statistics ) .
we note here that both _
spectroscopically_-identified agns in the ms105403 field have dropped out of the drg sample after recalibration of their photometry ( see van dokkum et al .
interestingly only one of the two ( object 1035 , with @xmath54 ; labeled in fig .
1 ) is detected with chandra , suggesting that a multi - wavelength approach is necessary to accurately determine the agn fraction among red galaxies at @xmath0 .
the low x - ray background allows very efficient stacking analyses of sources that individually are too faint to be detected ( e.g. , alexander et al .
2003 ; brandt et al . 2001 ; malhotra et al . 2003 ;
reddy & steidel 2004 ) .
we averaged the 38 flux measurements of the undetected drgs , effectively increasing the exposure time to 2.8ms .
we also averaged each of the 30 sets of background measurements .
these 30 random stacks have the same properties ( background and point spread function ) as the red galaxy stack , and are used to assess the significance of the flux in the averaged red galaxy apertures .
the results are listed in table 1 .
the average background and significance in each band was determined in the following way .
each of the 30 sets of random apertures was averaged in the same way as the drg apertures , providing 30 independent measurements of the mean of 38 random apertures .
the mean of these 30 measurements gives the background value , and their rms is used to calculate the significance of the mean detection .
after background subtraction , the mean emission from drgs is 0.54 counts ( 3.2@xmath55 significance ) in the full band , 0.41 counts ( 3.4@xmath55 ) in the soft band , and 0.14 counts ( 1@xmath55 ) in the hard band .
images of the average detections are shown in fig .
[ stack.plot ] .
a concern is that the mean flux is dominated by one or two luminous sources which are just below the detection threshold .
figure [ fluxhist.plot ] shows the distribution of the x - ray fluxes of our undetected drg sample after background subtraction , as well as the average distribution of the 30 random samples .
the drgs show a tail of positive fluctuations and a deficit of negative fluctuations , consistent with the hypothesis that most sources contribute to the mean detection .
we empirically test how much flux is lost by our adopted @xmath40 apertures by repeating our measurements of the mean x - ray emission using differently sized apertures .
the results are listed in table 1 . as expected , the flux is a function of aperture size .
based on the values in table 1 and the acis documentation we estimate that the @xmath40 apertures miss @xmath56% of the flux at the average location of the drgs ( @xmath57 from the optical axis ) . this aperture effect is partially countered by weak lensing of the background drgs by the foreground @xmath18 cluster .
this effect is @xmath58% on average ( see frster schreiber et al .
the aperture- and lensing - corrected mean flux of the drgs is @xmath59 ( random ) @xmath60 ( systematic ) counts in the full band .
the systematic error reflects the uncertainties in the two corrections .
the average x - ray detection of the drgs provides a constraint on their instantaneous star formation rate . in general , x - ray emission of galaxies is due to a combination of hot gas ( e.g. , from supernova bubbles and winds ) , low mass x - ray binaries , high mass x - ray binaries ( hmxbs ) , and possibly an active nucleus .
hmxbs are short - lived as they trace the prompt formation rate of massive stars with @xmath61 , and it appears that for galaxies with star formation rates in excess of @xmath62 they dominate the x - ray luminosity , particularly above 2kev ( grimm et al .
several recent studies have therefore explored the relation between the instantaneous star formation rate and the rest - frame 210kev x - ray luminosity ( e.g. , ranalli et al .
2003 , grimm et al .
2003 ) . here
we use the grimm et al.(2003 ) relation , sfr@xmath63 .
the grimm et al .
calibration gives @xmath64% lower star formation rates than the ranalli et al .
calibration .
we convert the observed counts in the @xmath6kev energy band to a rest - frame @xmath51kev luminosity in the following way .
first we determined the average conversion from counts to flux from the sources listed in johnson et al .
the conversion varies by less than 10% over the acis field .
their conversion assumes a spectrum with photon index @xmath65 , and we use the portable interactive multi - mission simulator ( pimms ) to calculate the effect of changing @xmath66 from 1.7 to 2.0 , which is the appropriate value for the @xmath67kev spectrum of hmxbs . applying this @xmath58% correction to the empirically determined conversion factor we obtain @xmath68 , with @xmath69 in units of @xmath70 .
the lensing- and aperture - corrected mean drg flux is then @xmath71 in the 0.58kev band .
next , the flux is k - corrected to the rest - frame 210kev band ( again assuming @xmath72 ) , using @xmath73 . converting the rest - frame flux to a rest - frame luminosity ( for @xmath73 ) we find @xmath74 , and the grimm et al . ( 2003 ) calibration gives an instantaneous star formation rate of @xmath11 ( random ) @xmath12 ( systematic ) .
the systematic error is due to uncertainties in the aperture- and lensing correction , uncertainties in the conversion of x - ray luminosity to star formation rate , and the scatter in the relation between x - ray luminosity and star formation rate .
it was assumed that these uncertainties can be added in quadrature .
we note that pimms predicts a @xmath75 ratio for the soft vs.hard band counts , consistent with our non - detection in the hard band .
our estimate of the star formation rate in drgs based on x - ray emission can be compared to previous estimates based on sed fitting and emission lines ( van dokkum et al .
2004 ; frster schreiber et al .
the main uncertainty in the sed fits is the assumed star formation history .
frster schreiber et al . ( 2004a ) find a median ( mean ) instantaneous star formation rate of @xmath76 ( 190 ) for drgs with @xmath77 in the ms105403 field assuming continuous star formation models , and only 23 ( 69 ) for declining models with @xmath78myr . as shown in van dokkum et al .
( 2004 ) the inclusion of h@xmath79 measurements does not break the degeneracy between star formation history , dust , and instantaneous star formation rate .
our result is in excellent agreement with the high star formation rates derived from continuous ( dusty ) models , and may imply that drgs contribute significantly to the cosmic star formation rate at @xmath13 .
it is also interesting to compare the mean x - ray luminosity of drgs with that of lyman break galaxies ( lbgs ) .
reddy & steidel ( 2004 ) find a mean x - ray flux of @xmath80 in the @xmath8kev band and an implied sfr of @xmath81 for lbgs at @xmath82 in the chandra deep field north . as the redshift range of these lbgs and
our drgs is roughly similar we can compare the fluxes directly , limiting systematic uncertainties .
the mean flux of the drgs is @xmath83 in the @xmath8kev band , a factor @xmath62 higher than that of lbgs .
the interpretation is unclear , as the two samples have not been matched in observed @xmath84 magnitude , their luminosity functions may be different , and the space density of drgs is still uncertain .
our estimate does not include one significant systematic uncertainty , likely to be present in most attempts to infer high redshift star formation rates from x - ray luminosities .
for nearby galaxies the angular resolution of chandra can be used to subtract out the possible contamination from a central active nucleus , but this is not possible for high redshift objects .
all the x - ray luminosity attributed to hmxbs could in fact come from a population of low luminosity agn . since we do not have optical spectra for most of our obects and we do not have the x - ray counts and count rates to impose constraints based on x - ray spectral and variability information
, we can not directly rule this possibility out . as a general caution ,
we remark that even if we did have optical spectra , it might be quite difficult to rule out the presence of a low luminosity agn .
if we assume that the relation between the mass of a galaxy bulge and the mass of the bulge s central black hole ( ferrarese & merritt 2000 ; gebhardt et al . 2000 ) is already established for these objects , then
the red galaxies in our sample have typical black hole masses @xmath85 the mean x - ray luminosity of @xmath86 is a factor @xmath87 below the typical x - ray luminosity for unobscured agn with similar black hole masses , i.e. , the nuclei of these galaxies are heavily obscured or the possible black holes in these galaxes are accreting at low , sub - eddington rates . in either case , if one does not have the angular resolution to separate out the nucleus , the starlight from the bulge can completely dominate the agn light ( e.g. , moran et al .
the lack of obvious agn features in a spectrum is thus not conclusive evidence against an agn contribution if the aperture used contains most of the galaxy and the signal to noise ratio of the spectrum is low .
a better ( but still circumstantial ) argument that the x - ray luminosity represents star formation might instead be the concordance of several independent star formation indicators , e.g. , between the radio , de - reddened uv , and x - ray in reddy & steidel ( 2004 ) .
further independent constraints on the star formation rate in drgs are thus urgently needed .
studies of larger samples will provide a better measurements of the number of _ luminous _ agns : whereas our initial spectroscopy indicates agn fractions of @xmath88% ( van dokkum et al .
2003 , 2004 ) the chandra data suggest that this is largely due to the strong spectroscopic selection bias toward galaxies with bright emission lines .
it is also important to directly assess the effect of agns on the broad - band seds ( and hence estimates of star formation rates , ages , and dust content ) .
planned observations with nicmos and acs on hst of ms105403 will provide constraints on the point - source contribution to the continuum from the rest - frame uv to the optical .
finally , deeper x - ray observations are needed to further constrain the hardness ratio of the mean detection and to detect more individual x - ray sources ; our results and those of barger et al .
( 2003 ) suggest that in 12ms exposures many of the @xmath0 red galaxies may be ( faint ) chandra sources . | very deep near - infrared observations in the faint infrared extragalactic survey ( fires ) have recently uncovered a significant population of red galaxies at redshifts @xmath0 . these distant red galaxies ( drgs )
are efficiently selected by the criterion @xmath1 .
we use chandra data to examine the x - ray emission from drgs in the @xmath2 fires ms105403 field .
two of 40 drgs with @xmath3 are detected by chandra , and we infer that @xmath4% of drgs host active nuclei with @xmath5 . this fraction is smaller than that inferred from optical and near - ir spectroscopy , probably largely due to strong spectroscopic selection biases . by stacking all undetected drgs
we find that their average x - ray flux in the @xmath6kev band is @xmath7 .
the detection is only significant in the soft ( @xmath8kev ; @xmath9 ) and full ( @xmath6kev ; @xmath10 ) energy bands .
the mean detection may result from star formation , the presence of low luminosity agn , or a combination of both .
assuming the detection is due exclusively to star formation , we find an average instantaneous star formation rate of @xmath11 ( random ) @xmath12 ( systematic ) , in excellent agreement with previous results from spectral energy distribution fitting when constant star formation histories are assumed .
these results may imply that drgs contribute significantly to the cosmic star formation rate at @xmath13 .
however , the mean x - ray flux strictly provides only an upper limit to the star formation rate due to the uncertain contribution of low luminosity , possibly obscured agns .
observations at other wavelengths are needed to provide independent estimates of the star formation rate of drgs .
50min@xmath14 727[oii]3727 007[oiii]5007 1054ms105403 ang4000 |
the increasing importance of micro- and nanotechnology initiated an intensive research in heat conduction @xcite .
experiments show deviations from the classical fourier theory @xcite and there are several theoretical developments to understand the nature of the deviations @xcite .
the starting point of all these approaches is the balance of internal energy @xmath0 where @xmath1 is the density , @xmath2 is the specific internal energy , and @xmath3 is the conductive part of the current density of internal energy , i.e. , the heat flow .
we are dealing with rigid heat conductors , the dot denotes the partial time derivative , @xmath4 is the space derivative of the corresponding physical quantity , and index notation with the einstein summation convention is applied . in the phenomenological generalizations of fourier s law of classical heat conduction , the constitutive equation ( [ four ] )
is modified by additional terms .
the most important modifications are the following : @xmath5 here , ( [ four ] ) is the classical fourier s law @xcite , ( [ catver ] ) is the maxwell - cattaneo - vernotte ( mcv ) equation @xcite , eq .
( [ guykru ] ) is the guyer - krumhansl ( gk ) equation @xcite , ( [ jef ] ) is known as the jeffreys type or lagging heat equation @xcite , and ( [ grenag ] ) leads to the green - naghdi ( gn ) equation @xcite .
the heat conduction coefficient is denoted by @xmath6 , the relaxation time by @xmath7 , and @xmath8 , @xmath9 and @xmath10 are further material parameters . the origin the motivation and derivation of these equations is manifold .
kinetic theory offers several different approaches for fourier s law and for mcv equation .
for instance , moment series expansion of the boltzmann equation results in fourier s law in first order and in the mcv equation in second order @xcite .
also , the gk equation was originally derived with the help of special collision terms in the boltzmann equation characterizing phonon - lattice interaction @xcite .
the phenomenological theories are also diverse regarding the origin of the above equations @xcite
. fourier s law is the consequence of non - negative entropy production in classical irreversible thermodynamics @xcite , and the mcv equation is the result of the deviation from local equilibrium , characterized by an internal variable @xcite .
the heat flow is the well - established candidate of that internal variable to obtain compatibility with kinetic theory @xcite .
the origin of weakly nonlocal extensions is a question in a phenomenological approach , too .
there are indications that the guyer - krumhansl equation is connected to nonclassical entropy current density @xcite .
the lagging heat equation ( [ jef ] ) was suggested according to an analogy to rheological models and its relation to thermodynamics is not clear @xcite .
one may obtain it by diverse simple mechanisms of temperature equilibration , like heat flow splitting @xcite , or two - step relaxation @xcite .
the jeffreys type heat conduction is free from several problematic aspects of the mcv equation @xcite . at last , the theory of green and naghdi is peculiar .
they have introduced a special scalar internal variable , the time derivative of which is the temperature , and have analysed the thermodynamic consequences of the deviations from local equilibrium in a nonstandard way @xcite .
they have obtained a model of heat conduction that can not be reduced easily to fourier s law , and which may result in zero entropy production : heat conduction without dissipation .
we can see from the previous concise and not complete survey that the derivation and the motivation of the weakly nonlocal generalizations of fourier s theory are manifold .
most of them are considered valid models of physical phenomena if their microscopic derivation provides a clear mechanism of the modifications . in some important cases like the jeffreys type or lagging heat equation
the role of the second law is not clarified .
all kinds of thermodynamically consistent justifications introduce a deviation from the local equilibrium entropy density . in case of nonlocal extensions , a deviation from the classical form
the entropy current density is also considered . in this paper , we derive a unified model of the previous constitutive relations with the help of irreversible thermodynamics .
we need only two simple and general assumptions : * there is a deviation from the equilibrium state , conveniently and generally characterized by a vectorial _ internal variable _
@xcite , and * there is a deviation from the classical form of the entropy current , conveniently and generally characterized by an arbitrary function that we will call _ current multiplier _
@xcite . in a linear approximation
, both the internal variable and the current multiplier can be completely eliminated and one obtains a general constitutive equation for the heat conduction .
we will see that the restrictions from the second law of thermodynamics are nontrivial and reduce the number of independent coefficients . in this paper
, we apply the simple heuristic constitutive theory of irreversible thermodynamics in order to consider the restrictions of the second law .
a detailed second law analysis of rigid heat conductors with the help of liu procedure in @xcite shows that current multipliers require second order weak nonlocality .
there is an important advantage of a phenomenological thermodynamic approach based only on general assumptions .
as long as the general conditions regarding the deviation from the local equilibrium are fulfilled by any microscopic or mesoscopic assumption regarding the mechanism and the structure of the deviation , the consequences will be the same : this is the universality of nonequilibrium thermodynamics . to demonstrate this property , we show , that if the deviation from fourier s law is characterised by a gradient of a scalar field , then the model equations reduce to the parabolic two - step model , and that , if the deviation from fourier s law is characterised by a general vector field , then the vector field should fulfil an mcv equation in order to satisfy the thermodynamic constraints .
for modeling phenomena beyond local equilibrium , we introduce a single vectorial internal variable denoted by @xmath11 , and through this paper we consider rigid isotropic materials .
the second law is given in the following form : @xmath12 where @xmath13 is the specific entropy , @xmath14 is the conductive current density of entropy and @xmath15 is the entropy production .
then we introduce two basic constitutive hypotheses , which will be crucial in the following : 1 .
we assume that nonequilibrium entropy depends on the internal variable @xmath16 quadratically : @xmath17 + here , @xmath18 is a scalar material coefficient , sometimes called thermodynamic inductivity @xcite .
the quadratic dependence , with @xmath19 can be justified by the morse lemma and the requirement of entropy maximum at the nonequilibrium part of the basic state space , spanned by @xmath16 @xcite .
@xmath20 because of the concavity of entropy .
if @xmath21 is constant then we obtain the following partial derivatives : @xmath22 where @xmath23 is the equilibrium temperature .
the thermodynamic relations are conveniently expressed by the following gibbs relation : @xmath24 + in what follows , we make the assumption @xmath25 , which is convenient in the following calculations , although it restricts the applicability of the theory .
then , the temperature is the function of the internal energy only and does not depend on the internal variable , as one can see from the first equation of ( [ intensives ] ) .
2 . our second assumption requires the generalization of the classical conductive current density of the entropy .
we demand that , without energy current , there is no entropy current in pure heat conduction .
therefore , in the following we assume that , instead of the classical @xmath26 form , the current density of the entropy is @xmath27[nyiriec ] where @xmath28 is a _ constitutive function _ that will be constrained by the second law .
this assumption was first introduced by nyri @xcite , and applied for heat conduction in @xcite , where the generalized multiplier of the current density of the internal energy was called current intensity factor . the recent terminology ( current multiplier )
was suggested in @xcite .
the above two conditions lead to the following form of the entropy production : @xmath29 this formula is a quadratic expression of thermodynamic fluxes and forces , where the following characterisation is applied : [ cols="^,^,^,^",options="header " , ] + .21 cm table 1 .
thermodynamic fluxes and forces the heuristic method of classical irreversible thermodynamics introduces a functional relationship between these quantities , assuming that the fluxes depend on the forces .
then in simple cases and for twice differentiable functions the mean value theorem of lagrange ensures the validity of linear relations @xcite .
however , in our case the bilinear expression above does not permit such a simple characterization , the natural choice of the constitutive state space spanned by the internal energy @xmath2 and the vectorial internal variable @xmath16 and their gradients is seemingly not without problems .
however , in @xcite it was shown that , in case of a second order weakly nonlocal state space , the above choice of the entropy flux in the form of ( [ nyiriec ] ) does not lead to a contradiction and , considering also that composite functions can appear in constitutive relations ( see e.g. @xcite ) , the consequent simple heuristic exploitation can be corroborated .
now we introduce such linear relationships that can be easily transformed into a convenient form . in isotropic
continua the entropy production is an isotropic function of the forces , therefore , the most general linear relationship @xcite between them introduces seven material parameters : @xmath30 here , @xmath31 , @xmath32 , @xmath33 , @xmath34 , @xmath35 , @xmath36 , @xmath37 are the isotropic scalar conductivity coefficients , and @xmath38 is the kronecker symbol . the non - negative entropy production requires the following inequalities for the material parameters : @xmath39 we do not assume onsager or casimir type reciprocal relations for the vectorial thermodynamic interactions characterized by the last two terms of the entropy production .
our internal variable may be a function of several of microscopic variables of different nature regarding time reversibility , some @xmath40 type and others @xmath41 .
this general approach was shown to be fruitful in formulating the thermodynamic framework of generalized continuum mechanics @xcite .
we can eliminate the current multiplier from ( [ o1 ] ) and ( [ o2 ] ) with the help of ( [ o3 ] ) .
moreover , after some simple steps , we can eliminate the internal variable from eq.([o1 ] ) and eq.([o2 ] ) , too .
if @xmath42 and @xmath43 are constants then we obtain the following constitutive relationship of the derivatives of temperature and of energy current density : @xmath44 here , we have denoted the partial time derivative by @xmath45 and introduced the shorthands @xmath46 we can see that ( [ geneq ] ) contains only five independent material parameter , as @xmath47 and @xmath48 , @xmath32 and @xmath33 can not be distinguished .
every coefficient in ( [ ujpar ] ) is positive according to ( [ poz ] ) .
the heat conduction coefficient deserves a special attention because @xmath49 where @xmath50 and @xmath51 are the symmetric and anti symmetric parts of the matrix in ( [ o1])-([o2 ] ) . as a consequence if @xmath52 , then @xmath53 follows .
we derive the following important special cases : 1 .
_ fourier .
_ if @xmath54 and @xmath55 then directly from ( [ o1])-([o3 ] ) we obtain fourier s law , in the following form : @xmath56 where @xmath57 is the fourier heat conduction coefficient .
this is not apparent from ( [ geneq ] ) because ( [ ujpar ] ) and ( [ poz ] ) exclude the straightforward choice @xmath58 , @xmath59 , @xmath60 , @xmath61 .
maxwell - cattaneo - vernotte_. it is frequently mentioned @xcite that extended irreversible thermodynamics , and the mcv equation , arises as a special choice of a vectorial internal variable as the conductive current density of the internal energy @xmath62 @xcite .
however , in our framework , the choice @xmath63 does not automatically lead to the mcv equation .
in fact , ( [ geneq ] ) shows that the choice of @xmath59 , @xmath60 , @xmath61 , leads to equation thence @xmath64=0 , and therefore , the internal variable is proportional to the heat flow according to ( [ o1 ] ) .
@xmath65 because of the last inequality of ( [ poz ] ) .
since in ( [ o1 ] ) the first term of the right hand side is responsible for the classical fourier type heat conduction , we may say he mcv equation is obtained if heat conduction is dominated by a casimir type antisymmetric cross effect . 3 .
_ jeffreys type .
_ if @xmath60 , @xmath61 , we obtain the thermodynamic version of the jeffreys type equation in the following form : @xmath66 if @xmath67 then @xmath68 follows , and the mcv equation is completed to a jeffreys type equation .
the emerging nonlinearity can not be circumvented by assuming temperature dependent coefficients .
guyer - krumhansl .
_ if @xmath59 , @xmath61 , and @xmath69 , then the gk equation ( [ guykru ] ) is obtained . as @xmath70 is the simplest condition for @xmath59 , we see that the gk equation requires a casimir type coupling of the terms in ( [ o1])([o2 ] ) .
_ general green - naghdi type . _
a gn type equation can be obtained if @xmath71 . in this case , the casimir type reciprocity @xmath72 is the consequence of the last inequality of ( [ poz ] ) , and we obtain @xmath73 here , we have introduced similar notation of the coefficients as in ( [ geneq ] ) , but they are combinations of the thermodynamic parameters different from that was given in ( [ ujpar ] ) : @xmath74 6 . _
green - naghdi type .
_ a pure gn type equation requires @xmath71 , @xmath70 , and the casimir type reciprocity @xmath72 .
then we obtain @xmath75 the gn type equation may be nondissipative with zero entropy production when the entropy current density is classical , i.e. @xmath76 .
an observation common to all the above cases is that heat conduction coefficient @xmath77 is always nonnegative .
in order to obtain an impression of the specific mechanisms that may lead to weakly nonlocal deviation from the classical entropy current , we consider here a simple form of the gk equation , where , in addition to the conditions @xmath59 , @xmath61 , we introduce @xmath78 . in this case
, ( [ geneq ] ) simplifies to @xmath79 with @xmath80 , @xmath81 and @xmath82 . as a first example , let us suppose that the deviation from fourier s heat conduction is characterized by a vector field @xmath83 : @xmath84 with @xmath85 being a constant parameter .
we may substitute @xmath86 into the simplified gk equation ( [ sgk ] ) and obtain the following condition for the unknown vector field : @xmath87 here we applied ( [ ebal ] ) , together with the equation of state @xmath88 with constant specific heat @xmath89 .
hence , with the choice of @xmath90 , our vector field @xmath83 fulfils an mcv equation with the @xmath91 fourier heat conduction coefficient .
therefore , the total heat flow @xmath3 is divided into two parts : @xmath92 fulfils fourier s law and @xmath83 an mcv equation .
apparently , there are two different channels of heat conduction characterized by a fourier s law and an mcv equation , respectively @xcite .
we can introduce a different characterization of the deviation from fourier s law by a scalar field @xmath93 in the following way : @xmath94 where @xmath95 and @xmath41 are constant coefficients .
substituting @xmath86 to the simplified gk equation ( [ sgk ] ) again , one obtains the following condition:@xmath96 therefore , with the choice of @xmath97 , @xmath98 one can find the condition that our scalar field @xmath99 fulfils a heat exchange equation .
apparently , there are two different temperatures that equilibrate by heat exchange .
this is a well - known mechanism of heat conduction in complex materials where one of the components can be thermally excited independently of the other , e.g. in case of metals with different electron and lattice temperatures @xcite .
therefore we can see that both heat splitting and heat conduction of two - temperature materials are special cases of weakly nonlocal heat conduction and our approach with a general internal variable can be well interpreted in terms of the general weakly nonlocal heat conduction law ( [ geneq ] ) .
the current multiplier @xmath100 is connected to the coefficients of the specific relaxation mechanism .
these two examples show the real advantage of our weakly nonlocal generalization of heat conduction : it is based on minimal and general assumptions regarding the current density of the entropy and therefore the results are _ robust _ and _
universal_. a straightforward microscopic interpretation of the internal variable could be a moment of the phonon distribution function .
that view is suggested by the quadratic dependence of the entropy density of @xmath101 .
however , a detailed kinetic interpretation of current multipliers in the generalized entropy flux is an open question in the light of the above mentioned macroscopic ones .
in this section we solve a special form of the above equations for a one dimensional space variable @xmath102 . assuming a constant specific heat , @xmath88 we can eliminate the heat flux from eq.([geneq ] ) with the help of the energy balance and
obtain @xmath103 where @xmath104 .
we can see that , due to the appearance of the reciprocal temperature , the equation is nonlinear and the jeffreys type and gk equations do not coincide . in this equation , we have introduced constant conduction coefficients based on ( [ o1])([o3 ] ) as it is natural in a thermodynamic approach .
we can observe that , in this case , the fourier heat conduction coefficient is temperature dependent , @xmath105 . in the following ,
we show some characteristic solutions of the differential equation ( [ hceq ] ) in case of jump boundary conditions .
initially we assume that a rod with length @xmath106 has a fixed temperature @xmath107 and also @xmath108 .
the far end of the rod is kept at this initial temperature , therefore , @xmath109 .
the other boundary has the same temperature at the beginning , but after a short time the temperature jumps to a higher value @xmath110 . for the sake of calculation convenience , we approximate the jump with the following smooth function : @xmath111 .
the used parameters are those of the naf crystal experiments of jackson and walker @xcite .
therefore the specific heat is @xmath112 , the density @xmath113 , the fourier heat conduction coefficient @xmath114 , therefore , @xmath115 , the length of the rod is @xmath116 , and the parameters of the initial condition are @xmath117 .
figure [ fourierfig ] shows the solution of the fourier equation .
the linear stationary solution @xmath118 is reached at the end . in the figure [ mcvfig ] a solution of the maxwell - cattaneo equation
is shown whith @xmath119 and the other parameters are the same , @xmath120 and @xmath121 are zero . the relaxation time is extracted from the measured @xmath122 . here one may observe a ridge due to the finite propagation speed , and the damping of the ripples and of the reflected pulse due to the dissipative effect of heat conduction .
figures [ jeffreysfig ] and [ gkfig ] show the solutions of the jeffreys type and the guyer - krumhansl equations , where @xmath7 is the same as in figure [ mcvfig ] , and @xmath123 ( jeffreys case ) and @xmath124 ( guyer - krumhansl case ) respectively . in both cases ,
the relaxation to the stationary solution is faster . in figure
[ jeffreysgkfig ] , one can see a solution where none of the coefficients of equation ( [ hceq ] ) is zero .
there the coefficients are @xmath125 , @xmath126 in this case , one may observe the change of the speed of the signal propagation .
the solution of the nondissipative gn type equation is given in figure [ gnfig ] with the same coefficient values as in case of mcv equation , @xmath127 .
figures [ gnjeffreysfig ] and [ gngkfig ] show that jeffreys or guyer - krumhansl dissipation do not damp the reflected travelling wave ( parameter values : @xmath125 , @xmath128 respectively @xmath124 in figure [ gngkfig ] .
finally figure [ dgnfig ] shows that , for the general gn equation , the damping effects is due to the additional terms are different , from which can be observed for the gk - jeffreys type one . using the @xmath125 , @xmath129 parameter values
, one gets a qualitatively different damping than with the same parameter values for the heat conduction equation ( [ ggneq ] ) . fourier equation , @xmath130.,scaledwidth=70.0% ] mcv equation , @xmath131.,scaledwidth=70.0% ] jeffreys type equation , @xmath131 , @xmath132.,scaledwidth=70.0% ] guyer - krumhansl equation , @xmath133 , @xmath134.,scaledwidth=70.0% ] jeffreys - guyer - krumhansl equation , @xmath135.,scaledwidth=70.0% ] nondissipative green - naghdi type equation @xmath131.,scaledwidth=70.0% ] green - naghdi type equation with jeffreys damping @xmath131 , @xmath136.,scaledwidth=70.0% ] green - naghdi type equation with guyer - krumhansl damping @xmath137 , @xmath134.,scaledwidth=70.0% ] green - naghdi with both kinds of damping , @xmath138 , @xmath139 , longer duration.,scaledwidth=70.0% ]
we have introduced a vectorial internal variable and a current multiplier as a simple and general characterization of the deviation from local equilibrium of the entropy and the entropy current density functions , respectively .
then we have derived the resulting connection between the heat flow and temperature , and obtained a common generalization of several known heat conduction equations .
consequently , the emerging heat conduction equation ( [ geneq ] ) is independent of the micro- or mesoscopic mechanism of heat conduction as long as the general conditions of this approach
namely , the second law and the existence of extra field quantities are fulfilled . in this sense ,
the found heat conduction equation is _
universal_. we have demonstrated this kind of universality by the derivation of two common simple heat conduction mechanisms in our framework .
the analysis of the different specific examples in the light of the thermodynamic constraints , stemming from the non - negative entropy production , revealed that : * if the main fourier coefficient @xmath140 is not zero then mcv is always extended by the characteristic nonlocal term of the jeffreys type equation . *
pure mcv , gk and gn type equations are related to a casimir type cross effect between the internal variable and thermal parts of the entropy production . *
the nondissipative wave equation of the gn type model was derived and interpreted in the framework of non - equilibrium thermodynamics . * the strictly linear thermodynamic constitutive equations ( [ o1])([o3 ] ) with constant coefficients
lead to a nonlinear evolution equation of the heat flow , where the nonlinearity can not be absorbed in special temperature dependent coefficients . *
the general heat conduction equation ( [ geneq ] ) is remarkably stable in numerical calculations , whenever the inequalities ( [ poz ] ) are fulfilled .
( for the link between stability and the second law of thermodynamic , see e.g. @xcite ) * with the natural choice of the internal variable as the non - fourier part of the heat flow , a physical interpretation of our internal variable was given .
we found that two - temperature relaxation and heat flow splitting models can be interpreted in terms of the current multiplier .
it is worth mentioning that ( [ nyiriec ] ) trivially corresponds to the usual generalisation of the entropy current density , when an additive constitutive term , the so - called mller s @xmath141 vector is introduced ( see e.g. @xcite , p441 ) . another comparison that one can perform is related to the generic approach in @xcite .
equations ( 42)-(49 ) in @xcite are in principle more general than ours from several points of view . on the other hand ,
the structure and derivation of the equations is rather different , therefore , the comparison is not straightforward .
for example , the suggested evolution equations are balance like and the particular form of the entropy flux ( first part of eq .
( 46 ) in @xcite ) looks structurally different from our simple generalization , preventing a direct comparison .
the authors are grateful to arkadi berezovski , jri engelbrecht , gyula grf , balzs czl and joe verhs for the discussions .
the work was supported by the grant otka k81161 and tt 10 - 1 - 2011 - 0061/za-15 - 2009 . | a linear irreversible thermodynamic framework of heat conduction in rigid conductors is introduced . the deviation from local equilibrium
is characterized by a single internal variable and a current multiplier .
a general constitutive evolution equation of the current density of the internal energy is derived by introducing linear relationship between the thermodynamic forces and fluxes .
the well - known fourier , maxwell - cattaneo - vernotte , guyer - krumhansl , jeffreys type and green - naghdi type equations of heat conduction are obtained as special cases . the universal character of the approach is demonstrated by two examples . solutions illustrating the properties of the equation with jump boundary conditions are given . |
boundaries and interfaces are omnipresent in the colloidal world @xcite .
geometric confinement introduces anisotropy in the diffusive motion of sub - micron particles , and the presence of neighbouring walls leads to a general slow - down of brownian motion due to hydrodynamic interactions of the diffusing particle with boundaries @xcite .
the central quantity in this context is the near - wall hydrodynamic mobility tensor @xmath0 which is related to the diffusion tensor @xmath1 by the fluctuation - dissipation theorem @xmath2 recent years have brought significant advancement in experimental techniques which allow to explore near - wall dynamics in more detail , including optical microscopy @xcite and scattering techniques , such as evanescent wave dynamic light scattering @xcite .
the latter is now a well - established tool which has profitably been used to investigate translational @xcite and rotational diffusion @xcite of spherical colloids in dilute suspensions .
due to the complex nature of the experiments , available experimental data for non - spherical particles such as dumbbells @xcite or rods are still lacking proper interpretation .
it is therefore particularly important in this context to understand the nature of hydrodynamic interactions of an axially symmetric particle with a wall , and has partially motivated this work .
axisymmetric particles moving close to a boundary experience an additional anisotropic drag force on top of their own friction anisotropy stemming from their non - spherical shape .
this coupling leads to a complicated behaviour , observed e.g. in simulations of such particles sedimenting next to a vertical wall @xcite , with the mobility of the particle depending on its position and orientation .
available predictions for the near - wall mobility of an axisymmetric particle mostly feature a slender - body approach , yielding quite complex results for general wall - particle orientations near a wall @xcite or a fluid - fluid interface @xcite , analysed in detail in the context of sedimentation in several special alignments @xcite . on the other hand , previous numerical works involve the boundary integral method @xcite , finite element method @xcite or stochastic rotation dynamics @xcite from which empirical relations are extracted .
the lack of theoretical predictions for the near - wall mobility of a rod - shaped and non - slender particle in an arbitrary configuration requires the use of more precise numerical methods .
a possible way is to use advanced algorithms involving bead - models which take into account lubrication when the particles come close to the interface@xcite , which are rather costly . in order to fill this gap , in this work
we derive a general form of the dominant correction to the bulk friction tensor due to the presence of a nearby no - slip wall from which the mobility tensor is calculated .
this allows to verify and correct earlier predictions in terms of distance and orientation of the particle .
importantly , the correction is valid for all axisymmetric particles , not just slender ones , provided that their bulk hydrodynamic properties are known .
our analysis leads to a convenient representation of the mobility tensor in situations when the particle is moderately far from the wall .
for the characteristic length of the body @xmath3 , the relative correction scales as @xmath4 , where @xmath5 is the wall - particle distance , and the exponent @xmath6 depends on the component of the friction matrix ( translational , rotational , or coupling terms ) .
we provide explicit analytical expressions for the dominant correction to bulk translational and rotational parts of the friction tensor which are the main result of the paper .
we use them to calculate the corrections to the friction tensor of an axially symmetric particle explicitly in terms of @xmath5 and the particle s inclination angle @xmath7 . by inverting the friction tensor
, we then calculate the near - wall mobility tensor .
the paper is organised as follows .
first , we introduce the notion of friction and mobility tensors for a colloidal particle in sec .
[ sec1 ] . in sec .
[ sec : correction ] , we sketch the idea behind the derivation of the correction , which is then given explicitly in sec . [ sec : axisymm ] for axially symmetric particles .
the theoretical predictions are compared to numerical simulations for an exemplary case in sec .
[ simulation ] , followed by conclusions in sec .
[ conclusions ] .
appendix [ multipole ] contains the details of the multipole method and a description of the simulation method .
details of the derivation of the correction are given in appendix [ derivation ] .
we consider a single colloidal particle immersed in an incompressible newtonian solvent of shear viscosity @xmath8 .
the configuration of the system is described by the position of the centre of the particle @xmath9 and its orientation which , for an axially symmetric particle , is specified by the unit vector @xmath10 pointing along the particle s symmetry axis . on the colloidal length scales and for time scales typical e.g. for scattering experiments , inertia of the fluid and the particle can be neglected .
the flow field @xmath11 around the particle is then described by the stationary stokes equations@xcite @xmath12 where @xmath13 is the force density the particle exerts on the fluid when subjected to flow , and @xmath14 stands for modified pressure field which includes the effect of gravity .
the flow disturbances caused by the presence of the particle in confined geometry are affecting the motion of the particle itself .
these dynamic , solvent - mediated hydrodynamic interactions ( hi ) are long - ranged and have a pronounced effect on the dynamics of colloidal systems .
this flow field may be superposed with an ambient linear flow @xmath15 satisfying the homogeneous stokes equations , with the vorticity and rate of strain defined at a point @xmath16 as @xmath17 with the bar denoting the symmetric and traceless part .
given the force density , one can calculate the force , torque , and symmetric dipole moment ( stresslet ) exerted by the fluid on the particle according to @xmath18 where the integrals are performed over the particle surface @xmath19 .
higher - order moments are defined in an analogous way . in result of the external flow , motion
is induced , and the particle gains linear and angular velocities , @xmath20 and @xmath21 , respectively .
owing to linearity of the stokes equations , the force moments @xmath22 , @xmath23 and @xmath24 , are linearly related to the velocity moments via the generalised friction ( or resistance ) tensor@xcite @xmath25 above we have decomposed the generalised friction tensor into 9 sub - matrices .
the indices @xmath26 and @xmath27 denote the translational and rotational parts , respectively .
the tensors @xmath28 and @xmath29 describe the translation - rotation coupling , and the tensors with superscript @xmath30 describe the response of the particle to an external elongational flow . in most cases , it is sufficient to consider only the @xmath31 friction matrix @xmath32 relating the force and torque to linear and angular velocities . here , we extend the friction matrix to the symmetric dipole moment subspace , since these elements turn out to be essential for the calculation of the correction to the friction matrix in the presence of a wall . in a complementary problem ,
if the forces and torques are known , the particle motion may be resolved by determining the mobility tensor @xmath33 which is related to the friction tensor by inversion @xmath34 using the lorentz reciprocal theorem @xcite , one may prove the symmetry properties of the mobility tensors . in a bulk system , the mobility tensor and the friction tensor , denoted by @xmath35 and @xmath36 , respectively ,
do not depend on the position of the particle due to translational invariance .
the situation is different if a confining boundary is present , since symmetry is broken and the hydrodynamic tensors depend both on the distance to the boundary , and on the relative orientation of the particle with respect to the surface .
the friction tensors of a near - wall particle , @xmath37 , may be written as @xmath38 in the course of this work , we derive analytic formulae for the first - order approximation to @xmath39 , with the expansion parameter being @xmath40 , the ratio of the characteristic size of the particle , @xmath3 , to the wall - particle distance @xmath5 . by inverting @xmath37 from eq . , we arrive at a convenient approximation to the near - wall mobility @xmath41 .
in order to determine the flow around a particle in a half - space bounded by an infinite , planar wall at @xmath42 , one has to solve the stokes equations ( [ stokes ] ) with the no - slip boundary condition @xmath43 at the surface . due to linearity , eq .
can be transformed into the integral form @xmath44 for an unbounded fluid , the green s function @xmath45 is the oseen tensor @xcite @xmath46 , with @xmath47 and @xmath48 . in the presence of boundaries ,
the full green s tensor contains and additional part describing the flow reflected from interfaces . for a hard no - slip wall
, the green s tensor has been first found by lorentz@xcite in 1907 as @xmath49 , where the wall contribution reads @xmath50 for a point force at a distance @xmath51 from a wall at @xmath42 . for a free surface ,
the wall - interaction part only contains the first term of the rhs of . here , @xmath52_{\alpha\beta } = { \frac{\partial } { \partial r_\beta } } { a}_\alpha $ ] and @xmath53 denotes the reflection operator which transforms any point into its mirror image with respect to the wall . the asterisk denotes the mirror image , i.e. @xmath54 .
this expression has been interpreted by blake@xcite in terms of the method of images for stokes flows .
the image system in this case involves three fundamental singularities : the reflection of the original stokeslet and the so - called stokeslet doublet and source doublet .
however , since their amplitude is proportional to the wall - particle distance @xmath51 , asymptotically they die out with distance @xmath55 as @xmath56 .
the tensor @xmath57 is often referred to as blake s tensor .
it can also be recast in different forms@xcite .
the idea of the derivation relies on the expansion of the blake s tensor about the line connecting the centre of the particle and its hydrodynamic image .
thus the interaction between any two points of the particle may be represented by the vertical component of the blake s tensor . by considering the action of higher force moments , higher - order flows incident on the particle can be found . performing a multipole expansion of the resulting flow field@xcite
, we project it onto the force multipole space and eventually find the explicit expressions for elements of the resistance matrix @xmath37 which involve the elements of the bulk friction tensor @xmath36 and the higher multipole elements of the blake s tensor .
the details of the derivation are presented in appendices [ multipole ] and [ derivation ] .
below we present specific results for an axially symmetric particle .
it follows from symmetry properties that the bulk friction matrix of a general axisymmetric particle has a particular structure@xcite . moreover , if the particle has both axial and inversional symmetry @xmath58 ( i.e. it is rod - like ) , its bulk friction matrix in eq .
simplifies , since @xmath59 , @xmath60 , @xmath61 and @xmath62 vanish , i.e. translational motion is not coupled to torque or elongational flow in the bulk case .
the correction has the following form @xmath63 the form of the tensors @xmath64 , @xmath65 , @xmath66 is derived from the multipole expansion of the vertical ( axial ) component of the blake s tensor in appendix [ derivation ] . as we have mentioned in sec .
[ sec1 ] , they depend on the @xmath67-components of the bulk generalised friction tensor of the particle and the particle s orientation via the inclination angle @xmath7 . explicitly , we write them in the following for an axisymmetric particle . at contact ,
the elements of the friction matrix diverge which is expected physically , although no lubrication effects are included in this scheme .
the mobility matrix is then obtained by inversion @xmath68 , and by that we assure that the particle mobility decreases to a non - negative value at contact , but in a way different than with lubrication effects included .
we note that different elements of the friction matrix behave differently with distance .
the hydrodynamic effect of the wall is most pronounced for translational motion where @xmath69 .
rotational motion is least affected with @xmath70 .
this is in contrast with earlier numerical results due to padding and briels@xcite , whose empirical finding was that both corrections scale as @xmath71 , with a simple fitted angular dependence of the mobility components on @xmath7 .
our findings provide exact expressions for the angular dependence , given as low - order polynomials in @xmath72 and @xmath73 , correcting the previous empirical formulae .
the source of the discrepancy lies in the inaccuracies of the numerical methods used in ref . .
they are also in complete agreement with the slender body expressions obtained by de mestre _
et al._@xcite for the cases @xmath74 provided that the slender body results for @xmath75 and @xmath76 are used as input . in order to discuss in detail the dependence of the correction terms on the orientation of the particle , we introduce two coordinate systems exploiting the symmetries of the problem , as sketched in fig . [ rodimage ] .
the laboratory coordinate system ( lab ) consists of three basis vectors @xmath77 , with the @xmath78-axis normal to the wall and the normal vector @xmath79 .
the particle resides in the @xmath80-plane . the rod - wall ( rw ) system is a body - fixed set of basis vectors @xmath81 , where @xmath10 is the unit vector along the long axis of the particle , @xmath82 is parallel to the wall and perpendicular to the particle axis , and @xmath83 completes the orthonormal basis .
the basis vectors are then given by @xmath84 and @xmath85 .
we note that @xmath86 . for an axially symmetric particle , it is convenient to use the representation of the mobility matrix in the rw frame , in which the bulk tensors @xmath87 and @xmath88 are diagonal . the structure of the near - wall tensors is identical to that given in ref . .
in the body - fixed frame of reference rw , the correction tensors in eqs - may be explicitly written in terms of the inclination angle @xmath7 . in the formulae
, we find the elements of the bulk friction tensor of the particle , namely the coefficients of translational and rotational friction in the directions parallel and perpendicular to the body axis , given respectively by @xmath89 and @xmath90 , where @xmath91 and : denotes double contraction .
in addition , the correction terms for rotational motion and rotation - translation coupling contain the coefficient @xmath92 . as seen from eq .
, it quantifies the stresslet exerted on the particle rotating with a prescribed angular velocity .
these coefficients can be taken from bulk results for friction of axisymmetric particles .
for ellipsoidal particles , analytical expressions are available@xcite , whereas for more complex shapes the coefficient may be determined using bead-@xcite or shell - models@xcite .
now we can write the tensors in eqs explicitly in the body - fixed frame rw .
for the translational part , we find the correction s angular dependence as @xmath93 with @xmath94 and @xmath95 \sin \theta\cos\theta.\end{aligned}\ ] ] we note here that since the axis perpendicular to the rod and parallel to the wall ( in the direction of @xmath96 ) is invariant with respect to the lab to rw frame transformation , the middle element of the matrix above is angle - independent .
the translation - rotation coupling part reads @xmath97 finally , the rotational tensor @xmath66 in eq . is a sum of three contributions and has the form @xmath98 for completeness of the discussion , it is worth noting that it is possible to find analytically the leading order behaviour of the mobility functions by expanding the inverted friction matrix - . in this way
, the dominant terms of the correction may be evaluated as functions of @xmath7 and indicate a very simple angular relations for the components , namely low - order polynomials in @xmath72 and @xmath73 .
however , as we noted before , this is not an optimal strategy , since the mobility functions obtained by inversion may become negative when the particle approaches the wall ( @xmath5 is small compared to @xmath3 ) . therefore , it proves better to first calculate the wall - corrected friction tensor , and then invert it to obtain the mobility tensor .
in order to assess the applicability range of the correction , we compare our theoretical result to precise numerical simulations using the hydromultipole package@xcite . as a representative test example , we consider a rod - like particle of aspect ratio @xmath99 constructed out of @xmath100 spherical beads glued together along a straight line . for the needs of demonstration , we choose one inclination angle , @xmath101 , implying the minimal contact distance @xmath102 .
we plot the components of the mobility matrix calculated using the procedure outlined above , and compare them to the corresponding accurate numerical predictions . to this end
, we introduce the following notation in the body - fixed re frame . taking into account the invariant properties of the rod - wall system and the lorentz symmetry ,
we can write the translational part as @xmath103 where @xmath104 . the matrix @xmath105 has a similar structure with elements @xmath106 , @xmath107 , @xmath108 , and @xmath109 , respectively . in both cases , the elements @xmath110 , @xmath111 and @xmath30 are even functions of @xmath112 and the elements @xmath113 are odd ones . for the @xmath114 part , we have @xmath115 the elements @xmath116 and @xmath117 are even functions of @xmath112 , while @xmath118 and @xmath119 are odd ones . by taking the transposition of the above matrix we get the @xmath120 part .
the diagonal components of the translational and rotational diffusion tensor of the rod are plotted in fig .
[ comparison ] .
they reveal that for translational motion in this particular case the correction accurately represents the actual mobility even up to @xmath121 , and the asymptotic inverse - distance behaviour of the correction is evident . similarly , the rotational components are even less sensitive to the effect of the wall due to the rapid decay of the hi for rotational motion . for larger distances
, the mobility matrix obeys the necessary symmetries , with @xmath122 and @xmath123 asymptotically .
the presence of the wall introduces also non - diagonal component to the mobility tensors , which we depict in fig .
[ comparison_nondiag ] .
normalised by the appropriate combinations of bulk mobility coefficients , these elements are rather small
. nevertheless , they numerical results are again in agreement with theoretical predictions up to quite close wall - particle distances , both for translations , and rotations .
the translation - rotation coupling tensors become more significant as the particle approaches the wall , as seen from fig .
[ comparison_tr ] . compared to the characteristic bulk quantities ,
however , they seem to play a marginal role in this case . with the derived correction , we are able to reproduce them accurately again up to @xmath124 .
the comparison of the correction to numerical results in the case of a relatively long ( @xmath125 ) rod - like particle is quite favourable .
indeed , we expect the correction to work even better for more slender particles , since it can be shown analytically using the slender body results for the bulk friction that the relative correction to the translational mobility tensor ( e.g. @xmath126 etc . ) decreases slowly with increasing aspect ratio as @xmath127 .
comparison of the near - wall mobility of a rod of apect ratio @xmath125 at an angle @xmath128 to the wall , as predicted by the correction ( solid lines ) and precise hydromultipole numerical simulations ( data points ) .
the coefficients are normalised by their corresponding bulk values , to that the all tend to unity at @xmath129 .
top : diagonal elements of the translational mobility matrix in the rw frame .
bottom : rotational diagonal elements . , title="fig:",scaledwidth=50.0% ] comparison of the near - wall mobility of a rod of apect ratio @xmath125 at an angle @xmath128 to the wall , as predicted by the correction ( solid lines ) and precise hydromultipole numerical simulations ( data points ) .
the coefficients are normalised by their corresponding bulk values , to that the all tend to unity at @xmath129 .
top : diagonal elements of the translational mobility matrix in the rw frame .
bottom : rotational diagonal elements .
, title="fig:",scaledwidth=50.0% ] non - diagonal components of the near - wall translational and rotational mobility tensors for the inclination angle @xmath128 .
the data points are predictions of multipole simulations with lubrication included , while the solid lines are predicted by our analytical formulae for the correction .
the coefficients are normalised by bulk average values of the diagonal terms , e.g. @xmath130 .
, title="fig:",scaledwidth=50.0% ] non - diagonal components of the near - wall translational and rotational mobility tensors for the inclination angle @xmath128 .
the data points are predictions of multipole simulations with lubrication included , while the solid lines are predicted by our analytical formulae for the correction .
the coefficients are normalised by bulk average values of the diagonal terms , e.g. @xmath130 . ,
title="fig:",scaledwidth=50.0% ] components of the translation - rotation coupling mobility tensor @xmath131 for the inclination angle @xmath128 .
the data points are predictions of multipole simulations with lubrication included , while the solid lines are predicted by our analytical formulae for the correction .
the coefficients are normalised by a combination of bulk coefficients @xmath132 .
the deviations are most pronounced in the coupling tensor .
the overall values are , however , rather small , not exceeding 5% for @xmath133.,scaledwidth=50.0% ]
we have presented a simple analytical scheme which allows for the representation of the near - wall friction and mobility tensors of a rod - like colloid close to a planar no - slip wall .
the correction to bulk mobility , expressed in terms of the bulk hydrodynamic properties of the particle , is valid for general axially symmetric colloids , which need not be slender .
our results show that the distance dependence varies between the types of motion in focus ( translational , rotational , and @xmath114-coupling ) .
moreover , we have demonstrated by analytical formulae that near - wall friction and mobility for particles at moderate distances from the wall indeed depends on their orientation via simple polynomials in sine and cosines of the inclination angle @xmath7 , as conjectured by padding _
et al._@xcite . by
that we have also verified earlier theoretical developments and recent numerical predictions@xcite .
our results are in agreement with numerical calculations even in the case when @xmath134 , rendering the results practical for large and moderate wall - particle distances .
ml acknowledges support from the national center of science grant no .
2012/07/n / st3/03120 .
part of the research has been conducted under a david crighton fellowship awarded to ml at the university of cambridge , and within the mobility plus fellowship awarded to ml by the polish ministry of science and higher education .
the idea of the multipole method relies on expressing the force densities and velocities on the surfaces of many spheres immersed in the fluid in the form of a boundary integral equation , which is then projected onto a complete set of multipolar solutions of the stokes equations
. the resulting system of linear equations may then be truncated and solved numerically for a conglomerate of spheres moving together . by projecting the many - particle friction matrix obtained in this way onto the subspace of rigid body motions of the conglomerate
, the friction tensor of a complex - shaped particle is found .
the method has been greatly developed over the last decades , and is presented in more details , e.g. in refs . .
with the use of the concept of induced forces due to bedeaux and mazur @xcite , the validity of the stokes equations ( [ stokes ] ) may be formally extended inside the particles by taking an appropriate surface distribution of the forces @xmath135 on the surfaces of the particles @xmath136 . for the stick
boundary conditions , the velocities on the surfaces read @xmath137 and eq . on the surfaces of the particles
takes the form @xmath138 where @xmath15 represents an ambient flow in the absence of the spheres .
we now separate the second term on the rhs of into the contribution from distinct particles and the self - contribution .
the self part is found by considering a single particle @xmath139 in an ambient flow @xmath140 .
the force density @xmath141 it exerts on the fluid is linearly related to the relative velocity at the surface , viz . @xmath142 where the integral operator @xmath143 is called the single - particle resistance operator , or the friction kernel @xcite , and depends solely on the internal composition and surface properties of the particle @xcite . for the distinct part ( @xmath144 )
, we introduce the green s integral operator ( propagator ) @xmath145 : @xmath146({\mathbf{r } } ) & \equiv { \int_{}^ { } \!\ ! \mathrm{d } { { \mathbf{r}}'}\ , } { \mathbf{t}}({\mathbf{r}},{\mathbf{r}}')\cdot { \mathbf{f}}_j({\mathbf{r } } ' ) , \qquad { \mathbf{r}}\in\sigma_i,\end{aligned}\ ] ] which allows eq . to be written as @xmath147
the above equations can be transformed into an infinite set of algebraic equations by expanding the velocity field and induced force densities in a basic set of irreducible multipoles developed by felderhof and co - workers @xcite .
for the velocity , the irreducible multipoles are linear combinations of lamb s solution of the homogeneous stokes equations@xcite .
they are labelled by three numbers : @xmath148 , @xmath149 and @xmath150 .
the force multipoles can be likewise be described by the labels @xmath151 .
the details of the expansion , along with explicit form are given in refs . .
we include the expansion coefficients of the velocities @xmath152 in an infinite - dimensional vector @xmath153 which encompasses all the velocity multipoles for all the particles . in a similar manner
, we arrange the force multipole moments in the vector @xmath154 . after the multipole expansion
is performed , the integral operators @xmath155 and @xmath156 become matrices and eq . is transformed into an algebraic equation @xmath157 the multipole matrix elements of @xmath155 for different particle models are given in ref . , while the elements of @xmath156 have been calculated in ref . for the case of an unbounded fluid , a fluid bounded by a free surface and a fluid bounded by a hard wall . in the friction problem ,
the forces acting on the particles are sought , given their velocities . upon inverting the above relation
, the grand resistance matrix @xmath158 is found as @xmath159 the friction matrix defined in eq . can be found by projecting @xmath158 on the subspaces @xmath67 , corresponding to @xmath151 equal to @xmath160 , @xmath161 , and @xmath162 , respectively .
for example , the force multipole @xmath163 with @xmath164 and @xmath165 has three spherical components @xmath166 corresponding to three components of the total force . in numerical computations , infinite matrices @xmath155 and @xmath156 in eq .
( [ multipoleeqset ] ) are truncated at the multipole order @xmath167 , so that only the elements with @xmath168 are considered @xcite .
after such a truncation , the matrix @xmath169 is inverted , and the force multipoles are determined . to improve numerical convergence of this scheme , the obtained grand friction matrix @xmath158 in eq .
( [ 3_frictionproblem ] ) is additionally corrected for lubrication effects @xcite .
the matrix @xmath158 constructed in the multipole method is not pairwise additive , and accounts fully for many - body hydrodynamic interactions .
the approximation is introduced at the level of truncation of the multipoles , and its error may be controlled .
the procedures outlined above have been implemented in a fortran code hydromultipole @xcite by wajnryb and collaborators .
the method for calculating the near - wall hydrodynamic tensors has been laid out by cichocki _
we employ these codes to calculate the friction tensors of non - spherical particles represented by their bead - models .
once the friction matrix @xmath32 is known , the mobility matrix @xmath0 is found by inversion .
for a wall - bounded fluid , it follows from the form of eq .
that the propagator @xmath156 can be decomposed as @xmath170 where the part @xmath171 consists of the multipole elements of the oseen tensor@xcite , while @xmath172 describes the wall contribution , the multipole matrix elements of which are calculated for a free surface and a hard wall in ref .
( see also ref . ) . in order to find the asymptotic correction to the bulk friction of a particle moving close to a wall
, we employ the scattering expansion @xcite .
we start from rewriting the grand resistance matrix in eq .
( [ 3_frictionproblem ] ) in the following form @xmath173 when the wall - particle distance is considerably larger than the particle itself , so we expect the wall contribution @xmath172 to be a small correction . expanding eq .
yields the form of the correction to the bulk resistance matrix @xmath174 with @xmath175 being short for the bulk resistance matrix of the particle @xmath176 .
further on , we evaluate the dominant terms of the correction for all the elements of the friction matrix in the @xmath177 subspace . the propagator @xmath178 connects the beads building up the particles with the beads of the image particle .
consider two interacting beads @xmath179 and @xmath180 building up the conglomerate and its image , respectively , as illustrated in fig .
[ rodimage ] .
introducing coordinates relative to the centre of each conglomerate , we have @xmath181 and @xmath182 .
hence the distance between the particles may be written as @xmath183 where we have used the fact that @xmath184 .
for the wall - particle distance @xmath5 large compared to the particle size @xmath3 , and thus for @xmath185 , we may expand the distance between each pair around the direction normal to the wall .
then , in leading order , the propagator takes the form @xmath186 .
due to this fact , its multipole elements have the axial symmetry around the normal direction @xmath187 . in this case
, the bulk grand resistance matrix @xmath175 reduces to the single - particle bulk friction matrix @xmath36 as in eq .
( [ friction_single ] ) . the relevant coordinate systems and a schematic illustration of the expansion of the distance between interacting points of the particle ( e.g. a rod ) and its image .
we depict the interaction between the point @xmath179 with the image @xmath180 of point @xmath188 . for large wall - particle distances
, it may be expanded around the vertical line connecting the particles centres lying at a distance @xmath189 apart , so that @xmath190 . ] the dominant correction may thus be looked upon as interaction of a particle of a given bulk friction matrix @xmath36 with an image particle via the propagator @xmath172 , which accounts for the flow reflected by the wall .
the matrix elements of @xmath178 decay according to their multipole indices as @xmath191 with the multipolar indices denoted by @xmath192 .
the indices @xmath193 refer to the superscript @xmath110 , while @xmath194 refer to @xmath111 .
the directional tensors @xmath195 depend only on the direction of the normal vector @xmath187 . to derive the correction for an axisymmetric particle ,
we need the following elements @xmath196 the bar indicates the symmetric and traceless part of the respective tensors with respect to the indices in brackets .
the symbol @xmath197 indicates the symmetric and traceless part in the index pair @xmath198 .
the appropriate reductions read @xmath199 taking into account the symmetries of the bulk friction matrix @xmath36 of an axisymmetric particle ( i.e. the lack of @xmath114 and @xmath200 elements ) , we find explicit expression for the correction terms in eqs . - as @xmath201 where appropriate contractions of the tensors are taken .
the evaluation of these expressions using eqs . and
the general form of @xmath36 ( cf . ref . ) leads to the expressions - . | hydrodynamic interactions with confining boundaries often lead to drastic changes in the diffusive behaviour of microparticles in suspensions . for axially symmetric particles ,
earlier numerical studies have suggested a simple form of the near - wall diffusion matrix which depends on the distance and orientation of the particle with respect to the wall , which is usually calculated numerically . in this work ,
we derive explicit analytical formulae for the dominant correction to the bulk diffusion tensor of an axially symmetric colloidal particle due to the presence of a nearby no - slip wall .
the relative correction scales as powers of inverse wall - particle distance and its angular structure is represented by simple polynomials in sines and cosines of the particle s inclination angle to the wall .
we analyse the correction for translational and rotational motion , as well as the translation - rotation coupling .
our findings provide a simple approximation to the anisotropic diffusion tensor near a wall , which completes and corrects relations known from earlier numerical and theoretical findings .
* published in : * j. chem .
phys * 145 * , 034904 ( 2016 ) . |
the study of nuclear properties at high temperature , spin and isospin has gained much of interest in recent times .
thanks to the recent developments in experimental facilities , these highly - excited nuclei are becoming more accessible and provide theorists with a challenging task . apart from these extremes
, there are still some unexplored regimes of hot nuclei .
the properties of nuclei at very low temperatures and the phase transitions associated with that belong to such area where conclusive experimental results are scarce . at such low temperatures , the shell ( quantal ) and pairing effects are quite active though being modified by thermal effects . among the famous and open questions in this regime
are the existence of pairing phase transition , the order of it if it exists , the role of fluctuations , etc . in hot nuclei ,
thermal fluctuations are expected to be large since the nucleus is a tiny finite system .
thermal shape fluctuations and fluctuations in the pairing field are the dominating fluctuations and they have been so far studied separately within different models @xcite . both of these fluctuations are expected to be present at low temperatures .
however , the interplay between them has not been investigated so far .
the present work addresses this subject and we study the influence of this interplay on the experimental observables , namely the width of giant dipole resonance ( gdr ) .
gdr is a fundamental mode of excitation of nuclei caused by the out of phase oscillations between the proton and neutron fluids under the influence of the electromagnetic field induced by the emitted / absorbed photon . in general , the resonance parameters of any resonating object are related to the geometry of the object . in this way , the gdr width and its cross - section could yield direct information about the shape of the nucleus .
this is only a macroscopic description of gdr and there are microscopic approaches , which couple the gdr to particle - hole , particle - particle and hole - hole excitations @xcite . for hot nuclei , which are not accessible by the discrete @xmath3-ray spectroscopy and other conventional techniques , the measurement of gdr is considered to be a major probe to obtain the details of nuclear structure .
the importance of measuring the gdr width at low temperatures was insisted by one among the authors @xcite and the first results were reported in ref .
@xcite , where it was found that the gdr width in @xmath1sn at @xmath4 1 mev is nearly the same as that in the ground state .
this data point was successfully explained only after treating properly the pairing correlations with the phonon damping model ( pdm ) @xcite .
similarly , it was found in ref .
@xcite that the gdr width in @xmath5au extracted at @xmath4 0.7 mev is almost the same as its ground state value .
however , it was misattributed to the shell effects as in the case of @xmath2pb .
this was clarified later @xcite where the proper inclusion of shell effect was found to act in the opposite direction to raise the width and the pairing fluctuations were speculated to explain this anomaly .
preliminary results in this regard were reported in refs .
@xcite where the importance of considering pairing in the thermal shape fluctuation model has been emphasized .
the recent low temperature gdr measurements done at variable energy cyclotron centre , kolkata @xcite highlight the interesting features of the gdr width in heavy nuclei observed at low temperatures .
the thermal shape fluctuation model ( tsfm ) @xcite , which is often used by experimentalists , describes the increase of the gdr width with temperature by averaging the gdr cross section over all the quadrupole shapes .
however this model is known to largely overestimate the gdr width in open - shell nuclei at low temperatures .
the success of a proper treatment of pairing within the pdm @xcite suggests the necessity of including pairing correlations to cure this shortcoming of the tsfm .
the pdm is a microscopic model , whose mechanism is different from that of the tsfm .
the pairing has never been taken into account within the tsfm so far because of the ( incorrect ) assumption that the pairing gap disappears at @xmath6 1 mev .
given the popularity of the tsfm the inclusion of pairing in the tsfm is quite important .
this is done for the first time in the present work .
apart from the tsfm , two phenomenological parameterizations have been reported @xcite which are very successful in approximating global behaviour of the gdr widths . in recent literature @xcite
these parameterizations referred to as , phenomenological tsfm ( ptsfm ) @xcite and its modification to take into account the quenching of width at low @xmath7 has been referred to as a critical temperature included fluctuation model ( ctfm ) @xcite .
it has to be noted that these empirical formulae constructed to mimic the results of tsfm should not be confused with the tsfm itself .
the ptsfm and ctfm are merely a phenomenological parameterizations based on empirical data , which has neither microscopic nor macroscopic foundation .
it is interesting to note , although both the pdm and ctfm give results consistent with the measurements of the gdr width in @xmath0tc @xcite , these two models are quite different from the tsfm .
it is indeed mentioned in ref .
@xcite that it would be interesting to compare the data with tsfm by including the effect of thermal pairing . here
we employ the thermal fluctuation model built on nilsson - strutinsky calculations with a macroscopic approach to gdr and examine the inclusion of the fluctuations in the pairing field . in refs .
@xcite the same hamiltonian was used to generate the free energy as well as the gdr observables in a consistent way , which led to a slow thermal damping of gdr width when compared to the experimental results . the formalism adopted in the present work
is well tested to reproduced several gdr observations at higher temperatures @xcite .
this formalism is extended to include thermal pairing and our results are discussed in the forthcoming sections .
the present theoretical approach is explained below in threefold within the models for a ) deformation energy calculations , b ) relating the shapes to gdr observables and c ) considering the fluctuations due to thermal effects in finite systems .
here we follow the finite temperature nilsson - strutinsky method @xcite .
the total free energy ( @xmath8 ) at a fixed deformation is calculated using the expression @xmath9the liquid - drop energy ( @xmath10 ) is calculated by summing up the coulomb and surface energies corresponding to a triaxially deformed shape defined by the deformation parameters @xmath11 and @xmath12 . the shell correction ( @xmath13 )
is obtained with exact temperature dependence @xcite using the single - particle energies given by the triaxial nilsson model . while considering the pairing fluctuations ,
the nucleus is assumed to behave as a grand canonical ensemble ( gce ) , which allows fluctuations in particle number @xcite .
the corresponding free energy is determined as @xmath14where @xmath15 is the nuclear hamiltonian which is independent of temperature , @xmath16 is the chemical potential , @xmath17 is the particle number , @xmath7 is the temperature and @xmath18 is the entropy .
the above expression can be expanded to@xmath19+\frac{% \delta ^{2}}{g}\ ] ] where @xmath20 are the single - particle energies obtained by diagonalizing @xmath15 with a harmonic oscillator basis comprising the first 12 major shells , @xmath21 are the quasiparticle energies , @xmath22 is the pairing gap obtained by solving the temperature dependent bcs equations @xcite by assuming a constant pairing strength given by @xcite @xmath23/a^2 $ ] . the smoothed free energy , in the strutinsky way , can be written as @xmath24 with the third term included to give better plateau conditions @xcite .
here @xmath25 is the averaging function given by@xmath26@xmath27 if @xmath28 is even and @xmath29 if @xmath28 is odd ; @xmath30 , @xmath31 is the smearing parameter satisfying the plateau condition @xmath32 ; @xmath33 is the order of smearing and @xmath34 are the hermite polynomials .
the averaged occupation numbers and single - particle entropies are given by @xmath35 and @xmath36 , respectively .
the quasiparticle occupation numbers at finite temperature @xmath7 is given by @xmath37^{-1}\;$ ] , so that the total entropy can be written as@xmath38.\end{aligned}\ ] ] for calculations without pairing ( @xmath39 ) , we consider the canonical ensemble ( ce ) for which the expression for free energy reduces to those given in ref . @xcite .
the nuclear shapes are related to the gdr observables using a model @xcite comprising an anisotropic harmonic oscillator potential with separable dipole - dipole interaction . in this formalism
the gdr hamiltonian can be written as @xmath40 where @xmath41 stands for the anisotropic harmonic oscillator hamiltonian , the parameter @xmath42 characterizes the isovector component of the neutron and proton average fields and @xmath43 denotes the strength of the pairing interaction .
the pairing interaction changes the oscillator frequencies [ @xmath44(@xmath45 ) ] , resulting in the new set of frequencies @xmath46 , where @xmath47 with @xmath48 having the units of mev@xmath49 .
alternatively , the role of pairing is to renormalize the dipole - dipole interaction strength such that , @xmath50 , with @xmath51 having the units of mev@xmath52 . including the dipole - dipole and pairing interactions , the gdr frequencies in the laboratory frame are obtained as @xmath53 @xmath54 with the pairing field
, the gdr hamiltonian has to be redefined , which will affect the oscillator frequencies .
the gdr cross section is constructed as a sum of lorentzians given by @xmath55 where @xmath56 , @xmath57 and @xmath58 are the resonance energy , peak cross - section and full width at half maximum , respectively . here
@xmath59 represents the number of components of the gdr and is determined from the shape of the nucleus @xcite .
@xmath60 is assumed to depend on the centroid energy through the relation @xcite @xmath61 the peak cross section @xmath57 is given by @xcite @xmath62 the parameter @xmath63 , which takes care of the sum rule is fixed at 0.3 for all the nuclei . in most of the cases
we normalize the peak with the experimental data and hence the choice of @xmath64 has a negligible effect on the results .
the other parameters @xmath65 ( or @xmath42 ) and @xmath51 ( or @xmath48 ) vary with nuclei so that the experimental width of the gdr built on the ground state is reproduced .
the choice for @xmath1sn is @xmath66 , @xmath67 mev@xmath52 and for @xmath0tc it is @xmath68 , @xmath69 mev@xmath52 .
when the nucleus is observed at finite temperature , the effective gdr cross - sections carry information on the relative time scales for shape rearrangements @xcite , which lead to shape fluctuations .
the general expression for the expectation value of an observable @xmath70 incorporating such thermal shape fluctuations is given by @xcite@xmath71\exp[-f(t;\beta , \gamma ) /t]\mathcal{o}}{\int
\mathcal{d}[\alpha ] \exp[-f(t;\beta , \gamma ) /t]}\;\]]with @xmath72=\beta ^{4}|\sin
3\gamma |\,d\beta \,d\gamma $ ] . with the inclusion of pairing fluctuations
, we have @xmath73\exp[-f(t;\beta , \gamma , \delta _ { p},\delta _ { n})/t]% \mathcal{o}}{\int \mathcal{d}[\alpha ] \exp[-f(t;\beta , \gamma , \delta _ { p},\delta _ { n})/t]}\;\]]with @xmath72=\beta ^{4}|\sin 3\gamma |\,d\beta \,d\gamma \ \delta _ { p}\,\delta _ { n}\,d\delta _ { p}\,d\delta _ { n}$ ] .
we perform the tsf calculations exactly by numerically computing the integrations in eq .
( [ average_2 ] ) with the free energy and the observables calculated at every mesh point ( deformations and pairing gaps ) , utilizing the microscopic - macroscopic approach outlined in sec .
[ sec_defene ] .
sn with @xmath74 calculated from the simple bcs approach and with the pairing fluctuations ( without shape fluctuations ) .
full results for both protons and neutrons are also shown : ( c ) average pairing gap , ( d ) gdr width in the case of @xmath1sn , as a function of temperature .
the calculations done without pairing utilize a ce approach ( ce ) , whereas the calculations with simple bcs pairing ( bcs ) and with pairing fluctuations ( pf ) utilize a gce approach ( gce ) .
the results obtained using the liquid drop model ( ldm ) are also presented .
experimental values for @xmath1sn are taken from refs .
@xcite are shown by solid squares .
for comparison , data for @xmath75sb , taken from ref .
@xcite , are also shown with open circles . ]
sn at @xmath76 mev plotted without pairing and with pairing where the calculations are with ce and gce , respectively .
the contour line spacing is 0.5 mev , the most probable shape is marked by a solid ( red ) circle and the first two minima are shaded . ] ( c ) and [ fig1](d ) but for the nucleus @xmath0tc .
experimental values are taken from ref .
@xcite are shown with solid squares . ]
( d ) but for the nucleus @xmath2pb .
experimental values are taken from ref .
@xcite are shown with solid squares . ]
we categorize our different approaches into the following cases : 1 .
* ldm * - the free energies correspond to that of the liquid drop model and hence no pairing is included .
2 . * without pairing ( ce ) * - the free energies are calculated assuming a ce and the pairing correlations are neglected .
3 . * bcs ( gce ) *
- the free energies are calculated assuming a gce and the pairing has been taken care of within the simple bcs approach .
* pf ( gce ) * - the free energies are calculated assuming a gce and along with the pairing and its fluctuations are also has been taken care of ( [ average_2 ] ) .
the first three cases are with tsf only ( [ average_1 ] ) .
it is convenient to consider the following important factors which can affect the gdr width ( @xmath77 ) in our calculations : 1 .
the shell effects through the modification of free energy surfaces . 2 .
the pairing effects : 1 .
modification of free energy surfaces . 2 .
damping of the gdr through the term @xmath78 we start with the study of temperature dependence of gdr width in the case of @xmath1sn where the measured values @xcite have been used as a benchmark for several theories .
the results from our calculations are presented in fig .
[ fig1 ] . in fig .
[ fig1](a ) we see that with the inclusion of fluctuations in the pairing field , even for a closed shell , the average pairing gap is finite and also sustained at high @xmath7 .
the trend for @xmath79 ( fig . [ fig1](b ) ) is in accordance with earlier works @xcite with the average value converging to the bcs value as @xmath80 .
once we include the tsf along with the pf , the average pairing gaps acquire a larger value as seen in fig .
[ fig1](c ) .
the gdr widths obtained within various approaches are presented in fig . [
fig1](d ) .
we observe that the shell effects have a very little role on @xmath77 because the results with ldm and those without pairing are almost the same .
shell effects can contribute to change in @xmath77 only by modifying the free energy surfaces .
the modification in free energy surface can be of two types viz .
change in the minimum with @xmath77 proportional to the quadrupole deformation ( @xmath81 ) corresponding to the minimum .
2 . change in surface around the minimum : a deeper or crisper minimum will lead to sampling over a smaller region of deformations ( lesser tsf ) and a shallow or well - spread minimum will lead to contributions from larger deformations ( stronger tsf ) . with @xmath82 ,
lesser tsf leads to a quenched @xmath77 . the free energy surfaces for @xmath1sn at @xmath76 mev are presented in fig .
[ fig2 ] where we can see that the one without pairing resembles that of a ldm and hence the shell effects are subtle . with pairing , the surface around minimum energy becomes much crisper and leads to a quenched @xmath77 .
we can notice from fig .
[ fig1](c ) that , while treating pairing within the bcs approach there is no proton pairing and the averaged neutron pairing gap vanishes at @xmath83 mev . with the inclusion of pf ,
the proton pairing develops , the neutron pairing strengthens and it is sustained even at @xmath84 mev .
subsequently the quenching of @xmath77 is also sustained at higher @xmath7 and the results agree with the experimental observations .
very recently , experimental data at low @xmath7 for the nuclei @xmath85tl @xcite and @xmath0tc @xcite are reported and our results for @xmath0tc nucleus is presented here . in @xmath0tc , with bcs calculations , the proton pairing gap and neutron pairing gap vanish at @xmath86 mev and @xmath87 mev respectively as shown in fig .
[ fig3](a ) . while considering the pf along with tsf , the average pairing gaps continue to be strong even @xmath84 mev . at low @xmath7 ,
the discrepancy between ldm " and without pairing " cases are due to the fact that the shell effects drive the shape of this nucleus to a deformed one ( and hence a larger @xmath77 ) .
the inclusion of pairing leads to a deeper minimum and hence quenched back @xmath77 as shown by the dashed line in fig .
[ fig3](b ) . however , this quenching in @xmath77 owing to bcs pairing is not sufficient to explain the experimental data .
the latter can be explained only after the pf along with tsf are included as shown by the solid line in fig .
[ fig3](b ) .
here we have demonstrated that with the inclusion of not just pairing but its fluctuations also is crucial to explain the observed quenching . at low @xmath7 , in @xmath1sn
we have seen that the shell effects have no role on @xmath77 and they increase @xmath77 in the case of @xmath0tc at very low @xmath7 . in @xmath2pb , it is known @xcite that the shell effects quench the gdr width as a consequence of a deeper minimum in the free energy surface .
our results in this case are shown in fig .
[ fig4 ] , where we see that the quenching in @xmath77 occurs once we include the shell effects .
even for such a doubly magic nucleus , we can see a significant pairing gap with the inclusion of pf . however , the pf has no contribution to @xmath77 , which is dominated by the shell effects only .
more experimental data at lower @xmath7 could be useful to validate this argument .
however , our results for @xmath0tc , @xmath1sn and @xmath2pb clearly demonstrate that the tsfm can be quite successful , if the shell effects ( with explicit temperature dependence ) and the pairing effects are properly incorporated in the free energy . to strengthen our arguments
, it would be nice to have more experimental data for a single nucleus but at several temperatures along with a precise data at @xmath88 .
we have constructed a theoretical framework to study the gdr with proper treatment of pairing and its fluctuations along with the thermal shape fluctuations .
our study reveals that the observed quenching of gdr width at low temperature in the open - shell nuclei @xmath0tc and @xmath1sn can be understood in terms of simple shape effects caused by the pairing correlations . for a precise match with the experimental data ,
the consideration of pairing fluctuations is crucial .
our results for @xmath0tc , @xmath1sn and @xmath2pb clearly demonstrate that the tsfm can be quite successful if the shell effects ( with explicit temperature dependence ) and the pairing ones are properly incorporated in the free energy .
it has to be noted that often the empirical formula in ref .
@xcite ( constructed to mimic the results of tsfm at higher temperatures ) is quoted as the phenomenological tsfm .
the subsequent failure of this empirical formula as well as its modifications in describing gdr width at low temperature should not be confused with that of the tsfm ( also called the adiabatic tsfm ) because of the absence of thermal pairing in the latter .
this work is supported by the department of science and technology , government of india .
project no .
sr / ftp / ps-086/2011 . a part of this work was completed at riken and the numerical calculations were carried out using riken integrated cluster of clusters ( ricc ) . | we present an approach that includes temperature - dependent shell effects and fluctuations of the pairing field in the thermal shape fluctuation model ( tsfm ) .
we apply this approach to study the width of giant dipole resonance ( gdr ) in @xmath0tc , @xmath1sn and @xmath2pb .
our results demonstrate that the tsfm that includes pairing fluctuations can explain the recently observed quenching in the gdr width .
we also show that to validate pairing prescriptions and the parameters involved , we require more and precise data . |
one of the central problems in astronomy is that of galaxy formation and evolution : when were the visible parts of galaxies assembled and when were the stars formed ?
the merging of galaxies leads to changes in their masses , and stellar evolution changes their luminosities .
the evolution of the mass - to - light ( @xmath3 ) ratio relates the mass and luminosity evolution .
galaxy mass measurements are notoriously difficult .
fortunately , empirical relations such as the tully - fisher relation for spiral galaxies ( tully & fisher 1977 ) and the fundamental plane ( fp ) for early - type galaxies ( dressler et al . 1987 ; djorgovski & davis 1987 ) can provide us with information on the masses and mass evolution of galaxies .
the fp is a tight relation between the structural parameters and velocity dispersion , which under the assumption of homology implies that @xmath14 with low scatter ( faber et al .
1987 ) .
due to stellar evolution the @xmath3 ratio of a stellar population changes with redshift , and hence also the fp will change .
the redshift evolution of the intercept of the fp is proportional to the evolution of the mean @xmath3 ratio .
hence , the tightness of the fp relation makes it a very sensitive indicator of the mean age of the stellar population of early - type galaxies ( van dokkum & franx 1996 ) . for cluster galaxies
, the @xmath3 ratio evolves very slowly , indicating that the stars were formed at redshifts @xmath15 ( e.g. kelson et al .
1997 ; bender et al .
1998 ; van dokkum et al . 1998 ) .
current semi - analytical hierarchical models place the assembly time of typical early - type galaxies at much lower redshifts ( e.g. kauffmann 1996 ; kauffmann & charlot 1998 ; diaferio et al .
2001 ) . however , in hierarchical galaxy formation models the assembly time and star formation epoch are strongly dependent on the environment , with cluster early - type galaxies forming much earlier than those in the general field ( e.g. kauffmann 1996 ) . hence cluster
early - type galaxies do not provide the most stringent tests of these models .
moreover , the measured evolution of early - type galaxies may underestimate the true evolution because of the effects of morphological evolution .
if many early - type galaxies evolved from late - type galaxies at @xmath16 , the sample of early - type galaxies at high redshift is only a subset of all progenitors of present - day early - type galaxies .
this would lead us to underestimate the luminosity evolution , and hence overestimate the stellar formation redshift ( see van dokkum & franx 2001 , hereafter vdf01 ) .
recent studies have started to explore the fp and @xmath3 ratios of high redshift early - type galaxies in the general field .
interestingly , the fp of field early - type galaxies appears to be quite similar to that of cluster early - type galaxies out to @xmath17 ( treu et al .
2001 ; van dokkum et al .
2001 ) , in apparent conflict with current semi - analytical models ( see van dokkum et al .
2001 ) . however , there are indications for a significant offset between the two populations at higher redshift ( treu et al .
2002 ) .
kochanek et al .
( 2000 , hereafter k00 ) analysed the fp and color evolution of strong gravitational lens galaxies up to @xmath18 .
the lensing cross section is dominated by galaxies with high central velocity dispersions , and the lens galaxies effectively provide a mass - selected sample of field early - type galaxies which can be compared to optically selected samples of field and cluster galaxies . the mass - selection is important because it limits malmquist - type biases , and it is much less sensitive to selection effects caused by morphological evolution .
k00 find that the stars constituting the lens galaxies must have formed at @xmath19 for a flat cosmology with @xmath7 and @xmath20 , and conclude from their analysis that there are no significant differences between field and cluster early - type galaxies .
k00 analyse the evolution of the fp in the observed photometric bands , and their modeling necessarily includes the large variation in observed magnitudes and colors due to redshift ( the `` k''-correction ) .
this complicates the measurement of the smaller effect due to evolution of ( the stellar populations of ) the lens galaxies .
rusin et al .
( 2003 , hereafter r03 ) study the same sample of lens galaxies , with recent photometric observations included .
they convert the data from observed filters into magnitudes in standard restframe bands , so that the evolution of the @xmath3 ratio that follows from the fp ( under the usual assumption that early - type galaxies are a homologous family ) can be investigated instead of the evolution within the fp .
this clarifies the analysis and allows for direct comparison of the results with those of other fp studies . from their @xmath3 evolution analysis r03
find a ( 2@xmath21 ) lower limit @xmath22 for the formation of the stars in lens galaxies .
although this limit is more precise , it is similar to that of k00 .
hence , r03 also conclude that the evolution measurements favor old stellar populations among field galaxies , like those of cluster galaxies , and argue against significant episodes of star formation at @xmath23 , as predicted by the semi - analytical hierarchical models . for the analysis in this paper we use the tabulated data of k00 , extended with the recent photometric measurements as given by r03 .
we convert the observed surface brightnesses and colors to the restframe johnson @xmath1 and gunn @xmath2 bands .
our transformation to restframe differs from that of r03 . in both approaches the modeled color between restframe band and observed filter is used to convert the observed magnitude into an estimate of the restframe magnitude . r03
obtain the modeled color for a given spectral energy distribution , whereas we use the observed color between a pair of filters to interpolate between the modeled colors for four different spectral types ( fig .
[ fig : magtrafo ] ) .
moreover , while r03 use the ( weighted ) contribution of all observed magnitudes to determine the restframe magnitude , we select the best one with small observational error and filter close to the ( redshifted ) restframe band .
hence , the model dependence of our correction is small and we exclude observed magnitudes with large uncertainties , minimizing the error in the resulting restframe magnitude .
we analyze the @xmath3 evolution of the lens galaxies and compare our results with those of r03 and with results from previous studies
. additionally , we test whether age differences between the lens galaxies are significant .
to this end , we study the scatter in both the @xmath3 evolution and the restframe colors of the lens galaxies , and investigate whether the deviations are correlated .
this paper is organized as follows . in section [ sec : fppar ] we determine the fp parameters , using the lensing geometry to estimate the velocity dispersion .
the transformation from observed to restframe bands is described in section [ sec : trafo2restrame ] . in section [ sec : fpandmlevo ] , we construct the fp of the lens galaxies and present the @xmath3 evolution derived from the fp .
we estimate the age of the stellar populations of the lens galaxies by fitting single burst models to the @xmath3 evolution in section [ sec : ages ] . in section [ sec :
colors ] we study the colors of the lens galaxies .
the results are summarized and discussed in section [ sec : discconcl ] . unless stated otherwise we assume @xmath24 and a flat cosmology with @xmath7 and @xmath8 .
we note that our results are not dependent on the value of the hubble constant .
[ cols="<,^,^,^,^,^,^ " , ] we have demonstrated that the @xmath3 evolution of lens galaxies can not be fitted with models that provide good fits to cluster galaxies .
the fit clearly improves if lens galaxies are allowed to be systematically offset from cluster galaxies due to metallicity differences or other systematic effects . if we consider both this offset and the stellar formation epoch of the lens galaxies as free parameters , the best - fit single burst model in the @xmath1 and @xmath2 band are consistent .
the resulting offset is not significantly different from zero , and although the best - fit stellar formation redshift implies for lens galaxies on average a younger stellar population than for cluster galaxies , the resulting ( @xmath25 ) range of @xmath26 is not significantly different from the formation epoch for cluster galaxies of @xmath27 for the full range from no to maximum progenitor bias correction .
if we impose the constraint that galaxies of the same age have the same @xmath3 ratio irrespective of their environment , we find stellar formation redshifts for the lens galaxies of @xmath28 if the stars in cluster galaxies formed at @xmath29 ( maximum progenitor bias ) , and @xmath30 if the stars in cluster galaxies formed at @xmath31 ( no progenitor bias ) . in the latter case
the stellar populations of the lens galaxies are significant younger ( 1015@xmath12 at the present epoch ) than those of the cluster galaxies .
if the local @xmath3 offset between lens and cluster galaxies is allowed to vary , r03 find for the stellar formation redshift of the lens galaxies a ( @xmath25 ) range of @xmath32 from their @xmath1 band analysis .
although we find a range of @xmath33 ( model a , @xmath1 band ) which implies on average a somewhat younger stellar population , the results are consistent .
r03 conclude that the @xmath3 evolution rates they measure favor old stellar populations for the lens galaxies with a mean formation redshift @xmath34 at a @xmath35 confidence level .
we find a lower @xmath35 confidence limit of @xmath36 .
to test the hypothesis whether the @xmath3 evolution of the lens galaxies can be fitted with a single burst model , we calculate the reduced @xmath37 .
we also compare the measured scatter around the fit with the expected scatter from the uncertainties in the @xmath3 ratios of the lens galaxies .
if we allow more freedom in the single burst models , the fit improves ( lower reduced @xmath37 ) and the measured scatter decreases .
however , even the over - all best - fit single burst model is rejected with @xmath38 confidence and the measured scatter of @xmath39 and @xmath40 in @xmath1 and @xmath2 band is significant higher than the expected scatter of @xmath41 from the observational errors . to establish whether the additional scatter is due to a significant spread in ages among field galaxies , we study in the next section the ( restframe ) colors of the lens galaxies , and investigate if there is also a significant scatter in the color and if it is correlated with the scatter in their @xmath3 ratios .
if the stellar populations of lens galaxies are on average younger than those of cluster galaxies ( section [ sec : ages ] ) we expect their colors to evolve more rapidly and to be on average bluer than those of cluster galaxies . for single burst stellar populations
the @xmath42 color evolves as @xmath43 with @xmath44 ( worthey 1994 ) and @xmath45 a normalization constant .
the left panel of fig .
[ fig : color ] shows the restframe @xmath42 colors of the lens galaxies versus redshift . for the lens galaxies with open circles ,
a modest extrapolation was required in the transformation to either the restframe @xmath1 or @xmath2 band .
the red outlier ( at @xmath46 ) is the lens bri0952 - 0115 , for which the observed @xmath47 ( f675w - f160w ) color is significantly redder than the modeled e / s0 color , and for which we had to extrapolate in the transformation to restframe @xmath1 band .
the single burst evolution of a stellar population formed at the mean cluster formation epoch of @xmath29 ( maximum progenitor bias ) , is drawn ( solid line ) through the averaged color of the coma cluster galaxies ( small dot ) .
a decline in the colors of the lens galaxies with increasing redshift is visible , in spite of the large scatter .
if we fit single burst models as in section [ sec : ages ] ( model a and b ) , we find on average a younger stellar formation epoch for the lens galaxies , but the difference with the cluster galaxies is never significant . in case
the stellar populations of field and cluster galaxies are assumed to evolve in the same way ( model b ) , the evolution model with the best - fit stellar formation redshift of @xmath48 is shown in the left panel of fig .
[ fig : color ] ( dashed line ) .
similarly as for the single burst model fits to the @xmath3 evolution , the fits to the color evolution are rejected with @xmath38 confidence , and the measured scatter is in all cases significantly higher than the expected scatter from the observational errors in the restframe colors of the lens galaxies . to further investigate the scatter in color ,
we subtract the predicted colors of cluster galaxies with @xmath49 , from the restframe colors of the lens galaxies . note that this choice of formation redshift corresponds to the _ minimum _ age difference between field and cluster galaxies . in the right panel of fig .
[ fig : color ] we show the resulting residual colors plotted against velocity dispersion , with a linear fit to the coma cluster galaxies .
the lens galaxies show a large spread in their reduced colors , and are on average bluer by @xmath50 mag than the cluster galaxies .
the bluer average color is qualitatively consistent with the on average younger ages of lens galaxies derived from the single burst model fits above and from the fits to the @xmath3 evolution ( section [ sec : summaryresults ] ) . for the total sample of 26 lens galaxies , we measure a scatter in the residual colors of @xmath51 , that is significant higher than the expected scatter of @xmath52 . from the left panel of fig .
[ fig : color ] , we observe that the color scatter increases beyond @xmath53 . for the 15 lens galaxies with @xmath54 we measure a scatter of @xmath55 , whereas for the lens galaxies with @xmath56 the measured scatter is only @xmath57 and nearly identical to the expected scatter of @xmath40 from the errors in the colors of these galaxies
. the additional measured scatter in the residual colors of the lens galaxies may indicate a significant spread in the ages of the stellar populations of the lens galaxies .
we test whether the color spread is caused by a spread in ages or other effects , by investigating whether the residual colors correlate with the residual @xmath3 ratios . in fig .
[ fig : colvsml ] we plot the color of the lens galaxies against the logarithm of their @xmath3 ratio . for both quantities
the expected evolution of a stellar population formed at @xmath49 was subtracted ; hence a galaxy with the average color and @xmath3 ratio of galaxies in coma would be located at the origin . from eq . and
it follows that a linear relation is expected in fig .
[ fig : colvsml ] , with a ( time - independent ) slope of @xmath58 ( solid line ) .
the data appear to be correlated in the expected sense , albeit with large scatter .
to test whether the correlation is significant we used the spearman s rank - order correlation coefficient @xmath59 .
we find that @xmath60 , so that with @xmath61 lenses , we can reject the hypothesis that the two quantities are uncorrelated with @xmath62 confidence .
the only viable explanation for the correlation is age variation .
if the correlation would be caused by metallicity variations , ( some ) field galaxies would be much less metal rich than cluster galaxies , opposite to the result by kuntschner et al .
( 2002 ) . the colors and @xmath3 ratios of the bluest lens galaxies are best fitted with stellar formation redshifts as low as @xmath13 .
about half the lens galaxies are consistent with an old cluster - like stellar population with stellar formation redshift @xmath63 .
if galaxies form in a sequence of bursts , formation redshifts are indicative of the last prominent epoch of star formation .
the galaxies can have older underlying stellar populations .
we studied the evolution of the @xmath3 ratios of lens galaxies in the restframe johnson @xmath1 and gunn @xmath2 bands . for an flat cosmology with @xmath7 and @xmath8
, we obtained an evolution rate @xmath4 of @xmath5 in restframe @xmath1 and @xmath6 in restframe @xmath2 .
due to differences in the determination of the fp parameters and the corresponding errors , the evolution rate obtained by r03 of @xmath64 is slightly slower but not significant different .
our results are consistent with determinations for field galaxies based on direct spectroscopic measurements of the velocity dispersions . for the @xmath1 band , van dokkum et al .
( 2001 ) arrive at an evolution rate of @xmath65 , and treu et al.(2002 ) find a value of @xmath66 .
the weighted mean of these results and our result yields an evolution rate for field early - type galaxies of @xmath67 in the @xmath1 band . for cluster galaxies van dokkum et al .
( 1998 ) and vdf01 find an evolution rate between @xmath68 and @xmath69 , for minimum and maximum progenitor bias respectively .
the latter value is not significantly different from the average evolution rate for the field galaxies .
we further investigated the @xmath3 evolution in section [ sec : ages ] , where we related it to stellar population ages by fitting simple single burst models for a salpeter ( 1955 ) imf and solar metallicity .
the @xmath3 evolution of cluster galaxies is well approximated by a mean single burst formation redshift of @xmath70 , and @xmath71 after maximum progenitor bias correction ( vdf01 ) .
we first tested if these cluster models could fit the @xmath3 evolution of the lens galaxies .
this is not the case , but the fits improve if there is a systematic offset between lens and cluster galaxies of @xmath72 in @xmath73 such an offset could be caused by , e.g. metallicity variations , systematic differences in the velocity dispersions due to the different measurement techniques , or other effects .
it is interesting to note that hierarchical models have predicted such a constant offset with redshift between cluster and field galaxies ( see van dokkum et al .
however , similar to van dokkum et al.(2001 ) , we conclude that the observed offset is much smaller than the predicted offset of @xmath74 .
if we next allow the stellar formation redshift @xmath75 also to vary , we find for the best - fit single burst model that the offset is not significant from zero , together with a mean stellar formation epoch @xmath76 for the lens galaxies of @xmath77 in the @xmath1 band and @xmath10 in the @xmath2 band .
this means that on average the stellar populations of the lens galaxies are younger , but the difference with the cluster galaxies is not significant .
however , if we impose the constraint that galaxies of the same age have the same @xmath3 ratio irrespective of their environment , we find without correction for progenitor bias that the stellar populations of the lens galaxies are significant younger ( 1015@xmath12 at the present epoch ) than those of the cluster galaxies . in the case of maximum progenitor bias the average difference ( @xmath78 )
is not anymore significant . from their analysis
r03 obtain a ( @xmath35 ) lower limit for the mean stellar formation epoch of lens galaxies of @xmath34 , whereas we find that lower stellar formation redshifts are allowed , with @xmath36 as a @xmath35 lower limit . nevertheless , these results disagree with the prediction of semi - analytical hierarchical galaxy formation models ( e.g. kauffmann 1996 ; kauffmann & charlot 1998 ; diaferio et al .
2001 ) that early - type field galaxies in general have very late star formation with @xmath79 .
whereas the @xmath3 evolution of cluster galaxies is well approximated by a single burst evolution model , we found that this is not the case for the lens galaxies .
all single burst models are rejected with @xmath38 based on the reduced @xmath37 , and the measured scatter is in all cases significant higher than the expected scatter from the observational errors .
the additional scatter is most likely caused by differences in the stellar population of the lens galaxies . to investigate whether there is a significant spread in ages , we studied in section [ sec : colors ] the restframe colors of the lens galaxies and their evolution .
the colors of the lens galaxies are on average bluer than those of the cluster galaxies , consistent with on a younger average stellar population .
we found that the measured scatter in the single burst fits to the colors is much larger than the expected scatter from the observational errors .
moreover , we showed that there is a significant correlation between the colors and the @xmath80 ratios of the lens galaxies .
we interpret this as evidence for a significant spread in the stellar population ages of the lens galaxies . whereas about half of the lens galaxies are consistent with old cluster - like stellar populations , the bluest galaxies are best fit by single burst models with young stellar formation redshifts @xmath13 .
for the seven blue lens galaxies with residual colors below @xmath81 mag in the right panel of fig .
[ fig : color ] , we found ( as a byproduct of our transformation to restframe @xmath1 band ) that only two of them are best fitted by the e / s0 spectral type , i.e. @xmath82 , whereas for the total sample we found @xmath83 .
for the @xmath2 band even six of the seven blue galaxies are best fitted by the scd spectral type , instead of the es/0 spectral type .
three of the seven blue galaxies indeed show ( some ) deviations from early - type morphology . fbq0951 +
2635 is an edge - on disk galaxy , sbs1520 + 530 is slightly irregular and b1608 + 656 is an apparently dusty galaxy with star forming regions .
however , a homogeneous sample of quasar subtracted ( ` nicmos ` ) images of the lens galaxies is needed to do a more detailed and systematic study of their morphologies
. such a study will be valuable to determine the cause of apparently younger populations in a fraction of the lens galaxies .
we thank the anonymous referee for constructive and detailed comments , which improved the paper significantly . | we analyse the fundamental plane ( fp ) of 26 strong gravitational lens galaxies with redshifts up to @xmath0 , using tabulated data from kochanek et al .
( 2000 ) and rusin et al .
( 2003 ) .
the lens galaxies effectively form a mass - selected sample of early - type galaxies in environments of relatively low density .
we analyse the fp and its evolution in the restframe johnson @xmath1 and gunn @xmath2 bands . assuming that early - type galaxies are a homologous family , the fp then provides a direct measurement of the @xmath3 ratio evolution .
if we assume that the @xmath3 ratios of field early - type galaxies evolve as power - laws , we find for the lens galaxies an evolution rate @xmath4 of @xmath5 in restframe @xmath1 and @xmath6 in restframe gunn @xmath2 for a flat cosmology with @xmath7 and @xmath8 . for a salpeter ( 1955 )
initial mass function and solar metallicity these results correspond to mean stellar formation redshifts of @xmath9 and @xmath10 respectively . after correction for maximum progenitor bias ,
van dokkum & franx ( 2001 ) find a mean stellar formation redshift for cluster galaxies of @xmath11 , which is not significantly different from that found for the lens galaxies .
however , if we impose the constraint that lens and cluster galaxies that are of the same age have equal @xmath3 ratios and we do not correct for progenitor bias , the difference is significant and we find that the stellar populations of the lens galaxies are 1015@xmath12 younger than those of the cluster galaxies .
we find that both the @xmath3 ratios as well as the restframe colors of the lens galaxies show significant scatter .
about half of the lens galaxies are consistent with an old cluster - like stellar populations , but the other galaxies are bluer and best fit by single burst models with younger stellar formation redshifts as low as @xmath13 . moreover ,
the scatter in color is correlated with the scatter in @xmath3 ratio .
we interpret this as evidence of a significant age spread among the stellar populations of lens galaxies , whereas the ages of the stellar populations of the cluster galaxies are well approximated by a single formation epoch .
[ firstpage ] galaxies : elliptical and lenticular , cd galaxies : evolution galaxies : formation gravitational lensing |
the wave functions satisfying the dirac equation @xmath128 is a four component spinor as , @xmath129 where @xmath38 and @xmath130 correspond to the conduction and valence band , respectively .
also , @xmath78 correspons to the spin directions of carrier with respect to @xmath79-axis .
the hole part of dbdg equation is obtained by operating with the charge conjugation on the hamiltonina and its spinor of eq.([e.01 ] ) .
it consists in the complex conjugation operator @xmath41 and multiplication by some @xmath9 matrxi the form of which depends on the representation of @xmath131 matrices and in our case is @xcite , @xmath132 such that , @xmath133 which can be recast into the form , @xmath134 where the charge conjugated state @xmath135 is explicitely given by @xcite , @xmath136 from @xmath137 and @xmath135 we can construct a nambu spinor @xmath138 that satisfies dbdg equation in the direct product space of eqs .
( [ e.04 ] ) and ( [ e.06 ] ) , @xmath139 the explicit representation of the dbdg hamiltonian matrix in the pseudo - scalar pairing channel becomes , @xmath140 to clarify the total spin of the pairing symmetry let extract the pairing part of @xmath141 which in terms of @xmath142 becomes , @xmath143 and simplifies to , @xmath144 which clearly shows the singlet nature of pairing as is expected from singlet character of a pseudo - scalar pairing .
to calculate the topological charge of our hamiltonian we need to construct the particle - hole ( @xmath145 ) , time - reversal ( @xmath146 ) , and sublattice ( @xmath147 ) transformations in the nambu space . in the main text we introduces the above three operators the explicit form of which
is given by , @xmath149 @xmath150 and @xmath151 where @xmath152 , @xmath153 and @xmath154 with @xmath155 are pauli matrices which act on nambu , band and spin spaces , respectively .
also , @xmath156 corresponds to 2 by 2 unit matrix .
a hamiltonian with chiral symmetry can be rotated to block off - diagonal form .
the unitary operator constructed from bases of chiral matrix can convert the original dbdg to block off - diagonal matrix ( chiral dbdg ) .
the eigenvalues and eigenvectors of eq .
( [ eq.03 ] ) are , @xmath158 which gives the unitary matrix @xmath159 as , @xmath160 this transformation brings the dbdg hamiltonian , eq .
( [ e.07 ] ) to the canonical form , @xmath161 in the presence of translational invariance , the ground states of 3d dirac superconductor can be constructed as a fermi sea that gets then gapped out by superconducting pairing which separates the filled and empty states in 3d brillouin zone ( bz ) .
this enables to define a projector operator as , @xmath163 to construct the projector operator , at first step , we must calculate the wave functions of eq .
( [ eq.07 ] ) for energy eigenvalues of the filled states which are given by , @xmath164 the wave functions corresponding to the eigenvalues in eq .
( [ eq.09 ] ) are , @xmath165 where we have used the notations , @xmath166 the polar and azimuthal angels in the @xmath167-space are denoted by @xmath168 .
the projector operator can be calculated by plugging the wave functions of eq .
( [ eq.10 ] ) into eq .
( [ eq.08 ] ) which gives , @xmath169 where @xmath170 and @xmath171-matrix are 4 by 4 matrices .
then the @xmath172-matrix is obtained as : @xmath173 here @xmath170 is a unit matrix with 8 by 8 dimension .
the @xmath171-matrix which is the off diagonal block of projector is essential for calculation of the topologial index , and is given by , @xmath174 we are now set to calculate the topological invariant from @xmath176 matrix that is given by the integral @xmath177=\int \frac{d^3k}{24 \pi^2}\epsilon^{\mu\nu\rho}tr\left[\left(q^{-1}\partial_\mu q\right)\left(q^{-1}\partial_\nu q\right)\left(q^{-1}\partial_\rho q\right)\right ] , \label{eq.15}\ ] ] here the @xmath178 is the levi - civita tensor and @xmath179 run over 3 directions @xmath180 . by plugging eq .
( [ eq.14 ] ) into eq .
( [ eq.15 ] ) we find , @xmath181 for 3d dirac materials ( non superconducting ) case where @xmath182 , the winding number will be @xmath183 while for 3d dirac superconductor @xmath184 the winding number will be @xmath185 .
the above topological index therefore shows a clear change by movign from @xmath0 dominated gapped state to @xmath1 dominated gapped state in a @xmath21 region .
in this section we obtain the transport coefficients in normal and andreev channels . in 3ddm region ,
the superconducting gap is zero @xmath187 , which reduces the dirac - bogoliubov - degennes to , @xmath188 the electron part of eq .
( [ 3ddmbdg ] ) is completely decoupled from the hole one .
each part has four components which can be defined as right or left mover with respect to @xmath79-axis .
also , each states has two spin orientations , @xmath189 .
for electron part with wave vector @xmath190 , we can define the longitudinal and transverse components of wave vector as : @xmath191 as mentioned in the main text , the transverse components of wave vector are quantized due to the finite width of junction and are labaled by @xmath192 .
so the wave functions can be written as , @xmath193 where @xmath194 , is the @xmath195 vector with zero entities , and @xmath196 . in order to conserve thei current , the wave functions of eq .
( [ wavefunctions1 ] ) need the normalization factor , @xmath197 . for the wave functions of hole part we obtain , @xmath198 where @xmath199 and @xmath200 .
the polar angle of propagation for hole - like quasi - particle is @xmath201 .
owing to translation invariance in the transverse directions , @xmath202 is conserved at the scattering processes .
the @xmath203 can be written as , @xmath204 this equation implies that for incident angle ( @xmath90 ) beyond critical value of , @xmath205 the perpendicular component of the wave vector for hole - like wave functions will be imaginary .
this means hole - like states beyond these propagation angles can not contribute to transport .
the normalization factor for wave functions of eq .
( [ wavefunction2 ] ) is @xmath206 . in superconducting region the pairing matrix @xmath208
and we have @xmath209 . in this regime
the dbdg hamiltonian will be : @xmath210 the eigenvalues of eq.([eq.dbdg ] ) are , @xmath211 each eigenvalues are doubly degenerate because of spin .
the wave vectors of electron - like and hole - like quasi - particles corresponding to the eigenvalues of eq.([eq.seigenvalues ] ) are , @xmath212 with corresponding wave functions : @xmath213 note that all wave functions in eq .
( [ wavefunctions3 ] ) must be multiplied by a the bloch phase of parallel component of wave vector , @xmath214 and the normalization factor @xmath215 . using wave functions of eqs .
( [ wavefunctions1 ] ) , ( [ wavefunction2 ] and ( [ wavefunctions3 ] ) , one can solve the boundary conditions of eq .
( 4 ) in the main text to complete the construction of scattering states .
the reflection amplitudes for an @xmath83-spin incident electron are derived as , @xmath217 where the notation is : @xmath218 in zero mode regime , @xmath219 , these reflection amplitudes reduces to eq .
( 6 ) of the main text .
in order to obtain the supercurrent for @xmath220 josephson junction we use transfer matrix method to calculate the energy dependence of the andreev bound states ( abs)@xcite . in the short junction limit , @xmath114 , when an @xmath83-spin electron hits the right interface of @xmath221 junction at @xmath222 , the reflection amplitudes of eq .
( [ eq.reflectionamplitudes ] ) reduce to : @xmath223 here , @xmath224 .
we use superscript @xmath225 when an @xmath84-spin electron hits the interface . similarly , @xmath226 and
@xmath227 are using for @xmath83-spin and @xmath84-spin hole , respectively .
the first index in the doulbe spin indices indicates the spin configuration of incoming particle and the second one denotes the outgoing one .
the primarily matrices @xmath228 and @xmath229 can be constructed by these 16 amplitudes as @xcite : @xmath230 using these two matrices , the transfer matrix at the right interface which convert electron to hole wave functions can be calculated as : @xmath232 with the similar procedure , one can construct the transfer matrix in the left interface which can convect hole to electron wave function , @xmath233 . also , there are two another transfer matrices , @xmath234 , which can be used for transferring electron or hole wave function from left interface to the right one
. the energy of abss is then obtained from the condition , @xmath235=0 .
\label{eq.abs01}\ ] ] after some straightforward algebra we obtain the energy of abss as given in eq .
( 8) of the main text .
the derivation of josephson current from abss is explained in the main text as well . | we find that _ singlet _ superconducting pairing can lead to majorana fermions in three dimensional dirac superconductors ( 3dds ) if the pairing order parameter is a pseudo - scalar , i.e. it changes sign under mirror reflection .
the pseudo - scalar superconducting order parameter , @xmath0 can close and reopen the spectral gap caused by the scalar dirac mass @xmath1 in a three - dimensional dirac material ( 3ddm ) , giving rise to a two - dimensional majorana sea ( 2dms ) at the plane of the gap kink . by bringing the hamiltonian into a canonical form which then gives the winding number , we show that this system belongs to the diii class of topological superconductors .
we calculate the transport signature of 2dms , namely a perfect andreev - klein transmission that manifests in a robust peak in the differential conductance .
further , we find the @xmath2 periodicity in the @xmath3 josephson junctions .
gauge transformed version of the present scenario implies that the interface of a conventional s - wave superconductor with a peculiar 3ddm whose mass is a pseudo - scalar , @xmath4 also hosts a 2dms . _
introduction_. band topology in insulators and superconductors is connected with the change in the sign of the gap parameter which in turn gives rise to zero energy states at the location of gap kink .
this mechanism in the case of insulators gives rise to gapless surface modes protected by a topological invariant @xcite .
when the spectral gap is of the superconducting ( pairing ) nature , these topologically protected modes will be majorana zero modes , which are their own anti - particles @xcite . to realize majorana fermions ( mfs ) various scenarios
have been proposed which involve closing and re - opening the superconducting gap in one way or another .
gaping chiral modes of topological insulators ( tis ) by zeeman field and superconducting pairing gives rise to mfs in the interface region where these two gapping mechanisms are of comparable strength @xcite . in one - dimensional nano - wires this can be achieved by the competition between a polarizing zeeman field , and depolarizing spin - orbit interaction @xcite .
alternative scenario is to generate mfs involve driving supercurrent through a superconductor ( sc ) next to a ti in a pairing dominated gap regime where twist due to supercurrent reduces the phase space for cooper pairing and gives way to zeeman dominated gap @xcite . in two - dimensions ,
the vortex core of a p - wave sc binds a mf @xcite .
the required p - wave sc can be engineered on the surface of a ti @xcite by proximity to a conventional s - wave sc .
the above scenarios are first of all limited to low dimensions , and secondly require a tr breaking by a zeeman field . in this letter
we propose yet another _ three dimensional _ system that admits a two dimensional sheet of mfs without requiring a zeeman field .
we find that a peculiar _ pseudo - scalar _
( odd - parity ) superconducting order parameter , @xmath0 can give rise to a two - dimensional sheet of mfs when it is interfaced with a three - dimensional dirac material ( 3ddm ) . to set the stages for our finding ,
let us start by noting that in a one - band situation described by a parabolic band dispersion , the strength of the gap is characterized by a ( scalar ) gap parameter .
however in 3ddm where the relevant degrees of freedom are described by the dirac equation @xcite , new possibilities can arise .
we assume a 3ddm which has a single dirac point in the @xmath5 point of it s brillouin zone @xcite .
for such a 3ddm , the low - energy degrees of freedom are described by 4-component dirac spinors and hence the most general pairing is @xmath6 where @xmath7 satisfies the same dirac equation as @xmath8 , but with opposite charge and is covariant under lorentz transformation .
the @xmath9 pairing matrix @xmath10 can be expanded in terms of a basis composed of @xmath11 , four @xmath12 , @xmath13 , four @xmath14 and six anti - commutators @xmath15/2 $ ] @xcite as @xmath16
. then the pairing @xmath17 will be the conventional scalar pairing , while @xmath0 will be a pseudo - scalar pairing under the lorentz transformation .
similarly @xmath18 , will be vector , pseudo - vector and tensor superconducting pairings .
we have examined all the above @xmath19 pairing channels and find that the pseudo - scalar superconducting gap @xmath0 works in an opposite direction to the normal dirac gap @xmath20 . therefore placing a @xmath1 dominated region
( i.e. a 3ddm ) next to a @xmath0 dominated region denoted as @xmath21 , gives rise to a two - dimensional sheet of majorana fermions .
this mechanism , _ does not require p - wave pairing , nor a magnet to break time - reversal_. instead , it requires a peculiar form of pseudo - scalar superconducting order parameter that breaks the mirror symmetry .
we then proceed to show that such a system belongs to the diii topological class allowing @xmath22-number classification which in turn guarantees the existence of two - dimensional majorana sea ( 2dms ) on the region where the strength of @xmath1 and @xmath0 are equal .
we further corroborate our results by showing a perfect andreev reflection and robust zero - mode resonance peak in differential conductance and fractional supercurrent in @xmath3 josephson junction . _
model : _ we consider a dirac material with a single dirac cone , @xmath23 we use the representation @xmath24 and @xmath25 for the clifford algebra where @xmath26 and @xmath27 are pauli matrices acting on spin and band spaces , respectively .
also , @xmath28 and @xmath29 are the @xmath30 unit matrices .
the @xmath31 is the wave vector of excitations . in four space - time dimensions one can also construct @xmath32 which will be very essential for our discussion in this paper . in eq.([eq.01 ] ) , the mass term is of the ordinary @xmath20 nature and is responsible for the band gap , and @xmath33 is the fermi velocity .
the pairing hamiltonian for such a system is : @xmath34 corresponding to which the dirac - bogoliubov - degennes ( dbdg ) equation in * k*-space is , @xmath35 where @xmath36 is the energy of eigenstate with respect to the chemical potential @xmath37 . here
@xmath38 is the macroscopic phase of superconductor , @xmath39 is the electron ( hole ) part of bdg wave function in nambu space .
the anti - unitary operator @xmath40 is the charge - conjugation of eq.([eq.01 ] ) , where @xmath41 is the complex - conjugation : this means that lorentz transformation for @xmath42 is the inverse of the lorentz transformation for @xmath8 . therefore requiring the superconducting pairing to be lorentz invariant @xcite
, the scalar ( conventional ) superconducting pairing is given by @xmath43 .
similarly pseudo - scalar pairing is @xmath44 , etc .
we have examined all of the above 16 possible superconducting pairing channels .
we find that only the @xmath0 parameter gives rise to a gap closing when both @xmath1 and @xmath0 are present .
to see this in @xmath0 channel , let us for clarity of notation set @xmath45 and @xmath33 to @xmath46 which gives the eigenvalues of eq .
( [ eq.dbdg ] ) as @xmath47 , where @xmath48 and @xmath49 .
when the dirac mass @xmath1 dominates the spectral gap ( @xmath50 ) the above eigenvalues reduce to @xmath51 . in this case
each eigenvalue is doubly ( spin ) degenerate @xcite .
deep in the 3dds region where @xmath0 dominates the eigenvalues acquire the following structure @xmath52 which corresponds to a pairing gap at the fermi level . in a region where both @xmath1 and @xmath0 are non - zero the nature of gap
can be more clearly seen if we look at @xmath53 where dispersion becomes @xmath54 and the band gap is determined by @xmath55 which clearly indicates the competition between the dirac mass @xmath1 and the pseudo - scalar superconducting parameter @xmath0 .
fig .
[ fig.topologicalphasetransition ] summarizes the closing and reopening of the spectral gap as one moves from @xmath1 dominated region to @xmath0 dominated region . .
the length scale @xmath56 is set by the energy gap of the 3ddm . ]
the closing of the dirac gap and re - opening of it in the form of a pseudo - scalar superconducting gap in this system does not require any magnetic field .
this implies that when a @xmath0 3dds is brought next to a normal 3ddm with gap parameter @xmath1 , such that @xmath57 , the induced @xmath0 on the normal 3ddm side decays towards the bulk of 3ddm and crosses @xmath1 somewhere in the interface where the excitations become gapless .
let us now prove that when @xmath58 a two dimensional majorana sea appears .
the pseudo - scalar character of @xmath0 pairing can be interpreted as a _ singlet _
superconductor whose mirror image has an opposite sign . to understand further the properties of such a pairing ,
let us now explicitly construct the anti - unitary operators corresponding to particle - hole , and time - reversal transformation for eq . : let @xmath59 be set of pauli matrices in the nambu space .
then the particle - hole and time - reversal symmetries in this space can be defined as @xmath60 and @xmath61 , respectively .
owing to @xmath62 and @xmath63 , the chiral symmetry @xmath64 satisfies @xmath65 which places the present system in the diii class of topological superconductors which can be classified with winding number .
we explicitely obtain the topoligical charge @xcite , @xmath66 which clearly shows that @xmath67 can be @xmath68 , and hence a @xmath22 number classification @xcite . 3dds junction ( @xmath21 ) .
we assume a 3ddm can be superconductor by proximity effect .
( b ) the josephson junction of 3dds@xmath693ddm@xmath693dds ( @xmath3 ) . ] _ transport signatures of majorana sea_. to further corroborate our central result concerning a sea of majorana fermions in the present system , let us now focus on the transport properties arising from 2dms which can be directly accessed in experiments .
let us begin by looking into a single @xmath21 junction .
the lateral coordinates in the plane of junction are @xmath70 confined within a lateral dimensions @xmath71 , which means the corresponding components of wave vector in these directions are quantized as @xmath72 .
each mode can be identified with a set of quantum number @xmath73 . in the 3ddm side of the junction ( @xmath74 ) ,
there are eight components of wave functions , @xmath75 where the indices @xmath76 ( @xmath77 ) characterize electron ( hole)-like quasi - particles , @xmath78 denotes its spin direction with respect to @xmath79-axis and @xmath80 indicates the right or left mover character of carriers @xcite .
the interface between @xmath1 , ( @xmath74 ) and @xmath0 ( @xmath81 ) regions can reflect an incident electron as a hole . in this andreev reflection process ,
the missing charge of @xmath82 is absorbed by superconductor as cooper pair @xcite .
typically the s - wave assumption of superconductivity imposes that the reflected hole from an @xmath83-spin incident electron must be in a @xmath84-spin state and vice versa .
however , due to strong spin - orbit coupling encoded in the dirac hamiltonian of a 3ddm , when an electron in a given spin direction hits a 3ddm , the spin of the electron transmitted into 3ddm can be flipped @xcite . therefore in 3ddm@xmath693dds junction
we also take into account the possibility of an anomalous i.e. spin flipped
andreev reflection of the reflected hole .
the boundary condition for an @xmath83 spin electron hitting the interface is given by , @xmath85 the left hand side is related to the wave function in 3ddm side , and the right hand side describes the wave equation in 3dds side . here
@xmath86 and @xmath87 are the amplitudes of conventional and spin - flipped normal reflection , respectively .
the @xmath88 and @xmath89 are the amplitudes of conventional and anomalous andreev reflection , respectively .
similar processes take place for @xmath84 spin and hole carriers as well @xcite . the probability of these reflections vs. @xmath90 , the polar angle of incidence ,
are depicted in fig .
( [ pvsangle ] ) .
because of the conservation of parallel component of wave vector , @xmath91 at the scattering process , the angle of propagation for reflected hole ( @xmath92 ) has a critical value @xmath93 beyond which the reflected hole can not contribute to transport . for zero modes , @xmath94 ,
the conventional and spin - flipped normal reflections would disappear and we are left with conventional and anomalous andreev reflection given by , @xmath95 where , @xmath96 is azimuthal angle . from eq .
( [ eq.08zero - energyar ] ) , it is obvious that for a zero - energy incident electron at any angle of propagation we have a perfect andreev reflection , @xmath97 .
this is a transport signature of majorana fermions @xcite .
this effect is robust against changing the chemical potential and angles of incidence ( @xmath98 ) . the btk formula for the differential conductance of the junction @xcite will be ,
, the andreev - klein tunneling occurs in @xmath99 .
( b ) in @xmath100 , the perfect andreev reflection occurs for all angles of incidence .
the input values are @xmath101.,width=6,height=2 ] @xmath102 \label{eq.conductance}\ ] ] where @xmath103 is the quantum of conductance .
note that the summation over @xmath104 includes spin - flipped contributions as well . when the linear dimensions of the interface are much larger than the superconducting coherence length , @xmath105 , i.e. @xmath106 , the summation over mode indices @xmath107 in eq .
( [ eq.conductance ] ) can be replaced by an integral .
the conductance for various values of chemical potential has been plotted in fig .
( [ fig.conductance ] ) as a function of energy .
note that the height of the resonance peak at zero energy is pinned at @xmath108 which indeed arises from the 2dms . system .
we set @xmath109 and @xmath110 for @xmath111 curves respectively.,width=188,height=122 ] now let us move to the next transport signature of 2dms , namely the fractional josephson current .
consider the geometry depicted in part(b ) of fig .
( [ fig.structure ] ) and assume that @xmath112 is the length of junction and @xmath113 is the phase difference between the two @xmath0 superconductors . in th short junction limit , @xmath114 ,
the andreev bound states ( abs ) responsible for carrying the super - current between two 3dds region are given by @xcite , @xmath115 where @xmath116 is the normal transmission probability of the junction , @xmath117 , with @xmath118 . despite that the hamiltonian in eq .
( [ eq.dbdg ] ) is invariant under @xmath119 phase shift , @xmath120 , the abs in eq.([eq.12 ] ) clearly has a @xmath2 period . using eq .
( [ eq.12 ] ) , the corresponding josephson current becomes , @xmath121 , where @xmath122 is the critical value of the josephson current . in the absence of perturbations which violate fermion parity conservation , such a form of fractional
josephson current is a signature of 2dms on the surface of 3dds @xcite . ) .
these states show @xmath2 periodicity with respect to @xmath123 which is an another direct signature of 2dms . ] to summarize we have shown that a 2dms can be obtained from singlet pairing as well , provided the superconducting order is pseudo - scalar , i.e. it changes sign under mirror reflection . in this extent ,
the present system is a singlet cousin in the family of odd - parity superconductors @xcite .
our scenario for majorana surface states does not require proximity to any magnet , as no triplet pairing is involved here .
perfect andreev reflection and fractional josephson current as two hallmarks of the ensuing majorana sea leave clear transport footprints .
the junction between such a superconductor and conventional @xmath124 and @xmath125 wave superconductors similar to the odd - parity p - wave superconductors may provide anomalous flux quantization in units of @xmath126 @xcite .
our analysis depends on the existence of the @xmath13 matrix which is a unique opportunity in even space - time dimensions .
therefore in 1 + 1 dimension as well , majorana bound states are expected from a competition between the dirac ( peierls ) mass @xmath1 and @xmath0 _ singlet _ superconductivity .
moreover , gauge transformation can place the odd - parity on the dirac mass .
therefore for a peculiar dirac hamiltonians where the mass term @xmath4 carries the minus sign needed in making of the odd - parity , majorana fermions can arise in @xmath127 interfaces where @xmath17 is a conventional s - wave superconductors .
_ acknowledgment : _ saj was supported by alexander von humboldt fellowship for experienced researchers .
99 b. bernevig and t. hughes , _ topological insulators and topological superconductors _ ( princeton university press , 2013 ) . x .- l
. qi and s .- c .
zhang , rev .
mod .
phys . * 83 * , 1057 ( 2011 ) .
s. shen , _ topological insulators : dirac equation in condensed matters _ ( springer london , limited , 2013 ) .
j. alicea , rep .
. phys . * 75 * , 076501 ( 2012 ) . c. w. j. beenakker , ann .
rev .
cond .
matt . * 4 * , 113 ( 2013 ) .
l. fu and c. l. kane , phys .
rev .
lett . * 100 * , 096407 ( 2008 ) .
l. fu and c. l. kane , phys .
rev .
b. * 79 * , 161408 ( 2009 ) .
j. d. sau , r. m. lutchyn , s. tewari , and s. das sarma , phys .
rev .
lett . * 104 * , 040502 ( 2010 ) .
y. oreg , g. refael , and f. von oppen , phys .
rev .
lett . * 105 * , 177002 ( 2010 ) .
a. romito , j. alicea , g. refael , and f. von oppen , phys .
rev .
b * 85 * , 020502 ( 2012 ) . m. stone and r. roy , phys . rev .
b * 69 * , 184511 ( 2004 ) .
y. fuseya , m. ogata , and h. fukuyama , phys .
rev .
lett . * 102 * , 066601 ( 2009 ) .
h. zhang , c .- x .
liu , x .-
l .
qi , x. dai , z. fang , and s .- c .
zhang , nat .
phys . * 5 * , 438 ( 2009 ) .
a. narayan , d. di sante , s. picozzi , and s. sanvito , phys .
rev .
lett . * 113 * , 256403 ( 2014 ) .
z. k. liu , b. zhou , y. zhang , z. j. wang , h. m. weng , d. prabhakaran , s .- k .
mo , z. x. shen , z. fang , x. dai , z. hussain , and y. l. chen , science * 343 * , 864 ( 2014 ) .
a. zee , _ quantum field theory in a nutshell : _ ( second edition ) ( princeton university press , 2010 ) .
k. capelle and e. k. u. gross , phys .
rev .
b. * 59 * , 7140 ( 1999 ) .
_ please see supplementary article .
_ a. p. schnyder , s. ryu , a. furusaki , and a. w. w. ludwig , phys .
rev .
b. * 78 * , 195125 ( 2008 ) .
c .- k .
chiu , j. c. y. teo , a. p. schnyder , and s. ryu , rev .
mod .
phys .
* 88 * , 035005 ( 2016 ) .
a. andreev , sov .
phys .
jetp * 19 * , 1228 ( 1964 ) .
m. salehi and s. a. jafari , ann . phys . * 359 * , 64 ( 2015 ) .
i. knez , r .-
r .
du , and g. sullivan , phys .
rev .
lett . * 109 * , 186603 ( 2012 ) . g. e. blonder , m. tinkham , and t. m. klapwijk , phys . rev .
b. * 25 * , 4515 ( 1982 ) .
l. fu and e. berg , phys .
rev .
lett . * 105 * , 097001 ( 2010 ) . * supplementary material * in this supplementary material we expand the details of calculations referenced in the main letter text . |
t dwarfs are low - temperature ( t@xmath5 @xmath6 13001500 k ) brown dwarfs exhibiting distinct absorption bands of ch@xmath3 in the near - infrared h- and k - bands @xcite .
they are distinguished from warmer l - type @xcite and m - type dwarfs in the near - infrared by the presence of these molecular features , in addition to significant h@xmath2o and collision - induced ( cia ) h@xmath2 absorption @xcite . since the discovery of the prototype of this class , gliese 229b @xcite , over 30 t dwarfs have been identified , most in the wide - field two micron all - sky survey ( * ? ? ?
* hereafter 2mass ) and sloan digital sky survey ( * ? ? ?
* hereafter sdss ) .
classification schemes for these objects based on near - infrared spectral morphology have been defined by @xcite and @xcite .
t dwarfs identified in wide - field surveys are all likely within 25 pc of the sun , the limiting distance for the nasa / nsf nstars project .
indeed , brown dwarfs in general may constitute a significant population in the solar neighborhood , roughly equal in number to the stellar population @xcite .
they have generally eluded detection , however , because of their intrinsic faintness . as the solar neighborhood is the basis for many studies of stellar atmospheres , star formation and evolution , initial mass function , and galactic structure , characterizing
the local substellar population is of some importance .
this is especially true at closer distances , such as the 10 pc recons sample @xcite , where the stellar census is far more complete , but few l- or t - type stars and brown dwarfs are currently known . in this article
, we present the discovery of a bright t dwarf , 2mass j15031961 + 2525196ss@xmath0ddmmss[.]s '' .
the suffix conforms to iau nomenclature convention and is the sexagesimal right ascension and declination at j2000 equinox . ]
( hereafter 2mass 1503 + 2525 ) , identified in a new search of the all - sky 2mass working point source database ( wpsd ) .
the spectrophotometric distance of this object places it within 10 pc of the sun . in @xmath72
we describe our new wide - field t dwarf search and compare it to previous search efforts . in @xmath73
we discuss the identification of 2mass 1503 + 2525 in our search and its subsequent confirmation as a t dwarf using the lick 3 m gemini infrared camera . in @xmath74
we derive the spectral type and spectrophotometric distance of 2mass 1503 + 2525 .
we discuss our results in @xmath75 .
our current program is a continuation of a previous 2mass search @xcite conducted when this survey was roughly 50% complete .
2mass is ideal for identifying t dwarfs , as the jhk@xmath8 filters it has employed sample the peak of the spectral energy distribution of cool brown dwarfs , enabling us to probe relatively deeply and over a large area of sky .
a total of 17 t dwarfs were detected in our earlier search in three separate samples : a primary sample covering 16,620 deg@xmath9 , requiring j @xmath10 16 , j
h @xmath10 0.3 , and h
k@xmath8 @xmath10 0.3 for candidates selected from the 2mass wpsd ; and two auxiliary samples selected from 2mass incremental release data with smaller search areas but somewhat different color criteria .
however , the overall stringent color cuts used in this search prevented our detecting early- and mid - type t dwarfs @xcite , while our small sample resulted in large statistical uncertainties for population statistics such as the t dwarf space density and substellar mass function ( burgasser 2001 ; burgasser et al . ; in preparation ) .
our current search effort , using the 2mass wpsd , attempts to address these shortcomings in two ways .
first , we have expanded our color criteria to j
h @xmath4 0.3 _ or _ h
k@xmath8 @xmath4 0 , allowing detection of objects with redder near - infrared colors ( and hence earlier spectral type ) , or faint objects with larger photometric uncertainties that were not identified in our first primary sample .
these color constraints are required to exclude an overwhelming number of background stars , as discussed below , and effectively limit our search to spectral types roughly t4 and later @xcite .
we continue to impose signal - to - noise ( s / n ) , photometric completeness , and reliability constraints by selecting only those candidates with j @xmath4 16 ( s / n @xmath11 10 completeness limit ) and both j- and h - band detections .
second , with the recent completion of 2mass survey operations , we are now able to search the entire sky for t dwarfs , excluding only those regions with very high source densities and hence likelihood of source confusion : the galactic plane ( @xmath12 ) , the large and small magellanic clouds , and various dense or obscured stellar regions such as 47 tuc and the orion nebular cluster .
we have also excluded the equatorial poles ( @xmath13 ) because of restrictions on observational followup . in total
, our current search encompasses 30,400 deg@xmath9 , or 74% of the sky , increasing our previous sample area by over 80% .
we expect to find roughly 15 new t dwarfs in this sample based on the increased areal coverage alone , and up to 10 additional red and/or faint t dwarfs based on the number of objects found in the @xcite auxiliary samples . in figure 1
we plot the j
k@xmath8 hess diagram @xcite of our 264,068 2mass - selected candidates , selected from over 1.3 billion sources in the 2mass wpsd , along with the 2mass colors of t dwarfs detected by 2mass and sdss ( table 7 in burgasser et al . , 2002 ; and table 4 in geballe et al . , 2002 ; all photometry from 2mass ) .
the density of candidates is highest near our color constraints , particularly at the early - type end of the @xcite dwarf track and near the turnoff between the dwarf ( late - k to early - m type ) and giant tracks .
the vast majority of these candidates are likely faint background stars , which are densely centered at and around the @xcite tracks . while our 2mass - selected candidates are required to have no optical counterparts within 5@xmath14 in the usno a2.0 catalog @xcite , based on the observed extreme red colors of known t dwarfs ( r j @xmath15 9 ; golimowski et al .
1998 ) , a substantial number of faint sources , proper motion stars , closely - separated visual binaries , and optically red sources ( the usno a2.0 catalog requires detections in both r and b bands ) remain .
hence , we visually examine multiple - epoch digital sky survey ( dss ) images of all candidate fields to exclude these contaminants . of the roughly 70,000 candidates
so far examined , 99.5% have been eliminated in this manner .
the remaining steps of our search process , as discussed in @xcite , include follow - up near - infrared imaging for the elimination of minor planet contaminants , optical imaging to exclude faint background stars , and near - infrared spectroscopy for final identification .
these follow - up observations are currently underway .
our color and magnitude constraints encompass 20 t dwarfs previously identified by 2mass and sdss , listed in table 18 @xcite it is obscured by its brighter primary in the 2mass imaging data .
distant t dwarfs , such as nttdf 1205@xmath160744 @xcite , ifa 0231@xmath16z1 @xcite , and sori 0538@xmath160236 @xcite are too faint to satisfy our magnitude constraints . ] .
this sample is already larger than that studied by @xcite .
furthermore , seven of these sources have multiple detections in the 2mass wpsd due to overlapping or repeated scans , as listed in table 2 .
these observations , separated by as much as 2 years , enable us to derive mean proper motions for six of the t dwarfs , assuming positional uncertainties of 0@xmath173 .
note that these values have not been corrected for solar motion .
four of the objects , 2mass 0937 + 2937 , sdss 1346@xmath160031 , gliese 570d , and 2mass 2339@xmath161352 , have motions of at least 0@xmath175 yr@xmath18 .
the substantial motions of these objects are not unexpected , given that t dwarfs detected by 2mass are all likely within 20 pc of the sun .
a comparison of the multiple - epoch photometry for the objects listed in table 2 gives some hint of photometric variability .
reasonably significant ( maximum @xmath19 m @xmath15 2@xmath20 ) variations are seen at j - band for sdss 0926 + 5847 ( 2.8@xmath20 ) , 2mass 1237 + 6526 ( 2.1@xmath20 ) , gliese 570d ( 2.8@xmath20 ) , and 2mass 2339 + 1352 ( 3.6@xmath20 ) ; and at h - band for sdss 1346@xmath160031 ( 2.3@xmath20 ) .
it is possible that these photometric variations are intrinsic , as has now been observed for a number of brown dwarfs ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* enoch , brown , & burgasser , in preparation ) .
@xcite have hypothesized that these variations arise from the motions and evolution of patchy condensate clouds in the upper atmospheres of early- and mid - type t dwarfs ; indeed , the variations appear to be greatest at j - band , as predicted , and possibly more pronounced in the t4.5 sdss 0924 + 5847 and t5.5 2mass 2339 + 1352 ( note the equally strong variation in the t8 gliese 570d , however . )
such low - level photometric variability as measured by the 2mass j - band filter must be considered carefully , however , as the filter bandpass extends well into the 1.4 @xmath1 telluric h@xmath2o band .
this results in substantial variations in the j - band zeropoint ( up to 0.1 mag ) from night - to - night and within a single night ( cutri et al . 2002 ) .
the source of these fluctuations , likely changes in observing conditions ( terrestrial weather and water vapor ) , can potentially affect the photometric calibration of late - l and t dwarfs , as their spectral energy distributions show far more structure in the j - band than earlier - type calibration stars .
such second - order extinction effects could therefore be responsible for the apparent variability seen , and a more directed program is required to validate these observations .
regardless , our complete sample will provide time - resolved astrometric and potentially photometric data for a few individual t dwarfs , in addition to improved population statistics .
2mass 1503 + 2525 is one of the first objects examined in our new wide - field search campaign .
astrometry and photometry for this object are listed in table 1 , while figure 2 shows the 2mass 1503 + 2525 field as imaged by poss - i @xcite , poss - ii @xcite , and 2mass .
this object is one of our brightest candidates , just fainter than the brightest known t dwarf 2mass 0559@xmath161404 @xcite .
the absence of an optical source at the position of 2mass 1503 + 2525 implies r@xmath16j @xmath15 6 @xcite , while its near - infrared colors ( j h = [email protected] , h
k@xmath8 = @[email protected] ) are similar to the t6.5 gliese 229b @xcite .
the optical data also rule out the possibility that this candidate is an earlier - type high - velocity proper motion star , as no such counterpart is seen at either epoch .
note that the bright star 18@xmath14 north and 103@xmath14 east of 2mass 1503 + 2525 is the m5 variable star ex boo ( a.k.a .
iras 15012 + 2537 , bd+25 2864 , hip 73662 ) , classified as a normal giant / supergiant by @xcite . a lower limit on this object s hipparcos distance , d @xmath21 680 pc @xcite , eliminates the possibility that it is associated with the t dwarf .
we imaged 2mass 1503 + 2525 in the near - infrared on 27 june 2002 ( ut ) using the gemini twin - arrays infrared camera @xcite , mounted on the lick observatory shane 3 m telescope .
light cirrus was present and seeing was 1@xmath172 .
short integration ( 20 sec ) , j- and k - band images unambiguously confirmed the presence of 2mass 1503 + 2525 at its 2mass position , and no secondary component was seen at the seeing spatial resolution of the data .
we subsequently obtained low - resolution ( @xmath22 @xmath11 500 ) jhk spectra using the gemini camera s grisms on two separate nights , 28 june 2002 and 19 august 2002 ( ut ) .
conditions for the first night ( j- and k - band spectra ) were affected by heavy cirrus and 2@xmath14 seeing , while smoke and light cirrus ( 1@xmath175 seeing ) were present on the second night ( h - band spectrum ) . for both sets of observations ,
2mass 1503 + 2525 was first acquired in imaging mode and placed into a 1@xmath174 slit .
sets of nodded ( 30@xmath14 ) pairs were obtained with individual integration times of 180300 sec , for total integrations of 1800 sec at j- and k - bands , and 2000 sec at h - band .
nod pair sets of the f8 v hd 133460 were obtained for calibration immediately after the target observations and at similar airmass .
spectral images were pairwise subtracted ; divided by a normalized , median - combined set of flat - field quartz lamp observations ; and corrected for bad pixels by linear interpolation .
spatial curvature was measured using the calibrator star dispersion lines , and spectral curvature was measured from argon lamp spectra obtained during each of the observing runs .
these lamp spectra were also used for wavelength calibration , using the identifications of @xcite .
target and calibrator spectra at each nod position were then individually extracted and interpolated onto a common wavelength scale .
relative flux calibrations and telluric corrections were made by dividing target spectra by calibrator spectra at each nod position , interpolating over paschen and brackett hydrogen lines .
the resulting individual spectra for each band were multiplied by a 6095 k blackbody ( * ? ? ?
* corresponding to spectral type f8 v ) and averaged .
2mass magnitudes were used to flux calibrate the separate bands by the prescription of @xcite .
figure 3 plots the reduced gemini camera grism spectrum of 2mass 1505 + 2525 .
signal - to - noise ratios are roughly 17 at j , 13 at h , and 9 at k. ch@xmath3 bands at 1.15 , 1.4 , and 2.2 @xmath1 , and h@xmath2o bands at 1.1 , 1.3 , and 1.9 @xmath1 are clearly evident , as are the 1.25 @xmath1 k i doublet lines .
the 1.17 @xmath1 k i doublet lines are obscured by noise at the bottom of the 1.1 @xmath1 h@xmath2o and ch@xmath3 absorption trough .
cia h@xmath2 is responsible for the slope on the red side of the k - band peak .
these features are characteristic of mid- and late - type t dwarfs @xcite , and confirm 2mass 1503 + 2525 as a bona - fide t dwarf .
we derived the spectral type of 2mass 1503 + 2525 using the independent classification schemes of @xcite and @xcite .
spectral indices for both schemes sample the major h@xmath2o and ch@xmath3 bands in t dwarf spectra , while the scheme of burgasser et al
. also includes color indices between the various spectral peaks .
measured ratio values and individual subtype determinations are listed in table 3 , following the procedures outlined by these two classification schemes .
the substantial scatter amongst the geballe et al .
indices is caused by sampling low signal - to - noise regions in the gemini camera spectrum ; similarly , the 2.2@xmath1 ch@xmath3 , k / j 2.11@xmath1/2.07@xmath1 indices in the burgasser et al.scheme are affected by the poor signal - to - noise in the k - band spectrum .
nonetheless , both yield a common spectral type of t5.5 .
we confirm this spectral type in figure 4 by comparing the j- and h - band spectra of 2mass 1503 + 2525 to ukirt cgs4 @xcite spectral data of the t5 2mass 0559@xmath161404 @xcite and the t6 sdss 1624 + 0029 @xcite .
adequate comparison can be made between these data , as their spectral resolutions are similar ( @xmath22 @xmath11 400500 ) and all have been corrected for telluric absorption and flux calibrated .
2mass 1503 + 2525 has stronger h@xmath2o and ch@xmath3 absorption bands than 2mass 0559@xmath161404 , but weaker bands than sdss 1624 + 0029 , consistent with a median type of t5.5 . as noted above ,
the colors of 2mass 1503 + 2525 are quite blue , for its assigned spectral type of t5.5 , suggestive of either a later spectral type or suppression of k - band flux , as is observed in the possibly high surface gravity or metal - poor t dwarf 2mass 0937 + 2931 @xcite .
however , within the photometric uncertainties these colors are still consistent with other mid - type t dwarfs such as the t5 2mass 0755 + 2212 ( j h = [email protected] , h
k@xmath8 = @[email protected] ) and the t5.5 2mass
1534@xmath162952ab ( j h = [email protected] , h
k@xmath8 = [email protected] ; burgasser et al . 2002 , 2003 ) .
higher accuracy photometry of a sample of mid - type t dwarfs is required to determine if the colors of 2mass 1503 + 2525 are unusually blue for its spectral type .
both of the comparison objects in figure 4 have known distances from trigonometric parallax measurements @xcite .
therefore , we can estimate the spectrophotometric distance of 2mass 1503 + 2525 from the absolute 2mass magnitudes of these objects and data from table 3 . if we assume this object has a spectral type and hence luminosity similar to 2mass 0559@xmath161404
, it lies between 10.712.0 pc from the sun .
this estimate may be inflated if 2mass 0559@xmath161404 is a multiple system , as suggested by @xcite . on the other hand ,
if 2mass 1503 + 2525 is a later - type , less luminous t dwarf similar to sdss 1624 + 0029 , it lies only 4.95.4 pc from the sun .
the mean spectrophotometric distance between these extremes is 8@xmath03 pc ; hence , if it is single , 2mass 1503 + 2525 is likely within the 10 pc recons distance , and is potentially a member of the 8 pc sample of @xcite .
trigonometric parallax observations are currently underway for 2mass 1503 + 2525 through the usno parallax program ( c. dahn , priv . comm . ) . given its brightness , it is possible that 2mass 1503 + 2525 could be a binary brown dwarf , as roughly 20% of late - m , l , and t dwarfs identified in magnitude - limited searches are found to have binary companions with separations @xmath23 10 au @xcite .
the presence of a companion would increase our distance estimate by up to a factor of 1.4 .
gemini images rule out a bright companion beyond 1@xmath172 , or 613 au assuming on a distance of 511 pc . because the apparent separation limit of brown dwarf binaries is similar to this observational limit @xcite ,
we can not rule out the multiplicity of 2mass 1503 + 2525 with any confidence ; higher - resolution imaging ( ao or space - based ) or radial velocity monitoring are required .
the possibility that 2mass 1503 + 2525 lies within 10 pc of the sun raises the question of how many late - type dwarfs are currently known in this volume .
table 4 lists all l and t dwarfs identified to date which are or may potentially be within the recons 10 pc horizon , based on trigonometric @xcite or spectrophotometric parallax , respectively .
we have used published spectrophotometric parallaxes for the l dwarfs @xcite or computed them using the spectral type / absolute magnitude relations of @xcite , 2mass photometry , and spectral classifications by @xcite .
we derived distances for the t dwarfs by interpolating over spectral type ( using the burgasser et al .
scheme ) between the 2mass j , h , and k@xmath8 absolute magnitudes of the l8 gliese 584c @xcite , t2 sdss 1254@xmath160122 @xcite , t5 2mass 0559@xmath161404 @xcite , t6 sdss 1624 + 0029 @xcite , t6.5 gliese 229b ( * ? ? ? * ; * ? ? ?
* ukirt jhk photometry ) , and t8 gliese 570d @xcite .
the average of spectrophotometric distances derived from 2mass jhk@xmath8 photometry are reported . while there are currently 313 known stars spectral type m or earlier in 227 systems within 10 pc of the sun ( t. henry , priv.comm .
) , only 5 l and 10 t dwarfs meet the same criteria , if we assume that the spectrophotometric distances are accurate .
in contrast , mass function simulations by @xcite , assuming conservatively @xmath24(m ) @xmath25 dn / dm @xmath26 m@xmath27 for 0.01 m@xmath28 m @xmath29 m@xmath30(m ) @xmath26 m@xmath18 , and @xcite find @xmath24(m ) @xmath26 m@xmath31 , over the same mass range .
] , predict 12 l dwarfs ( 1300 @xmath6 t@xmath5 @xmath6 2200 k ) and 21 t dwarfs ( 800 @xmath6 t@xmath5 @xmath6 1300 k ) in the same spatial volume , more than twice the number so far identified .
the dearth of observed l and t dwarfs worsens dramatically if the substellar mass function is steeper .
the primary reason for this discrepancy is quite simple - searches for these objects using large - scale surveys have not yet examined the entire sky .
sdss and denis programs have only searched a fraction of their final survey area , which ultimately will cover one - quarter and one - half of the sky , respectively .
searches for l and t dwarfs in the all - sky 2mass survey have only examined roughly one - half of the sky .
hence , the expanded areal coverage and color criteria of our current search will likely identify quite a few of the missing nearby t dwarfs , particularly those with spectral types later than t4 .
however , some of the latest - type t dwarfs within 10 pc , and objects significantly less luminous than gliese 570d , may be too faint to be detected by 2mass , sdss , or denis , and will remain to be uncovered in future surveys .
2mass 1503 + 2525 is a key t dwarf discovery not just because of its potential proximity to the sun , but also because it is sufficiently bright to permit both higher - resolution spectroscopic study and investigations outside the 12.5 @xmath1 spectral energy window . in particular
, this object is a prime sirtf target for mid - infrared observations , a spectral region containing diagnostics of temperature ( e.g. , nh@xmath32 at 10.5 @xmath1 ) and atmospheric composition ( e.g. , 69 @xmath1 silicate bands ) that is largely inaccessible due to the intrinsic faintness of these cool brown dwarfs .
we thank our referee sandy leggett for her useful criticisms of our manuscript and for providing the ukirt spectral data of 2mass 0559 - 1404 and sdss 1624 + 0029 used in our analysis .
we acknowledge useful discussions on the nearby star sample with kelle cruz and todd henry .
we also thank the gemini camera instrument specialist elinor gates and telescope operator andy tullis at lick for assistance in the observations , and the uco tac for its allocation of time for this project .
adam j. b. acknowledges support by nasa through hubble fellowship grant hst - hf-01137.01 awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , inc . , for nasa , under contract nas 5 - 26555 . this research has made use of the simbad database , operated at cds , strasbourg , france .
poss - i and poss - ii images were obtained from the digitized sky survey image server maintained by the canadian astronomy data centre , which is operated by the herzberg institute of astrophysics , national research council of canada .
this research has made use of the nasa / ipac infrared science archive , which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
this publication makes use of data from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center , funded by the national aeronautics and space administration and the national science foundation .
adam j. b. dedicates this publication to the memory of dr .
charles kincaid witham .
llcccc 2mass j02431371@xmath162453298 & t6 & [email protected] & [email protected] & [email protected] & 1 + 2mass j04151954@xmath160935066 & t8 & [email protected] & [email protected] & [email protected] & 1 + 2mass j05591914@xmath161404488 & t5 & [email protected] & [email protected] & [email protected] & 2 + 2mass j07271824 + 1710012 & t7 & [email protected] & [email protected] & [email protected] & 1 + 2mass j07554795 + 2212169 & t5 : & [email protected] & [email protected] & [email protected] & 1 + sdssp j092615.38 + 584720.9 & t4.5 & [email protected] & [email protected] & [email protected] & 3 + 2mass j09373487 + 2931409 & t6p & [email protected] & [email protected] & [email protected] & 1 + 2mass j10475385 + 2124234 & t6.5 & [email protected] & [email protected] & @xmath21 16.4 & 4 + 2mass j12171110@xmath160311131 & t7.5 & [email protected] & [email protected] & @xmath21 15.9 & 4 + 2mass j12255432@xmath162739466ab & t6 & [email protected] & [email protected] & [email protected] & 4,5 + 2mass j12373919 + 6526148 & t6.5 & [email protected] & [email protected] & @xmath21 15.9 & 4 + sdssp [email protected] & t6 & [email protected] & [email protected] & [email protected] & 6 + gliese 570d & t8 & [email protected] & [email protected] & [email protected] & 7 + 2mass j15031961 + 2525196 & t5.5 & [email protected] & [email protected] & [email protected] & 8 + 2mass j15344984@xmath162952274ab & t5.5 & [email protected] & [email protected] & [email protected] & 1,5 + 2mass j15462718@xmath163325111 & t5.5 & [email protected] & [email protected] & [email protected] & 1 + 2mass j15530228 + 1532369 & t7 & [email protected] & [email protected] & [email protected] & 1 + sdssp j162414.37 + 002915.6 & t6 & [email protected] & [email protected] & @xmath21 15.5 & 9 + 2mass j22541892 + 3123498 & t5 & [email protected] & [email protected] & [email protected] & 1 + 2mass j23391025 + 1352284 & t5.5 & [email protected] & [email protected] & @xmath21 15.5 & 1 + 2mass j23565477@xmath161553111 & t6 & [email protected] & [email protected] & [email protected] & 1 + lccccccccc sdss 0926 + 5847 & t4.5 & 990422 & 09@xmath3426@xmath3515@xmath3639 & + 58@xmath3747@xmath3821@xmath1719 & [email protected] & [email protected] & [email protected] & 0@xmath173@xmath00@xmath173 & 220@xmath040 + & & 990422 & 09@xmath3426@xmath3515@xmath3643 & + 58@xmath3747@xmath3821@xmath1734 & [email protected] & [email protected] & @xmath21 16.0 & & + & & 000320 & 09@xmath3426@xmath3515@xmath3640 & + 58@xmath3747@xmath3820@xmath1790 & [email protected] & [email protected] & [email protected] & & + & & 000320 & 09@xmath3426@xmath3515@xmath3638 & + 58@xmath3747@xmath3821@xmath1723 & [email protected] & [email protected] & [email protected] & & + 2mass 0937 + 2931 & t6p & 981124 & 09@xmath3437@xmath3534@xmath3681 & + 29@xmath3731@xmath3842@xmath1766 & [email protected] & [email protected] & [email protected] & 1@[email protected] & 150@xmath04 + & & 000402 & 09@xmath3437@xmath3534@xmath3688 & + 29@xmath3731@xmath3840@xmath1799 & [email protected] & [email protected] & [email protected] & & + 2mass 1237 + 6526 & t6.5 & 990313 & 12@xmath3437@xmath3539@xmath3618 & + 65@xmath3726@xmath3814@xmath1759 & [email protected] & [email protected] & @xmath21 15.9 & & + & & 990313 & 12@xmath3437@xmath3539@xmath3620 & + 65@xmath3726@xmath3814@xmath1781 & [email protected] & [email protected] & @xmath21 16.1 & & + sdss 1346@xmath160031 & t6 & 990122 & 13@xmath3446@xmath3546@xmath3643 & @xmath1600@xmath3731@xmath3850@xmath1743 & [email protected] & [email protected] & @xmath21 15.7 & 0@xmath175@xmath00@xmath173 & 275@xmath015 + & & 000604 & 13@xmath3446@xmath3546@xmath3643 & @xmath1600@xmath3731@xmath3850@xmath1733 & [email protected] & [email protected] & [email protected] & & + & & 010203 & 13@xmath3446@xmath3546@xmath3634 & @xmath1600@xmath3731@xmath3850@xmath1713 & [email protected] & [email protected] & [email protected] & & + & & 010208 & 13@xmath3446@xmath3546@xmath3638 & @xmath1600@xmath3731@xmath3850@xmath1755 & [email protected] & [email protected] & @xmath21 15.6 & & + gliese 570d & t8 & 980516 & 14@xmath3457@xmath3514@xmath3696 & @xmath1621@xmath3721@xmath3847@xmath1775 & [email protected] & [email protected] & [email protected] & 1@xmath179@xmath00@xmath172 & 149@xmath03 + & & 990729 & 14@xmath3457@xmath3515@xmath3604 & @xmath1621@xmath3721@xmath3849@xmath1772 & [email protected] & [email protected] & [email protected] & & + 2mass 1534@xmath162952ab & t5.5 & 980702 & 15@xmath3434@xmath3549@xmath3684 & @xmath1629@xmath3752@xmath3827@xmath1740 & [email protected] & [email protected] & [email protected] & 0@xmath1731@xmath00@xmath1712 & 137@xmath011 + & & 980708 & 15@xmath3434@xmath3549@xmath3684 & @xmath1629@xmath3752@xmath3827@xmath1742 & [email protected] & [email protected] & [email protected] & & + & & 000430 & 15@xmath3434@xmath3549@xmath3687 & @xmath1629@xmath3752@xmath3827@xmath1782 & [email protected] & [email protected] & [email protected] & & + 2mass 2339 + 1352 & t5.5 & 980929 & 23@xmath3439@xmath3510@xmath3621 & + 13@xmath3752@xmath3830@xmath1720 & [email protected] & [email protected] & @xmath21 15.5 & 0@xmath1783@xmath00@xmath1711 & 159@xmath011 + & & 001029 & 23@xmath3439@xmath3510@xmath3623 & + 13@xmath3752@xmath3828@xmath1756 & [email protected] & [email protected] & @xmath21 15.6 & & + & & 001110 & 23@xmath3439@xmath3510@xmath3628 & + 13@xmath3752@xmath3828@xmath1764 & [email protected] & [email protected] & @xmath21 15.8 & & + & & 001129 & 23@xmath3439@xmath3510@xmath3625 & + 13@xmath3752@xmath3828@xmath1748 & [email protected] & [email protected] & [email protected] & & + lcc
1.1@xmath1 h@xmath2o & 0.274 ( 56 ) & 4.62 ( 6 ) + 1.3@xmath1 ch@xmath3 & 0.784 ( 5 ) & + 1.5@xmath1 h@xmath2o & 0.432 ( 5 ) & 3.28 ( 3 ) + 1.6@xmath1 ch@xmath3 & 0.364 ( 56 ) & 2.60 ( 6 ) + 2.2@xmath1 ch@xmath3 & 0.094
( 7 ) & 10.8 ( 7 ) + h / j & 0.383 ( 5 ) & + k / j & 0.131 ( 8) & + 2.11@xmath1/2.07@xmath1 & 1.12 ( ) & + final spectral type & t5.5 & t5.5 + llccccc 2mass 0036 + 1821 & l3.5 & [email protected] & [email protected] & [email protected] & [email protected] & 1 + 2mass 0835@xmath160819 & l4.5 & [email protected] & [email protected] & [email protected] & _ 9.1 _ & 2 + 2mass 1507@xmath161627 & l5 & [email protected] & [email protected] & [email protected] & [email protected] & 1 + gj 1001b & l5 & [email protected] & [email protected] & [email protected] & [email protected] & 3 + denis 0255@xmath164700 & l8 & [email protected] & [email protected] & [email protected] & _ 5.0 _ & 4 + sdss 0423@xmath160414 & t0 & [email protected] & [email protected] & [email protected] & _ 8.9 _ & 5 + 2mass 1503 + 2525 & t5.5 & [email protected] & [email protected] & [email protected] & _ 7.8 _ & 5 + 2mass 0937 + 2931 & t6p & [email protected] & [email protected] & [email protected] & _ 8.8 _ & 5 + 2mass 0243@xmath162453 & t6 & [email protected] & [email protected] & [email protected] & _ 9.8 _ & 5 + gliese 229b & t6.5 & [email protected] & [email protected] & [email protected] & [email protected] & 3 + 2mass 0727 + 1710 & t7 & [email protected] & [email protected] & [email protected] & _ 9.1 _ & 5 + 2mass 1553 + 1532 & t7 & [email protected] & [email protected] & [email protected] & _ 9.6 _ & 5 + 2mass 1217@xmath160311 & t7.5 & [email protected] & [email protected] & @xmath21 15.9 & _ 8.6 _ & 5 + gliese 570d & t8 & [email protected] & [email protected] & [email protected] & [email protected] & 3 + 2mass 0415@xmath160935 & t8 & [email protected] & [email protected] & [email protected] & _ 6.7 _ & 5 + | we present the discovery of a bright ( j = [email protected] ) t dwarf , 2mass 1503 + 2525 , identified in a new , wide - field search for t dwarfs using the recently completed two micron all sky survey ( 2mass ) .
the 12.5 @xmath1 spectrum of this object exhibits the strong h@xmath2o and ch@xmath3 bands characteristic of mid- and late - type t dwarfs , and we derive a spectral type of t5.5 using both the burgasser et al . and the geballe et al .
classification schemes . based on its spectral type and the absolute magnitudes of known t dwarfs
, we estimate the distance of this object as 8@xmath03 pc if it is single , likely within 10 pc of the sun .
our new 2mass search , which covers 74% of the sky and greatly expands on earlier color constraints , should identify 1525 new t dwarfs with j @xmath4 16 . combined with the 20 known members of this class that already fall within our search criteria , our new sample will provide improved statistics for such key quantities as the binary fraction and the field substellar mass function .
furthermore , multiple detections from overlapping 2mass scans provide multiple epoch astrometry and photometry , and we present proper motions for five t dwarfs in our sample . |
mesoscopic systems are characterized by spatial dimensions smaller than the phase breaking length @xmath4 so that the phase of an electron is not destroyed by inelastic processes . along their optical paths , electrons pick up phases from propagation and scattering . even for negligible dephasing , certain circumstances can lead to an _ averaging _ of these phases , a phenomenon which has been analyzed in connection with localization , see e.g. refs . .
it remains however a fundamental task to compare in detail the effects of phase averaging and dephasing . for a noninteracting model system
it was recently shown that a single one - channel dephasing probe gives the same full counting statistics as a single phase - averaging probe.@xcite however , it was noted that this result does no longer hold for several probes , i.e. phase averaging over many independent phases seems to be fundamentally different from dephasing . in the present paper we address this issue in detail .
we discuss a generic one - dimensional interacting system coupled to two reservoirs with few barriers at arbitrary but fixed positions .
we show that for incommensurate barrier positions an independent average over several phases can be induced by finite temperature . for more than two barriers
we find that phase averaging drastically differs from dephasing and show that the commonly made assumption that barriers with mutual distances larger than the thermal length @xmath5 can be treated independently @xcite ( note that in experiment contacts to leads constitute additional barriers ) does not hold in the absence of dephasing .
this means that even in the high - temperature regime interference effects are still important , as also shown recently for quantum dots in the sequential tunneling regime.@xcite for three strong barriers with individual conductances @xmath0 , we obtain @xmath6 for the total conductance , in drastic contrast to the addition of resistances which follows from dephasing . for an interacting quantum wire ( qw )
this implies power - law scaling of @xmath3 with novel exponents which can be used as an experimental fingerprint for phase averaging . in section [ sec : noninteracting ]
we introduce a noninteracting model for a quantum wire with barriers which allows a simple discussion of temperature induced phase averaging . in section [ sec : interacting ] we study the influence of temperature induced phase averaging on the scaling beghaviour of interacting quantum wires and propose an experimental setup to distinguish phase averaging from dephasing . while sections [ sec : noninteracting ] and [ sec : interacting ] focus primarily on a wire with three barriers , section [ sec:4barrier ] treats phase averaging for a wire with four barriers .
for a discussion of the basic physical idea consider first the noninteracting case .
we model the qw coupled to leads by an infinite tight - binding chain for spinless electrons at half - filling ( @xmath7 ) @xmath8 with hopping matrix elements equal to 1 ( defining the energy unit ) except for @xmath9 , @xmath10 , where the barriers are situated . without barriers the dispersion relation is @xmath11 with band width @xmath12 , where @xmath13 is the momentum and @xmath14 the lattice spacing .
the distance between subsequent barriers is @xmath15 with @xmath16 . we characterize the barriers by their transmission @xmath17 and reflection @xmath18 follows from @xcite @xmath19 where @xmath20 , and @xmath21 is the transmission amplitude .
thus , the effect of temperature is an average of the transmission probability @xmath22 over an energy range @xmath23 .
this is fundamentally different from dephasing which destroys the phase information for each individual path of the electron .
if one takes account of dephasing by summing up the classical probabilities of all paths , one obtains @xmath24 , with @xmath25.@xcite since @xmath26 is proportional to the resistance of a barrier , this gives the classical law of adding up resistances in series .
in contrast , we now show that temperature induced energy averaging leads to completely different results .
since @xmath27 and @xmath28 vary slowly with energy on the scale @xmath29 , temperature averages only over the rapidly varying phases an electron acquires by bouncing back and forth between two subsequent barriers @xmath30 . as @xmath31 depends weakly on energy ,
the change of @xmath32 over the energy range @xmath23 near the fermi energy is given by @xmath33 .
thus , for @xmath34 , @xmath35 is roughly given by @xmath36 .
for all @xmath37 and incommensurate length @xmath38 ( i.e. , @xmath39 is only valid for large coprime integers @xmath40 ) , this implies that the average over energy will cover a representative part of the multidimensional phase space defined by @xmath41 ( see fig .
[ fig : phase_averaging ] ) . in this case , temperature induced energy averaging is equivalent to an _ independent _ average over all phases @xmath32 , and we can replace @xmath42 in eq . by the phase - averaged quantity @xmath43
provided that @xmath44 .
note that this result does not hold for commensurate lengths @xmath38 ( where the path in phase space closes very quickly ; see below ) and becomes less relevant for a large number @xmath45 of barriers ( since @xmath46 has to be chosen too large in order to cover a representative part of the high dimensional phase space ) .
as function of the phases @xmath47 for a noninteracting qw with three barriers @xmath48 . lines : path which the integral in eq .
traces in the @xmath49-plane for @xmath50 , @xmath51 .
the grayscale of the line indicates the relative amplitude of @xmath52 for @xmath53 , the same grayscale applies as for the density plot .
for sufficiently high temperature the line covers a representative part of the plane . ]
the phase average can be calculated easily for a small number of barriers . for two barriers we use @xmath54 and for three barriers ( forming two dots which we refer to as left and right ) @xmath55^{-1}\ , , \label{eq : t}\end{aligned}\ ] ] with @xmath56 . by accident ,
phase averaging yields the addition of resistances for two barriers .
in contrast , for three barriers we obtain @xmath57 with @xmath58 . for weak barriers , @xmath59 ,
this gives @xmath60 leading to exponentially small transmission when scaling it up to many barriers , in analogy to ref . .
for strong barriers , @xmath61 , we obtain the surprising result @xmath62 leading to @xmath1 when inserted in eq . .
[ fig : nonint_incommens ] shows that the phase - averaged @xmath2 through three barriers at incommensurate positions agrees precisely with the exact one for sufficiently large @xmath63 , whereas summing up the individual resistances is incorrect for all @xmath63 . in the inset of fig .
[ fig : nonint_incommens ] we present the energy dependence of the total transmission probability .
it forms energetically large superstructures , as the height and area of the peaks are very sensitive to the mutual distance of the resonance positions of the left and right dot .
those are roughly given by even ( odd ) multiples of @xmath64 for odd ( even ) @xmath65 .
thus , if @xmath66 is valid only for large coprime integers @xmath67 ( incommensurate case @xcite ) , the mutual energetic distance of left- and right - dot resonance peaks of @xmath22 shifts slowly along the energy axis , leading to a modulation of the peak height and area .
this provides another picture for temperature induced energy averaging : if @xmath63 is large enough to average over a sufficient part of a superstructure , it averages over many mutual distances between left- and right - dot resonances , which is equivalent to phase averaging . and incommensurate dot length , @xmath68 , @xmath69 .
the solid line corresponds to eq . .
for higher @xmath63 , the dashed line which results from using the phase averaged transmission given by eq . in eq .
matches perfectly .
summing up the individual resistances ( dotted line ) yields an incorrect result .
_ inset : _ @xmath22 as a function of energy .
the changing mutual energetic distance of resonances of the left and right dot leads to energetically large superstructures of the total transmission probability . ] and commensurate dot length , @xmath70 , @xmath71 , hence @xmath72 .
the solid line results from eq . .
for temperatures greater than the distance of coinciding peaks ( compare inset ) , the dashed line corresponding to eq . is a valid approximation .
summing up the individual resistances ( dotted line ) yields an incorrect result .
_ inset : _ @xmath22 as function of energy .
labels indicate well - separated left - dot ( l ) and right - dot ( r ) resonances as well as coinciding ( c ) ones .
the distance of the coinciding resonances is @xmath73 . ] in contrast , for dots of commensurate length , i.e. , when @xmath66 is fulfilled for small coprime integers @xmath67 , every @xmath74th left - dot and @xmath75th right - dot resonance coincide , whereas the other resonances are well separated in energy ( see inset in fig .
[ fig : nonint_commens ] ) .
it can be shown that coinciding resonances ( avoided crossings ) give rise to two peaks of the total transmission probability , each one with area @xmath76^{-1 } = { \mathcal o } \left ( \tau^2 \right)$ ] .
note that while @xmath77 is proportional to the distance of the two peaks , it does not influence their area .
the well - separated peaks can be neglected when calculating the total conductance through the qw because they have a much smaller area of @xmath78 than the coinciding peaks and they are not in great majority ( as @xmath67 are small ) . for @xmath79 , the energy - integral in eq .
averages over the large coinciding peaks . for the total conductance it follows ( see fig . [
fig : nonint_commens ] ) @xmath80 with @xmath81 .
since the smaller peaks have been neglected , this formula slightly underestimates @xmath2 , but becomes more and more accurate for stronger barriers .
eq . in obviously inconsistent with summing up the single resistances , and
the total conductance is independent of the strength of the barrier in the middle .
we now turn to the more realistic case of an interacting qw ( luttinger liquid , ll ) on sites @xmath82 by adding a short ranged interaction @xmath83 with @xmath84 .
abrupt contacts to the leads follow if the interaction is chosen to be the same on all sites , @xmath85 . for smooth contacts
@xmath86 rises slowly from zero at the leads to its full value within the wire over roughly @xmath87 sites .
the interaction is treated using the truncated functional renormalization group.@xcite this approach leads to a real and frequency independent self - energy , which then serves as an effective single - particle potential so that eq .
is still applicable .
inelastic processes mediated by the interaction are neglected as they would just contribute subleading corrections .
a single strong impurity in a ll generates a slowly decaying oscillatory effective potential of range @xmath88 which makes the local density of states in the vicinity of a large impurity scale as @xmath89.@xcite the boundary exponent @xmath90 can be computed from the ll parameter @xmath91 which in turn is known from bethe ansatz @xcite , @xmath92 at half - filling . for @xmath93 , as taken in the following , our approach is known to produce the exponent @xmath94 in good agreement with @xmath95.@xcite if a strong barrier is placed within the interacting part of the qw , the density of states scales with @xmath63 on both sides of the barrier yielding a power - law @xmath96 .
if a strong barrier separates an interacting part of the wire from a noninteracting part as do tunneling barriers to the leads , @xmath97 .
we are interested in possible power laws if three barriers of mixed types are present . given three barriers in the wire , we find that the effective potential formed at one barrier is not affected by the presence of the other barriers , as long as @xmath98 .
consequently , the scaling law of the transmission probabilities of the individual barriers @xmath99 ( with @xmath100 or @xmath101 depending on the type of the barrier ) is not affected by the presence of the other barriers and thus @xmath102 .
interpreting a barrier and its surrounding oscillations as an effective barrier , we can use eq . and the picture of two dots in series developed for the noninteracting case .
thus , for incommensurate @xmath38 and sufficiently large @xmath63 , phase averaging as given by eq . can as well be applied to an interacting wire ( see fig . [
fig : int_incommens ] ) .
( lines ) and position of resonance energies of @xmath103 ( circles ) for @xmath93 and barriers at the two contacts and a barrier within the wire , @xmath104 , @xmath105 , @xmath106 .
the solid line corresponds to eq . .
for higher temperatures , the dashed line which results from using the phase averaged transmission given by eq . in eq .
matches perfectly . as left - dot resonances ( open circles )
are shifted a different amount in energy with changing temperature than right - dot resonances ( filled circles ) , an avoided crossing can occur ( arrow ) leading to a drastically enhanced conductance if temperature is low . ]
this implies power - law scaling of @xmath107 where the exponent depends on _ all _ exponents of the individual barriers @xmath108 therefore , combinations of different barriers will produce exponents equal to different integer multiples of @xmath109 . for three barriers within the interacting part of the wire ( @xmath110 ) , eq .
yields @xmath111 .
if one of the three barriers is a contact to a noninteracting lead , it follows @xmath112 .
two contacts and one barrier within the interacting part imply @xmath96 ( see fig .
[ fig : int_incommens_exp ] ) . for @xmath93 and three barriers , @xmath104 , @xmath113 , @xmath114 .
( a ) all three barriers within the wire .
( b ) one barrier at end - contact to lead , two barriers within the wire .
( c ) two barriers at end - contacts , one barrier within the wire .
( d ) one barrier at bulk - contact , one barrier at end - contact , one barrier within the wire .
( e ) two barriers at bulk - contacts , one barrier within the wire .
the solid lines correspond to eq .
, the dashed lines result from using the phase averaged transmission given by eq . in eq . .
the different levels of the exponents ( multiples of @xmath115 ) are clearly visible . in the bulk - contacted cases ( d ) and
( e ) , phase averaging is not as good an approximation as in the other cases . ] in addition , in fig .
[ fig : int_incommens_exp ] we show the effect of bulk contacts ( in contrast to end contacts discussed so far ) .
also for this setup @xmath3 at sufficiently large @xmath63 is given approximately by @xmath6 . since bulk contacts
do not change significantly the density of states in the wire @xcite , their contribution to the exponent in eq .
is @xmath116 .
this provides a concrete experimental setup to measure fingerprints of phase averaging by considering a quantum wire with a single impurity contacted by two end contacts and , in situ , also by two scanning tunneling microscopy contacts situated left and right to the impurity but far away from the end contacts compared to @xmath88 . by either measuring transport through the two end contacts , or one end and one bulk contact , or two bulk contact , the situation ( c ) , ( d ) , and ( e ) of fig .
[ fig : int_incommens_exp ] are realized , respectively . if phase averaging dominates different exponents
will be observed , whereas in the case of dominant dephasing the exponent will always be given by the impurity exponent @xmath117.@xcite an additional feature in fig .
[ fig : int_incommens ] is the occurrence of peaks in @xmath3 at certain moderate temperatures . they arise when left- and right - dot resonances coincide leading to an avoided crossing with a strongly enhanced total transmission .
as the transmission and reflection amplitudes of the single effective barriers depend on @xmath63 via the range @xmath88 of the oscillations of the effective potential , the dot resonances shift energetically with @xmath63 .
the peak shifts occur in two steps , when temperature is lowered from far larger than the bandwidth to zero : the first step occurs when @xmath63 sweeps over the largest part of the band and is due to a renormalization of the hopping in the chain .
the second step occurs when @xmath63 sweeps over the very peak under consideration .
the size in energy of this second step is proportional to the level spacing of the corresponding dot and therefore in general different for left- and right - dot resonances making possible an avoided crossing ( see fig .
[ fig : int_incommens ] ) . at higher temperatures , there are so many transmission resonances contributing to the conductance that the enlargement of single peaks due to their avoided crossing is not visible in the conductance .
the temperature dependence of the dot resonances has also the consequence that only two dots of exactly the same length represent a truly commensurate configuration because only for @xmath118 regularly coinciding left- and right - dot resonances occur at every temperature .
in this special case our data confirms that eq . , which now reads @xmath119 , is generalizable to interacting qws .
depending on whether the single barriers are within the wire or at the contacts , this means @xmath96 , @xmath120 or a combination of both .
finally we mention the case of four barriers .
. then leads to @xmath121 where @xmath122 , @xmath123 , @xmath124 and the arithmetic - geometric mean @xmath125 of @xmath14 and @xmath126 is defined as the limit of the sequence given by @xmath127 , @xmath128 , @xmath129 , @xmath130 . for weak barriers , @xmath59 ,
this gives again the product law @xmath131 .
for strong barriers , @xmath61 , we get @xmath132 the logarithmic factor appearing now causes small deviations from exact power - laws of @xmath3 .
however , as for an increased number of barriers higher temperatures are necessary for phase averaging to become a reasonable approximation , this formula might not be as relevant as in the case of three barriers .
we have compared the effects of phase averaging and dephasing in an one - dimensional mesoscopic system .
investigating the scaling behavior of the conductance we have proposed that the two phenomena can be distinguished experimentally by analyzing different kinds and numbers of barriers .
this is an important issue for the interpretation of past and future experiments , especially due to the fact that also the contacts to the leads form barriers . in cases where phase averaging dominates
, those contact barriers will influence the scaling behavior even if their resistance is lower than the one of barriers within the wire .
this might be connected with still unexplained power - law exponents in transport measurements through carbon nanotubes.@xcite whether phase averaging or dephasing is dominant in these experiments is still an open question and needs further investigations .
the authors thank m. bttiker , s. de franceschi , y. gefen , y. imry , f. reininghaus and k. schnhammer for helpful discussions .
this work has been supported by the vw foundation and the forschungszentrum jlich via the virtual institute ifmit ( s.j . and h.s .
) , the deutsche forschungsgemeinschaft via sfb 602 ( v.m . ) , and the alexander von humboldt foundation and the infm smc
cnr ( t.e . ) .
r. landauer , ibm j. res .
dev . * 1 * , 223 ( 1957 ) ; m. bttiker , phys .
lett . * 57 * , 1761 ( 1986 ) ; phys .
46 * , 12485 ( 1992 ) ; y. imry , in _ directions in condensed matter physics _ , g. grinstein and g. mazenko , eds .
( world scientific , singapore , 1986 ) , p.101 . for strong hopping barriers ,
the reflection phases @xmath133 contribute @xmath134 to @xmath32 , so that the condition of commensurability is to be formulated with @xmath135 or @xmath136 instead of @xmath38 or @xmath137 .
m. thorwart , m. grifoni , g. cuniberti , h. w. ch .
postma , c. dekker , phys .
lett . * 89 * , 196402 ( 2002 ) ; yu .
v. nazarov and l. i. glazman , phys .
91 * , 126804 ( 2003 ) ; d. g. polyakov and i. v. gornyi , phys .
b * 68 * , 035421 ( 2003 ) . | we analyse phase averaging in one - dimensional interacting mesoscopic systems with several barriers and show that for incommensurate positions an independent average over several phases can be induced by finite temperature . for three strong barriers with conductances @xmath0 and mutual distances larger than the thermal length , we obtain @xmath1 for the total conductance @xmath2 . for an interacting wire ,
this implies power laws in @xmath3 with novel exponents , which we propose as an experimental fingerprint to distinguish temperature induced phase averaging from dephasing . |
the three spitzer space telescope focal plane instruments were designed to investigate four major scientific topics : ( 1 ) early universe , ( 2 ) brown dwarfs and superplanets , ( 3 ) active galactic nuclei , and ( 4 ) protoplanetary and planetary debris disks . of these topics ,
the most important in defining the infrared array camera ( irac ) design was the study of the early universe , and in particular , the study of the evolution of normal galaxies to z @xmath2 3 by means of deep , large - area surveys .
the 3 to 10 @xmath0 m wavelength range was selected because stars have a peak emission at a wavelength of 1.6 @xmath0 m , at the minimum of the h@xmath3 opacity @xcite .
the emission peak is an ubiquitous feature of stellar atmospheres and can be used to determine a photometric redshift for 1 @xmath4 z @xmath5 5 @xcite .
the irac sensitivity requirement was set such that irac could achieve a 10@xmath6 detection of an @xmath7 galaxy at z = 3 .
this in turn required the measurement of a flux density at 8 @xmath0 m of 8 @xmath0jy ( 10@xmath6 ; * ? ? ? * ) .
channel 1 ( 3.6 @xmath0 m ) was selected to be at the minimum of the zodiacal background radiation @xcite , and to avoid the water ice absorption band at 3.1 @xmath0 m .
the central wavelengths of the remaining irac filters and their bandwidths ( approximately 25% ) were then optimized ( considering the detector materials available ) to reach the sensitivity requirement and to permit the measurement of a photometric redshift from 1 @xmath5 z @xmath5 5 @xcite .
the number of individual channels in irac was limited by the number of array cameras that could fit in the specified volume . to reduce costs and maximize reliability , moving parts had to be minimized ; hence no filter wheels , only fixed filters , were used .
transmissive rather than reflective optics were used because of limited space .
the pixel size was optimized to achieve the best point source sensitivity for weak sources while maximizing the survey efficiency .
the shutter at the entrance aperture was the only moving part allowed in irac .
although the design was optimized for the study of the early universe , irac is a general - purpose , wide - field camera that can be used for a large range of astronomical investigations . in -
flight observations with irac have already demonstrated that irac s sensitivity , pixel size , field of view , and filter selection are excellent for studying galaxy structure and morphology , active galactic nuclei , the early stages of star formation and evolution , and for identifying brown dwarfs .
irac was built by the nasa goddard space flight center ( nasa / gsfc ) with the smithsonian astrophysical observatory ( sao ) having management and scientific responsibility .
additional information on the operation and performance of the irac instrument can be found at the irac web site .
irac consists of two parts : the cryogenic assembly ( ca ) installed in the multiple instrument chamber ( mic ) within the cryogenic telescope assembly ( cta ) , and the warm electronics assembly ( wea ) mounted below the cta in the payload assembly area @xcite .
the irac cryogenic assembly , depicted in figure 1 , consists of the following major subassemblies : two pickoff mirrors ; the shutter ; two optics housings , which hold the doublet lenses , beamsplitters , filters , and cold stops ; four focal plane assemblies ( fpas ) that include the detector arrays and associated components ; the transmission calibrator with its source and integrating spheres ; and the housing structure , consisting of the main housing assembly and the wedge - shaped mic adapter plate .
the ca is cooled to the temperature of the mic base plate ( @xmath81.2 k ) .
the ca has a mass of 11.1 kg , is 0.15 m high and 0.28 m wide at its outer edge , and consumes an average of 3.0 mw of power during nominal operation .
the irac wea resides in bay 5 of the spacecraft , operating at near room temperatures ( @xmath910c ) .
all electrical interfaces between the spacecraft and irac pass through the wea .
the wea provides all power and data interfaces to both the spacecraft and to the ca as necessary to conduct the operation of irac .
the wea is connected to the ca via an extensive set of conventional and cryogenic cables , which all pass through a junction box located on the spacecraft near the cta .
the junction box serves as the transition between the conventional and cryogenic cables .
the principal function of the wea is the support and control of science data taking : the generation of bias voltages to the detectors ; the timing of readout sequences ; the amplification and digitization of the analog science signals ; the digital signal processing of the images ; and the transmission of the digital data to the spacecraft c&dh solid state recorders for mass storage , to be followed by downlinks to ground stations at twelve hour intervals . apart from autonomous fault protection , the wea internal software responds only to commands sent by the spitzer command & data handling ( c&dh ) computer .
the wea has a mass of 24.5 kg , and its dimensions are 0.46 m in length , 0.38 m in width , and 0.23 m in height .
it consumes 56 w of power during nominal operation .
the irac optical path is shown in figure 2 .
light enters the ca via two pickoff mirrors located near the telescope focal plane .
the two mirrors are slightly displaced and tilted to physically separate the optical components of the channel pairs .
thus the pickoff mirrors project nearly adjacent [email protected] arcmin images of the sky to the channel pairs .
the centers of the two images are separated by 6.8 arcmin .
[ fig2 ] the lower mirror selects the field for channels 1 and 3 .
the reflected beam is incident upon a mgf@xmath10 zns ( cleartran ) vacuum - spaced doublet that re - images the spitzer focal plane onto the detectors .
after the lenses , a germanium substrate beamsplitter reflects the channel 1 beam and transmits the channel 3 beam .
the beamsplitter is tilted by 45 degrees and is in the converging beam .
the channel 1 light is reflected from the beamsplitter , passes through a bandpass filter , and then a lyot stop , which rejects stray light . for channel 3 ,
after the beam goes through the beamsplitter , a filter selects the proper wavelength bandpass .
the filter is tilted at 45 degrees , in the plane opposite the beamsplitter , to correct for the astigmatism introduced by the beamsplitter .
a lyot stop rejects stray light .
the situation is similar for channels 2 and 4 , which are selected by the upper pickoff mirror . in these channels the doublet lens materials are znse and baf@xmath10 .
the channel 2 beam is reflected by the beamsplitter , and the channel 4 beam is transmitted .
all filters are mounted on germanium substrates .
the lyot stops are oversized by 15% .
barr associates , inc . , westford ,
ma , produced the filters , ocli - ajds uniphase co. , santa rosa , ca , the beamsplitters , and spectral systems , inc .
, dayton , ohio , the lenses .
a shutter at the entrance to the ca was designed to block the incoming beams and to permit the measurement of the dark count rate of the detector arrays .
a mirror , mounted on the rear surface of the shutter , reflects the light from the transmission calibration source into the two optical paths . to eliminate the possibility that the shutter could get stuck in the closed position
, it has not been operated in flight to date .
irac has two internal calibration systems .
the transmission calibration system measures irac optical throughput , and the flood calibrators test for detector responsivity and stability . since the transmission calibration system requires that the shutter be closed , it has not been used in flight .
the irac detector arrays were developed by raytheon / sbrc , goleta , ca , under contract to sao @xcite .
channels 1 ( 3.6 @xmath0 m ) and 2 ( 4.5 @xmath0 m ) insb detector arrays operate at a temperature of 15 k , and channels 3 ( 5.8 @xmath0 m ) and 4 ( 8.0 @xmath0 m ) si : as detector arrays operate at 6 k. to help achieve the relative photometric accuracy requirement of 2% , the array temperatures are controlled by an active feedback circuit to @xmath5 10 mk peak - to - peak variation from their set points .
both array types are 256@xmath1256 pixels in size , and have the same physical pixel size of 30 @xmath0 m .
cryo - cmos technology was used in the readout multiplexer design @xcite .
the arrays were anti - reflection coated with sio for channels 1 , 2 , and 3 , and with zns for channel 4 .
the power dissipation for each array during nominal operations is @xmath5 1 mw . table 1 lists some of the detector properties .
the operability is the percentage of the pixels in an array that meet specifications .
llcccc 1 & 48534/34 ( ur ) & 8.1 & 87 & 145,000 & 99.97 + 2 & 48975/66 ( gsfc ) & 6.8 & 86 & 140,000 & 99.99 + 3 & 30052/41 ( arc ) & 13.0 & 45 & 170,000 & 99.99 + 4 & 30219/64 ( arc ) & 6.6 & 70 & 200,000 & 99.75 + [ tab1 ] the arrays can also be heated above their nominal operating temperatures to perform a thermal anneal " of the fpas to remove hot pixels or residual images . in the in - flight anneal operation ,
the temperatures of the fpas are raised to 23 k ( channel 1 ) and 30 k ( channels 2,3,4 ) for approximately 2 minutes and then returned to their nominal operating temperatures in approximately 4 minutes .
the insb arrays were tested and characterized at the university of rochester @xcite and the si : as arrays at the nasa / arc @xcite .
the flight arrays were evaluated at the nasa / gsfc .
irac has one method of operation on the spitzer space telescope : stare and integrate .
the field - of - view is determined by use of the full array ( 256@xmath1256 pixels ) or the sub - array ( 32@xmath132 pixels ) readout mode .
frame times for the full array can vary from 0.4 to 500 sec in 0.2 sec steps , but the following times are standard : 2 , 12 , 30 , 100 , and 200 sec , except that channel 4 uses two or four 50-sec frames instead of 100 and 200 sec frames , respectively . in the subarray mode , a 32@xmath132 pixel region is read out at a faster rate .
the standard frame times are : 0.02 , 0.1 , and 0.4 sec ; other times can be commanded in steps of 10 msec . each data collection command in subarray mode produces 64 image frames , taken consecutively with no dead time between frames .
this mode is primarily for observing bright sources that would otherwise saturate the array , or for observing time - critical events where a finer time resolution is required .
this mode was used to measure the pointing jitter of the telescope . during an integration data
are taken using a fowler sampling method @xcite to reduce the effective read noise .
this mode of sampling consists of taking n non - destructive reads immediately after the reset ( pedestal levels ) and another n non - destructive reads near the end of the integration ( signal levels ) .
the average of the pedestal levels is subtracted from the average of the signal levels to derive the signal .
the averaging and subtraction of the two sets of reads is performed in the irac on - board electronics to generate a single image from each array , which is stored and transmitted to the ground .
n can vary from 1 to 64 .
the fowler n used for an observation will depend on integration time and has been selected to maximize the s / n at each frame time , based on in - flight performance tests .
irac is often used in a high dynamic range mode where a long exposure is preceded by one or more short exposures : 0.6 , 12 sec ; 1.2 , 30 sec ; 0.6 , 12 , 100 sec ; and 0.6 , 12 , 200 sec .
irac is capable of operating each of its four arrays independently and/or simultaneously . in normal flight operation
, all four arrays are operated together .
the point source sensitivities for the frame times available to irac are shown in table 2 . for each frame time and channel ,
the 1@xmath6 sensitivity in @xmath0jy is given for the low background case ( near the ecliptic poles ) .
the values were calculated based on the sensitivity model given by @xcite , using measured in - flight values for the read noise , pixel size , noise pixels , background , and total system throughput .
the numbers were compared to observations at several frame times to confirm the validity of the calculations .
the values in table 2 are close to pre - flight predictions of 0.5 , 0.9 , 3.1 , and 5.0 @xmath0jy ( 1@xmath6 , 200 sec ) for channels 1 - 4 , respectively , except for the influence of two factors : the lower ( better ) than expected noise pixel values for all channels , and the lower ( worse ) throughput in channels 3 and 4 .
irac meets its required sensitivity limits of 0.92 , 1.22 , 6 , and 9 @xmath0jy ( 1@xmath6 , 200 sec ) for channels 1 - 4 , respectively .
crrrr 200 & 0.40 & 0.84 & 5.5 & 6.9 + 100 & 0.60 & 1.2 & 8.0 & 9.8 + 30 & 1.4 & 2.4 & 16 & 18 + 12 & 3.3 & 4.8 & 27 & 29 + 2 & 32 & 38 & 150 & 92 + 0.6 & 180 & 210 & 630 & 250 + 0.4 & 86 & 75 & 270 & 140 + 0.1 & 510 & 470 & 910 & 420 + 0.02 & 7700 & 7200 & 11000 & 4900 + [ tab2 ] the noise measurements in channels 3 and 4 scale as expected with the inverse square root of time to the limit measured , 20,000 seconds .
channels 1 and 2 deviate from this rule in approximately 3000 seconds , which indicates that these channels are approaching the source completeness confusion limit ( see also * ? ? ?
the deviation occurs at about 1 @xmath0jy ( 5@xmath6 ) , which is slightly fainter than the predictions for the irac confusion limit @xcite . for extended emission in channels 3 and 4 ,
the calibration was found to be significantly different than expected from the calibration derived from a point source .
the calibration aperture ( 10 pixel radius ; 122 aperture radius ) does not capture all of the light from the calibration sources so the extended sky emission appears too bright . for photometry using different apertures , the estimated correction is listed in table 6.4 of the spitzer observer s manual .
irac s optics provide diffraction - limited imaging , with wavefront errors @xmath11 in each channel .
the in - flight irac image quality is limited by a combination of the optical quality of the spitzer telescope , which is diffraction limited at 5.4 @xmath0 m , and the size of the irac pixels .
the predicted fwhm of the point spread function ( psf ) , based on the preflight telescope and irac optical models , was 1.6 , 1.6 , 1.8 , and 1.9 arcsec at 3.6 , 4.5 , 5.8 , and 8.0 @xmath0 m , respectively .
table 3 shows the properties of the irac psf , derived from in - flight measurements of bright stars .
the fwhm of the psf will differ slightly for non - stellar spectra , e.g. dust .
the psf was monitored throughout the early mission while the telescope cooled down , and while the secondary mirror was moved in two steps to achieve a near - optimal focus @xcite .
the `` noise pixels '' column in table 2 gives the equivalent number of pixels whose noise contributes to noise in the analysis when an image is spatially filtered for optimum faint point - source detection @xcite .
ccccccc 1 & 7.0 & 1.66 & 1.44 & 42 & 1.221 & 1.3 + 2 & 7.2 & 1.72 & 1.43 & 43 & 1.213 & 1.6 + 3 & 10.8 & 1.88 & 1.49 & 29 & 1.221 & 1.4 + 4 & 13.4 & 1.98 & 1.71 & 22 & 1.220 & 2.2 + there are two columns for the full width at half - maximum ( fwhm ) of the psf .
the mean fwhm is from observations of a star at 25 different location on the array .
the fwhm for `` centered psf '' is for cases where the star was most closely centered in a pixel .
the fifth column in table 2 is the percentage of the total flux in the central pixel for a point source that is well - centered in a pixel . the flux in the central pixel for a random observation will be lower , because the psf of the telescope is rather undersampled at the irac pixel scale except in channel 4 .
the strehl ratios observed are 0.37 , 0.58 , 0.62 , and 0.95 for channels 1 - 4 respectively .
the strehl ratio for channel 3 , after correction for scattered light in the array detector , is 0.71 .
there is a small amount of distortion over the irac fov ( table 3 ) . a quadratic distortion model is provided in the spitzer science center ( ssc ) basic calibrated data ( bcd ) pipeline which fits in - flight data at rms @xmath502 ( channels 1 and 2 ) and @xmath503 ( channels 3 and 4 ) across the full arrays .
diffuse stray light from outside the irac fov is scattered into the active region of the irac detectors in all four channels .
the problem is more significant in channels 1 and 2 than in channels 3 and 4 . in channels 1 and 2
the stray light pattern due to diffuse sources resembles a `` butterfly '' , while in channels 3 and 4 it resembles a `` tic - tac - toe '' board .
a discussion of these effects and example images are given by @xcite .
these artifacts are due to zodiacal light scattered onto the arrays , possibly reflected from a hole in the fpa covers in channels 1 and 2 , and from reflective surfaces outside the edges of channel 3 and 4 arrays . the stray light scales with zodiacal light , which is the light source for the flat fields , so the stray pattern contaminates the flats and all irac images .
the `` butterfly wings '' have amplitudes of about 5% of the background intensity .
the pipeline subtracts a model for the tic - tac - toe and butterfly patterns for the data , scaled based on a model for the zodiacal light intensity @xcite at the location and time of the observation . to the extent the model is accurate , this corrects for the stray light contamination .
stars that fall in specific regions just outside the array edges scatter light into the detectors and produce distinctive patterns of scattered light on the array . scattered light avoidance zones have been identified in each channel .
observers should avoid placing bright stars in these zones if their observations are sensitive to scattered light .
typically , in channels 1 and 2 about 2% of the light from a star is scattered into a `` splatter pattern '' , which has a peak intensity of about 0.2% of the light from the star .
the avoidance zones for channels 3 and 4 are a narrow strip about 3 pixels wide , 16 pixels outside of the array and surrounding it .
the overall requirement for the irac mission is that the system photometric responsivity be calibrated to a relative accuracy of 2% and that the absolute accuracy of the data set be determined to better than 10% .
we expect that both of these requirements will be achieved .
currently 25% of the time that irac is on is devoted to calibration , but the excellent stability of irac will result in a significant decrease in that fraction .
the irac system throughput and optical performance are governed by a combination of the system components , including the lenses , beamsplitters , filters , mirrors , and detectors .
figure 3 gives the system throughput , including transmission of the telescope , irac optics , and quantum efficiency of the detectors @xcite .
for each irac channel , both polarizations were measured @xcite , and the average polarization is shown as a solid curve .
[ fig3 ] the system parameters are summarized in table 4 .
the isophotal wavelength @xmath12 is the wavelength which must be assigned to the monochromatic flux density derived from a broadband measurement .
the isophotal wavelength is defined as @xmath13 where @xmath14 is the measured in - band flux , @xmath15 is the irradiance at the entrance aperture of the telescope , @xmath16 is the spectral response of the system , and @xmath17 and @xmath18 define the wavelength range of interest .
we calibrate the flux densities for a nominal spectrum @xmath19 = constant .
for source spectra with a range of slopes , @xmath20 from 2 to + 2 , the color corrections are very close to unity ( 0.99 to 1.02 ) .
a color correction table for various values of the spectral index will be given in the spitzer observer s manual .
cccccccc 1 & 3.535 & 3.58 & 0.750 & 21 & 0.676 & 0.563 & 0.748 + 2 & 4.502 & 4.52 & 1.015 & 23 & 0.731 & 0.540 & 0.859 + 3 & 5.650 & 5.72 & 1.425 & 25 & 0.589 & 0.552 & 0.653 + 4 & 7.735 & 7.90 & 2.905 & 36 & 0.556 & 0.450 & 0.637 + a number of astronomical standard stars are observed in each instrument campaign to obtain a relative flux calibration @xcite .
stars with a range of spectral indices and fluxes are observed at a number of positions across the array and many times throughout the mission to monitor any changes that may occur . to calculate vega magnitudes from the calibrated irac images , use the following zero - magnitude fluxes : 277.5 , 179.5 , 116.6 , and 63.1 jy for channels 1 , 2 , 3 , and 4 , respectively .
these fluxes were calculated from the irac throughput values and the absolute spectra at irac wavelengths @xcite . in - flight data
indicate that irac is very stable ; observations of calibration stars taken over two months are repeatable to about 2% .
both types of detectors have measurable non - linearity . to linearize the data , a quadratic fit is used for the insb arrays , and a cubic fit is used for the si : as arrays .
all arrays have been linearized to better than 1% up to approximately 90% of their full - well capacity .
the detector linearity was measured during ground testing and verified during flight .
the flat field for each channel is defined as the factors by which one must correct each pixel to give the same value at a particular flux for uniform illumination .
the factors include any field - dependent optical transmission effects as well as the responsivity of the pixel and output electronics gain .
sky flats are obtained using a network of 24 high zodiacal background regions of the sky in the ecliptic plane , ensuring a relatively uniform illumination with a reasonable amount of flux .
if the data are taken with an appropriate dither pattern , it is possible to relate the total response of each pixel to that of all others .
one such region is observed at the beginning and end of each irac campaign .
the data are reduced and combined to construct a flat field image for each channel , which is then divided into the science data .
the detector dark currents are generally insignificant compared to the sky background .
however , there is a significant offset or bias ( which can be positive or negative ) in a dark frame , which must be subtracted from the observations . especially for shorter frames ,
the `` dark '' images are mostly due to electronic bias differences , rather than true dark current .
therefore , the number of electrons in a dark image does not scale linearly with exposure time .
the shutter has never been used in flight , and therefore isolated dark / bias data can not be taken . instead ,
ground - based `` lab '' darks are used , in conjunction with sky darks which measure changes to the darks that have occurred since launch . as part of routine operations a dark region of the sky near the north ecliptic pole
is observed at the beginning and end of an irac campaign , and every three days during the campaign .
these data are reduced and combined in such a way as to reject stars and other astronomical objects with size scales smaller than the irac array .
the resulting image of the minimal uniform sky background contains both the bias and the dark current . when subtracted from the routine science data
, this will eliminate both of these instrumental signatures .
the dark frames , when corrected for changes in the zodiacal background flux , remain very stable from campaign to campaign . in general
, the first frame of any sequence of images tends to have a different median value from other frames in the sequence , hence the name `` first frame effect . ''
the largest offsets occur when a frame is taken with a very short interval from the preceding image , which occurs when multiple frames are commanded at once .
such commands have been eliminated from irac science observations , so that there is always a finite interval between one frame and its predecessor . since many irac observations will take only one image at each sky position , and there is an 8 to 60 second delay while spitzer points and settles to a new position , the `` first frame effect '' is reduced significantly .
the correction for the first frame effect was derived from pre - flight ground - based data . in flight ,
cosmic rays excite approximately 3 pixels per second in channels 1 and 2 and approximately 5 pixels per second in channels 3 and 4 .
this is close to the predicted rate @xcite .
the most common cosmic rays do not affect the pixel performance in subsequent frames and are confined to a few pixels around the peak .
the cosmic ray effects in channel 1 and 2 are more compact than those in channels 3 and 4 , which have thicker detectors . in these latter channels
the cosmic ray effects appear as streaks and blobs .
however , other events can cause streaks or other multiple pixel structures in the array , and less common energetic and high - z events can cause image artifacts in subsequent frames .
most of the cosmic ray effects are removed by the post - bcd pipeline processing for observations that are well dithered . on 2003
october 28 , spitzer encountered a large solar proton flare with an integrated dose of 1.6 x 10@xmath21 p/@xmath22 for proton energies greater than 50 mev .
irac was powered off at the time .
after several thermal anneals of the arrays following the flare , only a few ( @xmath920 ) new hot pixels were detected , all in the channel 4 array .
array detector artifacts known before launch or discovered in flight include : persistent images in channels 1 and 4 , multiplexer bleed , column pulldown , and banding .
these effects are described in more detail in the ssc observer s manual and in articles by @xcite . to erase persistent images both channel 1 and 4
are annealed simultaneously every 12 hours ( after each downlink ) and bright source observations are scheduled only near the end of an irac campaign .
column pulldown , bright source multiplexer bleed , and banding will be removed in the post - bcd pipeline corrections .
the effects of persistent images and cosmic rays can be avoided or greatly reduced in observations that are well dithered .
we recommend taking a minimum of three images of a science target in each channel as a best practice for observing with irac .
in flight , irac meets all of the science requirements which were established before launch , and in many cases the performance is better than the requirement .
irac is a powerful survey instrument because of its high sensitivity , large field of view , and four - color imaging from 3.2 to 9.5 microns wavelength .
for example , in channel 1 ( 3.6 @xmath0 m ) , irac can reach the same point source sensitivity ( 19 magnitude , 5@xmath6 , 1 hour ) as the w. m. keck telescope at 2.9 - 3.2 @xmath0 m in only one - hundredth the exposure time and does so with a far larger field of view : 5@xmath15 versus 40@xmath140 . in channel 4 ( 8.0 @xmath0 m ) , irac can achieve the ultimate sensitivity limit reached by isocam ( 15 @xmath0jy , 1@xmath6 , 1 hour , 3@xmath13 field ) on the infrared space observatory ( iso ) in the lw2 filter ( 6.7 @xmath0 m ) in approximately one - hundreth the frame time @xcite .
irac continues to function extremely well since the first images were produced , seven days after launch , on 2003 september 1 .
the first scientific results of irac , which are presented in this issue , are examples of irac s great potential for producing exciting new science with the spitzer space telescope .
lynne deutsch , a co - author of this paper , died on 2 april 2004 after a long illness .
lynne was a dear friend and a close colleague .
the irac team will deeply miss her presence .
we dedicate this paper in her memory for all her contributions to infrared astronomy .
this work is based on observations made with the spitzer space telescope , which is operated by the jet propulsion laboratory , california institute of technology under nasa contract 1407 .
support for this work was provided by nasa through contract number 125790 issued by jpl / caltech .
fazio , g. g. , ashby , m. l. n. , barmby , p. , hora , j. l. , huang , j .- s .
, pahre , m. a. , wang , z. , willner , s. p. , arendt , r. g. , moseley , s. h. , brodwin , m. , eisenhardt , p. , stern , d. , tollestrup , e. v. , $ wright , e. l. 2004 , , in press | [ sec : abstract ] the infrared array camera ( irac ) is one of three focal plane instruments in the spitzer space telescope .
irac is a four - channel camera that obtains simultaneous broad - band images at 3.6 , 4.5 , 5.8 , and 8.0 @xmath0 m .
two nearly adjacent [email protected] arcmin fields of view in the focal plane are viewed by the four channels in pairs ( 3.6 and 5.8 @xmath0 m ; 4.5 and 8 @xmath0 m ) .
all four detector arrays in the camera are 256@xmath1256 pixels in size , with the two shorter wavelength channels using insb and the two longer wavelength channels using si : as ibc detectors .
irac is a powerful survey instrument because of its high sensitivity , large field of view , and four - color imaging .
this paper summarizes the in - flight scientific , technical , and operational performance of irac . |
although the basic explosion mechanism of supernovae of type ia ( henceforth sn ia ) has been established to be the thermonuclear combustion of degenerate c / o white dwarfs ( henceforth wd ) , many aspects of the progenitor systems of remain to be understood .
nearly all viable progenitor models involve mass transfer in binary systems ( howell 2011 ; wang & han 2102 ; maoz , mannucci & nelemans 2014 ; but see chiosi et al .
one idea is the initiation of carbon ignition as the mass approaches the classical chandrasekhar limit of stability , @xmath4 , for non rotating wds by means of accretion in a binary system .
this classic model is most closely associated with mass transfer from a non degenerate companion , the single degenerate ( sd ) scenario .
variations on this theme allow for explosions with less than @xmath5when accretion of helium from a companion leads to the accumulation of an explosive degenerate layer of helium on top of the c / o core that generates compression waves that can trigger a central carbon detonation ( fink et al .
2010 ; woosley & kasen 2011 ; shen & moore 2014 ) .
this might occur by accretion from a non degenerate companion , or from a degenerate companion in one variety of the double degenerate ( dd ) scenario .
other dd models involve the tidal disruption of one wd , thus adding mass to the other , resulting in explosion ( dan et al .
2014 ) , or the violent merger or collision of two wds ( pakmor et al . 2012 ; kushnir et al .
the dd scenarios typically invoke wds with mass less than @xmath5 , but the total mass might exceed this amount .
other scenarios manifested in either the sd or dd context involve spinning up a wd until it is rotationally supported , with a mass exceeding @xmath5 , the explosion being postponed until sufficient angular momentum is lost from the star .
these are loosely called _ spin up
/ spin down _ models ( yoon & langer 2005 ; di stefano et al . 2011 ; justham 2011 ; tornambe & piersante 2013 ) . in these models ,
the central density rises to the point of carbon ignition because angular momentum is lost , not because mass is gained .
yet another variation invokes the merger of a wd with a stellar core in the context of common envelope evolution ( livio & riess 2003 ; ilkov & soker 2013 ) . in the violent merger models ,
the explosion occurs so quickly that an issue of the quasi static rotational state of the wd does not arise , but in most models the physical context requires the accumulation of mass and angular momentum , and hence that the wd should rotate .
the rotational state is ignored in many models of the progenitor evolution and explosion , including many computationally demanding multi dimensional models ; however , it is obvious that , in the absence of a specific mechanism to lose angular momentum , it must accumulate .
a generic question that persists throughout the various sn ia progenitor models is how the internal rotation profile of the accreting wd influences the essential dynamical and secular processes that operate in sn ia progenitor environments .
one basic influence of the internal rotation has been known now for four to five decades , , that ( a ) uniform rotation can increase the maximum mass of stable wds above the chandra limit by only about 3 to 4 percent , and ( b ) differential rotation can increase this maximum mass substantially , exceeding the chandra limit by factors of up to @xmath6 , and so resulting in maximum wd masses up to @xmath7 ( see ostriker & bodenheimer 1968 and references therein ) .
we shall henceforth refer to @xmath5 as the `` chandra limit , '' and wds with masses above this limit as `` super chandra . ''
most typical sn ia are consistent with explosion at @xmath5 ( but see scalzo et al .
polarization observations suggest that most typical sn ia may not rotate significantly , but that subluminous events ( sn 1991bg like ) may do so ( patat et al . 2012 ) . observations of some very bright sn ia have led to inferred values of the ejecta mass and hence the mass of the pre sn wd in the range ( 2.1 2.8 ) @xmath8 .
examples are sn 2003fg , sn 2006gz , sn 2007if , sn 2009dc @xcite .
these masses are much in excess of @xmath5 , and possible implications for the internal rotation of the pre
sn wd have been widely discussed . in this context ,
one may ask what evolutionary pathway would lead to a wd in a slowly rotating state , and so in a slightly super chandra condition , as opposed to a pathway that would lead to a rapidly and differentially rotating wd in a substantially super chandra condition . if the wd rotates rapidly , then issues also arise about the onset of bar mode instabilities that might affect the onset of the explosion and the subsequent dynamics . in this paper , we focus on clarifying the origins of the different regimes of internal rotation that can exist inside accreting wds .
we consider wds without strong , permanent magnetic fields .
accordingly , all angular - momentum transport processes considered here are necessarily described by viscosities generated only by non
magnetic mechanisms , , hydrodynamic mechanisms such as kelvin helmholtz and baroclinic instabilities , and secular mechanisms such as the zahn instability ( zahn 1992 ) .
we formulate these viscosities with the aid of well known prescriptions given in the literature .
we thus describe angular - momentum transport inside accreting wds in terms of the standard angular momentum transport equation applied in a background wd model , which is pre - specified ( [ elltransport ] , [ angmomtransport ] ) .
we apply boundary conditions at the wd surface to describe the deposition of angular momentum by accreting matter . for low viscosity ( zahn ) transport ( [ rothydro ] ) , we recover the differential rotation profile found earlier @xcite . for high viscosity transport , which includes viscosities due to both kelvin - helmholtz and baroclinic instabilities , we find both ( a ) the nearly - uniform rotation profile found earlier @xcite , and , ( b ) a new , differential rotation profile , which is qualitatively similar to the inner parts of the differential rotation profile found earlier for viscosity due to kelvin - helmholtz instability alone @xcite , but different in detail .
all of the above regimes of rotation appear to be self - consistent , corresponding to different viscosities and different regimes of operation in terms of the richardson number .
all regimes described have profiles of specific angular momentum ( henceforth @xmath9profile ) that increase monotonically outward , as they must to be rayleigh stable , but the profile of the angular velocity , @xmath10 , ( henceforth the @xmath10profile ) can have a maximum inside the wd , from which it decreases both outward to the surface and inward to the center
. a general and comparative discussion of these rotation profiles are given in
[ rotzahn ] , [ rotyl ] and [ discussregime ] . a discussion , our conclusions and the prospects for future work are given [ discuss ] .
mechanisms of angular momentum transport inside rotating stars can be classified into two general types , , ( a ) those that transport angular momentum in the @xmath11-direction , often called the _
vertical _ direction , the direction of local gravity , and ( b ) those that transport angular momentum in the @xmath12-direction , often called the _ horizontal _ direction .
the angular momentum is about the stellar rotation axis of the star , defined as the @xmath13-direction , or equivalently the @xmath14-direction , in this coordinate system .
our concern in this work on accreting wds is primarily with mechanisms of the first type , as these determine the internal rotation profile
. mechanisms of the second type work through processes such as eddington - sweet ( henceforth es ) meridional circulation , and the goldreich - schubert - fricke ( henceforth gsf ) instability , the primary influence of which is on the distribution of angular velocity / momentum in the meridional plane and along the @xmath13-axis ( , @xmath15 ) . we describe the mechanisms of vertical or @xmath11transport of angular momentum in this section , and refer to horizontal or @xmath12-transport only briefly at appropriate places .
vertical transport mechanisms involve turbulent viscosities that can be of two basic kinds : ( a ) those that are generated by hydrodynamical instabilities that grow on short , dynamical timescales , , the kelvin - helmholtz instability ( henceforth khi ) and the baroclinic instability ( henceforth bci ) , and are generally high , and ( b ) those that are generated by secular instabilities that grow on long , thermal timescales , , the zahn instability ( henceforth zi ) , and are generally low .
we now discuss these two types of instabilities . among the best known and most studied hydrodynamic instabilities that grow on dynamical timescales @xmath16 are khi and bci .
the former occurs when there is a ( vertical ) gradient in @xmath10 , and the latter when isobaric and isodensity surfaces do not coincide , as is the case when hydrostatic balance has to be maintained under differential rotation .
a vertical gradient in the angular velocity , as measured by the shear @xmath17 causes khi when this shear is sufficiently strong to overcome the stabilizing effect of stratification as measured by the brunt - visl frequency , @xmath18 , given by @xmath19 .
\label{bvfreq}\ ] ] in equation ( [ bvfreq ] ) , @xmath20 , @xmath21 is the pressure scale height , and @xmath22 is the adiabatic temperature gradient , the second term in the square brackets being the actual temperature gradient at the relevant point inside the star .
we neglect the effect of composition gradients .
the criterion for the onset of khi is expressed in terms of the richardson number ri , which is defined as @xmath23 and which compares destabilizing shear with stabilizing buoyancy .
when ri falls below a critical value , which is widely taken to be 1/4 , khi sets in .
we shall adopt this critical value of the richardson number throughout our work here .
khi leads to a turbulent viscosity , @xmath24 , that can be formulated @xcite as : @xmath25 this formulation has been used widely in the literature .
an alternative formulation by @xcite has also been used widely .
it is given by : @xmath26 here , @xmath27 , is the local keplerian frequency inside the star and , @xmath28 is the mass interior to radius @xmath29 .
we use both formulations in this work and do not find significant differences in the final results .
surfaces of constant pressure and density may no longer coincide when hydrostatic equilibrium is maintained under differential rotation .
their misalignment causes bci , the characteristics of which depend on both the richardson number , ri , introduced above , and a critical baroclinic richardson number , ri@xmath30 , given by @xmath31 above which coriolis forces limit the scale of the horizontal perturbations and so reduce the strength of the bci somewhat .
bci leads to a turbulent viscosity , @xmath32 , which has been given @xcite as : @xmath33 we use the prescription of equation ( [ bcvisc ] ) in this work .
angular momentum transport by bci - generated viscosity for accreting degenerate stars has been studied by fujimoto ( 1993 ) , and detailed results for accreting wds with bc viscosity included have been obtained by saio & nomoto ( 2004 ) and by piro ( 2008 ) .
the work of yoon & langer ( 2004 ) is relevant in this context for having excluded bci .
we have included here both kh and bc viscosities ( , @xmath24 and @xmath32 ) , in order to have a complete discussion of all possible rotation regimes found so far .
the rotation regime found by yoon & langer ( 2004 ) , who included only the kh viscosity among the hydrodynamic ones , together with the secular zahn viscosity and processes like the es circulation and the gsf instability , is described in [ rotyl ] . the inclusion of the bci yields the result that there are two asymptotic , disparate rotation regimes .
although the overall prescription for the baroclinic viscosity is customarily given a single label bc , we see from equation ( [ bcvisc ] ) that there are two different asymptotic regimes depending on the richardson number , ri .
the actual viscosities are quite _ different _ in the two regimes , with different scalings . according to equation ( [ bcvisc ] ) ,
the bc viscosity , @xmath32 , scales with the richardson number ri as @xmath34 in the regime of small ri , while it scales as @xmath35 in the regime of large ri .
since ri scales with the shear as ri @xmath36 , we see that @xmath32 scales as @xmath1 in the small ri regime , but as @xmath3 in the regime of large ri .
the addition of the kh viscosity to the bc viscosity does not change these scalings significantly , since @xmath24 also scales with ri roughly as @xmath34 ( equation ( [ khvisc1 ] ) ) in the small ri regime while @xmath24 vanishes in the regime of high ri - values .
omission of the bc viscosity yields a regime dominated by the kh viscosity .
this regime is roughly comparable to the small ri bc regime and corresponds only to differential rotation . from these considerations
, we see that in the presence of the bci the viscous stress , @xmath37 , scales _
nonlinearly _ with the shear in the regimes of both small and large ri , with @xmath38 in the small ri regime and @xmath39 in the large ri regime .
this difference leads to a duality in the rotation profiles , essentially solid body and substantially differential , even with the same initial and boundary conditions , depending on the effective value of ri that in turn depends on the shear in the self - consistent solution . in practice
, these two different regimes were reproduced by implementing equation ( [ bcvisc ] ) in different , but formally equivalent , ways in the two different regimes .
one set of solutions began with no rotation , a given set of boundary conditions and ri @xmath40 1 .
this formulation evolved ( in @xmath41 ) a differential rotation solution , new to this work , in which both the kh and bc viscosities contributed .
this solution remained in the small ri regime , specifically ri @xmath42 , where the viscosity scales as @xmath1 and @xmath43 scales as @xmath44 ( [ diff ] ) .
the second formulation again adopted equation ( [ bcvisc ] ) , but recast in terms of the variable 1/ri , which is a small quantity in the regime of large ri . solving the same equations with the same boundary conditions but beginning the evolution with
1/ri @xmath40 1 evolved a solution that , after short transient phase , attained and remained in a state of nearly solid - body rotation ( [ uniform ] ) in the regime of large ri , ri @xmath45 , with viscosity scaling as @xmath3 , and @xmath43 scaling as @xmath46 .
this solution was dominated by the bc viscosity and had negligible contribution from the kh viscosity .
this procedure is sufficient to establish that equation ( [ bcvisc ] ) supports at least two very different , self - consistent , rotation states .
the question of how the solutions evolve for the large range of intermediate values of ri with yet different scalings with ri and @xmath1 remains open .
omitting bci , the formulation of yoon & langer ( 2004 ) is dominated by the kh viscosity in the inner regions and corresponds to ri very slightly below 0.25 in the inner regions of their models , near the low ri regime as we define it here ( [ rotyl ] ) .
uniqueness theorems say that the solution in each power - law limit , @xmath47 and @xmath48 is unique but do not constrain the behavior in intermediate regimes .
we return to a discussion of these multiple solutions in [ discussregime ] and [ discuss ] .
even when stable stratification prevents dynamical instabilities , secular instabilities are still possible if there is sufficiently strong thermal diffusion that reduces the buoyancy force through radiative leakage .
this phenomenon is described in terms of the pclet number , pe , which in this context is essentially the ratio of the rate of advection of momentum by the turbulent flow to the rate of thermal diffusion .
the pclet number is given for this particular problem by @xmath49 , where @xmath50 and @xmath51 are respectively the eddy velocity and the size of turbulent eddies , and @xmath52 is the thermal diffusivity .
thermal diffusion modifies the criterion for the onset of instability from ri @xmath53 1/4 to ri @xmath54 pe @xmath53 1/4 . the resultant secular instability ,
, the zahn instability ( zi ) leads to a turbulent viscosity , @xmath55 , that was calculated by zahn ( 1992 ) , and can be expressed as : @xmath56 here , @xmath57 is the thermal diffusivity .
we use the above prescription for zi viscosity in this work .
in this work , we describe angular momentum transport inside accreting wds by solving the transport equation in the background of a wd model with specified structure , thus neglecting the effect of changing rotation on wd structure .
we adopt this `` toy '' approach here for the purpose of clarifying the possible basic regimes of rotation profiles in accreting wds .
we do emphasize at this point that the effects of accretion are explicitly taken into account in our work by ( a ) keeping track of the changes in the mass and radius of the white dwarf as a result of accretion , as described below , and , ( b ) representing the inward transport of angular momentum by the inward motion of mass - shells in an accreting , contracting wd by a suitable `` advection '' term , as explained below and detailed in appendix a. the equation for angular momentum transport inside the wd can be expressed as @xmath58 = { 1\over 4\pi\rho r^2}{\partial\sigma\over\partial r}. \label{transport1}\ ] ] here , @xmath59 is the total rate of angular momentum transport , through radius @xmath29 at time @xmath60 , , the total torque at that radius and time . if this transport were entirely viscous , the rate would simply be @xmath61 ; however , accreting wds contract in response to their increase in mass .
consequently , each mass shell moves inward , carrying angular momentum with it .
this behavior constitutes , in effect , a slow , inward advection of angular momentum through the wd on the accretion timescale @xcite .
this advection is automatically taken into account if one recalculates the wd model at each step as mass and angular momentum accretion proceeds .
since we are working here with a backgound wd structure unaffected by changing rotation , we have to account for this advection effect explicitly .
this is easily done , as we do take into account the contraction of the wd as it accretes mass , as described below .
we have chosen the term `` advection '' for this effect to stress that this part of the transport is actually due to the slow inward motion of mass shells ( in eulerian co - ordinates ) as an accreting white dwarf adjusts its internal structure , and not due to viscosity . alternatively ,
if we visualize the situation in terms of lagrangian co - ordinates , we can look upon this effect as being due to local conservation of angular momentum .
we emphasize that if we define a formal advection velocity @xmath62 ( see below ) for this effect , it will be much smaller than the velocities that occur , , in advective accretion disks .
we now model the above effect in a straightforward manner in terms of a very slow , inward flux of matter through the wd at a rate @xmath63 , which depends on the eulerian radius @xmath29 and the time @xmath60 .
we note first that we can express this slow advection as @xmath64 to define a formal advection velocity @xmath65 , which is actually determined by the slow contraction of the wd as its mass increases due to accretion , expressed through the mass - radius relation @xmath66 , with @xmath67 decreasing as @xmath68 increases .
power - law prescriptions , @xmath69 , have been widely used as approximations to numerical results for mass - radius relations obtained from detailed models . for low - mass wds
, the non - relativistic result @xmath70 holds well , while the power steepens as the chandra limit is approached . as necessary , we have considered in the rest of this work both the non - relativistic limit of @xmath71 and @xmath71-values @xmath72 inferred from numerical wd models in the literature @xcite . to proceed further , we require a prescription for the contraction of the interior mass - shells as the whole wd contracts .
we describe our prescription in appendix a , where we show that the advection rate , @xmath63 , can be reasonably approximated by a relation of the form @xmath73 where @xmath74 is the mass accretion rate onto the wd . with this term included ,
the complete expression for the angular momentum transport rate , or the total torque , becomes @xmath75 with @xmath76(r , t ) given by equation([advrate ] ) .
the equation for angular momentum transport inside the wd is obtained by combining equations ( [ transport1 ] ) , ( [ advrate ] ) , and ( [ totrate ] ) .
we then change the dependent variable to the specific angular momentum , @xmath77 , and the independent variable to the mass co - ordinate , @xmath28 , which is related to the radial co - ordinate , @xmath29 , through @xmath78 .
the final result is : @xmath79 .
\label{transport2}\ ] ] here , @xmath80 is the accretion timescale .
we solve equation ( [ transport2 ] ) numerically for the various types of viscosities described above .
for the background wd structure in which this transport occurs , we adopt the numerical wd models supplied to us by montgomery ( 2010 ) .
the initial model , on which accretion begins , is a c / o - core wd described by the following parameters : a mass of @xmath81 , an effective surface gravity ( in cgs units ) of @xmath82 , and an effective surface temperature of @xmath83 k. the model has an appropriate gradient in the c / o ratio in the core , and there are layers of he and h on top of this core . as accretion proceeds , the wd contracts according to scheme described above , which is reflected in the outer boundary condition indicated below . for our numerical solutions ,
we need appropriate initial conditions .
we also need appropriate boundary conditions at the center and surface of the accreting wd .
for the former , we have taken wds that are nonrotating initially .
for the latter , consider the center of the star first , where the boundary condition we apply is @xmath84 in all cases , remembering that the transport equation ( equation ( [ transport2 ] ) ) is formulated in terms of @xmath9 .
we emphasize here that the angular velocity at the center , @xmath85 , is not directly constrained in our solutions .
rather , it evolves along with the solution throughout the star according to the particular prescription for the viscosity .
now consider the surface of the star , where we have explored the effects of applying the two types of boundary condition suggested previously in the literature .
these are : ( 1 ) the boundary condition of saio & nomoto ( 2004 ) , ( henceforth referred to as the _ sn boundary condition _ ) , wherein the angular velocity at the surface of the wd is set to the _ current _ keplerian value @xmath86 there , and , ( 2 ) the boundary condition of yoon & langer ( 2004 ) , ( henceforth referred to as the _ yl boundary condition _ ) , wherein each element of accreted mass @xmath87 deposits an amount of angular momentum on the wd that is ( a ) @xmath88 if the surface angular velocity @xmath89 of the wd is less than @xmath86 , and ( b ) zero if @xmath90 . here
, @xmath91 is the specific angular momentum in a keplerian orbit at @xmath67 .
we emphasize that quantities such as @xmath92 and @xmath86 are understood here to be the _ current _ values , corresponding to the current values of @xmath93 and @xmath94 during the process of accretion .
although the sn and yl boundary conditions work with the same or closely related variables , their stipulations are not exactly the same .
we explore the effects of sn and yl boundary conditions in our work , using a simple version of the latter condition adequate for our purposes , which we describe in appendix b. we find only minor differences in the final outcome .
where the differences are negligible , we quote only one result .
we now present the internal rotation profiles for the viscosities generated by hydrodynamic instabilities , , khi and bci . with both kh and bc
viscosities included in our calculations , we find two regimes of rotation as sketched in 2.1.2 , , ( a ) nearly - uniform rotation , and , ( b ) strongly differential rotation . the former regime has been discussed earlier @xcite .
we consider each regime in turn .
the regime of nearly - uniform rotation corresponds to the asymptotic regime of large ri ( ri @xmath45 in our solutions ) , @xmath48 and @xmath95 described in [ bci ] .
the solution is illustrated in figure [ uniformomega ] .
the left panel is a 3-d surface plot of the evolution of the @xmath10profile as accretion proceeds , while the right panel shows this evolution as stacked profiles in a 2-d plot as time elapses , the topmost curve giving the final time - step .
note that the mass co - ordinate for the profiles in this figure , and all subsequent evolutionary figures in this work , is @xmath96 , with @xmath28 as defined earlier , and @xmath93 the current mass of the wd .
also note that time axis is marked in this figure in units of the accretion timescale as defined earlier , which is much longer than the viscous timescale for the high kh and bc viscosities ( see [ discussregime ] ) .
evolution of the @xmath10-profile in this case is basically the increase of this nearly - uniform @xmath10-value as accretion proceeds and the wd contracts . in practice ,
our solutions in this regime start with no rotation .
as soon as angular momentum is added by means of a boundary condition , strong gradients , and hence finite viscosities , are temporarily generated in the outer layers .
the initially large viscosity quickly leads to nearly solid - body rotation with lower viscosity and slightly , but monotonically , increasing angular velocity with radius .
the viscosity in these solutions is @xmath97 @xmath98s@xmath99 , which we label an _ intermediate viscosity _
, in contrast with other solutions described below .
we emphasize that , for this regime as well as that of strong differential rotation described in the next subsection , the angular momentum transport equation was always solved self - consistently by evaluating equation 7 according to the regime of locally large or small ri at each location at a given time .
we also note that the simple " solid - body solution described in this subsection is not possible in a formulation that neglects the bci and hence is dominated by the khi . ) in the nearly - uniform rotation case with kh and bc viscosities .
the time axis is marked in units of the accretion timescale .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at various time - steps from the left panel are color coded , with time going up vertically.__,title="fig : " ] ) in the nearly - uniform rotation case with kh and bc viscosities .
the time axis is marked in units of the accretion timescale .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at various time - steps from the left panel are color coded , with time going up vertically.__,title="fig : " ] we find that strong differential rotation is also possible when viscous transport includes both kh and bc viscosities or either viscosity prescription alone.the regime of strong differential rotation corresponds to the asymptotic regime of small ri where both @xmath24 and @xmath32 scale approximately proportionally to @xmath1 as described in [ bci ] .
we find the effective viscosity in this regime to be @xmath100 to @xmath101 @xmath98s@xmath99 ( see figure [ gwdetail ] below ) that we label a regime of _ high viscosity_. a typical example is shown in detail in figures [ gwomega1 ] , [ gwomega2 ] , [ gwomega3 ] , and [ gwomega4 ] . in each figure , the left panel displays in a 3-d surface plot the evolution of the @xmath10profile as accretion proceeds . _
the time axis is marked in these figures in units of the viscous timescale , in order to study how the profile `` heals '' to an asymptotic form on this timescale , and how that form evolves as accretion proceeds on a much longer timescale .
_ the right panel of the same figure shows the same evolution as stacked profiles in a 2-d plot as time elapses , the topmost curve giving the final time - step .
we first explore the effect of sn boundary condition .
figure [ gwomega1 ] shows the approach to the asymptotic state , wherein we follow the evolution to a maximum time @xmath102 in units of the viscous time .
on the 2d plot in the right panel , the results are displayed in 10 equal steps over the time span 0 to @xmath103 , the topmost curve corresponding to @xmath103 . ) in the differential rotation case with kh and bc viscosities , and sn boundary condition ( see text ) .
the time axis is marked in units of the viscous timescale , the maximum time @xmath103 being 4 in these units .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at 10 equal time - steps are color coded , with time going up vertically.__,title="fig : " ] ) in the differential rotation case with kh and bc viscosities , and sn boundary condition ( see text ) .
the time axis is marked in units of the viscous timescale , the maximum time @xmath103 being 4 in these units .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at 10 equal time - steps are color coded , with time going up vertically.__,title="fig : " ] we show longer evolution of this profile in figure [ gwomega2 ] , where we increase @xmath103 to 500 in the same units . on the 2d plot in the right panel , the results are displayed in 10 equal steps over the time span 0 to @xmath103 , the topmost curve corresponding to @xmath103 .
we see that the approach to the asymptotic profile is rapid , attaining it practically within the first time - step . hence the profiles are almost indistinguishable from one another in the 2d plot . ) in the differential rotation case with kh and bc viscosities , and sn boundary condition ( see text ) .
the time axis is marked in units of the viscous timescale , the maximum time @xmath103 being 500 in these units .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at 10 equal time - steps are color coded , with time going up vertically.__,title="fig : " ] ) in the differential rotation case with kh and bc viscosities , and sn boundary condition ( see text ) .
the time axis is marked in units of the viscous timescale , the maximum time @xmath103 being 500 in these units .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at 10 equal time - steps are color coded , with time going up vertically.__,title="fig : " ] we show even longer evolution of this profile in figure [ gwomega3 ] , where we increase @xmath103 to 50,000 in the same units to confirm the result .
we do not repeat the 2d plot in this case , as the rapid attainment of the asymptotic profile makes the 10 equal time - step profiles completely indistinguishable from one another on this plot .
_ evolution of the angular velocity profile ( in units of s@xmath99 ) in the differential rotation case with kh and bc viscosities , and sn boundary condition ( see text ) .
the time axis is marked in units of the viscous timescale , the maximum time @xmath103 being 50,000 in these units .
only the 3-d surface plot is shown in this case , for reasons explained in the text._gwomega3 we now explore the effect of the yl boundary condition .
we mimic this boundary condition by a simple prescription detailed in appendix b , adequate for our purposes . basically , as angular momentum
is added to a non - rotating wd , the specific angular momentum at its surface , @xmath104 , increases according to the prescription @xmath105 until the the angular velocity at the surface reaches the keplerian value there . at that point
, the surface angular velocity is maintained at the keplerian angular velocity ( which changes with time as accretion continues as the wd contracts ) . in equation ( [ ylbc ] ) , @xmath106 is the keplerian value of @xmath9 at the surface of the wd of mass @xmath68and radius @xmath67 . in reality , after the surface angular velocity reaches the keplerian value for the first time , the value of the keplerian angular velocity at the surface increases as the wd contracts due to accretion , so that the actual surface angular velocity becomes sub - keplerian , and so angular momentum is added to it by the accreting mass , spinning it to the keplerian value again , and so on @xcite .
this prescription is an adequate approximation for our purposes here . in figure
[ gwomega4 ] , we show the evolution of the rotation profile with yl boundary condition and with a wd mass - radius relation described by a power - law index @xmath107 ( see above ) , which roughly describes many of the computed wd models of yoon and langer ( 2004 ; see their table 2 ) .
we follow the evolution to a maximum time @xmath108 in the same units .
we model the approach to keplerian rotation at the surface by assuming that it is attained in a time @xmath109 and maintained at keplerian thereafter . as shown in appendix b , @xmath110 , so that our choice corresponds to @xmath111 in this example .
see further discussion below . as figure [ gwomega4 ] shows , the asymptotic profile is obtained soon after keplerian rotation is reached at the surface and changes little on further evolution . ) in the differential rotation case with kh and bc viscosities , and yl boundary condition , with a mass - radius relation with power - law index @xmath112 ( see text ) .
the time axis is marked in units of the viscous timescale , the maximum time @xmath103 being 5000 in these units .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at 10 equal time - steps are color coded , with time going up vertically.__,title="fig : " ] ) in the differential rotation case with kh and bc viscosities , and yl boundary condition , with a mass - radius relation with power - law index @xmath112 ( see text ) .
the time axis is marked in units of the viscous timescale , the maximum time @xmath103 being 5000 in these units .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at 10 equal time - steps are color coded , with time going up vertically.__,title="fig : " ] the evolutions shown in figures [ gwomega1 ] , [ gwomega2 ] , [ gwomega3 ] , and [ gwomega4 ] clarify the nature of the asymptotic rotation profile and demonstrate that sn or yl boundary conditions make little difference to the final profile .
we stress at this point that we have used rather large values of @xmath113 in some of these calculations in order to simulate the effect of an accretion timescale @xmath114 which is much longer than the viscous timescale @xmath115 and demonstrate how the asymptotic profile is approached . for the large hydrodynamic viscosities generated by khi and bci ,
the actual value of @xmath116 is much larger ( see [ discussregime ] ) and is not possible in practice to achieve computationally ; however , the asymptotic approach shown above makes this unnecessary .
the profiles of the viscosities @xmath24 and @xmath32 ( in cgs units ) are shown in the left panel of figure [ gwdetail ] for the asymptotic profile , calculated self - consistently during the computation .
the viscosity is itself dependent on the @xmath10profile through the shear and hence the richardson number ( see [ hydrovisc ] ) . shown in the right panel of the same figure are the profiles of the richardson number ri and its critical value @xmath117 defined in [ hydrovisc ] , for this asymptotic profile . and its critical value @xmath117 defined in [ hydrovisc ] .
color code : @xmath118 blue , @xmath117 green.__,title="fig : " ] and its critical value @xmath117 defined in [ hydrovisc ] .
color code : @xmath118 blue , @xmath117 green.__,title="fig : " ] figure [ gwdetail ] shows that the kh viscosity is comparable to or less than the bc viscosity throughout the structure .
this suggests that the solution would not be significantly different if we were to neglect the bc viscosity and employ only the kh viscosity .
we ran such models and , as expected , found a similar differentially - rotating solution .
see [ rotyl ] for a related discussion .
note from the right panel of figure 6 that ri @xmath42 in this solution , in strong contrast to the value ri @xmath45 found in the regime of nearly solid body rotation that had negligible kh viscosity , and an intermediate value of the bc viscosity .
the secular , low , zahn viscosity ( equation ( [ zvisc ] ) ) produces a regime of differential rotation @xcite .
we show the evolution of the @xmath10-profile as accretion proceeds for the case of zahn viscosity in figure [ zahnomega ] .
the left panel displays a 3-d surface plot and the right panel shows stacked profiles in a 2-d plot , as before .
we stress that the viscous timescale @xmath115 in this low - viscosity case is generally comparable to the accretion timescale @xmath114 , a point we discuss in detail in [ discussregime ] . the time axis in figure [ zahnomega ]
is marked in units of the viscous timescale . ) with zahn viscosity .
the time axis is marked in units of the viscous timescale , which is generally comparable to the accretion timescale for the zahn viscosity .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at 10 equal time - steps are color coded , with time going up vertically.__,title="fig : " ] ) with zahn viscosity .
the time axis is marked in units of the viscous timescale , which is generally comparable to the accretion timescale for the zahn viscosity .
left panel : 3-d surface plot .
right panel : stacked profiles in a 2-d plot , wherein the profiles at 10 equal time - steps are color coded , with time going up vertically.__,title="fig : " ] the evolved zahn profile has a characteristic maximum in @xmath10 in the outer parts of the wd , with @xmath10 decreasing with increasing @xmath29 or @xmath28 in those parts of the star beyond this maximum .
the left panel of figure [ zahndetail ] shows the profile of the zahn viscosity @xmath55 ( in cgs units ) and the right panel that of the richardson number , ri , at the final time step , calculated self - consistently during the computation .
note that ri is in the low - ri regime as defined in this work .
the viscosity in this case is also dependent on the shear , through the richardson number , as in the high - viscosity case described earlier .
figure [ zahndetail ] shows that the zahn viscosity falls in a typical range of @xmath119 to @xmath120 @xmath98s@xmath99 , defined here as a regime of _ low viscosity_. we re - emphasize here that @xmath115 is generally of the same order as @xmath114 for the low zahn viscosity .
accordingly , the picture presented in [ diff ] for the situation @xmath121 , wherein the rotation profile can be thought of as attaining an `` asymptotic '' shape quickly on the viscous timescale and that shape then evolving much more slowly on the accretion timescale , does not apply here .
instead , the whole profile evolves slowly on a common timescale @xmath122 .
we return to this and related matters in [ discussregime ] .
.__,title="fig : " ] .__,title="fig : " ]
yoon and langer ( 2004 , 2005 , henceforth yl ) studied internal rotation profiles of accreting white dwarfs in a pioneering study contemporaneous with that of saio and nomoto ( 2004 ) . among the high , hydrodynamic viscosities discussed in [ hydrovisc ]
, these authors included kh viscosity , but not bc viscosity , arguing that it was unlikely to be operational inside wds ( see [ bci ] and [ discuss ] ) .
yl also included the low , secular zahn viscosity described in [ secvisc ] and [ rotzahn ] .
these authors also included mechanisms we characterized in [ elltransport ] as transporters of angular momentum in the @xmath12-direction or horizontal direction , , es circulation and gsf instability .
a typical rotation profile found by yl is shown in figure [ ylomega ] .
( in units of s@xmath99 ) in the outer parts of the wd . from yoon and langer ( 2005 ) .
_ _ ] the yl profile has a characteristic maximum in @xmath10 in the outer parts of the wd , with @xmath10 decreasing with increasing @xmath29 or @xmath28 in those parts of the star beyond this maximum . at radii smaller than that corresponding to the maximum in @xmath10 , the angular momentum transport is dominated by the kh viscosity , with
the richardson number ri maintained very close to its critical value of 1/4 ( see [ khi ] ) throughout almost all this inner region .
the effective viscosity in the inner regions of this solution is @xmath123 and hence in the range of intermediate viscosity as defined here . at larger radii , ri exceeds its critical value , khi is shut off , and the radial angular momentum transport is dominated basically by the low zahn viscosity , with es and gsf processes ( which generate low viscosities ) being also operational .
the yl regime is thus a combination of two regions , one with intermediate viscosity and the other with low viscosity , as opposed to all other regimes described in earlier sections of this paper , which were governed either entirely by high viscosity , entirely by intermediate viscosity , or entirely by low viscosity .
this leads to interesting consequences , which we discuss in [ discussregime ] .
we now give a general overview of the multiple rotation regimes for accreting wds found so far , as described above , clarifying the current understanding of the origins of these regimes .
figure [ ri_nu ] gives a schematic summary of the regimes we summarize below .
we note first that three timescales are of crucial importance in this problem .
the first , and by far the shortest , timescale is the viscous timescale for angular momentum penetration into the whole wd , @xmath124 , due to the large hydrodynamic viscosity , @xmath125 , generated by the combined action of khi and bci , as detailed in [ hydrovisc ] .
since the scale of this viscosity is very large , @xmath126 @xmath98 s@xmath99 , as seen from figures [ gwdetail ] and [ ri_nu ] , the viscous timescale , given by @xmath127 is very short . here , @xmath128 is @xmath67 in units of @xmath129 cm .
the second , much longer timescale , is the viscous timescale for angular momentum penetration into the whole wd , @xmath130 , due to the low , secular , zahn viscosity , @xmath131 described in [ secvisc ] .
since the scale of this viscosity is very low , @xmath132 @xmath98 s@xmath99 , as seen from figures [ zahndetail ] and [ ri_nu ] , the viscous timescale , given by @xmath133 is very long .
the third timescale is that on which the advection effect , as detailed in
[ angmomtransport ] , transports angular momentum into the whole wd .
the timescale for this process is simply the accretion timescale , @xmath134 in this problem . in the subsequent discussion
, we take a representative value of @xmath135 s where required for explicit calculations . we note that there is a huge separation between @xmath124 and @xmath114 , while there is a large overlap between @xmath130 and @xmath114
. a straightforward characterization of the rotation regimes that emerge from the interplay of these timescales can be given as follows .
consider first the situation where the hydrodynamic viscosity is relevant .
then the viscous timescale , @xmath124 , determines the fast , viscous evolution , leading to an asymptotic profile that can be either nearly - uniform or strongly differential rotation depending on the operational regime of ri , as detailed in
[ rothydro ] .
this profile subsequently evolves very slowly on the accretion timescale @xmath114 .
we can call this situation the _ viscosity dominated _ transport of angular momentum ( figure [ ri_nu ] ) subject to the caveats below .
the two regimes of rotation found in [ uniform ] and [ diff ] , , nearly - uniform and strongly differential rotation , arise from the fact that there are two regimes of bc viscosity , as shown by equation ( [ bcvisc ] ) and illustrated in figure [ ri_nu ] , where the scalings of @xmath32 and the viscous stress , @xmath43 , with the shear , @xmath1 , are quite different .
the former regime is that of @xmath136 , which corresponds to strong differential rotation , and the latter that of @xmath137 , which corresponds to nearly - uniform rotation .
note that the viscous stress - shear relation is non - linear in both regimes , but the degree of non - linearity is different in the two regimes , with @xmath43 scaling as @xmath44 in the former regime , but as @xmath46 in the latter .
we emphasize that it is this difference in the degree of non - linearity that causes the two different regimes of rotation .
a similar argument holds for @xmath138 in the regime of small ri , in accord with equation ( [ khvisc1 ] ) . for large values of ri , @xmath138 vanishes , of course , as the khi shuts off .
a useful way of classifying these two regimes is obtained from considerations of angular momentum and torque introduced in [ angmomtransport ] through equations ( [ transport1 ] ) and ( [ totrate ] ) , when applied to the idea of the two - phase evolution of the @xmath10profile , first on the viscous then on the accretion timescale , for high hydrodynamic viscosities ( see the beginning of [ diff ] ) .
the profile first evolves on the short viscous timescale to an asymptotic , quasi - steady state that is described by the limit of equation ( [ transport1 ] ) where the left - hand side is approximately zero , which implies that @xmath139 , , a spatially - constant @xmath140 .
this asymptotic viscous state is thus one in which the total torque @xmath140 is roughly a constant inside the white dwarf , which we can label @xmath141 . as accretion proceeds on a much longer timescale , stellar conditions evolve slowly on this long accretion timescale , and @xmath141 also evolves on the accretion timescale , corresponding to the slow evolution of the asymptotic profile on the accretion timescale .
the two regimes can now be described in terms of the two constant - torque asymptotic states in a straightforward way .
note first that the total torque @xmath140 is given by equation ( [ totrate ] ) , the first term on the right - hand side being the viscous part and the second , the advective part . in the high - viscosity regime , the first term is normally dominant .
consider first the strong differential - rotation regime , where the shear is large , ri is small , the viscosity is high , and the viscous term is thus completely dominant . here the constant torque is given approximately by @xmath142 , a large constant with a finite shear .
thus , there is strong angular momentum transport in this asymptotic state , dominated by viscous transport .
consider next the nearly - uniform rotation regime , where the shear is small , ri is large , and the viscosity takes an intermediate value as illustrated in figure [ ri_nu ] . here , both viscous and advective components are small , and so , therefore , is the constant torque , @xmath141 . in this small - torque regime ,
the angular momentum transport is very small , but non - vanishing .
a constant - torque solution corresponding to the asymptotic viscous profile is thus a generic feature of the high - viscosity regimes .
the case where the constant , @xmath141 , is very small is the nearly - uniform rotation regime , which does not appear to be otherwise special .
we emphasize that a _ strictly _
constant-@xmath10 solution is not possible , as there would be no viscous transport at all in that case : the shear would then be exactly zero , and so would be both the kh and the bc viscosities , as equations ( [ khvisc1 ] ) and ( [ bcvisc ] ) show .
a final remark about the hydrodynamic viscosities used in this work is that there may be caveats about their prescriptions that need to be studied further .
we briefly consider here those relating to the bc viscosity .
the widely - used fujimoto prescription @xcite , as given by equation ( [ bcvisc ] ) , is obtained in the limit ri @xmath2 1 @xcite .
for the regime of rotation profile described by figure [ gwdetail ] , however , this condition is not satisfied . in order to check the self - consistency of this regime ,
we have therefore gone back to the original fujimoto ( 1988 ) prescription ( see equation ( 26 ) of that paper ) , which is more complicated but valid for general values of ri , and calculated values of @xmath32 corresponding to the range of ri - values shown in figure [ gwdetail ] .
we have then compared these @xmath32-values with those in figure [ gwdetail ] , obtained with the simple prescription of equation ( [ bcvisc ] ) .
we have found that the full prescription gives @xmath32-values that are smaller than those obtained from the simple prescription by factors that range between 2 and 5 over the entire range of ri - values given in figure [ gwdetail ] .
we recall now from [ diff ] that the rotation profile in this regime is determined by the sum of kh and bc viscosities , and that the two are generally comparable in value , with @xmath24 having a higher value over most of the star , as figure [ gwdetail ] shows .
therefore , the rotation profile obtained with the more complicated bc - prescription will be qualitatively very similar to that obtained with the simpler prescription , although it will have a higher dominance of kh viscosity .
thus this regime is roughly self - consistent .
we do not adopt the more complicated fujimoto prescription of bc viscosity in this work because of other , more general , concerns about the bc viscosity ( [ discuss ] ) .
consider now the situation when the secular zahn viscosity is relevant , figure [ ri_nu ] .
since @xmath143 over most of the phase space , viscosity and advection generally play comparable roles in angular momentum transport in this case .
the result is a regime of differential rotation where the penetration of angular momentum into the interior of the wd is due to both viscosity and advection , the two terms being roughly comparable in magnitude over most of the wd interior .
we call this situation the _ viscous advective _ transport of angular momentum ( figure [ ri_nu ] ) .
similar or related situations have been studied in other areas of physics and are sometimes called _ advection
diffusion_. consider , finally , the yl regime that requires a detailed discussion because of its special , interesting nature , as indicated in [ rotyl ] .
we show first that the hierarchy of timescales established at the beginning of this section also holds for the yl regime .
this is so because the viscous timescale for angular momentum penetration into the wd , @xmath144 , due to the large hydrodynamic viscosity @xmath145 generated by khi in the yl regime pertains in the inner regions of the wd .
this regime of intermediate viscosity and short viscous time occurs at radii smaller than that at which @xmath10 has a maximum ( this region covers roughly the inner @xmath146 of the wd mass ) .
the scale of @xmath147 @xmath98 s@xmath99 can be seen in figure 8 of yoon and langer ( 2004 ) .
this is quite large , although smaller than the high - viscosity scale associated with the regime of low ri ( figure [ ri_nu ] ) , and hence in the range of intermediate viscosity as defined here .
the reduction in viscosity is due to the fact that the richardson number is very close to its critical value 1/4 ( see equations and ( [ khvisc1 ] ) and ( [ khvisc2 ] ) ) over this entire region of the yl regime .
this value of the kh viscosity leads to a viscous timescale given by @xmath148 which can be compared with zahn timescale @xmath130 and the accretion timescale @xmath114 , which are still given by equations ( [ tz ] ) and ( [ tac ] ) respectively .
the nature of the yl regime can now be clarified in a straightforward way . as emphasized by these authors ,
although the inner region of their regime does have a large , hydrodynamic viscosity and so a rapid viscous transport , the low zahn viscosity in their outer region with its slow transport ( with a comparable contribution from advection ) represents a severe `` bottleneck '' in the angular momentum transport into the wd interior @xcite .
it could be argued , therefore , that overall penetration of the accreted angular momentum into the wd can only happen on the long timescale @xmath149 .
one could then argue further that the zahn and yl regimes are rather similar in this sense , and cite the general similarity between the shapes of the zahn and yl @xmath10profiles ( even to the extent that the maximum in @xmath10 comes in the range @xmath150 of the wd mass in both cases ) in support of this point of view .
strictly speaking , however , this is not correct because there is provision for fast viscous transport in the yl case in the deep interior of the wd , once the angular momentum reaches there .
indeed , in analogy with the discussion given earlier , the question of a possible `` asymptotic profile '' in this interior yl region merits detailed study , but is outside the scope of this paper .
we thus see that , among the multiple regimes of rotation described above , many possible regimes of differential rotation exist , in addition to that of uniform rotation .
indeed , it appears that the _ generic _ regime is one of differential rotation and the uniform rotation regime may be a special case .
we have attempted to clarify the nature of the different regimes of internal rotation that can exist inside accreting wds .
we consider wds without strong , permanent magnetic fields and hence consider angular - momentum transport processes described by viscosities generated by hydrodynamic mechanisms such as kelvin helmholtz and baroclinic instabilities , and secular mechanisms such as the zahn instability .
we formulated these viscosities with the aid of prescriptions given in the literature .
we elucidate that the bc viscosity as adopted here implies two asymptotic regimes depending on the richardson number , ri .
the bc viscosity , @xmath32 , scales as @xmath151 in the regime of small ri .
this yields a regime of high viscosity that nevertheless corresponds to differential rotation , a solution presented here for the first time .
the bc viscosity scales as @xmath152 in the regime of large ri , yielding an intermediate level of viscosity and solid - body rotation .
we thus classify the collection of rotation regimes explored here in terms of the richardson number , ri , in the following way . at small values of ri
@xmath153 0.1 , we have both the low - viscosity zahn regime and the new , high - viscosity regime found in this work . near the critical value of ri @xmath154 1/4
, we have the inner , intermediate - viscosity region of the yl regime , its outer , low - viscosity region being at a higher value of ri .
employment of kh viscosity alone yields differential rotation . finally , at large values of ri @xmath2 1 , we have the intermediate - viscosity nearly - uniform rotation regime .
the two regimes of the bc viscosity correspond to a viscosity that scales with the shear as @xmath1 in the small ri regime and as @xmath3 in the large ri regime ( [ bci ] ) .
this implies , in turn , that the viscous stress , @xmath155 , scales with the shear as @xmath44 in the small ri regime , but as @xmath46 in the large ri regime .
the form of nonlinearity is a power - law with an index @xmath156 , @xmath157 being 2 in the small ri regime , and 4 in the large ri regime .
it is interesting to consider what the situation would be in a more complete theory of bc viscosity . instead of the knowledge of @xmath32 only in the two limits @xmath136 and @xmath158
, we would then have a complete prescription over the whole range of ri , which would reduce to the above limits appropriately .
the non - linearity in the viscous stress will change from a power - law with @xmath159 to one with @xmath160 as ri increases , and the rotation regime will change from a strongly differential to nearly - uniform . accordingly , the two limiting regimes will have a _ transition region _ between them , as illustrated schematically in figure [ ri_nu ] .
an insight that is particularly useful for our problem in the regimes of power - law viscosities is available from the literature on the p - laplacian nonlinear diffusion equation @xcite , whose form , @xmath161 , is very similar to that of equation ( [ transport2 ] ) in the limit where viscous transport dominates over advective transport , as is the case for the large hydrodynamic viscosities .
this similarity becomes clear if the advective term in equation ( [ transport2 ] ) is neglected and the equation is rewritten back in terms of the angular velocity @xmath10 to read @xmath162 .
the power - law index @xmath157 for the viscous stress @xmath163 is related to the index @xmath164 by @xmath165 , so that that the two asymptotic regimes correspond to @xmath166 and @xmath167 , respectively .
the non - trivial asymptotic solution of the p - laplacian pde for a _ given _ value of @xmath164 can be shown to be unique , but the solutions for _ different _ values of @xmath164 are quite different @xcite .
this is consistent with the two viscosity regimes found here .
the implication is that when @xmath164 changes from one value to another through the transition region , the solution will undergo a corresponding transition .
the specific issue of the behavior of the bc viscosity at such intermediate @xmath164- or @xmath157-values ( or , equivalently , the corresponding values of ri ) are beyond the scope of this paper , but of definite interest .
there has been discussion in the literature about the applicability of bci in the electron - degenerate matter inside wds since completely degenerate matter is barotropic , pressure is only a function of density and their gradients can not be skew to one another ( kippenhahn & mlenhoff 1974 ; tassoul 1984 ; yoon & langer 2004 ) .
realistic wds have finite temperature so the issue of baroclinicity is a quantitative one .
the key point is that buoyancy may become very small in degenerate matter , with consequences for ri and other parameters of the problem , in which case the standard bci formulation , including the fujimoto prescription , may not apply .
a first - principles reformulation of the problem may be necessary . while our immediate goal in this work has been to better understand the physics that drives the differential rotation regimes of accreting white dwarfs , our larger , long term goal is to elucidate the progenitor evolution of sn ia
. currently viable models of sn ia involve either accretion onto a wd from a non degenerate star ( the sd scenario ) , from a degenerate companion ( the dd scenario ) , or the violent impact of two degenerate stars ( maoz , mannucci & nelemans 2014 ) .
all these possibilities demand that the wd progenitor of the sn ia be rotating .
it is an interesting task to use the physics of angular momentum transport described here to elucidate and guide our understanding of the possible rotation state of the progenitors of sn ia .
a major challenge remains to determine what rotational states nature accommodates , under what circumstances , and to what effect on the subsequent explosion .
the reverse problem , , using observations of sn ia to constrain the physics of the progenitor evolution , is equally interesting and perhaps even more challenging .
while acknowledging the definite variety among typical sn ia , we must recognize that the majority of these events display a remarkable degree of homogeneity .
the fact that most sn ia are consistent with values of ejecta mass not remarkably different from @xmath5 suggests that most sn ia can not be in a regime of strong differential rotation , neither that of low ri we have presented here nor in the regime propounded by yoon & langer ( 2004 ) .
thus , it is important to understand the conditions that prevent regimes of differential rotation from occurring in common circumstances .
the answer can not be so simple as saying that the viscosity , even magnetic viscosity , is large , because we have shown here that there can be differential rotation in a high - viscosity regime .
likewise , some sn ia fall in the super - chandrasekhar regime that would seem to demand differential rotation , implying that the nearly solid - body rotation solutions of saio & nomoto ( 2004 ) and of piro ( 2008 ) do not pertain .
if many sn ia involve sub - chandra masses , as suggested by scalzo et al .
( 2014 ) and others , then the state of rotation is currently unconstrained .
the nature of internal rotational profiles in all these circumstances requires deeper understanding .
we thank michael montgomery for providing the basic white dwarf model and for helpful discussion of white dwarf structure and the referee for helpful comments which improved the paper .
some work on this paper was done in the hospitable environment of the aspen center for physics that is supported by nsf grant phy-1066293 .
g. thanks the university of texas at austin for its warm hospitality during part of this work .
the authors are also grateful for the posse east establishment where much of the conceptual development of this work was done .
this work was supported in part by nsf grant nsf ast-1109801 .
in modeling the inward mass advection rate @xmath63 in the interior of the wd , we first note that , due to mass conservation inside the wd , we can express the advection rate as @xmath168 in terms of the mass co - ordinate @xmath28 introduced in [ angmomtransport ] , and defined by @xmath169 we must then model @xmath170 for accreting wds in terms of the accretion rate @xmath74 on the wd , or , equivalently , model @xmath171 in terms of @xmath172 . for the simple case of a uniform density profile , it is trivial to show that @xmath173 for approximately self - similar contraction of the wd as its mass increases , , a contraction during which the density profile remains roughly self - similar , equation ( [ unicase ] ) still holds approximately , with the constant of proportionality replaced by a slowly varying function of radius .
the value of this function varies by a factor @xmath72 between the center and the surface of the star , so that equation ( [ unicase ] ) is approximately valid over the entire extent of the wd .
this means that @xmath171 is roughly independent of radius and roughly proportional to @xmath172 . in other words , the dominant radial variation in @xmath170
is given by that in @xmath28 ; detailed structure effects give only a slow variation .
we illustrate this point with polytropic models , used widely in studies of wds . a poytrope with polytropic index @xmath157 has an equation of state @xmath174 , and a density profile @xmath175 , where @xmath176 is the lane - emden function of index @xmath157 . here ,
@xmath177 is the central density , and the scale , @xmath178 , used to define the dimensionless radial co - ordinate @xmath179 inside the polytrope is given by @xmath180^{1/2}. \label{adefn}\ ] ] the mass of the polytropic wd is given by @xmath181^{3/2}\rho_c^{3-n\over2n}\xi_1 ^ 2 \mid\theta ' ( \xi_1)\mid , \label{mpoly}\ ] ] and its radius by @xmath182 .
here , @xmath183 is the value of @xmath184 at the surface of the polytrope ( where the density is zero ) , and @xmath185 is the value of the derivative of @xmath186 at @xmath183 .
tables of @xmath183 and @xmath187 for various values of @xmath157 are widely available .
the value of @xmath28 , as given by equation ( [ masscoord ] ) , can be readily calculated with the aid of the equations given in the previous paragraph , and this value can be related to @xmath68 as given by equation ( [ mpoly ] ) to obtain @xmath188 here , @xmath189 is an appropriate moment of the lane - emden function @xmath186
. equation ( [ mrpoly ] ) allows us to calculate the left - hand side of equation ( [ unicase ] ) for the polytropic case as @xmath190 the second term on the right - hand side of equation ( [ polycase ] ) can be evaluated with the aid of equations ( [ phidefn ] ) and ( [ xidefn ] ) as @xmath191 the relation between the total mass , @xmath68 , and the radius scale , @xmath178 , for a polytrope of index @xmath157 is well known to be @xmath192 and can be derived by combining equations ( [ adefn ] ) and ( [ mpoly ] ) , the constant in the above equation coming out to be @xmath193^{n\over n-1 } \xi_1 ^ 2\mid\theta ' ( \xi_1)\mid$ ] . equation ( [ poly2 ] ) is closely related to the mass - radius relation for polytropes . a logarithmic differentiation of it yields @xmath194 combining equations ( [ polycase ] ) , ( [ poly1 ] ) , and ( [ poly3 ] ) , we arrive at the final result : @xmath195 where the function @xmath196 is defined by @xmath197 for @xmath198 polytropes .
note that the surface of the polytrope ( where the density falls to zero ) corresponds to @xmath199 for this value of the polytropic index .
blue line : exact calculation .
green line : taylor approximation .
_ _ ] the function @xmath200 for the polytropic index @xmath198 , often used in model wd calculations , is shown in figure [ advec ] .
it is a slowly varying functon , increasing from unity at the surface of the polytrope to a value of 2 at the center . for comparison
, we have used a taylor expansion of the lane - emden function of polytropic index @xmath157 , which is given by @xmath201 to derive a series representation of @xmath202 , the result being @xmath203 , \label{taylor2}\ ] ] which reduces in the special case @xmath198 to @xmath204 this taylor approximation to @xmath200 is also shown in fig.[advec ] , keeping only the terms given in equation ( [ taylor3 ] ) , and showing thereby how close this approximation is to the exact value .
similar results are obtained for other polytropic indices , leading us to use the approximation for the advection rate given by equation ( [ advrate ] ) in [ angmomtransport ] .
as explained in [ diff ] , we envisage angular momentum being added to a non - rotating wd , so that the specific angular momentum at its surface , @xmath104 , increases according to the prescription @xmath205 where @xmath206 is the keplerian value of @xmath9 at the surface of the wd of mass @xmath68 and radius @xmath67 .
we emphasize that both @xmath68 and @xmath67here are functions of time as accretion goes on , which is of crucial importance in the relations given below .
we use the mass - radius relation @xmath207 introduced in [ angmomtransport ] , and describe the mass increase due to accretion by @xmath208 @xmath209 being the initial mass of the wd , and @xmath210 its initial radius . combining equations
( [ yls1 ] ) , ( [ ellksurf ] ) , ( [ massradius ] ) , and ( [ mincrease ] ) , and integrating with respect to time , we obtain the dimensionless specific angular momentum at the surface , @xmath211 in terms of the fiducial value @xmath212 , as a function of time in the form @xmath213 \label{ellsurf}\ ] ] here , @xmath214 is a dimensionless time , in units of the accretion timescale .
the surface angular velocity , @xmath215 is similarly given in terms of the fiducial value @xmath216 as @xmath217 .
\label{omsurf}\ ] ] the above increase with time continues until the the angular velocity at the wd surface reaches the _ instantaneous _ keplerian value @xmath218 there .
this happens when the ratio @xmath219 . \label{omsurf}\ ] ] reaches unity , which occurs at a critical value of the ( dimensionless ) time @xmath220 given by the solution of the equation @xmath221 for values of @xmath71 in the range @xmath222 , as is the case for some actual wd models in the literature ( see [ transport1 ] ) , the value of @xmath220 is close to 1 , as can be easily verified . beyond this point , the surface angular velocity is maintained at the current keplerian angular velocity , , @xmath223 is maintained at a value of unity .
since the ratio of @xmath224 to @xmath225 is seen to be @xmath226 by combining equations ( [ ellsurf ] ) and ( [ omsurf ] ) , it follows that , under these circumstances , @xmath104 is given as a function of time by @xmath227 | most viable models of type ia supernovae ( sn ia ) require the thermonuclear explosion of a carbon / oxygen white dwarf that has evolved in a binary system .
rotation could be an important aspect of any model for sn ia , whether single or double degenerate , with the white dwarf mass at , below , or above the chandrasekhar limit .
_ differential rotation _ is specifically invoked in attempts to account for the apparent excess mass in the super chandrasekhar events .
some earlier work has suggested that only uniform rotation is consistent with the expected mechanisms of angular momentum transport in white dwarfs , while others have found pronounced differential rotation .
we show that if the baroclinic instability is active in degenerate matter and the effects of magnetic fields are neglected , both nearly - uniform and strongly - differential rotation are possible .
we classify rotation regimes in terms of the richardson number , ri . at small values of ri @xmath0 0.1 , we find both the low - viscosity zahn regime with a non - monotonic angular velocity profile and a new differential rotation regime for which the viscosity is high and scales linearly with the shear , @xmath1 .
employment of kelvin - helmholtz viscosity alone yields differential rotation .
large values of ri @xmath2 1 produce a regime of nearly - uniform rotation for which the baroclinic viscosity is of intermediate value and scales as @xmath3 .
we discuss the gap in understanding of the behavior at intermediate values of ri and how observations may constrain the rotation regimes attained by nature . |
one of the most exciting developments of the past decade has been the resolution of the cosmic far - infrared background into discrete sources , providing a first glimpse of the rapid build - up of massive galaxies in the early universe long predicted by theory .
deep , wide blank field surveys at millimetre ( mm ) and submillimetre ( submm ) wavelengths @xcite have shown that ultraluminous infrared galaxies ( ulirgs ) at @xmath3 contribute significantly to the observed far - ir background .
wavelength follow up studies of these so - called submillimetre galaxies ( smgs ) suggest that they are massive , young galaxies seen during the period of rapid stellar mass build - up , with very high specific star formation rates at @xmath4 ( see review by * ? ? ?
a major obstacle in understanding the nature of this luminous dusty galaxy population is the limited angular resolution ( @xmath5 ) of current instrumentation , which prevents unambiguous identification of their counterparts at other wavelengths .
deep interferometric radio imaging and _ _
s__pitzer 24 mips imaging are shown to be effective for identifying a subset of smgs at @xmath6 @xcite , and our understanding of their nature and evolution is based almost entirely on some 100 smgs identified this way .
recent 890 continuum imaging of the seven brightest aztec @xcite 1100 sources in the cosmos field @xcite using the smithsonian submillimeter array ( sma ) by @xcite has shown that there may be a substantial population of higher redshift ( @xmath7 ) smgs that are extremely faint or undetected at radio and/or 24 mips bands .
this discovery accentuates the need for a new method to identify and investigate the higher redshift smgs in order to quantify their contribution to the cosmic energy budget and star formation history , and to map their evolution over time .
one common feature among all seven cosmos aztec / sma sources and the four other smgs detected with the sma by @xcite , @xcite , and @xcite is that they are all detected in the _ _
s__pitzer irac bands at @xmath8jy level , raising the exciting possibility that deep irac imaging may provide a powerful new tool for identifying and obtaining deeper understanding of the smg phenomenon .
the irac bands cover at @xmath9 the rest - frame optical to near - ir portion of the spectral energy distribution ( sed ) and offer direct insight to the properties of their stellar component .
a detailed analysis of the multiwavelength properties of the aztec / sma sources is presented elsewhere ( yun et al . , in prep . ) . in this paper
, we examine the rest - frame optical / near - ir properties of a large sample of well - studied smgs and the utility of the irac colour colour plots for identifying and investigating the nature of the smg population in general .
a colour colour diagram is a powerful tool for analyzing seds sampled coarsely with just a few broad - band measurements .
the four _ _ s__pitzer irac bands ( 3.6 , 4.5 , 5.8 , and 8.0 ) probe
the portion of the galaxy sed that includes photospheric emission from cool stars ( at @xmath10 ) and power - law continuum from hot dust surrounding young stars and/or agn .
polycyclic aromatic hydrocarbon ( pah ) and other spectral features are also important for the @xmath11 galaxies .
the 1.6 stellar photospheric pah feature is prominent in stellar systems older than 10 myr , and galaxies with substantial cool stellar populations appear blue in an irac colour colour diagram as a result ( see * ? ? ?
* ; * ? ? ?
sources dominated by stellar photospheric emission form a densely concentrated cloud near ( @xmath120.4 , @xmath120.4 ) in the @xmath13 vs. @xmath14 colour colour diagram for a sample of field galaxies shown by @xcite .
the @xmath11 late type galaxies with varying amounts of pah emission in the 8.0 band form a distinct branch of galaxies with a constant @xmath13 ratio emerging from this cloud . noting that 54 quasars identified in the sloan data release 1 quasar survey @xcite are associated with a second branch with red ir colours , lacy et al . identified a wedge - shaped area in the colour colour diagram as the characteristic region populated by obscured and unobscured agns . to explore the importance of agn activity among smgs
, we examine the ir properties of a sample of 47 well studied smgs and compare them to agns using the agn diagnostic irac colour colour plot by @xcite .
our smg sample is constructed from the literature primarily for their secure identification using high angular resolution interferometric imaging at radio wavelengths and deep _ _
s__pitzer mips imaging @xcite .
a small subset of this sample has been observed with the _ _
s__pitzer infrared spectrograph ( irs ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and those classified as `` starburst '' ( strong pah features ) and `` starburst+agn '' ( pah plus power - law continuum ) are shown using different symbols in figure [ fig : colourcolourplot ] . of these 20 submm - selected galaxies
, none shows a pure power - law - dominated irs spectrum characteristic of an agn .
we compare the ir colours of our smg sample to those of the 19 @xmath15 agns identified by their power - law continuum in the first look survey ( fls ) field by @xcite , and to the ir colours of the foreground galaxy population , using the 4000 irac sources randomly selected from the cosmos field @xcite .
the irac colours of these foreground sources are slightly offset from the fls centroid ( indicated in figure [ fig : colourcolourplot ] by a large circle ) , probably because the cosmos sources are on average fainter and at higher redshift .
a surprising result is that more than 90% of the smgs are located within the region designated for agns by @xcite .
it is tempting to interpret this result as an indication that the majority of smgs host a luminous agn , as @xcite concluded for their mambo - selected smgs in the field surrounding the @xmath16 radio galaxy 4c 41.17 .
however , the majority ( 13 out of 20 , or 65% ) of smgs observed with _ _
s__pitzer irs show spectra dominated by pah features , characteristic of pure starburst systems .
the remaining seven smgs show starburst+agn hybrid spectra , suggesting that the power - law agn emission is not the dominant contributor to the total ir luminosity of these smgs either .
four of the smgs targeted by @xcite were undetected by their _ _
s__pitzer irs observations .
deep chandra x - ray data in the goods - north field suggest that only a small fraction of smgs are detected in the hard x - ray band @xcite , further supporting the notion that energetically - dominant , luminous agns are not common among the smgs ( also see * ? ? ? * ) .
another commonly used diagnostic for agn activity is the _
spitzer_/irac [ 5.8][8.0 ] vs. [ 3.6][4.5 ] colour colour diagram , first introduced by @xcite .
in addition to red colours , stern et al . utilize empirical colour tracks of star - forming galaxies ( e.g. , m82 ) to further differentiate starbursts from agns .
we reproduce the stern et al .
plot in figure [ fig : colourcolourplot2 ] using the same samples as used in figure [ fig : colourcolourplot ] .
about 1/3 of the smgs ( 16/47 ) , including 5 out of 13 `` starburst '' irs spectrum sources , are now found outside the agn region outlined by stern et al . however , a large fraction of the smgs ( 31/47 ) , including many with `` starburst '' irs spectrum , are within the agn region .
a natural explanation for the observed red ir colours of smgs emerges when empirical and theoretical sed models of dusty starbursts are examined in the context of these irac colour colour plots .
the extreme ir luminosity of the mm / submm detected sources ( @xmath17 ) indicates that a dust - obscured extreme starburst or a luminous agn dominates their energy output , and the entire observed sed can be described by a relatively simple model ( see * ? ? ?
theoretically motivated radiative transfer models for dust - enshrouded starbursts and agns @xcite are successful in reproducing the observed uv - to - radio seds , including those of extreme objects such as smgs and `` hyperluminous '' ir galaxies not commonly found in the local universe ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
a large number of free parameters ( e.g. , source geometry , initial mass function , metallicity ) and some degeneracy among them limit the utility of these sed models for unique quantitative interpretations of the photometric data . nevertheless , they are highly valuable tools for evaluating the effects of model parameters such as starburst age and extinction . for this work ,
we adopt the model seds computed by @xcite for their ease of use and because they have already been widely tested ( e.g. * ? ? ?
* ; * ? ? ?
qso sstxfl j172253.9 + 582955 @xcite and the @xmath18 smg smm j123707.7 + 621411 ( gn19 ; * ? ? ?
the model seds are normalized to roughly match the measured flux densities of gn19 in the irac bands .
the sed of sstxfl j172253.9 + 582955 is redshifted to @xmath18 to match that of gn19 for an easier comparison . ]
the efstathiou et al .
sed models are shown in figure [ fig : seds ] along with the observed seds of the @xmath18 smg j123707.7 + 621411 @xcite and the @xmath19 qso sstxfl j172253.9 + 582955 @xcite in order to illustrate the irac colour evolution with starburst age .
the shape of the model sed changes quickly during the first 100 myr , primarily driven by the rapid evolution of the young stellar population .
as the stellar population ages , the combined effect of decreasing radiation intensity and increasing photospheric emission ( `` 1.6 bump '' ) makes the seds flatten ( become bluer ) in the irac bands .
these are generic features of all sed models , only dependent on the details of the input stellar population synthesis model .
even though the build - up of cool stars and the 1.6 photospheric component become evident as early as @xmath2030 myr after the initial starburst , the overall colour of starburst systems remains _ red _ even at 64 myr after the burst . in comparison , the red continuum of the ir agn increases monotonically across the irac bands into the mid - ir ( mips 24 ) band , with a steeper slope than that of the smg and most of the theoretical starburst seds .
the comparison shown in figure [ fig : seds ] nicely demonstrates the clear difference in the origin of their near - ir emission and the spectral slope between ir agns and smgs , despite the broad similarity in their red irac colours .
three colour tracks for a single starburst population sed model with different ages are shown in figures [ fig : colourcolourplot]c & [ fig : colourcolourplot2]c .
they cover the full range of irac colours associated with both agns and smgs , with the older starburst tracks showing successively _ bluer _ irac colours .
the majority of smgs appear scattered about the model sed colour tracks , consistent with their _ _
s__pitzer irs spectra being characteristic of starburst - dominated systems . for a given model
sed , the irac colour becomes monotonically redder with increasing redshift at @xmath21 , and most smgs have colours consistent with model seds redshifted to @xmath22 .
the colour dependencies on starburst age and redshift are nearly parallel to each other , leading to some degeneracy between the two quantities .
nevertheless , the observed red irac colours of smgs can arise _ only if smgs are at high redshift _ ( @xmath21 ) .
the effects of extinction ( column density along the line of sight ) are not as important in determining the observed irac colours of smgs , unlike the case at shorter , optical wavelengths .
the model colour tracks for two different visual extinctions ( @xmath23 and 200 ) shown in figures [ fig : colourcolourplot]d & [ fig : colourcolourplot2]d track each other closely at @xmath24 , despite the large opacity difference between the two models .
optical depth is greatly reduced at these long wavelengths ( @xmath25 for the irac bands ; see * ? ? ?
* ; * ? ? ?
* ) , and the weak dependence on extinction @xmath26 can be naturally understood . as redshift increases ,
the irac bands begin to probe the near - ir to optical bands , and an increasing dependence on extinction is expected .
indeed the irac colours of the two model seds diverge at @xmath1 , with the higher extinction @xmath27 model predicting redder irac colours as expected ( see fig .
[ fig : colourcolourplot]d ) , and some degeneracy between redshift and extinction is also identified .
overall , extinction still plays a minor role compared with starburst age , and we identify starburst age and redshift as the dominant physical parameters that affect the observed irac colours . accounting for smgs with the reddest observed colours ( e.g. , log(@xmath13)@xmath280.3 , log(@xmath14)@xmath280.3 ) requires young ( @xmath29 myr old ) starbursts at high redshifts ( @xmath30 ) or a power - law agn dominating the rest - frame near - ir sed .
an agn - like warm ir colour is a generic feature for _ all _ stellar systems whose near - ir luminosity is dominated by a dust - obscured young stellar population . in analyzing the `` mid - ir excess '' of galaxies selected using the infrared astronomy satellite ( iras ) data , @xcite noted that the youngest dusty starbursts can display warm mid - ir colours , exceeding the classic seyfert division at log @xmath31 @xcite . while exploring the presence of power - law agn candidates in the chandra deep field north region
, @xcite also noted the incursions by their empirical ulirg colour tracks ( derived from the observed seds of arp 220 , iras 17208@xmath120014 , and mrk 273 ) into the agn region in their irac colour colour diagram .
the incursion of m82-like objects into the agn boundary has lead @xcite to refine the selection boundary , and @xcite question the completeness and reliability of agn identification using irac colours for a similar reason . using theoretical and empirical starburst sed models
, we demonstrate that young , dusty starbursts exhibit red irac colours , and red irac colour is _ not _ unique to power - law agns .
the popular agn identification methods using red irac colours , such as by @xcite and @xcite , should be used with caution and a clear understanding of this important caveat .
both smgs and agns appear within the broadly defined red irac colour regions by lacy et al . and stern et al . , but smgs and agns are distinguished by clear systematic difference in their mean irac colours .
the smgs as a group are systematically bluer than the agns , with only about 20% of smgs appearing mixed in among the power - law agns in both diagnostic colour colour plots .
all agns show log(@xmath14 ) @xmath32 in figure [ fig : colourcolourplot ] while the overwhelming majority ( @xmath33 ) of the smgs have a smaller flux ratio .
similarly , all agns have [ 5.8][8.0 ] colours redder than + 0.2 in figure [ fig : colourcolourplot2 ] .
the most significant colour difference is between the 4.5 and 8.0 bands with an average difference of @xmath34-[8.0]\rangle\ga 1 $ ] mag .
the [ 3.6][4.5 ] colour difference is the smallest , with @xmath35-[4.5]\rangle \sim 0.25 $ ] mag .
the comparison of the full irac band seds shown in figure [ fig : iracseds ] offers a visually compelling demonstration of a steeper spectral slope for the agns .
the average irac sed of the agn sample is in excellent agreement with the empirical type-2 qso sed template by @xcite .
the template sed for a type-1 qso is nearly identical ( not shown ) to type-2 qso at these wavelengths , and the steep spectral slope is characteristic of all obscured and unobscured agns . in comparison ,
the average irac sed of the smg sample is significantly flatter , with only a few outliers with an agn - like steeper spectral slope .
the ulirg arp 220 sed redshift to @xmath36 offers a far better match to the smgs , lending further support for the dust obscured starburst interpretation .
furthermore , the flatter spectral slope ( bluer irac colours ) suggests that smgs in general do not host a dust - obscured , energetically dominant agn in most cases .
we note that rigorous diagnostics of an agn include the detection of high - ionization , high - excitation emission lines and copious amounts of hard x - ray emission , and an ir power - law spectrum is only indirect evidence for the presence of an agn , and does not directly measure the agn accretion power .
the distinct difference in the observed seds between agns and smgs also offers an important constraint on the possible smg - qso evolutionary scenario . the apparent correlation observed between the black hole mass ( @xmath37 ) and the host spheroid mass ( velocity dispersion @xmath38 ;
see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) has raised considerable interests in the process that leads to the build - up of the stellar mass in galaxies and the growth of the central super - massive black hole .
the evolutionary scenario that a massive nuclear starburst associated with an ultraluminous infrared galaxy leads to a qso phase @xcite offers an attractive explanation for the @xmath39 relation if the smg phase corresponds to the period of rapid stellar mass build - up .
however , only the minor overlap between the agns and the smgs we find here suggests that the duration of the transition period should be much shorter than the smg or the ir agn phase in such an evolutionary scenario . &
[ fig : colourcolourplot2 ] .
all seds are normalized at 4.5 band , and the agns and smgs are displaced vertically by an arbitrary amount for ease of comparison .
the @xmath36 seds of type-2 qso and arp 220 ( m82 sed is nearly identical ) compiled by @xcite are shown for comparison , with an arbitrary offset in the flux scaling . ]
as discussed above , every smg identified with high angular resolution interferometric imaging using the sma has a clear irac counterpart , including those undetected in the radio and in the mips 24 bands . _ _
s__pitzer irac data may therefore offer a powerful new method for identification of smgs that is otherwise extremely difficult because of their faintness at optical and near - ir wavelengths .
the uncertainty in the mm - wave positions of smgs is typically 5 - 10 , and the high irac source density ( @xmath40 arcmin@xmath41 at the 1.4 @xmath42jy level in the 3.6 band ; * ? ? ?
* ) is too high to enable unique identification of smg counterparts in general ( see discussions by * ? ? ?
the new knowledge obtained from figures [ fig : colourcolourplot ] & [ fig : colourcolourplot2 ] that smgs as a population have red ir colours , similar to agns and distinct from the foreground field population , offers the exciting possibility of identifying or significantly narrowing down the counterpart candidates using the irac data alone . to evaluate the utility of these new irac colour selection criteria for the smg counterpart candidate selection , we examine the colours of the irac sources in the fields surrounding the nine smgs securely identified through submm interferometric imaging @xcite . the 20 @xmath46 20 irac 3.6 images ( not shown ) centered on the original submillimetre source
coordinates contain up to a dozen irac sources each , demonstrating the difficulty of using the irac source catalog alone for the smg counterpart identification .
there are a total of 20 irac sources located within a 6 radius ( typical positional accuracy ) of the nine nominal smg positions ; their colours are shown in figures [ fig : colourcolourplot]b & [ fig : colourcolourplot2]b .
colours of the nine smgs , shown as filled circles , fall within our proposed smg colour boundaries in both plots .
these smgs were not used in defining the irac colour selection criteria and they thus represent an independent and successful test of our method .
the remaining irac sources have colours consistent with the background sources , and about 1/2 of them fall within the new smg candidate identification boundaries .
the new colour selection technique does not completely resolve the confusion problem .
however , the situation is now greatly improved , as the candidate counterparts are reduced to a single unique candidate or two potential candidates , making expensive follow - up observations more efficient .
other identification methods such as deep radio or mips 24 imaging do not greatly improve the situation in these test cases since many of these smgs are undetected at other wavelengths .
three of the securely identified sources have irac colours typical of @xmath47 starburst systems while the remaining six cluster near the power - law agns ( see fig . [
fig : colourcolourplot]b & [ fig : colourcolourplot2]b ) .
in addition to being undetected in the radio , these six smgs with agn - like colour are undetected in the _ _
s__pitzer mips 24 bands , ruling out the presence of a power - law agn component that extends into the mid - ir bands , as seen for the qso spectrum shown in figure [ fig : seds ] .
therefore these smgs with very red irac colours may be the youngest , highly - obscured starburst systems at very high redshifts ( @xmath30 ; also see the discussions in * ? ? ?
the irac colour selection method described here is qualitatively similar to the irac 8.0 band selection method used to identify the scuba galaxies in the cudss 14 hour field by @xcite . noting the red colour of the smgs and reduced foreground confusion compared with the shorter wavelength irac bands , ashby et al .
argued that 8.0 selection offers a better means for identifying smg counterparts than near - ir or optical selection .
their 8.0 selection resulted in a different counterpart from the optical or @xmath48-band selection methods in ten out of 17 cases , casting some doubt on the earlier identifications of smgs .
the main difference of the new irac colour selection method outlined here is that we fully quantify the red colour of the smg population using eqs . 1 & 2 for easy implementation and calibrate them using a large , well - studied sample in conjunction with empirical and theoretical dusty starburst seds .
high - redshift submillimetre - bright galaxies identified by blank field surveys at millimetre and submillimetre wavelengths appear in the region of the irac colour colour diagrams previously identified as the domain of luminous agns by @xcite and @xcite . rather than interpreting this as a sign that the majority of the smgs are powered by a luminous agn
, we have shown using empirical and theoretically motivated dusty starburst sed models that their ir colours can be interpreted as those of power - law continuum associated with hot dust heated by young ( @xmath0 myr old ) , extreme _ starbursts _ at @xmath1 .
these smgs fall along the branch of galaxies extending from the blue photospheric peak toward the red power - law region in the irac colour colour plot by @xcite , and our analysis suggests that this branch is a heterogeneous ensemble of power - law agns and dust obscured starbursts whose continuum near - ir luminosity exceeds that of the photospheric emission from cool stars . in fact , our analysis demonstrates that the popular red irac colour selection methods for agn identification , such as the criteria by lacy et al . and stern et al .
, should be used with caution because of the significant starburst contribution expected . while there is some overlap between smgs and agns in these irac colour colour plots , smgs are systematically bluer ( @xmath34-[8.0]\rangle\ga 1 $ ] mag ) , consistent with 30 to 70 myr old starbursts observed at redshifts between @xmath15 and @xmath49 .
our examination of the model dusty starburst seds of @xcite shows that the main physical parameters that determine the observed irac colours are starburst age and redshift , while extinction ( column density ) plays a less important role . for the models examined in figures [ fig : colourcolourplot ] and [ fig : colourcolourplot2 ] , starburst
age and redshift are nearly degenerate .
the use of irac colours as a redshift indicator was first proposed by @xcite and @xcite , and various empirical photometric redshift relations have been proposed recently with estimated accuracies of @xmath50 ( see * ? ? ?
* ; * ? ? ?
this photometric redshift method may not work well for galaxies with red power - law continuum and/or weak 1.6 photospheric feature that are more common at @xmath51 .
the degeneracy between the starburst age and redshift found with these theoretical model seds suggests that a careful calibration using a large sample of smgs is needed in order to quantify fully the systematic uncertainty of this method .
a detailed comparison of the irac colour colour plots and seds shows that agns and smgs are distinct from each other due to intrinsic differences in their energy source and dust distribution .
smgs as a group have a flatter sed ( bluer by @xmath34-[8.0]\rangle\ga 1 $ ] mag ) in comparison with agns .
only 20% of the objects overlap in the colour colour plots shown in figures [ fig : colourcolourplot ] & [ fig : colourcolourplot2 ] , and this suggests that smgs powered by an agn are not common . in the context of the ulirg - qso evolutionary scenario @xcite , the little overlap between the agns and the smgs population may indicate that the transition period is much shorter than the duration of the smg or the ir agn phase .
the red ir colours of the smgs are distinct from the colours of the majority of the foreground irac sources , which are both early and late type galaxies at @xmath24 .
we show that colour selection criteria similar to those of agns proposed by lacy et al . and stern et al .
can be used to pinpoint the irac counterpart to an smg uniquely or to pare down the candidates to just a few , improving the efficiency of expensive follow - up observations at other wavelengths .
the lack of high - quality multi - wavelength data for a large sample of smgs is currently the primary limiting factor that prevents a better understanding of the smg phenomenon .
the irac colour selection method discussed here appears to work even in cases when the radio and mips 24 counterparts are too faint to be detected at the present sensitivity , and an unbiased investigation of the entire smg population may become possible using this new identification method .
there are some important advantages and disadvantages to using the ( @xmath13 vs. @xmath14 ) colour colour plot ( figure [ fig : colourcolourplot ] ) versus the ( [ 5.8][8.0 ] vs. [ 3.6][4.5 ] ) colour colour plot ( figure [ fig : colourcolourplot2 ] ) for smg candidate identification .
the former uses ir colours with a longer baseline in wavelength , making the colour measurements more robust and the model colour tracks better behaved , as shown in figure [ fig : colourcolourplot ] . on the other hand ,
the instrumental sensitivity of the irac 3.6 and 4.5 bands are a factor of a few better than the 5.8 and 8.0 bands , and the 5.8 and 8.0 channels will be inoperable during the upcoming warm _ _
s__pitzer operation .
the model colour tracks shown in figure [ fig : colourcolourplot2 ] are more irregular because of shorter wavelength baselines in these colours , but the [ 3.6][4.5 ] colour is the best determined quantity among all irac colour combinations .
a far simpler and potentially more robust alternative smg candidate selection criterion is [ 3.6][4.5 ] @xmath52 .
this single colour selection method does only a slightly poorer job of rejecting foreground sources than the full colour section criteria described by eq . 2 .
given the superior sensitivity of the irac 3.6 and 4.5 bands and their availability during the warm _ _
s__pitzer mission , this simpler colour selection criterion may be the more effective method for identifying smg counterpart candidates in the longer term .
alternatively , colours of irac to optical or near - ir bands or to mips 24 bands have also been proposed previously for identifying agn activity ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and similar analysis may also prove fruitful for smg candidate identification .
we thank m. poletta for providing us with the library of empirical galaxy sed templates used in figure [ fig : iracseds ] . this work is based on observations made with the _ spitzer _ space telescope , which is operated by the jet propulsion laboratory , california institute of technology , under nasa contact 1407 .
this work is also partially funded by nsf grant ast 05 - 40852 to the five college radio astronomy observatory .
99 alexander , d. m. et al .
2003 , aj , 125 , 383 aretxaga , i. , et al .
2003 , mnras , 342 , 759 aretxaga , i. et al .
2007 , mnras , 379 , 1571 ashby , m.l.n .
, dye , s. , huang , j .- s . ,
eales , s. , willner , s.p . ,
et al . 2006 , apj , 644 , 778 barger , a.j . , et al .
1998 , nature , 394 , 248 barmby , p. et al . 2006 , apj , 642 , 126 bertoldi , f. , carilli , c. , aravena , m. , schinnerer , e. , voss , h. et al .
2007 , apjs , 172 , 132 blain , a.w . ,
smail , i. , ivison , r.j . , kneib , j .-
, frayer , d.t .
2002 , _ physics reports _ , 369 , 111 chakrabarti , s. , & whitney , b. 2008 , apj , submitted ( astro - ph/0711.4361 ) chapman , s.c . ,
windhorst , r. , odewahn , s. , yan , h. , conselice , c. 2003 , apj , 599 , 92 chapman , s.c . ,
blain , a.w .
, smail , i. , ivison , r.j .
2005 , apj , 622 , 772 dannerbauer , d. , walter , f. , & morrison , g. 2008 , apj , 673 , l127 de grijp , m. h. k. , miley , g. k. , lub , j. , de jong , t. 1985 , nature , 314 , 240 donley , j. l. , rieke , g. h. , prez - gonzlez , p. g. , rigby , j. r. , alonso - herrero , a. 2007 , apj , 660 , 167 eales , s. , lilly , s. , gear , w. , dunne , l. , bond , j.r . , et al . 1999 ,
apj , 515 , 518 efstathiou , a. , rowan - robinson , m. , siebenmorgan , r. 2000 , mnras , 313 , 734 efstathou , a. , & rowan - robinson , m. 2003 , mnras , 343 , 322 egami , e. , dole , h. , huang , j .- s .
, perez - gonzalez , p. , le floch , e. , et al .
2004 , apjs , 154 , 130 farrah , d. , et al .
2003 , mnras , 343 , 585 fazio , g. g. et al .
2004 , apjs , 154 , 39 ferrarese , l. , merritt , d. , 2000 , apj , 539 , 9 gebhardt , k. , bender , r. , bower , g. , dressler , a. , faber , s. m. et al .
2000 , apj , 539 , 13 greve , t.r .
2004 , mnras , 354 , 779 greve , t.r .
, et al . 2005 , mnras , 359 , 1165 greve , t. r. , et al .
2007 , mnras , 382 , 48 groves , b. , et al . 2008 , apjs , in press ( astro - ph/0712.1824 ) huang , j .- s . , et al .
2004 , apjs , 154 , 44 hughes , d.h .
, et al . 1998 , nature , 394 , 241 indebetouw , r. , mathis , j. s. , babler , b. l. , meade , m. r. , watson , c. et al .
2007 , apj , 619 , 931 iono , d. , peck , a.b . ,
pope , a. , borys , c. , scott , d. , et al .
2006 , apj , 640 , l1 ivison , r. j. et al .
2002 , mnras , 337 , 1 ivison , r. j. et al . 2005 , mnras , 364 , 1025 lacy , m. , et al .
2004 , apjs , 154 , 166 laurent , g. t. , aguirre , j. e. , glenn , j. , ade , p. a. r. , bock , j. j. , et al .
2006 , apj , 623 , 742 magorrian , j. , tremaine , s. , richstone , d. , bender , r. , bower , g. et al .
1998 , aj , 115 , 2285 martnez - sansigre , a. , lacy , m. , sajina , a. , rawlings , s. 2008 , apj , 674 , 676 menndez - delmestre , k. , et al .
2007 , apj , 655 , l65 mortier , a.m.j . ,
serjeant , s. , dunlop , j.s . , scott , s.e . , ade , p. , et al .
2005 , mnras , 363 , 563 norman , c. , scoville , n. z. , 1988 , apj , 332 , 124 polletta , m. , et al .
2007 , apj , 663 , 81 pope , a. , scott , dickinson , m. , chary , r.r . ,
morrison , g. , d. , et al .
2006 , mnras , 370 , 1185 pope , a. , chary , r .-
r . , alexander , d. m. , armus , l. , dickinson , m. , et al .
2008 , apj , 675 , 1171 rigby , j. r. , marcillac , d. , egami , e. , rieke , g. h. , richard , j. , et al .
2008 , apj , 675 , 262 romn - zniga , c. g. , lada , c. j. , muench , a. , alves , j. f. , 2007 , apj , 664 , 357 sawicki , m. 2002 , aj , 124 , 3050 sanders , d.b . ,
soifer , b. t. , elias , j. h. , madore , b. f. , matthews , k. , neugebauer , g. , scoville , n. z. , 1988 , apj , 325 , 74 sanders , d.b .
, salvato , m. , aussel , h. , ilbert , o. , scoville , n. et al .
2007 , apjs , 86 schneider , d. p. et al .
2003 , aj , 126 , 2579 scott , s. e. , fox , m. j. , dunlop , j. s. , serjeant , s. , peacock , j. a. , et al .
2002 , mnras , 331 , 817 scott , k. s. , et al .
2008 , mnras , in press ( astro - ph/0801.2779 ) scoville , n. z. , et al .
2007 , apjs , 172 , 1 siebenmorgen , r. , krgel , e. 2007 , a@xmath53a , 461 , 445 silva , l. , granato , g. l. , bressan , a. , danese , l. 1998 , apj , 509 , 103 simpson , c. , eisenhardt , p. 1999
, pasp , 111 , 691 smail , i. , ivison , r. j. , & blain , a. w. 1997 , apj , 490 , l5 stern , d. , eisenhardt , p. , gorjian , v. , kochanek , c. s. , caldwell , n. , et al .
2005 , apj , 631 , 163 tacconi , l.j . ,
neri , r. , chapman , s.c . ,
genzel , r. , smail , i. , et al .
2006 , apj , 640 , 228 valiante , e. , lutz , d. , sturm , e. , genzel , r. , tacconi , l. j. , et al .
2007 , apj , 660 , 1060 vega , o. , clemens , m. s. , bressan , a. , granato , g. l. , silva , l. , panuzzo , p. 2008 , a@xmath53a , submitted ( astro - ph/0712.1202 ) wang , w .- h . , cowie , l.l . ,
barger , a.j .
2004 , apj , 613 , 655 wang , w .- h . , cowie , l.l . ,
barger , a.j .
2006 , apj , 647 , 74 wang , w .- h . ,
2007 , apj , 670 , l89 webb , t. m. a. , et al .
2006 , apj , 636 , l17 wilson , g. w. , et al . 2008a , mnras , in press ( astro - ph/0801.2783 ) wilson , g. w. , et al .
2008b , mnras , submitted ( astro - ph/0803.3462 ) younger , j. d. , fazio , g. g. , huang , j .- s .
, yun , m. s. , wilson , g. w. et al . 2007 , apj , 671 , 1531 younger , j. d. , et al .
2008 , mnras , in press ( astro - ph/0801.2764 ) yun , m.s . ,
reddy , n. , & condon , j.j .
2001 , apj , 554 , 803 yun , m.s . , & carilli , c.l .
2002 , apj , 568 , 88 | high - redshift submillimetre - bright galaxies identified by blank field surveys at millimetre and submillimetre wavelengths appear in the region of the irac colour colour diagrams previously identified as the domain of luminous active galactic nuclei ( agns ) .
our analysis using a set of empirical and theoretical dusty starburst spectral energy distribution ( sed ) models shows that power - law continuum sources associated with hot dust heated by young ( @xmath0 myr old ) , extreme starbursts at @xmath1 also occupy the same general area as agns in the irac colour colour plots .
a detailed comparison of the irac colours and seds demonstrates that the two populations are distinct from each other , with submillimetre - bright galaxies having a systematically flatter irac spectrum ( @xmath2 mag bluer in the observed [ 4.5][8.0 ] colour ) .
only about 20% of the objects overlap in the colour colour plots , and this low fraction suggests that submillimetre galaxies powered by a dust - obscured agn are not common .
the red ir colours of the submillimetre galaxies are distinct from those of the ubiquitous foreground irac sources , and we propose a set of ir colour selection criteria for identifying smg counterparts that can be used even in the absence of radio or _ _
s__pitzer mips 24 data .
[ firstpage ] cosmology : observations galaxies : evolution galaxies : high redshift galaxies : starburst galaxies : active infrared : galaxies . |
quantum dots ( qds ) have long been discussed as a source for semiconductor - based quantum information processing @xcite . for a physical realization of a quantum computer , a quantum mechanical two - level system that can serve as a qubit is required .
this system must allow for preparation , manipulation , and readout of its quantum state @xcite .
furthermore , a coupling mechanism between the qubits is needed to allow for conditional operations on multiple qubits .
other interactions , especially damping mechanisms that affect the two - level system , have to be minimized . as representations of qubits , excitonic states in single or double qds as well as spin states have been proposed @xcite .
quantum information schemes in qd systems are often based on the coherent optical control of the electronic states , taking advantage of the femtosecond time scales common in current ultrafast laser optics @xcite .
the coulomb interaction is one of the coupling mechanisms that has been proven to allow for conditional operations on qd qubits @xcite .
however , it also influences the optical manipulation of the qd states , especially for small interdot distances where the coulomb interaction between neighbouring qds is an important coupling mechanism @xcite .
therefore , it affects the applicability of any quantum information scheme based on qds , even if the scheme does not utilize the coulomb interaction for conditional operations . here , we want to quantify this influence and discuss the implications for qd based quantum information processing .
the key quantity to describe successful quantum information processing in a physical system is the fidelity with which gates can be implemented on the considered set of qubits @xcite . for a basic evaluation of a given proposal for quantum information processing , the fidelity of a universal set of gates , consisting , e.g. , of single - qubit gates and the cnot gate , has to be investigated .
fidelity calculations have been presented for single - qubit gates @xcite and two - qubit gates based on spin states in single @xcite and coupled quantum dots @xcite as well as excitonic states in single qds @xcite .
the fidelity of two - qubit gates in coulomb - coupled qds , especially with entangled output states , has not been discussed in the literature . in this article , we present calculations of the fidelity of a single- and a two - qubit gate implemented on coulomb - coupled qds .
we focus on a scheme utilizing the coulomb interaction for the implementation of a controlled two - qubit gate , namely the cnot gate @xcite .
the quantum operations are implemented using coherent light pulses resonant to the excitonic transitions of the system . by applying a series of coherent laser pulses with fixed areas ,
special quantum information gates can be realized . in particular , we consider the generation of bell states in this system .
we consider a system consisting of two qds , interacting with a coherent light field and with each other via the coulomb interaction .
it is assumed that the wave functions of the electrons in different dots do not overlap , i.e. , we neglect interdot tunneling .
even though we focus on the low temperature regime ( a few kelvin ) , dephasing effects due to the interaction with phonons can be of importance , depending strongly on the pulse length and detuning of the dot levels @xcite . since the resulting errors are typically an order of magnitude smaller than the one arising from radiative dephasing , which is studied in this work , we do not considered electron - phonon coupling in the following . following the proposition in ref .
@xcite , the qubits are represented by the single - excitonic states in each qd , one state being the qd ground state , the other being the lowest exciton state ( exciton ground state ) .
the reason is as follows : due to the exchange interaction , the two heavy hole and two electronic states closest to the fermi edge typically show fine structure splitting into two bright and dark excitonic states each , which are energetically split depending on the symmetries of the qds @xcite .
one of the two bright states can be selectively excited via a suitable choice of pulse polarization .
we neglect the impact of spin - flip in the following , which is typically a good approximation in inas / gaas qds @xcite , where the spin - orbit coupling is weak @xcite .
then , due to spin preservation , coulomb - induced transitions to dark states are suppressed , and we can neglect the influence of dark states on the radiative lifetime of the considered exciton states @xcite .
therefore , we restrict the model of each qd to a two - level system in the following and neglect the spin degree of freedom , as was done , e.g. , in ref .
the states are denoted by @xmath2 and @xmath3 .
for the two qds , this leads to a four - level system with the basis @xmath4 , where @xmath5 denotes the presence (= 1 ) or lack (= 0 ) of an exciton in qd @xmath6 and @xmath7 , respectively .
the hamiltonian of the system then reads @xmath8 the noninteracting electrons in effective mass approximation are described by @xmath9 with the energy levels @xmath10 of the electrons in the valence ( @xmath11 ) and conduction band ( @xmath12 ) in qd @xmath13 and the electron creation and annihilation operators @xmath14 and @xmath15 , respectively .
the hamiltonian describing the interaction with the light field is given by @xmath16 with the rabi frequency @xmath17 , the classical light field @xmath18 , and the ( real ) dipole matrix element @xmath19 .
the pulse area is defined via the envelope @xmath20 of the pulse @xmath21 , where @xmath22 is the pulse frequency : @xmath23 the coulomb interaction between the electrons within the above approximations is described by @xcite @xmath24 with the coulomb matrix elements @xmath25 . when treated in dipole approximation
, the coulomb interaction reduces to two parts represented by a nondiagonal matrix element @xmath26 and a diagonal matrix element @xmath27@xcite . the nondiagonal matrix element @xmath28 couples the two excited single - excitonic states , leading to the well - known effect of frster energy transfer @xcite .
the diagonal element @xmath29 shifts the energy of the state @xmath30 and is usually denoted as the biexcitonic shift . defining the matrix elements of the hamiltonian with respect to the basis @xmath31 via @xmath32 ( cf . ref .
@xcite ) , the system hamiltonian in matrix representation takes the form @xmath33 where @xmath34 denotes the energy of the ground state @xmath35 and @xmath36 , @xmath37 are the energies of the single excitons in the two qds .
we consider a system with different qd resonance energies @xmath38 with the difference of the single - excitonic energies in the qds @xmath39 [ cf .
[ fig:2](left ) ] . with a suitable choice of @xmath40 ,
this allows for selective excitation of the individual qds which is necessary for the proposed quantum information schemes . in the following ,
we refer to the energetically lower qd as the first qd and the energetically higher qd as the second qd . starting with this hamiltonian
, we calculate the heisenberg equations of motion for the density matrix elements of the system in rotating wave approximation @xcite . to allow for radiative damping of the system , the interaction of the electrons with the quantized light field , treated in born - markov approximation ,
is included in the calculation .
this results in radiative damping terms of the electronic density matrix elements .
the full set of equations is discussed in ref .
@xcite .
the hamiltonian of the system can be diagonalized with respect to the coulomb interaction , leading to new eigenstates and eigenvalues : @xmath41 where @xmath42 .
the factors @xmath43 depend on @xmath44 via @xmath45 : @xmath46 the new resonance energies of the single excitons are the given by @xmath47 the energy difference between the two resonances changes from @xmath40 to @xmath48 in this basis , the hamiltonian of the system takes the form @xmath49 with the renormalization factors @xmath50 and @xmath51 .
the modification of the energy levels and the transition energies due to the coulomb interaction is depicted in fig .
[ fig:2 ] .
all transition energies in the four - level system are altered by the frster interaction and the biexcitonic shift . for optical gating ,
the influence of the coulomb interaction on the effective rabi frequencies during optical excitation is of utmost importance : via the factors @xmath52 and @xmath53 , the rabi frequencies now depend on the involved states of the coupled system . as an example , we consider the creation of an exciton in the second qd , corresponding to one of the transitions @xmath54 or @xmath55 . for the first transition , the pulse area @xmath56 is required , while the second transition requires the pulse area @xmath57 .
this is in contrast to the case of a single , uncoupled dot , where a pulse area of @xmath58 is needed to create full inversion .
this state - dependent renormalization effect plays an important role for the applicability of quantum gates in this system as will be seen in the next section .
of course , the applicability of the considered system for quantum information processing depends crucially on the size of the physical parameters . for qds in the low temperature regime , we assume a radiative damping constant of @xmath59 = ( 500 ps)@xmath60 @xcite .
the energy difference of the qds is taken to be to @xmath61 mev , which is well in the range of the inhomogeneous broadening for different quantum dot types @xcite .
the pulse duration is then strongly constricted , since on the one hand it has to be long enough to allow for spectral selective excitation of the qds , and on the other hand , it must be short enough to minimize the effects of damping during gating .
it is here taken to be @xmath62 fs , which corresponds to a spectral pulse width of @xmath63 mev .
for the implemetation of the cnot operation , this calls for a biexcitonic shift of a few mev , as will be seen below . the size of the coulomb matrix elements @xmath28 and @xmath29 depends strongly on the material parameters and geometry of the specific qd setup such as the qd material and size , the detuning , and the interdot distance .
the values can be as high as @xmath64 = 0.25 ( calculated for self - assembled inas / gaas qds with @xmath40 = 2 mev @xcite ) and @xmath65 = 1.6 ( calculated for cdse quantum dots with @xmath66 75 mev under the influence of a static electric field of @xmath67 kv / cm @xcite ) . on the other hand , very small values can always be obtained by a suitable choice of parameters ( for example , @xmath68 in cdse qds @xcite and @xmath69 in gaas interface qds @xcite ) .
the coupling constants can be further influenced by external electric fields @xcite .
instead of focusing on a specific parameter set , we investigate the system for a range of values of the matrix elements known to represent different qd systems .
the physical realization of a desired gate will always lead to an output state with a certain amount of error compared to the ideal result .
this error can be quantified with the fidelity of the process @xcite : the gate fidelity @xmath70 measures the overlap between the ideal final state and the actual final state which results by applying the gate @xmath71 .
if the result of the ideal gate is a pure state @xmath72 and the actual final state is given by the density matrix @xmath73 , the fidelity is defined as @xmath74 if the result @xmath73 is identical to the desired pure state , the fidelity @xmath75 .
for @xmath76 , one typically defines the error of the operation as @xmath77 . expanding the pure state @xmath78 in the basis @xmath79 , @xmath80 , eq .
( [ eq:12 ] ) takes the form @xmath81 the fidelity is therefore determined by the four - operator expectation values of the system , i.e. the elements of the two - electron density matrix . for the not gate on the first qubit , for example , the fidelity for the input state @xmath35 which is mapped to @xmath82 is just given by @xmath83 .
quantum operations on coulomb - coupled qds can be realized following the proposition in ref .
@xcite : under resonant coherent excitation of the excitons , the system can be driven from one state to the other , thus realizing gates on the computational states .
the coulomb interaction is used to implement conditional gates , in the following the cnot gate .
however , the resulting frster interaction mixes the basis states and therefore destroys the original computational basis as seen in eq .
( [ eq:10 ] ) .
as was seen above , the basis in the presence of frster coupling is given by @xmath84 ( cf .
@xcite ) . furthermore , the rabi frequencies for the single - excitonic transitions change from @xmath85 to @xmath86 and @xmath87 , respectively .
the consequences for potential quantum information processing are the following : 1 .
the computational subspace has to be adjusted . since the states @xmath82 and @xmath88 are no longer optically active when the coulomb interaction is considered , they can not be prepared and read out as proposed .
therefore , the states @xmath89 and @xmath90 have to be regarded as part of the new computational basis .
this has to be kept in mind when the system is treated theoretically , especially when calculating the fidelity of quantum operations .
frster energy transfer occurs when the system is not prepared in one of its new computational basis states @xmath91 . under the influence of decoherence caused by the electron - phonon interaction
, it may be irreversible @xcite , thus presenting an error that adds to the general error caused by decoherence . since we neglect electron - phonon coupling here and consider the system prepared in its computational basis , this has no influence in this work .
the state - dependent renormalizations of the rabi frequencies leads to an inevitable error in single - qubit operations since the condition for switching a qubit in one qd depends on the state of the other qd .
as discussed below , this error depends on the strength of the frster interaction and adds to the errors caused by decoherence . to illustrate the importance of pulse area renormalization and basis change , we investigate the following situation :
the state @xmath35 is excited by a pulse which is resonant on the first qd . for an excitation of the uncoupled system with a pulse area
@xmath58 , this causes a complete inversion of the first qd and therefore corresponds to a not operation on the first qubit with the resulting state @xmath82 . in the coupled case
one would have to adjust the computational subspace and calculate the fidelity with respect to the resulting state @xmath89 and excite with a pulse area @xmath57 , as can be seen in fig .
[ fig:3 ] : the fidelity is shown for an excitation with a pulse area @xmath58 ( red lines ) and the renormalized pulse area @xmath92 ( green lines ) , calculated with the resulting state @xmath82 ( solid lines ) and @xmath89 ( dotted lines ) , respectively . as expected , the highest fidelity is achieved when working with a renormalized pulse area and within the new computational basis which arises under coulomb coupling . in this case
, the fidelity can be optimized to high values even for large values of the frster coupling @xmath44 . for the input state @xmath35 .
the fidelity is shown for an excitation with a @xmath58-pulse ( red lines ) and a renormalized ( @xmath57- ) pulse ( green lines ) , both for the output state @xmath82 ( dotted lines ) and @xmath89 ( solid lines ) . ] as a consequence , we calculate the performance of the following quantum gates on the computational basis @xmath84 using pulse areas renormalized with the factors @xmath93 , respectively .
the fidelity is usually calculated by averaging over all possible computational input states . in our case , there are important qualitative differences in the performance of the four basis states of the computational subspace .
therefore , we investigate the fidelity for these four basis states separately . in this section , we investigate the not gate on the first qubit of our system .
it is implemented by applying an @xmath57-pulse which is resonant on the first qd ( see table [ table:1 ] for the corresponding truth table ) .
since the biexcitonic shift is not important for the realization of the not gate , we show the results for @xmath94 .
[ table:1 ] .truth table for qubit gates [ cols="<,<,<,<,<",options="header " , ] [ truthtable ] in fig .
[ fig:6](a ) , the calculated fidelity @xmath70 is shown for the system without radiative damping . for the states
@xmath35 and @xmath95 , the renormalized pulse area leads to a high fidelity , as is expected from the prior considerations . without damping , the states @xmath35 / @xmath95 and
@xmath96 / @xmath30 show the same fidelity due to the symmetry of excitation and deexcitation . for the states @xmath96 and @xmath30 , the renormalization leads to a diminished fidelity .
this is due to the fact that these states are excited with the `` wrong '' pulse area , as discussed above .
thus , the adaptation of the pulse area does not optimize the gate for all possible input states : on the average , the fidelity is diminished compared to the uncoupled case . in this way , the frster interaction presents a problem for the proposed quantum information schemes and has to be minimized . in the inset in fig .
[ fig:6](a ) , one can see that the error stays in the range of @xmath97 for small values ( up to hundreds of @xmath98ev ) of @xmath44 . in ref .
@xcite , a frster coupling constant of @xmath28 = 0.45 mev , corresponding to @xmath68 was calculated for a cdse double qd structure .
it was also shown that the frster coupling can be suppressed by applying a static electric field .
since this simultaneously allows an enhancement of the biexcitonic shift @xcite , this is one possibility to improve the gate performance of the not gate .
for the four input states constituting the computational basis : ( a ) without radiative damping , ( b ) with radiative damping [ @xmath59 = ( 500 ps)@xmath60 ] . the biexcitonic shift is set to @xmath94.,title="fig : " ] for the four input states constituting the computational basis : ( a ) without radiative damping , ( b ) with radiative damping [ @xmath59 = ( 500 ps)@xmath60 ] .
the biexcitonic shift is set to @xmath94.,title="fig : " ] it should be noted that there is an additional small fidelity loss with rising coupling strength that can not be seen in fig .
[ fig:6](a ) : for @xmath100 , it adds up to @xmath101 .
this has a noticeable impact on the generation of bell states as will be seen in sec .
[ bellstates ] . in a second step ,
we calculate the performance of the gate under the influence of radiative damping [ fig .
[ fig:6](b ) ] .
now the situation is more complex , since the damping affects the states differently : for @xmath90 and @xmath30 , where both input and output are damped states ( see table [ truthtable ] ) , damping causes a considerable error in the range of @xmath1 .
the situation is much better for the input states @xmath89 and @xmath35 , where the undamped state @xmath35 is involved either as input or output in the gating process .
here , the error stays in the range of @xmath0 , as can be seen in the inset of fig .
[ fig:6](b ) .
the cnot gate is implemented as suggested in ref .
@xcite : the system is excited with an @xmath57-pulse that is tuned to the exciton - biexciton transition of the second qd . this way , the resonance condition is only fulfilled when the first qd is in its excited state , thus representing a conditional not operation on the second qubit ( see table [ table:1 ] for the truth table of the single - qubit cnot operation on the second qubit ) . in fig .
[ fig:7 ] , the fidelity of the cnot operation is shown for a biexcitonic shift of @xmath102 mev as a function of @xmath44 . without damping [ fig .
[ fig:7](a ) ] , the performance is better than that of the not gate , since only the two input states @xmath90 and @xmath30 are affected by the gate . as seen in the truth table , the other two states remain unchanged , and therefore the pulse areas can be optimized for the mapping of @xmath90 and @xmath30 . in consequence , it does not show the strong dependence on the frster coupling constant @xmath28 that is observed for the not gate . as can be seen in the inset ,
the error is of the order of @xmath0 for all input states .
similar to the not gate , the situation changes drastically when damping is considered [ see fig .
[ fig:7](b ) ] : except for the ground state @xmath35 , the fidelity shows a significant reduction in the range of @xmath1 , the same order of magnitude as for the not gate .
for both cases ( with and without damping ) , there is a small increase of the fidelity for the input states @xmath35 and @xmath89 with increasing @xmath28 .
this is due to the corresponding increase of the energy difference of the single - excitonic transitions with @xmath28 [ see eq .
( [ eq:27 ] ) ] . even for an energy difference between the two qds of @xmath61 mev
, there is still a spectral overlap of the @xmath62 fs gaussian pulse with the first qd .
this overlap decreases when the energy difference increases ; therefore , the fidelity increases .
( a ) without and ( b ) with radiative damping [ @xmath59 = ( 500 ps)@xmath60 ] for a biexcitonic shift of @xmath29 = 5 mev ( @xmath103).,title="fig : " ] ( a ) without and ( b ) with radiative damping [ @xmath59 = ( 500 ps)@xmath60 ] for a biexcitonic shift of @xmath29 = 5 mev ( @xmath103).,title="fig : " ] of course , the performance of the cnot gate depends crucially on the size of the biexcitonic shift @xmath29 , since the energy shift is the condition for the selective excitation upon which the cnot gate bases . as shown in fig .
[ fig:7a ] for a vanishing frster coupling @xmath104 , for the pulse durations used here ( which are restricted by the damping constant of the system ) , the biexcitonic shift has to be at least @xmath105 mev to ensure an error of @xmath0 or less . in the presence of damping , this error rate
can only be reached for the input state @xmath35 ; the other input states result in higher errors .
the more complex cnot operation performs better than the not gate , especially for high values of @xmath28 , since , as explained above , it does not show such a strong dependence of the error on @xmath28 . as a result , the generation of entangled states works surprisingly well in this system , as will be seen in the next section .
( a ) without and ( b ) with radiative damping [ @xmath59 = ( 500 ps)@xmath60 ] .
the fidelity is shown for @xmath106 .
, title="fig : " ] ( a ) without and ( b ) with radiative damping [ @xmath59 = ( 500 ps)@xmath60 ] .
the fidelity is shown for @xmath106 .
, title="fig : " ] bell states are a key ingredient for basic quantum information processes such as teleportation @xcite . as maximally entangled states , they are also of general interest in quantum information theory .
a physical system that is to be used for quantum information processing must provide the possibility to generate and manipulate maximally entangled states . in our case , the cnot gate is utilized to create a maximally entangled state in the system , the bell state @xmath107 : one of the qds ( here , the first qd ) is excited with a @xmath108-pulse , leaving the qubit in an equal superposition of the ground state and @xmath89 .
then , the second qd is excited with an @xmath92-pulse resonant on the biexcitonic resonance , thus selectively exciting the part of the two - electron state where the first qubit is already excited . the resulting fidelity is shown in fig . [ fig:8 ] .
[ fig:8](top ) , a slight @xmath28 dependence of the fidelity can be seen which resembles the one observed for the single - qubit not gate .
figure [ fig:8](bottom ) shows the @xmath29 dependence which arises from the underlying cnot gate . as a combination of the two gates ,
the generation of bell states can not be more efficient than the separate not and cnot gate .
however , since a specific input state is used here , the pulse areas are optimized for this input state , and thus a high fidelity can be achieved for this process .
although the errors add up , the value is still acceptable : even when radiative damping is taken into account , the error stays in the range of @xmath0 . as a function of the coulomb coupling strength without ( red line ) and with ( green line )
radiative damping .
top : the dependence on @xmath28 for a biexcitonic shift of @xmath109 mev ( @xmath103 ) .
bottom : the dependence on @xmath29 for a frster coupling constant of @xmath106.,title="fig : " ] as a function of the coulomb coupling strength without ( red line ) and with ( green line ) radiative damping .
top : the dependence on @xmath28 for a biexcitonic shift of @xmath109 mev ( @xmath103 ) .
bottom : the dependence on @xmath29 for a frster coupling constant of @xmath106.,title="fig : " ]
we have investigated the influence of the coulomb interaction on the applicability of a single- and a two - qubit quantum gate .
we find that the frster interaction and radiative damping cause considerable error rates in the fidelity of the gates .
while the error caused by the frster interaction can be kept in the acceptable range of @xmath0 by a suitable choice of the system parameters , the fidelity loss due to radiative damping is typically larger and more difficult to deal with . despite the fidelity loss of the basic gates , the generation of a maximally entangled bell state is possible with error rates in the range of @xmath0 , even in the presence of radiative dephasing . in general , for using the considered process for successful quantum information processing , one has to think about ways of minimizing the dephasing due to radiative damping , for example by placing the quantum dots in a cavity @xcite or exploiting dark excitons @xcite .
future work should then also focus on the error caused by the interaction with phonons .
m. richter , k.j .
ahn , a. knorr , a. schliwa , d. bimberg , m.e.a .
madjet , and t. rengertheory of excitation transfer in coupled nanostructures - from quantum dots to light harvesting complexes , * 243 * , 2302 ( 2006 ) .
a. grodecka , c. weber , p. machnikowski , and a. knorrinterplay and optimization of decoherence mechanisms in the optical control of spin quantum bits implemented on a semiconductor quantum dot , * 76 * , 205305 ( 2007 ) .
a. grodecka , l. jacak , p. machnikowski , and k. roszak , quantum dots : research development , ( nova science publishers , new york , 2005 ) , chap .
phonon impact on the coherent control of quantum states in semiconductor quantum dots , pp.4788 .
s. fafard , r. leon , d. leonard , j.l .
merz , and p.m. petroffvisible photoluminescence from n - dot ensembles and the linewidth of ultrasmall @xmath110@xmath111as/@xmath112@xmath113as quantum dots , * 50 * , 8086 ( 1994 ) .
m. bayer , t.l .
reinecke , f. weidner , a. larionov , a. mcdonald , and a. forchelinhibition and enhancement of the spontaneous emission of quantum dots in structured microresonators , * 86*(14 ) , 31683171 ( 2001 ) . | we investigate the effect of the coulomb interaction on the applicability of quantum gates on a system of two coulomb - coupled quantum dots .
we calculate the fidelity for a single- and a two - qubit gate and the creation of bell states in the system . the influence of radiative damping is also studied .
we find that the application of quantum gates based on the coulomb interaction leads to significant input state - dependent errors which strongly depend on the coulomb coupling strength . by optimizing the coulomb matrix elements via the material and the external field parameters , error rates in the range of @xmath0
can be reached .
radiative dephasing is a more serious problem and typically leads to larger errors on the order of @xmath1 for the considered gates . in the specific case of the generation of a maximally entangled bell state , error rates in the range of @xmath0 can be achieved even in the presence of radiative dephasing . |
collisions between galaxy clusters are the most energetic events in the cosmos , with unique implications for structure - formation and the nature of dark matter .
the concordance @xmath5cdm model , predicts that the infall velocities of massive merging systems are typically less than 1000 @xmath2 @xcite .
however , a few massive merging clusters discovered recently suggest a significant tail in the velocity distribution at high infall velocities ( @xmath6@xmath73000 @xmath2 ) . the bullet cluster ( 1e0657 - 56 )
is readily identified as an extreme example by its clearcut mach cone found in its x - ray image @xcite .
velocities derived from x - ray observations using the shock jump conditions and n - body / hydrodynamical simulations support this interpretation @xcite , but with inferred velocities in the range 27004500 @xmath2 .
several other merging clusters have been subsequently found with a bullet - cluster - like morphology ( a2744 : @xcite ; a2146 : @xcite ) , and some also identified as high - infall - velocity systems ( cl j0152 - 1347 : using detailed hydrodynamical simulations ; @xcite ; macs j0717.5 + 3745 : using radial velocities and the kinematic sz effect ; @xcite ) .
these high - infall - velocity merging clusters provide a potential challenge for the concordance @xmath5cdm model .
however , to test the concordance model with these mergers , we need more accurate and robust determination of impact velocities based on multifrequency observations and detailed numerical simulations , as well as larger volume cosmological simulations to define the probability of these extreme merging systems @xcite .
the newly discovered `` el gordo '' galaxy cluster ( act - cl j0102 - 4915 , @xmath8 ) by the atacama cosmology telescope ( act ) sunyaev
zeldovich ( sz ) cluster survey ( marriage et al .
2011 ) has the largest sz decrement of any cluster reported and displays a bullet - like x - ray morphology in deep follow - up chandra images @xcite .
multifrequency observations of this cluster show a large offset between the x - ray and sz centroids , and between the x - ray and the centroid of mass - surface density of the main cluster component @xcite .
these large offsets are simple distinguishing features caused by high velocity encounters as shown by @xcite using n - body / hydrodynamical simulations , suggesting that el gordo is a high - infall - velocity merging cluster .
the lensing work identified el gordo as the most massive cluster known at z@xmath7@[email protected] , as of today @xcite . in this paper we present results from numerical simulations of idealized binary encounters between galaxy clusters containing dark matter and gas initially in equilibrium based on the adaptive mesh code , flash .
our main goal is to find a physical interpretation of the overall features of el gordo found in multifrequency observations .
the structure of this paper is the following . in section [ s : elgordo ] we summarize results from previous analyses of el gordo based on multifrequency observations and numerical simulations .
then , in section [ s : flash ] , we describe our flash simulations , and methods to obtain simulated x ray and sz images , and mass surface density maps .
we present our results and provide a physical interpretation of the morphology of multifrequency observations of el gordo in section [ s : results ] .
we discuss our results in comparison with previous investigations in section [ s : discussion ] .
we summarize our findings and their implications to cosmology in section [ s : conclusion ] . unless stated otherwise , errors quoted are 68% cl .
we adopt , as our concordance cosmological model , the 5-year wmap cosmology , @xmath9 , @xmath10 , and h@xmath11 km s@xmath12 mpc@xmath12 @xcite .
this model gives the scaling of 1@xmath13 corresponds to 459.5 kpc ( note , we do not use the dimensionless hubble parameter , @xmath14 , because numerical simulations work with physical distances ) .
el gordo has the largest sz decrement in the 455 deg@xmath15 act sz cluster survey ( marriage et al .
2011 ) , and the most significant detection in the overlapping 2500 deg@xmath15 south pole telescope ( spt ) sunyaev - survey ( williamson et al .
it was confirmed as a high redshift ( @xmath0 = 0.870 ) galaxy cluster by optical observations @xcite .
the amplitude of the sz decrement of el gordo is comparable to the massive bullet cluster , and it was found to be the most massive , x - ray , and sz right cluster at z@xmath7@[email protected] discovered as of today @xcite .
a detailed study of this cluster using multifrequency observations was presented by @xcite .
chandra acis - i observations of el gordo reveal a disturbed morphology .
the x - ray emission of the cluster is elongated in the south east to north
west direction .
one of the tails is following the peak emission at an angle of about 45@xmath16 towards north - west ( northern tail ) , below that is located the southern tail , which may be interpreted as a wake from a merging event .
menanteau et al .
found a high overall x - ray temperature of t@xmath17 = @xmath18 kev , and a peak temperature of @xmath19 kev around the shock region .
based on the x - ray morphology , the high x - ray temperature and sz amplitude , menanteau et al .
suggested that el gordo is a high - infall - velocity binary merger at high redshift , very similar to the bullet cluster located at a lower redshift .
the relative positions of the mass surface density , the x - ray and sz peaks are important in the physical interpretation of a merging galaxy cluster such as el gordo .
@xcite located two main peaks in the images of the galaxy number density , the rest frame @xmath20-band and 3.6 @xmath21 m luminosity density , and stellar mass density ( derived from infrared irac / spitzer imaging using a spectral energy distribution fitting procedure ) , which are correlated with the mass distribution .
they concluded that the nw mass component centered on the northern tail is more massive than the mass center in the se , in the vicinity of the x - ray peak .
the sz peak was found to be close to the nw component slightly offset toward the se component .
their results suggest a mass ratio of about 2:1 of the two components . based on the cometary morphology , the high x - ray temperature , the mass distribution , and the offsets between mass peaks and the x - ray and sz peaks , menanteau et al .
suggest that the bright x - ray peak marks the emission from an infalling cluster moving from the nw towards se plunging through the main cluster disrupted in its x - ray emission indicating a large infall velocity producing a wake with two tails . using mass scaling relations based on x - ray , sz observations and velocity dispersion , menanteau et al .
( 2012 ) estimated a total virial mass of el gordo to be @xmath22 @xmath23 .
they found that all of their mass estimates are less than 2@xmath24 away from the interval of 1.22.0@xmath25 @xmath23 .
this mass estimate for el gordo is compatible with subsequent analyses of strong and weak lensing data : @xmath26 @xmath23 @xcite , and @xmath3 @xmath23 @xcite , respectively . based on the x - ray morphology and that the location of the peak of the galaxy number distribution of the se component which precedes the x - ray peak by 173 @xmath27 kpc , menanteau et al .
suggest that el gordo is a high redhsift close analogue to the bullet cluster . in this scenario , as a consequence of the high infall velocity , the gas ( marked by the x - ray peak ) trails the dark matter associated with the infalling cluster ( marked by the peak of the galaxy number distribution ) due to ram pressure .
however , this close analogy with the bullet cluster is questionable , because another mass proxy , the peak of the @xmath20-band luminosity and the mass peak of the infalling cluster from weak lensing @xcite are , in fact , trailing the x - ray peak .
the positions of the mass peaks estimated using different methods scatter around the x - ray peak within a 173 kpc , thus we conclude that the offset between the x - ray peak and the center of the infalling cluster has not been established
. recently discovered massive high redshift clusters and large impact velocity merging clusters pose a potential challenge to the concordance @xmath5cdm model @xcite . @xcite and
@xcite estimated the probability to find a cluster with the mass and the redshift of el gordo based on cosmological numerical simulations using the exclusion curve method of @xcite , and concluded that , although a cluster such as el gordo is a rare massive cluster , it is within the allowed mass range predicted by the @xmath5cdm models .
however , @xcite noted that a more accurate mass measurement may put el gordo into the excluded region .
the discovery of the bullet cluster with its high inferred infall velocity provided the first challenge to the concordance @xmath5cdm model @xcite .
estimates based on cosmological simulations found that the probability of finding a cluster with the infall velocity implied by the bullet cluster is very unlikely @xcite .
@xcite used cosmological simulations with different box sizes ( 0.252 h@xmath12 gpc ) to improve the accuracy of the probability distribution at the high end of the velocity tail taking into account the effect of the box size .
they confirmed that it is very unlikely ( @xmath28 ) to find even one massive cluster merger with impact velocity of @xmath6@xmath73000 @xmath2 in a concordance @xmath5cdm model at the redshift of the bullet cluster .
the most probable infall velocity was found to be about 550 @xmath2 , ( see figure 15 of @xcite ) .
most recently @xcite carried out a simulation using large box size ( 6 h@xmath12 gpc ) to improve on the statistics of large clusters and rare cosmological events .
they found no merging system with the same large relative velocity , spatial separation , and redshift as the bullet cluster ( see their figure 7 ) .
they conclude , however , that this might be due to their low statistics of these rare major mergers at a fixed redshift .
clearly , more numerical work is necessary to quantify the probability distribution of relative velocities of mergers as a function of redshift , masses , and distances between the components . since the discovery of the bullet cluster several other merging clusters have been found with large offsets between the x - ray peak and the sz and dark matter surface density peaks ( macs j0025.4 - 1222 : @xcite ; cl j0152 - 1347 : @xcite ; a2163 : @xcite ; macs j0744.8 + 3927 and cl j1226.9 + 3332 : @xcite ; macs j0717.5 + 3745 : @xcite ; dlscl j0916.2 + 2951 : @xcite ; sl2s j08544 - 0121 : @xcite ) .
these offsets are generated by the different behavior of the collisionless component , the dark matter , and the gas during the merging process : the dark matter responds only to gravity , but the gas is subject to hydrodynamical effects , and that the x - ray emission and the sz effect depend differently on the gas density and temperature .
the size of the offsets between the dark matter and the gas peaks is determined by the relative strength of the gravitational force which is trying to keep the gas locked into the potential well of the dark matter , and the ram pressure which is removing gas from the infalling subcluster . using n / body hydrodynamical simulations of merging galaxy clusters , @xcite showed that a large impact velocity is necessary to explain hundreds of kpc scale offsets .
the relative velocity of el gordo was estimated by @xcite from the measured line - of - sight peculiar velocity difference between the se and nw galaxy concentrations , as 586@xmath2996 @xmath2 . however , for the bcg , they found a higher velocity difference : 731@xmath2966 @xmath2 . assuming a projection angle relative to the plane of the sky of @xmath30 , the relative velocity can be expressed as 586/@xmath31 .
based on the x - ray morphology , menanteau et al .
argue that @xmath30 should be small , and they obtain a relative velocity of 2300 @xmath2 and 1200 @xmath2 assuming 15@xmath16 and 30@xmath16 for @xmath30 , and use this to estimate the infall velocity .
the location of radio relics ahead and behind the two mass centers enclosed by the detected x - ray emission support this interpretation @xcite .
these estimated infall velocity values fall on the less likely to the unlikely region of the velocity probability distribution derived by @xcite .
menanteau et al .
conclude that a more accurate determination of the infall velocity is necessary for el gordo using numerical simulations to find out wether it is allowed by the concordance @xmath5cdm model . binary merger simulations to reproduce the observed properties of el gordo using a publicly available smoothed particle hydrodynamics ( sph ) code gadget-3 ( springel 2005 ; dolag & stasyszyn 2009 ) was carried out by @xcite .
the parameters of the initial setup were based on those derived by @xcite .
@xcite reproduced the x - ray luminosity of el gordo and the observed distances between the peaks of the dark matter surface density distribution and x - ray emission and sz effect .
however , the x - ray morphology has not been reproduced well : the leading x - ray peak followed only by one tail close to the collision axis ( the line connecting the two dark matter centers ) .
donnert is suggesting that the reason why the two - tailed x - ray morphology was not reproduced was the lack of substructure in the simulation .
we carried out self - consistent numerical simulations of binary galaxy cluster mergers including dark matter and intracluster gas using the publicly available eulerian parallel code , flash , developed at the center for astrophysical thermonuclear flashes at the university of chicago ( @xcite and @xcite ) .
flash , with its new modules , offers the possibility for the future to include many non - gravitational processes , such as thermal conduction , viscosity , radiative cooling , and magnetohydrodynamics , which might be relevant for certain aspects of cluster merging .
the box size of our simulations was 13.3 mpc on a side , reaching the highest resolution , i.e. , cell size , of 12.7 kpc at the merger shocks and the centers of the clusters .
our simulations were semi
adiabatic , i.e. , only shock heating was included . in order to establish our notation and
for the convenience of the reader , we briefly summarize our well established simulation method here , a detailed description and verifications of the method can be found in @xcite . as initial conditions for the two colliding clusters , for the distribution of the dark matter and the intracluster gas , we assumed spherical models with a cut off at the the virial radius of each cluster ( @xmath32 ) .
we adopted a nfw model @xcite for the dark matter distribution : @xmath33 where @xmath34 , and @xmath35 , @xmath36 are scaling parameters for the density and radius , and @xmath37 is the concentration parameter .
we assumed a truncated non - isothermal @xmath38 model for the gas density distribution , @xmath39 where @xmath40 , @xmath41 , is the density at the center , and @xmath42 is the scale radius for the gas distribution .
the temperature of the gas was determined from the equation of hydrostatic equilibrium assuming the ideal gas equation of state with @xmath43 .
we adopted @xmath44 for the large scale distribution of the intracluster gas ( suggested by cosmological numerical simulations ; e.g. , @xcite ) .
we represented the dark matter and the stellar matter in galaxies with particles since galaxies can also be considered collisionless for our purposes .
the number of particles at each amr cell was determined by the local density , normalized according the total number of particles , which were 5 million in our simulations , assuming a gas fraction of 0.14 .
the velocities of the dark matter particles were derived from sampling a maxwellian distribution with a velocity dispersion as a function of distance from the cluster center obtained from the jeans equation @xcite adopting an isotropic velocity dispersion .
the direction of the velocity vectors were assumed to be isotropic ( for more details on modeling the particle velocity distribution see @xcite ) .
we performed a series of numerical simulations with a total mass of @xmath3 @xmath4 and a mass ratio of 1.9 motivated by the previous analysis of el gordo based on observations @xcite .
we adopted concentration parameters of @xmath45 and @xmath46 , and core radii @xmath47 = 288 kpc and @xmath48 = 216 kpc , for the main and the infalling component motivated by our previous merging cluster simulations producing offsets between peaks of dark matter surface density and x - ray emission @xcite .
we used different impact parameters , @xmath49 , and infall velocities , @xmath50 , searching for initial parameters which could reproduce the observed morphology of the x - ray , sz and surface mass density of el gordo .
the initial parameters for our simulations are listed in table [ t : table1 ] .
the ids of our runs are listed in the first column as @xmath51 , where @xmath52 is the serial number of the run , the pairs of integers , @xmath53 and @xmath54 , indicate the impact parameter , @xmath49 ( in units of 10 kpc ) , and the infall velocity of the run , @xmath50 ( truncated at the 100s in @xmath2 ) .
the second column lists the relative initial velocities of the infalling cluster in @xmath2 at the distance where the two clusters truncated gas spheres touch .
the third column lists the impact parameter , defined as the perpendicular distance between the trajectory of the infalling cluster and the center of the main cluster ( in kpc ) .
the other columns show parameters related to the analysis of our simulations to be discussed in section [ s : results ] .
ccrcrrr r1p05v22 & 2250 & 50@xmath55 & 48@xmath16 & -60@xmath16 & 860@xmath56 & 655@xmath57 + r2p20v22 & 2250 & 200@xmath55 & 49@xmath16 & -160@xmath16 & 850 & 690@xmath57 + r3p30v22 & 2250 & 300@xmath55 & 45@xmath16 & 5@xmath16 & 914 & 712@xmath57 + r4p35v22 & 2250 & 350@xmath55 & 44@xmath16 & 10@xmath16 & 932 & 745@xmath57 + r5p25v20 & 2000 & 250@xmath55 & 50@xmath16 & -160@xmath16 & 760 & 497@xmath57 + r6p25v25 & 2500 & 250@xmath55 & 44@xmath16 & 10@xmath16 & 960 & 935@xmath57 + r7p25v30 & 3000 & 250@xmath55 & 42@xmath16 & 8@xmath16 & 960 & 1376@xmath57 + r8p10v25 & 2500 & 100@xmath55 & 48@xmath16 & -60@xmath16 & 860 & 1000@xmath57 + r9p20v25 & 2500 & 200@xmath55 & 42@xmath16 & 10@xmath16 & 960 & 951@xmath57 after each simulation , we generated images of the x ray surface brightness , sz amplitude , and total mass surface density distribution for a range of viewing angles at different phases of the collision ( i.e. different output times ) .
our starting point for each snapshot was a coordinate system with the @xmath0 coordinate axis aligned with the two dark matter centers pointing towards the infalling cluster , and the main plane of the collision ( the plane determined by the two cluster centers and the relative velocity vector parallel to the @xmath0 axis ) aligned with the @xmath58@xmath0 plane , the @xmath58 axis pointing towards the offset of the infall velocity from the main cluster center .
first , we rotated the cluster around the @xmath0 axis with a roll angle @xmath59 , which we changed freely , then around the @xmath58 axis with a polar angle , @xmath30 , determined by the constraint of the observed projected distance between the two mass centers ( active rotation : we rotated the cluster not the coordinate system ) . at the end , we project the cluster along the @xmath60 axis , assuming that the @xmath60 axis is parallel to the line of sight ( los ) . in this arrangement
the zero polar and roll angles ( @xmath61 ; @xmath62 ) mean that we assume that the collision happened in the plane of the sky . ]
once we choose the time for the output and the rotation angles , we generated x
ray surface brightness images by performing a los integral ( along the @xmath60 axis ) over the emissivity at fixed @xmath58 and @xmath0 using @xmath63 where @xmath64 is the x - ray emissivity at frequency @xmath65 , @xmath66 is the abundance , @xmath67 is the temperature of the intracluster gas , and @xmath68 is the effective area of the telescope .
we used the 4-chip averaged on axis effective area of acis - i @xcite as a sufficient approximation for our purposes for comparison with the chandra observations .
x - ray emission from el gordo was detected in the 0.39 kev frequency band ( see figure 3 of @xcite ) , which corresponds to 0.56116.8 kev at emission .
the compressed , shock heated gas due to merging can reach 4080 kev at large impact velocities . at frequencies
@xmath610 kev the relativistic corrections increase the emissivity by 1525% , as opposed to only a few percent increase at a gas temperature of 10 kev ; overall , in the 0.517 kev frequency band , the emissivity of an 80 kev temperature gas is about 20% higher than a 10 kev gas , as it can be seen from figure [ f : relatbrem ] . in this figure
we show the emissivity ratio of the relativistic to non - relativistic bremsstrahlung emissivity for gas temperatures of 10 , 20 , 30 , 40 , 60 and 80 kev as a function of frequency ( in kev ) . even though we are not interested in the normalization of the x - ray emission , relativistic corrections have to be included , because they change the relative normalization : they enhance the emission for high temperature gas .
this enhancement is due to an increase in the relativistic gaunt factor and the contribution from electron
electron bremsstrahlung ( for a detailed discussion of the soft and hard x - ray emission due to thermal bremsstrahlung in merging galaxy clusters see molnar et al .
( 2014 , in preparation ) .
however , the commonly used x - ray spectral analysis packages ( e.g. , xspec ) have been developed for line diagnostics .
the highest resolution x - ray emission package , apec includes some corrections , but , since it has been developed for line diagnostics , does not include all corrections .
therefore we choose to calculate the continuum , @xmath69 , and line x - ray emissivity , @xmath70 , at frequency @xmath65 separately : @xmath71 where , for the abundance of elements , @xmath66 , we used 30% of the solar abundance ( @xmath72 ) . for the continuum emissivity
, we used the relativistic calculations of @xcite : @xmath73 \ , d \nu , \ ] ] where @xmath74 and @xmath75 are the electron and ion number densities , and the first term on the right hand side is a sum of emissivities over ions for electron ion collisions , the second term represents the emissivity due to electron electron collisions .
the first term can be expressed as @xmath76 where @xmath77 , @xmath78 , is the charge of ion @xmath20 , the dimensionless temperature and frequency are @xmath79 , and @xmath80 , where @xmath81 and @xmath82 are the boltzmann and planck constants , @xmath83 is the mass of the electron , @xmath84 is the speed of light , and the thermally averaged gaunt factor is calculated as @xmath85 where @xmath86 is the classical electron radius , @xmath87 , ( where @xmath88 is the charge of the electron ) , and the lengthy expressions for the @xmath89 and @xmath90 integrals , which are easily done numerically , can be found in @xcite .
we used the fitting formula of @xcite for the relativistic electron
electron thermal bremsstrahlung emissivity , @xmath91 where @xmath92 , where @xmath93 is the fine structure constant , @xmath94 is the thomson cross section , and the fitting functions @xmath95 and @xmath96 can be found in @xcite . for the line emissivity , @xmath70 , we used the publicly available apec code developed for line diagnostics .
the sz surface brightness was calculated using relativistic corrections , following @xcite , @xmath97 , \ ] ] where @xmath98 is the non - relativistic frequency function , where the the dimensionless frequency is @xmath99 , @xmath100 is the temperature of the cosmic microwave background , and the second term is a sum of the relativistic corrections approximated as a polynomial with coefficients , @xmath101 , listed in @xcite .
we integrated the total ( dark matter and gas ) density along the @xmath60 axis to generate the mass surface density images at position , @xmath58 and @xmath0 : @xmath102 note that the total mass surface density maps are dominated by the main mass component , the dark matter .
when carrying out numerical simulations of binary cluster collisions , the observed mass surface density distribution provides a starting estimate for the initial masses of the two components .
although here we fix the masses to reduce the initial parameter space , they can also be determined along with other input parameters by comparing observations and numerical simulations . the observed projected distance between the two mass centers , by itself , does not provide constraints on the phase of the collision ( e.g. , the time relative to the first core passage ) due to possible projection effects .
the phase of the collision can be constrained by using multi - frequency observations comparing the morphology of the intracluster gas derived from high resolution x - ray and/or sz observations , and the positions of the mass centers ( in projection ) derived from optical / infrared observations using the method of gravitational lensing ( e.g. , @xcite ) . in the case of el gordo , the sz observations have a low angular resolution of about 1.@xmath134 ( act ) .
therefore , in order to constrain the morphology of the gas , we can only use the chandra x - ray observations which have a resolution of @xmath1031@xmath104 .
in @xcite we performed a set of simulations to study the offsets between the x - ray emission , the sz amplitude and mass surface density peaks .
based on a comparison with that study , we conclude that el gordo is most likely after the first core passage , and before the turn around aiming for the second core passage : before the first core passage there should have been two distinct , smooth x - ray peaks where the gas is in equilibrium around them .
it must be that the core has passed for there to be only one x - ray peak and the gas has a very disturbed , cometary morphology .
therefore , in order to reduce the parameter space , we fixed the initial masses of the two components as m@xmath105 @xmath4 and m@xmath106 @xmath4 ( mass ratio of 1.9 ) , and ran a set of simulations changing the other parameters looking for a strongly peaked x - ray emission with a cometary structure , and two tails in the wake behind the collision after the first core passage .
we found that a series of simulations with different infall velocities @xmath107 @xmath2 and in a relatively narrow range of impact parameters , about 200 to 350 kpc , were able to reproduce the x - ray morphology with the two tails in the wake .
simulations with small impact parameters produce three x - ray tails ( see section [ s : discussion ] ) , very large impact parameters result in only one tail . in figure
[ f : xrimages ] , as an illustration , we show examples of simulated x - ray surface brightness images from our merger simulations , displayed with projection angles that match best the cometary x - ray morphology with a " twin tailed morphology behind the infalling cluster as well as its physical size ( @xmath103860 kpc ) , except the first panel in the upper row which is included for our discussion presented in section [ s : discussion ] ( here we ignore very faint additional tails ) . in this figure we overlaid the contours from chandra observations of el gordo @xcite for an easy comparison . in the upper row
we show images from simulations with fixed infall velocity of @xmath50 = 2250 @xmath2 , but different impact parameters , @xmath108 , 200 , 300 , 350 kpc ( runs r1p05v22 , r2p20v22 , r3p30v22 , and r4p35v22 ; left to right ) . in the lower row
we instead fix the impact parameter to 250 kpc and vary the infall velocities , @xmath109 , 2500 , 3000 @xmath2 ( runs r5p25v20 , r6p25v25 , and r7p25v30 ; first 3 panels , left to right ) .
the last panel in the lower row is from a simulation with @xmath49 = 200 kpc and @xmath50 = 2500 @xmath2 ( run r9p20v25 ) .
the ids for the runs , the infall velocities , @xmath50 , the impact parameters , @xmath49 , and the polar and roll angles , @xmath30 and @xmath59 , of the corresponding rotations are shown in table [ t : table1 ] .
the other two parameters shown in this table , the projected distance between the two dark matter centers and the radial velocity associated with each projection ( columns 6 and 7 ) , are discussed in section [ ss : vrad ] .
we find that the phases of the collision for these massive clusters with large infall velocities ( @xmath62000 @xmath2 ) and small mass ratios are the following : first a large scale contact discontinuity forms when the gas of the infalling cluster presses against that of the main cluster .
as the infalling cluster plunges into the main cluster , adiabatically compressed layers of gas develop in both components around the contact discontinuity , and form a wedge - shaped region , and a shock develops on the two sides of the contact discontinuity .
this wedge - shaped contact discontinuity passes through the main cluster and visible after the first core passage as well ( upper row , first panel in figure [ f : xrimages ] ) .
however , in the case of el gordo , due to an impact parameter in the order of a few hundred kpc , and a high infall velocity , only one side of the wedge - shaped region is visible in the x - ray image , because there is not enough gas on the side of the collision to produce a visible signal ( e.g. , first row 3rd , 4th panel in figure [ f : xrimages ] ) .
after the first core passage , a bow shock forms in the main cluster gas ahead of the contact discontinuity as the infalling cluster makes its way outward from the main cluster ( clearly visible in all panels of figure [ f : xrimages ] ) .
this is the same phase as that of the bullet cluster . at this stage of the collision ,
tidally stretched gas produces the x - ray emission that connects the two mass centers , seen as the northern x - ray tail in the wake of el gordo .
this kind of tail can always be seen in simulations with smaller infall velocities , because in this case the gravitational force of the main cluster can retain enough gas to produce visible x - ray emission . in our simulated x - ray images
there is a hint of a third tail , above the tidally stretched gas , which is due to the compressed gas on that side of the collision axis .
the x - ray emission observed in el gordo suggests the presence of a possible third tail , indicated by enhanced emission to the north - west of the x - ray peak .
however , the photon counts are quite low in this direction .
our simulations suggest that a third x - ray tail above the two observed tails would be visible in an x - ray image of el gordo with a longer exposure time .
we have demonstrated that the observed cometary x - ray morphology and system size of el gordo can be reproduced by our simulations using several infall velocities and impact parameters values ( figure [ f : xrimages ] ) . as a next step ,
we use the observed projected distance between the two dark matter centers and the observed relative radial velocity of the infalling cluster to constrain the initial parameters of the system .
the mass center of the main ( nw ) component marked by the rest frame @xmath20-band and 3.6 @xmath21 m luminosity density , stellar mass density @xcite , and mass surface density derived from gravitational lensing @xcite are all very close to each other , so they seem to be reliable . on the other hand ,
the position of the mass peak of the infalling cluster , the se component , derived using different methods show about a 170 kpc scatter around the x - ray peak .
since our simulations suggest that the x - ray peak and the mass peak of the infalling cluster should be close , we adopt the position of the x - ray peak as the center of the infalling cluster .
thus we demand that the projected distance between the centers of the two components in our merging simulations to be @xmath103860 kpc .
based on galaxy redshift measurements , the observed relative radial ( los ) velocity between the infalling cluster center and the main component was found to be 586@xmath2996 @xmath2 @xcite .
however , the observed radial velocity of the bcg was 731@xmath2966 @xmath2 , so we may assume that the radial relative velocity in el gordo falls in the interval of 500800 @xmath2 .
the projected distances between the mass centers and the relative radial velocities for our simulations are listed in table [ t : table1 ] ( columns 6 and 7 ) . based on this table ,
we conclude , that simulations with an infall velocity of @xmath110 @xmath2 and impact parameters @xmath111 , 300 , and 350 kpc ( runs r2p20v22 , r3p30v22 , and r4p35v22 ) , and @xmath112 @xmath2 and @xmath113 kpc ( run r5p25v20 ) provide a good match with the x - ray morphology , the positions of the x - ray peak and mass peaks , and the constraints from the observed radial relative velocity .
we can exclude run r1p05v22 , because it has three tails due to the small impact parameter of @xmath114 kpc ( we included only for comparison with the simulation carried out by @xcite for el gordo ) . in general we can exclude simulations with small impact parameters because they produce either one or three tails depending on the infall velocities .
runs r6p25v25 , r7p25v30 , and r9p20v25 can be excluded as well , because these simulations with larger infall velocities produce either too large projected distances between the two mass centers , or too large radial velocities . in figure
[ f : xrayszdmcont ] we show the x - ray surface brightness , sz and mass surface density contours for these simulations ( runs r2p20v22 , r3p30v22 , r4p35v22 , and r5p25v20 ) in the same projection as in figure [ f : xrimages ] ( see table [ t : table1 ] for the parameters ) .
all four contour images show similar morphology .
the x - ray and mass contours show that the mass centers of the `` bullet '' coincide with the x - ray peak . at these assumed infall velocities , unlike in the case of the bullet cluster , the ram pressure is not sufficient to drag the gas behind the dark matter component as it passes through the main cluster gas cloud .
thus our simulations predict that we will not find an offset between the x - ray peak and the mass center of the `` bullet '' using more precise observations . on each image , the mass center of the main cluster
is located at the north - western end of the northern tail , as predicted by observations .
we further constrain the infall velocity and the impact parameter of the collision for el gordo comparing the shapes of our x - ray surface brightness profiles across the wake with that from chandra observations ( figure 1 of @xcite ) .
the observed profile is nearly symmetric with two maxima separated by about 35@xmath104 , and a local minimum between them .
first , we went through different snapshots for each run after the first core passage with a separation between the two dark matter centers of 6001200 kpc , and generated x - ray images by projecting them out with a set of different polar and roll angles on a grid , and kept those snapshots and projections which resembled el gordo ( showed a cometary morphology with one peak x - ray emission and two tails in the wake of the infalling cluster ) . as a next step , we extracted x - ray surface brightness profiles along cuts through the wake with different positions and rotation angles , and found the best fits to the chandra data .
the extracted x - ray surface brightness profiles of those projections of each run that produced the best match with el gordo observations are shown in figure [ f : xrayprofiles ] ( runs r2p20v22 , r3p30v22 , r4p35v22 , and r5p25v20 ) .
the x - ray images for these snapshots with the same projections are shown in figure [ f : xrimages ] ( upper row panels 2 , 3 and 4 , and lower row first panel ) .
the corresponding paths are displayed in figure [ f : xrayszdmcont ] ( red dashed lines ) . in figure
[ f : xrayprofiles ] the el gordo observations are also shown with points with error bars ( read off from figure 1 of @xcite ) . our simulation with @xmath110 @xmath2 , and @xmath115 kpc is the best match with the observed profile . our simulation with @xmath110 @xmath2 and a smaller impact parameter , @xmath111 kpc , and a simulation with @xmath109 @xmath2 and @xmath116 kpc ( runs r2p20v22 and r5p25v20 )
have the two peaks too far from each other , the profile from our simulation with @xmath110 @xmath2 and larger impact parameter , @xmath117 @xmath2 ( run r4p35v22 ) has three x - ray peaks in the wake . therefore , based on the observed x - ray morphology , the profile of the x - ray surface density across the wake , the relative radial velocity , and the relative positions of the x - ray , sz and mass peaks , subject to the limitations of our models , we derive an infall velocity of @xmath118 @xmath2 , and a large impact parameter of @xmath119 kpc for el gordo ( note , the errors shown are rough estimates only ) . , per arc sec@xmath15 ) across the wake of the cluster merger taken from the chandra observations ( data points with error bars ) overlaid on those extracted from our flash simulation ( solid lines ) .
the simulations were done with an infall velocity of @xmath50 = 2250 @xmath2 and different impact parameters , @xmath49 = 200 , 300 , 350 kpc ( runs r2p20v22 , r3p30v22 , and r4p35v22 ) and with @xmath50 = 2000 @xmath2 and @xmath49 = 250 kpc ( run r5p25v20 ) as a function of distance along the path , @xmath120 , in arc second ( top to bottom ) . the same projections of these simulations are shown in figure [ f : xrayszdmcont ] .
the extraction path for these profiles ( shown as red dashed lines in figure [ f : xrayszdmcont ] ) were chosen to provide the best match with the x - ray surface brightness profile across the wake of el gordo derived from chandra observations .
[ f : xrayprofiles ] ]
our results confirm the basic interpretation of the merging cluster el gordo by @xcite : this system is after the first core passage of a large infall velocity collision , the infalling cluster moving from the nw towards se , and the main cluster center coincides with the observed position of the mass concentration of the nw component .
based on the merger morphology , menanteau et al .
( 2012 ) proposed that the inclination angle ( between the collision axis and the plane of the sky ) should be shallow . assuming that the infall velocity is close to the velocity of the observation after the first core passage , menanteau et al . predicted an infall velocity of 2300 @xmath2 and 1200 @xmath2for polar angles of 15@xmath16 and 30@xmath16 .
our best result with an infall velocity of 2250 @xmath2 ( run r4p35v22 ) is in a good agreement with their prediction , however this is a coincidence .
the relative velocity of this run at the phase of best fit with observations is 1009 @xmath2 .
this shows that the instantaneous velocity at a phase after the first core passage could be 50% of the infall velocity ( or it could even be zero at the phase of the turnaround ) .
the rotation angle out of the sky ( the main plane of the collision ) , i.e. , the polar angle , @xmath30 , in our case is 45@xmath16 .
following menanteau et al.s procedure , using this rotation angle , we would find 600/@xmath121@xmath16@xmath122850 @xmath2for the infall velocity , and not 2250 @xmath2 . and 100 kpc , and infall velocities of @xmath110 and 2500 @xmath2 ( runs r1p05v22 and r8p10v25 ; left to right ) .
for other parameters see table [ t : table1 ] .
the contours of the same data are overlaid to emphasize the morphology .
[ f : xraycomp ] ] @xcite carried out simulations to reproduce the observed properties of el gordo .
most of the important initial conditions adopted were similar to ours ( we also based our simulations on the results of @xcite ) : large masses and the core radius for the main component : m@xmath123 @xmath4 , m@xmath124 @xmath4 , @xmath47 = 300 kpc , infall velocity : 2600 @xmath2 . however , the impact parameter , the concentration parameters , @xmath38 , and the core radius of the infalling cluster were significantly smaller than ours : 20 kpc , @xmath125 = 2.9 , @xmath126 = 3.2 , @xmath38 = 2/3 , @xmath48 = 25 kpc , as well as @xmath38 = 2/3 , and the initial distance , which does not affect the results , larger : 5231 kpc .
donnert used a hernquist profile @xcite for the distribution of dark matter without a cut off at the virial radius , but since the density profile of this model falls off as @xmath127 , the density is quite low at large radii , whereas we used a truncated nfw profile with zero density at large radii .
this difference should not have any significant effect on neither the dynamics of the collision nor the observable x - ray emission around the first core passage .
donnert s simulations reproduced the x - ray luminosity of el gordo , the distances between peaks of the x - ray emission , sz signal and the mass surface density well .
however , the x - ray morphology was not reproduced correctly : the cometary structure showed only one tail pointing to the center of the mass distribution of the main cluster .
our simulations were able to reproduce the two x - ray tails with a range of infall velocities ( 20003000 @xmath2 ) and impact parameters ( 200350 kpc ) .
turbulence plays an important role in structure formation , it facilitates the mixing of the intracluster gas during cluster merging .
comparisons between sph and eulerian amr codes suggest that artificial viscosity , necessary in sph codes to handle shocks , suppress dynamical instabilities , and thus turbulent mixing of the gas ( mitchell et al . 2009 ; agertz et al .
2007 ) . therefore ,
as opposed to donnert , we chose to use an amr code for our merging cluster simulations .
however , the main physical process which produces ordered large scale shock - heated and compressed gas is more likely linked to formation of contact discontinuities and not to turbulence . in figure [
f : xraycomp ] we show x - ray images from our merging cluster simulations with an infall velocity of @xmath110 @xmath2 and an impact parameter of @xmath108 kpc , and with @xmath128 @xmath2 and @xmath129 kpc ( runs r1p05v22 and r8p10v25 ; left to right ) .
we overlaid the contours of the same x - ray images to emphasize the morphology .
these images show a similar symmetric morphology to that of donnert , who used @xmath130 @xmath2 and @xmath131 kpc .
we may conclude that even an impact parameter of 50 or 100 kpc would not break the symmetry of the x - ray images significantly , and can not produce the asymmetric two - tail morphology of el gordo . in both of these panels of figure
[ f : xraycomp ] , three tails can be seen , a larger tail in the middle and two fainter tails in a wedge shape pointing towards the bullet .
our simulations suggest that the one tail seen in the wake of donnert s simulation is due to the tidally stretched gas between the two mass centers .
a similar wedge - shaped structure is generated in donnert s simulation , but it disappears after the first core passage ( figure 5 of @xcite ) .
we speculate that donnert s simulation with a small impact parameter produced only one x - ray tail and not three since they assumed the infalling cluster is in a bubble inside the main cluster atmosphere .
this causes their contact discontinuity to form only between the two cores and not on a large scale of the entire cluster , as in our simulations @xcite . in our simulations , we needed impact parameters of a few 100 kpc to produce two asymmetric x - ray tails .
we note that our simulation , r8p10v25 , with the same infall velocity , 2500 @xmath2 , used by donnert and with a small impact parameter , predicts a relative radial velocity of 1000 @xmath2 , much larger than observed ( table [ t : table1 ] ) .
using self - consistent n - body / hydrodynamical simulations we confirm the previous interpretation of el gordo , as a massive galaxy cluster in an early stage of merging , about 480 million years after a first core passage , well before the first turn around .
we have used the x - ray morphology , sz , gravitational lensing and radio observations of el gordo to constrain the initial conditions of this merging system .
we have found that our simulations within a relatively narrow range of infall velocities and impact parameters can reproduce the data well . in particular
, we can reproduce the distinctive `` twin - tailed '' cometary x - ray morphology of el gordo with a simple clear physical explanation .
we match the observations with an initial infall velocity of @xmath1 2250 @xmath2 , when the clusters are separated by the sum of their virial radii .
the impact parameter of the collision is well constrained at @xmath1300 kpc to provide the asymmetric second tail which is a projection of the compressed and shock heated gas front .
we quantify this using the profile of the x - ray surface brightness across the wake @xcite , by the relative radial velocity from spectroscopy of member galaxies , and from the relative positions of the x - ray , sz and mass peaks .
note that our best simulation does reconstruct all the observed morphological features of el gordo , as well as the x - ray surface brightness across the wake quantitatively , but the position of the extraction region through the wake is not an exact match with that of the chandra observations .
therefore , for a more accurate determination of the infall velocity , we need to perform a search in a larger parameter space including the masses of the components , their concentration parameters , core radii for the intracluster gas .
this task necessitates many more simulations , which we defer to a future study .
this massive merging cluster with an infall velocity of 2250 @xmath2 at the the time the universe was half of its age ( at the redshift of @xmath8 ) is apparently very unusual in the standard @xmath5cdm models , for which extensive simulations have been performed to determine the expected probability distribution of relative velocities @xcite . at face value , such extreme cases present a challenge to our understanding of structure formation and may lead to a better understanding of dark matter and/or alternative theories of gravity .
we aim to expand our grid of input parameters for a fuller investigation of the uniqueness of our solution and the uncertainties on the derived physical parameters using more powerful gpu accelerated architectures .
we will also benefit from deeper x - ray observations and higher resolution sz imaging to more accurately constrain our models .
this work will become increasingly more important for cosmology as new discoveries of extreme cluster collisions are uncovered in the new wide area sz surveys sensitive to high pressure shocked gas and also from future radio surveys sensitive to `` relic '' radio emission from merger shocks .
the code flash used in this work was in part developed by the doe - supported asc / alliance center for astrophysical thermonuclear flashes at the university of chicago .
we thank the theoretical institute for advanced research in astrophysics , academia sinica , for allowing us to use its high performance computer facility for our simulations .
we thank j. p. hughes for providing us the processed x - ray data from their chandra observations of el gordo , and the referee for a thorough reading of our paper and for suggestions , which helped to improve on the presentation of our results .
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 the national aeronautics and space administration . | the distinctive cometary x - ray morphology of the recently discovered massive galaxy cluster `` el gordo '' ( act - ct j0102 - 4915 ; @xmath0 = 0.87 ) indicates that an unusually high - speed collision is ongoing between two massive galaxy clusters . a bright x
- ray `` bullet '' leads a `` twin - tailed '' wake , with the sz centroid at the end of the northern tail .
we show how the physical properties of this system can be determined using our flash - based , n - body / hydrodynamic model , constrained by detailed x - ray , sunyaev
zeldovich ( sz ) , and hubble lensing and dynamical data .
the x - ray morphology and the location of the two dark matter components and the sz peak are accurately described by a simple binary collision viewed about 480 million years after the first core passage .
we derive an impact parameter of @xmath1300 kpc , and a relative initial infall velocity of @xmath12250 @xmath2 when separated by the sum of the two virial radii assuming an initial total mass of @xmath3 @xmath4 and a mass ratio of 1.9 .
our model demonstrates that tidally stretched gas accounts for the northern x - ray tail along the collision axis between the mass peaks , and that the southern tail lies off axis , comprising compressed and shock heated gas generated as the less massive component plunges through the main cluster .
the challenge for @xmath5cdm will be to find out if this physically extreme event can be plausibly accommodated when combined with the similarly massive , high infall velocity case of the bullet cluster and other such cases being uncovered in new sz based surveys . |
interacting galaxies like the antennae ( ngc 4038/4039 ; whitmore & schweizer 1995 ) or stephan s quintet ( hcg 92 ; hunsberger 1997 ) show much star - burst activity in their tidal features .
high resolution images from the hst resolve these regions into many compact groups of young massive star clusters ( i.e. super - clusters ) and/or tidal - tail dwarf galaxies , with typical radii of 100500 pc . here
we aim to study the future fate of these super - clusters .
+ we begin with an overview of the numerical method and explain the setup of our simulations .
we then show results obtained so far and conclude with an outlook on future work we intend to pursue .
_ superbox _ is a hierarchical particle - mesh code with high - resolution sub - grids focusing on the cores and the star - clusters as a whole , and moving with them through the simulation area ( fellhauer et al .
the code has , for particle - mesh codes , a highly - accurate force - calculation based on a nearest grid - point ( ngp ) scheme .
the main advantages of _ superbox _ are it s speed and the low memory requirement which makes it possible to use a high particle number with high grid - resolution on normal desktop computers .
as a model for our massive star - clusters we use , for each , plummer - spheres with 100,000 particles , a plummer - radius @xmath0 pc and a cutoff radius @xmath1 pc , giving a total mass of @xmath2 and crossing time of @xmath3 myrs .
twenty of these clusters are placed in a compact group orbiting in a logarithmic potential of the parent galaxy , @xmath4 with @xmath5 kpc and @xmath6 km / s . the case
@xmath7 is dealt with in kroupa ( 1998 ) .
the distribution of the super - cluster is also plummer - like with different plummer - radii ( table 1 ) .
the tidal radius is @xmath8 2.4 kpc at apo - galacticon and 1.2 kpc at peri - galacticon .
the orbits have the same eccentricity in all cases , and begin at apo - galacticon ( @xmath9 kpc , @xmath10 ) with @xmath11 km / s ( @xmath12 ) . c|r|r simulation & scale - length & @xmath13 + run6 & 300 pc & 108.4 myr + run5 & 150 pc & 38.3 myr + run7 & 75 pc & 13.5 myr + [ tab : runs ]
in all runs some of the clusters merge very rapidly within the first @xmath14 myrs as seen in fig . 1 . the number of surviving clusters drops with increasing concentration of the super - cluster ( sc ) .
the orbits of the clusters inside their super - cluster also change rapidly due to the very short relaxation - time of the sc ( @xmath15 ) .
thereafter the merger - object and the surviving clusters move on epicycles around the orbit of the original sc about the galactic centre , with an increasing amplitude and period due to tidal heating ( fig 2 ) .
the surviving star - clusters remain in the vicinity of the merger - object ( fig .
in the intermediate concentration case ( run5 ) we find 2 merger - objects . to check whether this was just a chance event , a second run with a different random number seed for the star - cluster positions also gave 2 objects .
interesting in this context is that theis ( 1996 ) found that two clusters form in a cold collapse of a stellar system in an external tidal field .
violent relaxation in an external tidal field needs further addressing in the present context . in those cases where the sc has a low concentration
initially , the resulting merger - system is an extended object ( several kpc ) with a low density and an off - centre nucleus ( fig . 3 ) .
the radial density profile follows an exponential distribution ( fig . 4 ) .
the velocity - dispersion in these objects drops to @xmath8 5 km / s ( fig . 5 ) , which is comparable with the measured dispersions of dsph - galaxies in the milky way , and is slightly anisotropic .
in the case of the most concentrated sc , the merger - object is a dense compact spheroidal object ( fig .
3 ) with a high density core ( @xmath16 ) . the density profile follows a power - law with @xmath17 out to 0.8 - 1.0 kpc ( fig . 4 ) .
the velocity - dispersion of this spheroidal dwarf galaxy is anisotropic and around 15 km / s ( fig . 5 ) .
it has a half - mass radius of about 300 pc .
the size of this object is too large in comparison with even the biggest of the globular clusters , and the central density is far too high to be comparable with any de- , dirr- or dsph - galaxy .
however , it has properties similar to the initial satellites studied by kroupa ( 1997 ) , and is thus likely to evolve to a dsph - like satellite . but
further investigation of this issue is necessary .
we found that even if new high resolution images show that most of the so - called tidal dwarf galaxies are clusters of young compact massive star clusters they are likely to merge within a short time - scale .
the properties of the merger - objects differ with the scale - length of their initial distribution .
we found large fluffy objects with similar properties as the local dsph - galaxies as well as very compact and massive spheroidal objects , which , however , may be similar to the progenitors of some of the local present - day dsph satellites .
+ in the course of future work we intend to investigate the influence of the choice of orbit around the parent galaxy and how this alters the results .
we will focus on the transition between bound and unbound objects , and look for a region in the space of parameters where 2 merger - objects ( binary system ) are more likely to form .
our further research will also address the future fate of the merger - objects and their possible counterparts in reality .
fellhauer , m. , et al .
2000 , submitted to newa hunsberger , s. d. 1997 , , 29 , 1406 kroupa , p. 1997 , newa , 2 , 139 - 164 kroupa , p. 1998
, , 300 , 200 - 204 theis , c. 1996 astron.soc.pac.conf.ser . , 112 , 35 - 44 whitmore , b. c. , schweizer , f. 1995 , , 109 , 960 - 980 , 1412 - 1416 : it would be interesting to try and see if the age distribution among clusters that are clustered is different ( younger ) from that of yscs distributed more homogeneously .
what do you expect to happen with the merged clusters that you compared to a dsph .
is nt it bound to sink into the core of the merger remnant by dynamical friction ?
the bulk of the yscs is at @xmath18 kpc from the nucleus of ngc 4038 , i.e. not further away than ngc 4039 ( nucleus ) ! +
* answer * : we assume that at least some of the tidal - tail dwarfs seen to form in outer ( @xmath19 kpc ) tidal arms are composed of clusters of young massive star clusters . for our models we use an analytic galactic potential , because in the mass range of the merger - object of about @xmath20 m@xmath21 , dynamical friction does not play a significant role . + * e. grebel * : could you comment on how the merged clusters will resemble a dsph galaxy in their properties ( e.g. dsph do nt show rotation , have very low density and surface brightness , etc . ) ? + * answer * : it is too early to quantify the reply in detail , but we expect the merged object to show properties that could make it look similar to the progenitors of some of the dsph satellites .
the merged object is spheroidal , and has a high specific frequency of globular clusters , and low angular momentum , which however depends on the initial conditions .
it s stellar population contains stars from the mother galaxy , as well as stars formed during the star burst , and maybe stars ( and clusters ) formed during a possible later accretion event of a co - moving gas cloud .
many of these issues are discussed in kroupa ( 1998 ) . + * j. gallagher * ( comment ) : since super - star - clusters are often born in groups
the luminous clumps destruction via cluster merging is of general interest .
for example , if this process reduces the survival rate of massive - star - clusters , it might help to explain why intermediate age examples seem to be rare .
+ * d. mclaughlin * : what evidence is there that the clustered clusters in the antennae will actually merge ?
there are @xmath22 m@xmath21 clouds of gas in this system , so it may be that young clusters are clustered because several form in any given cloud , but they disperse after gas - loss . with no information on either the cluster - cluster velocity dispersion or the time when the parent cloud was disrupted , it would be difficult to rule out this possibility . + * answer * : this is an important issue , and very similar to the problem of forming bound star clusters . while not disproving rapid dispersal entirely , the argument which makes rapid dispersal less likely is as follows : the cluster - cluster velocity dispersion , @xmath23 is either small , which will lead to a bound merger object .
if @xmath23 is near to virial for the _ stellar _ mass in the super - cluster , then our models take care of that . if , however , @xmath23 is virial for the stars and a much larger mass in gas , then @xmath24 km / s assuming a star - formation efficiency of 20 per cent and a pre - gas removal super - cluster configuration as in run5 here .
thus , within 10 myr , the object will have expanded to a radius of at least 350 pc .
since many of the super - clusters are still very concentrated , and about 10 myr old , rapid expansion does not appear to be taking place , especially so since there are at least about 10 observed super clusters and they would have had to start in unrealistically concentrated configurations for them to appear with the sizes they have now .
this is further discussed in ( kroupa 1998 ) . | we investigate the behaviour of a cluster of young massive star clusters ( hereafter super - cluster ) in the tidal field of a host galaxy with a high - resolution particle - mesh code , _
superbox_. specifically we want to establish if and how such super star - clusters merge and carry out a detailed study of the resulting merger - object .
this merger - object shows either the properties of a compact spherical object or the elongated ( ` fluffy ' ) shape of dsph - galaxies depending on the initial concentration of the super - cluster . |
the spin@xmath1orbit interaction is a relativistic interaction connecting the spin and orbital motion of charged elementary particles .
this interaction plays essential roles in conversion between electric signals and magnetic ones @xcite .
of particular interest is the rashba spin@xmath1orbit interaction@xmath1the one that arises when the inversion symmetry of the system is broken @xcite .
it is represented by the quantum mechanical hamiltonian @xmath2 where @xmath3 is the electron mass , @xmath4 is the planck constant divided by @xmath5 , @xmath6 is a vector representing the direction and strength of the rashba field and @xmath7 and @xmath8 are the momentum operators for the electron and vector of pauli matrices , respectively .
the interaction leads to various cross - correlation effects in electromagnetism , i.e. a mixing of electric and magnetic fields .
a straightforward consequence of the rashba interaction in the dc limit is the spin polarization of a conduction electron induced when an electric current is applied ( called the rashba@xmath1edelstein effect ) @xcite , and the electric current generated when a magnetic field is applied ( the inverse edelstein effect ) @xcite .
the rashba - induced spin polarization has been predicted to be highly useful for magnetization switching by electric current @xcite .
magnetoelectric effects have been intensively studied in insulating materials ( multiferroics ) , where magnetization and electric polarization are mutually coupled @xcite .
a conducting rashba system is unique in this sense , since magnetoelectric effects couple magnetization and electric current instead of electric polarization ; thus , the effect is expected to be smoothly integrated into conventional electronic devices .
although the rashba interaction realized in two - dimensional electron gases in semiconductors has generally been weak , a strong rashba interaction of @xmath9 ev was recently observed at surfaces of noble metals doped with bi @xcite .
metallic surfaces and interfaces have now become highly important in spintronics for magnetization control and spin - to - charge conversion with high efficiency @xcite .
it has been observed that topological insulators , the extreme cases of rashba systems where the quadratic term in the dispersion disappears , have intriguing features such as a coupled collective mode of spin and charge ( spin plasmon ) on the surface @xcite .
bulk bitei , a semiconductor where carrier electrons are introduced by bi vacancies @xcite , was recently observed to have a large rashba splitting of @xmath10 ev@xcite ; a large magneto - optical ( kerr ) effect @xcite and a giant spin - polarized photocurrent @xcite were also discovered .
the coupling of electric and magnetic degrees of freedom by the rashba spin@xmath1orbit interaction leads to intriguing optical responses , including anisotropic propagation ( directional dichroism ) and electromagnetic metamaterial behavior ; metamaterials are materials that exhibit extraordinary electromagnetic characteristics such as negative refraction @xcite .
negative refraction was originally discussed by veselago and was considered to occur if the electric permittivity and magnetic permeability were simultaneously negative @xcite .
it has in fact been achieved by designing resonators for electric and magnetic responses with the same resonant frequencies in ghz @xcite and thz @xcite ranges . for shorter wavelengths , small structures comparable to the wavelength are necessary , and thus , artificial structures would not be suitable . instead , natural materials become promising , since their structures can be designed down to the atomic size . in fact , oxide superconductors with layered structures , such as la@xmath11sr@xmath12cuo@xmath13 and bi@xmath14sr@xmath14cacu@xmath14o@xmath15 , behave as metallic materials in the direction parallel to the layers and as dielectric materials in the direction perpendicular to the layers in the infrared frequency range @xcite . such materials with strong anisotropic electric properties ( called hyperbolic materials )
exhibit negative refraction when an interface parallel to the metallic plane is subjected to an electromagnetic wave .
these materials are also useful for filters and for enabling hyperlenses to go beyond the diffraction limit @xcite .
tetradymites , such as bi@xmath14te@xmath16 and bi@xmath14se@xmath16 , have been observed to be a new class of natural materials exhibiting a a hyperbolic nature for visible light @xcite . in this paper , we present a quantitative and consistent theoretical analysis of the optical properties of a bulk rashba conductor by evaluating the correlation functions from a microscopic standpoint .
first , we shall demonstrate that bulk rashba systems such as bitei indeed exhibit naturally significant hyperbolic metamaterial properties in the infrared regime because of spin@xmath1charge mixing .
these unique behaviors of rashba systems are not observed in topological insulators , for which the bare plasma frequency is not defined by the quadratic dispersion .
second , we demonstrate that in the presence of magnetization or a magnetic field , the system exhibits directional dichroism because of breaking of both spatial inversion and time reversal symmetries .
the effect is strongly enhanced by a singularity at the interband transition edges , the stronger of which is predicted to occur in the infrared and visible light regime for bitei .
our analysis establishes a bridge from the theoretical side between the optical properties discussed mostly on phenomenological grounds and solid - state transport phenomena based on the microscopic quantum mechanical viewpoint .
the qualitative equivalence of the dc transport and optical responses is noted experimentally as well as from the symmetry viewpoints @xcite .
the equivalence is natural from our analysis , since both phenomena are governed by a same conductivity tensor evaluated at different frequencies .
when an electric field is applied , the rashba interaction generates electron spin polarization ( the so - called edelstein effect ) @xcite ( fig . 1 a ) whose density reads @xmath17 where @xmath18 is a unit vector along the rashba field and @xmath19 is a coefficient .
the induced magnetization is @xmath20 , where @xmath21 is the gyromagnetic ratio .
the interaction also induces the inverse effect , where an electric current density is induced by an applied magnetic field @xmath22 , namely @xmath23 where @xmath24 is a coefficient
. these cross - correlation effects also affect the optical properties .
the inverse edelstein current ( [ jie ] ) is written by using faraday s law as ( in the momentum and angular frequency representation ) @xmath25 $ ] .
the edelstein effect ( [ se ] ) also leads to a magnetization current density , @xmath26 .
the coefficients of the edelstein effect and its inverse effect are related as @xmath27 since both the spin density ( [ se ] ) and current density ( [ jie ] ) are written in terms of the uniform component of the spin@xmath1current correlation function @xmath28 ( the superscripts @xmath29 denote the direction ) .
we thus see that the sum of the two generated currents is @xmath30 $ ] and that the conductivity tensor has off - diagonal elements linear in @xmath31 .
the off - diagonal elements result in a rotation of the incident electric field around a rotation axis @xmath32 by an angle proportional to @xmath33 .
the dispersion reads @xmath34 ( @xmath35 is the magnetic permeability ) for a circular polarization @xmath36 , and hence , the light velocity depends on both the incident direction ( sign of @xmath37 ) and the circular polarization ( directional circular dichroism ) .
this result is natural from a symmetry viewpoint , since breaking of inversion symmetry alone leads to the so - called natural optical activity , which affects only circularly polarized light @xcite .
peculiar behaviors of unpolarized light arise if the diagonal components of the conductivity tensor are modified by the rashba interaction that occurs when the direct and inverse edelstein effects are combined .
in fact , the magnetization @xmath38 induced by the edelstein effect generates an electric current due to the inverse edelstein effect : @xmath39 $ ] .
the resultant polarization @xmath40 is opposite to the applied electric field when @xmath41 ; in other words , the mixing of spin and charge excitations reduces the plasma frequency ( fig .
1 a ) . microscopically , this softening effect is represented in the conductivity tensor in the long - wavelength ( @xmath42 ) limit : @xmath43 where @xmath44 is a function representing the magnitudes of the rashba@xmath1edelstein and inverse edelstein effects ( plotted in fig .
1 b ) . here , @xmath45 is the electron spin polarization ( @xmath46 denotes spin ) , @xmath47 , @xmath48 , @xmath49^{-1}$ ] is the fermi distribution function , ( @xmath50 and @xmath51 are the boltzmann constant and temperature , respectively ) , @xmath52 ( @xmath53 and @xmath54 are the fermi energy and fermi wave vector , respectively ) and @xmath55 is a damping parameter . in terms of @xmath56 ,
the rashba@xmath1edelstein coefficient is @xmath57 ( @xmath58 denotes the imaginary part ) .
the function @xmath59 is finite for a broad frequency range lower than the higher interband transition edge @xmath60 , where @xmath61 are the transition edges , with @xmath62 being a dimensionless rashba strength parameter and @xmath63 ( see supplementary information ) . for bitei , @xmath64 ,
the plasma frequency @xmath65 , and @xmath66 are in the infrared regime , whereas @xmath60 is in the visible frequency .
choosing @xmath6 along the @xmath67-axis , the dielectric tensor @xmath68 reads @xmath69 where @xmath70 is the bare plasma frequency ( @xmath71 is the electron density ) .
as observed in equation ( [ e - perp ] ) , the plasma frequency in the @xmath72 plane is indeed reduced by the rashba interaction to @xmath73 .
the frequency range between the bare plasma frequency @xmath65 and the reduced plasma frequency @xmath74 is of particular interest , since the system is insulating in the direction of the rashba field but metallic in the perpendicular direction ( fig .
2 a , b ) .
materials with such anisotropy are called hyperbolic because of their hyperbolic dispersion @xcite .
in fact , choosing the wavevector @xmath31 in the @xmath75 plane , the dispersion relation is @xmath76 for @xmath77 ( @xmath78 denotes the @xmath79-component of the electric field ) and @xmath80 for @xmath81 and @xmath82 . when both @xmath83 and @xmath84 are positive , the dispersion is elliptic ; however , it is hyperbolic if one of these values is negative ( fig
. 2 c , d ) . note that in equation ( [ disp1 ] ) the component @xmath85 is governed by @xmath86 with positive real part , whereas @xmath87 is by @xmath88 with negative real part ; thus , the wave propagation in the @xmath89-direction is insulating ( metallic ) .
hyperbolic materials exhibit the following peculiar optical properties .
when the interface is parallel to the metallic plane , the incident light tends to be refracted in the negative direction , since the metallic nature reflects the light ( fig .
this behavior is observed in the dispersion curve ( fig .
2 c ) , which reveals that the @xmath90-component of the group velocity is negative .
the poynting vector @xmath91 also has a negative @xmath90-component .
notably , negatively refracted light is focused along the insulating direction , as observed in the plot of the transmittance defined by the ratio of the flow of @xmath91 as @xmath92 ( @xmath93 is the incident poynting vector ) ( fig .
2 e , g ) .
in fact , the angle of refraction , determined by @xmath94 , does not exceed a threshold value @xmath95 and the absolute value of the threshold angle is small , since no solution exists for small @xmath96 ( fig .
2 c ) . for @xmath97 , @xmath98 ( fig
. 2 g ) , the incident light even for a large angle of incidence is concentrated in a direction within @xmath99 from the direction perpendicular to the interface .
this focusing effect is stronger for smaller frequencies close to @xmath74 , whereas the distribution is broader for larger frequencies ( fig . 2 c , g , i ) .
in contrast , when the interface is perpendicular to the metallic plane ( fig . 2 b ) , the light is mostly reflected when the angle of incidence is small ( fig . 2 f ) , since no solution exists for small @xmath100 ( fig .
2 d ) . for transmitted light , in contrast to the parallel case , the angle of refraction can not be small , and thus , the refracted waves propagate close to the interface ( fig
. 2 h , j ) .
an interesting feature of the refracted wave is the backward wave , where the wave vector @xmath31 and the poynting vector @xmath91 are in opposite directions when the propagation is perpendicular to the interface ( fig .
2 b , d)@xcite . for bitei , the maximum plasma frequency known so far is @xmath101 hz ( corresponding to a wavelength of @xmath102 m ) for @xmath103 m@xmath104 and @xmath105 ev @xcite . using @xmath10 evand @xmath106 @xmath107 , we have @xmath108 , and thus , @xmath109 for frequencies @xmath110 ( fig .
1 b ) , resulting in @xmath111 ( @xmath112 hz , corresponding to the wavelength of @xmath113 m ) .
thus , hyperbolic behavior arises in the infrared regime . for experimental realization , the penetration depth @xmath114 of the system , defined by @xmath115}$ ] , is also essential .
the imaginary part of the relative permittivity is on the order of @xmath116 in the hyperbolic regime , and the imaginary part of @xmath117 calculated from equation ( [ disp1 ] ) is @xmath118\simeq \frac{\omega}{c}{{\rm im}}[{\varepsilon}]$ ] . for @xmath119
, we have @xmath120 m , and the absorption of the light is neglected if the sample thickness is smaller .
so far , we have considered a pure rashba conductor where time reversal symmetry is maintained .
we now discuss the case without time reversal symmetry , as realized in a thin film of a rashba conductor attached to a ferromagnet with uniform magnetization or in the presence of an external magnetic field .
we thus consider the @xmath121 exchange interaction @xmath122 where @xmath123 is a dimensionless magnetization vector and @xmath124 represents the strength of the exchange interaction .
the conductivity tensor including the effect of @xmath123 to the linear order is ( the coefficients are defined in the supplementary information ) @xmath125 { \nonumber\\ } & -\frac{1}{{{\epsilon_{\rm f}}}{{k_{\rm f}}}\omega } \biggl ( e_{1}(\omega ) [ ( \hat{\bm \alpha}\times{{\bm m } } ) \cdot { { \bm q } } ] \left[\delta_{ij}- \hat\alpha_{i}\hat\alpha_{j}\right ] + \frac{e_{2}(\omega)}{2}\left [ ( \hat{\bm \alpha}\times { { \bm m}})_{i}q_{\perp , j } + ( \hat{\bm \alpha}\times{{\bm m}})_{j}q_{\perp , i } \right ] { \nonumber\\ } & - \frac{e_{3}(\omega)}{2}\left [ m^{\perp}_{i}(\hat{\bm \alpha}\times { { \bm q}})_{j } + ( \hat{\bm \alpha}\times{{\bm q}})_{i}m^{\perp}_{j } \right ] \biggr ) \biggr ] , \label{sigmamq}\end{aligned}\ ] ] where @xmath126 and @xmath127 .
the first term of equation ( [ sigmamq ] ) is proportional to @xmath128 and corresponds to the anomalous hall effect induced by the magnetization parallel to the rashba field , which leads to an enormous magneto - optical response ( kerr effect ) @xcite .
other terms having diagonal components linear in the wave vector @xmath31 induce directional dichroism when @xmath129 is finite .
we consider the case where the interface is in the @xmath72 plane with the magnetization along the @xmath130-axis and the light incident to the @xmath75 plane .
for the electric field in the @xmath130-direction , the dispersion relation including the effect of magnetization ( equation ( [ sigmamq ] ) ) leads to @xmath131 when @xmath132 , where @xmath133 , @xmath134 and @xmath135 ( @xmath136 ) is a moment that governs the dichroism . in the broad frequency range below @xmath60
, @xmath137 is on the order of unity , whereas it has singularities at the transition edges ( @xmath138 ) ( fig .
more precisely , the dominant contribution to the function @xmath139 behaves as @xmath140 , and , since @xmath141 ( @xmath142 is a step function ) close to the edge at @xmath143 , @xmath137 diverges at the transition edges if @xmath144 as @xmath145 .
considering a finite damping , the directional dichroism is strongly enhanced by a factor of @xmath146 at the transition edges .
let us define the non - reciprocity ratio as the ratio of the transmission rate for the angle @xmath147 and @xmath148 as @xmath149 .
for the system with thickness @xmath150 , @xmath151 , where @xmath152 .
for @xmath153 and @xmath154 m , @xmath155 in the infrared regime with @xmath156 for @xmath157 , which is already large enough for experimental detection . at the edge @xmath143 , @xmath158
is strongly enhanced to be approximately 0.24 if @xmath159 . mathematically , the edge singularity results from the dichroism arising from the correlation function linear in @xmath160 , which is obtained by expansion with respect to the angular frequency .
an additional condition for imposing finite @xmath161 was necessary for the singularity in the rashba case.the singularity does not exist in the coefficient @xmath162 , and only a cusp structure exists at @xmath66 ( fig .
3 b ) ( the singularity in ref .
@xcite is a cusp . ) . in the presence of of many interband transitions , the singularity is still expected to be present , since each transition contributes independently to the singularity .
we have demonstrated that the non - reciprocal optical property is determined by the mutual orientation of @xmath31 and a vector @xmath161 .
this vector is known as the rashba - induced effective vector potential that drives electron spin flow @xcite , and in this context , the directional dichroism in the magnetic rashba system is a result of the doppler shift of the light due to the flow induced by the vector potential @xcite .
the factor @xmath163 essential for directional dichroism is consistent with the observation of dc transport @xcite , where the electric resistance in the presence of applied electric and magnetic fields ( @xmath164 and @xmath22 ) was observed to be anisotropic depending on @xmath165 ( @xmath31 is the wave vector of electron ) , as observed by the replacement @xmath166 and @xmath167 .
the effective vector potential @xmath161 is also interpreted as a toroidal moment ( a vector that couples with the rotation of magnetic field ) and has been discussed intensively in multiferroics @xcite .
in fact , the isotropic contributions of our dichroism result ( [ sigmamq ] ) is described by an effective hamiltonian @xmath168 where @xmath169 and @xmath170 are coefficients and @xmath171 ( see supplementary information ) .
the interaction ( [ hd ] ) is the interaction hamiltonian proposed to describe cross correlation effects in insulator multiferroics , and the vector @xmath161 and @xmath172 in this context are toroidal and quadrupole moments , respectively @xcite .
the form of interaction between the toroidal moment and the electromagnetic fields in equation ( [ hd ] ) suggests that the edge singularity is a common feature of toroidal moment - induced directional dichroism .
in fact , the energies of lights having positive and negative @xmath163 get shifted according to the interaction ; this energy difference results in directional dichroism if the interband transition matrix element depends on the energy .
as a result , the dichroism occurs proportional to the energy derivative of the interband matrix element , and the edge singularity arises whenever the matrix element vanishes at the edge @xmath173 faster than linear , i.e. proportional to @xmath174 with @xmath175 . in the real - space representation ,
the isotropic part of the current responsible for the dichroism ( contribution of equation ( [ sigmamq ] ) ) is @xmath176 , where @xmath177 and @xmath178 are the effective magnetization and electric polarization induced in metals by the toroidal and quadrupole moments , respectively .
these expressions are the metallic analogues of those in multiferroics , and our study indicates that the cross - correlation physics discussed so far for insulators is universal and can be applied to metals without changing their mathematical structure .
10 hirsch , j. e. spin hall effect .
, 1834 - 1837 ( 1999 ) .
saitoh , e. , ueda , m. , miyajima , h. & tatara , g. conversion of spin current into charge current at room temperature : inverse spin - hall effect .
, 182509 ( 2006 ) .
rashba , e. i. , properties of semiconductors with an extremum loop .1 .
cyclotron and combinational resonance in a magnetic field perpendicular to the plane of the loop . , 1109 - 1122 ( 1960 ) .
edelstein , v. m. , spin polarization of conduction electrons induced by electric current in two - dimensional asymmetric electron systems .
, 233 - 235 ( 1990 ) .
shen , k. , vignale , g. & raimondi , r. microscopic theory of the inverse edelstein effect .
, 096601 ( 2014 ) .
obata , k. & tatara , g. current - induced domain wall motion in rashba spin@xmath1orbit system .
, 214429 ( 2008 ) .
manchon , a. & zhang , s. theory of spin torque due to spin@xmath1orbit coupling . , 094422 ( 2009 ) .
tokura , y. multiferroics as quantum electromagnets .
1481 - 1482 ( 2006 ) .
christian , r. a. _ et al . _ giant spin splitting through surface alloying . , 186807 ( 2007 ) .
miron , i. m. _ et al .
_ current - driven spin torque induced by the rashba effect in a ferromagnetic metal layer .
230 - 234 ( 2010 ) .
snchez , j. c. r. _ et al . _ spin - to - charge conversion using rashba coupling at the interface between non - magnetic materials .
, 2944 ( 2013 ) .
raghu , s. , chung , s. b. , qi , x. l. & zhang , s. c. collective modes of a helical liquid .
, 116401 ( 2010 ) .
hobk , j. , lotk , l. , jansa , l. & matouek , p. departures from stoichiometry of bitei crystals grown from the vapor phase .
, 493 - 498 ( 1985 ) .
ishizaka , k. _ et al .
_ giant rashba - type spin splitting in bulk bitei .
, 521 - 526 ( 2011 ) .
demk , l. _ et al .
_ enhanced infrared magneto - optical response of the nonmagnetic semiconductor bitei driven by bulk rashba splitting .
, 167401 ( 2012 ) .
ogawa , n. , bahramy , m. s. , kaneko , y. & tokura , y. photocontrol of dirac electrons in a bulk rashba semiconductor . , 125122 ( 2014 ) .
smith , d. r. , pendry , j. b. & wiltshire , m. c. k. metamaterials and negative refractive index .
, 788 - 792 ( 2004 ) .
veselago , v. g. the electrodynamics of substances with simultaneously negative values of @xmath179 and @xmath180 .
, 509 - 514 ( 1968 ) .
pendry , j. b. negative refraction makes a perfect lens .
, 3966 - 3969 ( 2000 ) .
moser , h. o. , casse , b. d. f. , wilhelmi , o. & saw , b. t. terahertz response of a microfabricated rod - split - ring - resonator electromagnetic metamaterial .
, 063901 ( 2005 ) .
narimanov , e. e. & kildishev , a. v. metamaterials : naturally hyperbolic . , 214 - 216 ( 2015 ) .
poddubny , a. , iorsh , i. , belov , p. & kivshar , y. hyperbolic metamaterials . , 948 - 957 ( 2013 ) .
esslinger , m. _ et al .
_ tetradymites as natural hyperbolic materials for the near - infrared to visible . , 1285 - 1289 ( 2014 ) .
rikken , g. l. j. a. , flling , j. , & wyder , p. electrical magnetochiral anisotropy .
, 236602 ( 2001 ) .
rikken , g. l. j. a. & wyder , p. magnetoelectric anisotropy in diffusive transport .
, 016601 ( 2005 ) .
wagnire , g. h. the magnetochiral effect and related optical phenomena .
, 165 - 173 ( 1999 ) .
j. s. et al
. optical response of relativistic electrons in the polar bitei semiconductor .
, 117401 ( 2011 ) .
takeuchi , a. & tatara , g. spin damping monopole .
, 033705 ( 2012 ) .
kim , k. w. , moon , j. h. , lee , k. j. & lee , h. w. prediction of giant spin motive force due to rashba spin@xmath1orbit coupling . , 217202 ( 2012 ) .
nakabayashi , n. & tatara , g. rashba - induced spin electromagnetic fields in the strong sd coupling regime . ,
015016 ( 2014 ) .
kawaguchi , h. , ( 2015 ) .
spaldin , n. a. , fiebig , m. & mostovoy , m. the toroidal moment in condensed - matter physics and its relation to the magnetoelectric effect .
, 434203 ( 2008 ) . *
figure 1 * figure 1 .
a : schematic figure showing the cross - correlation effects in the plane perpendicular to the rashba field @xmath6 .
the edelstein ( e ) effect generates spin polarization @xmath181 from the applied electric field @xmath164 , and the inverse edelstein ( ie ) effect generates electric current @xmath182 from the induced magnetization @xmath38 in the direction opposite to the applied field .
[ figeie ] b : the real ( dotted lines ) and imaginary ( solid lines ) parts of the function @xmath56 are plotted as functions of @xmath183 for @xmath184 .
the interband transition edges @xmath143 and imaginary part of @xmath185 governing directional dichroism in the presence of magnetization are shown for the case of @xmath186 .
[ figcomega ] c : real and imaginary parts of @xmath187 and @xmath188 as a function of @xmath183 for @xmath189 and @xmath190 .
the shaded region between @xmath74 and @xmath65 is the hyperbolic regime .
[ figepsilon ] * figure 2 * + + + + figure 2 . optical properties in the hyperbolic regime for two geometries , the interface perpendicular to the rashba field @xmath6 ( a , c , e , g , i ) and the rashba field within the interface plane ( b , d , f , h , j ) .
a , b : schematic illustration of the geometries .
the incident wave vector , the wave vector in the rashba medium and the poynting vector are denoted by @xmath191 , @xmath31 and @xmath91 , respectively .
[ figrashba_light ] c , d : dispersion curves for a fixed angular frequency and the relation between the wave vector @xmath31 and the group velocity @xmath192 .
it is seen that the group velocity is refracted to the negative direction in the case c , while the wave vector @xmath31 has a negative @xmath193 component ( backward wave ) in the case d. [ figdispersion ] e , f : transmittance @xmath51 ( solid lines ) and angle of refraction for the poynting vector @xmath194 ( dotted lines ) plotted as functions of the angle of incidence @xmath195 for several values of @xmath196 , corresponding to close to the two edges and the center of hyperbolic regime .
g , h : transmittance @xmath51 plotted as function of @xmath194 . focusing effect and its dependence on the frequency are schematically shown in i , j .
focusing effect is stronger for larger wavelength ( shown by red in i and j ) .
[ figtzvsthetaout ] * figure 3 * figure 3 . a : schematic illustration showing the configuration of the rashba field @xmath6 , magnetization @xmath123 and wave vector @xmath31 required for the directional dichroism to appear .
the transmittance depends on the orientation of @xmath31 with respect to @xmath197 .
b : the imaginary parts of @xmath56 and @xmath198 ( representing the dichroism ) and the real part of @xmath199 ( representing the strength of the kerr effect ) . the strength of the dichroism has a singularity at the transition edges @xmath143 , whereas the kerr effect has only a cusp .
[ figdichroism ] * acknowledgements *
+ the author would like to thank s. seki , n. ogawa , r. arita , h. kawaguchi and n. nagaosa for their valuable comments .
this work was supported by a grant - in - aid for scientific research ( c ) ( grant no .
25400344 ) , ( a ) ( grant no .
24244053 ) from the japan society for the promotion of science and grants - in - aid for scientific research on innovative areas ( grant no.26103006 ) from the ministry of education , culture , sports , science and technology ( mext ) , japan .
the work of j. s. was supported by the inoue enryo memorial foundation for promoting sciences in toyo university . + * author contributions * + all authors performed calculations and commented on the manuscript .
most of calculations was performed by j.s and a.t . with help from other authors .
j. s. and g. t. contributed to interpretation .
g. t. planned the investigation and prepared the manuscript . + * author information * + correspendence should be addressed to j. s. ( e - mail : [email protected] ) and g. t. ( e - mail : [email protected] ) .
+ * competing financial interests * + the authors declare no competing financial interests . *
supplementary information * + in this supplementary information , we present details on the calculations , which is relevant to the discussion of the theory in the main text .
in this section , we present the microscopic model hamiltonian . let us consider a bulk rashba - electron conductor coupled with the magnetization , where the rashba spin@xmath1orbit interaction breaks the space inversion symmetry and the magnetic coupling breaks the time reversal symmetry .
the hamiltonian of the conduction electron is given by @xmath200c_{{{\bm k } } } , \nonumber\\ & = \sum_{{{\bm k}}}c_{{{\bm k}}}^{\dagger}\left [ \epsilon_{{{\bm k}}}-\epsilon_{\rm f } + { \bm \gamma}_{{{\bm k}},{{\bm m}}}\cdot{\bm \sigma } \right]c_{{{\bm k } } } , \label{h0}\end{aligned}\ ] ] where @xmath201 is the electron creation operator with wave vector @xmath202 , the up - arrow(down - arrow)@xmath203 represents the spin component along the @xmath67-axis in the spin space , @xmath204 is the energy dispersion of free electrons , @xmath205 is the fermi energy , @xmath206 is a vector of pauli spin matrices , @xmath207 is the magnetization vector and @xmath208 represents the strength of exchange interaction and @xmath209 with @xmath210 .
the energy dispersion of the hamiltonian in equation ( [ h0 ] ) is given by @xmath211 where @xmath212 correspond to the spin splitting upper ( @xmath213 ) and lower ( @xmath214 ) bands , respectively , and @xmath215 where @xmath216 with @xmath217 being the unit vector in the direction of rashba field .
when @xmath218 the energy dispersion is symmetric with respect to @xmath219-direction in @xmath202-space . on the other hand , when @xmath220 , the energy dispersion becomes asymmetric in the @xmath219-direction .
this asymmetry due to the magnetic coupling implies that the light absorption for electronic interband transitions depends on whether the @xmath219 is parallel or antiparallel to @xmath221 , which will lead to the directional dichroism for electromagnetic wave propagation in the rashba system .
the one - particle green s function ( resolvent ) is given by @xmath222 where @xmath223 .
the retarded and advanced green s function are given by respectively , @xmath224 and @xmath225 , where @xmath226 is a positive infinitesimal .
in this paper , we consider a clean limit and use the following relation in the latter calculations : @xmath227 where @xmath228 is the delta function .
let us evaluate current and spin response induced by an external electromagnetic field based on the linear response theory .
the perturbative hamiltonian representing the linear coupling between current and spin of electrons and an external electromagnetic field is given by @xmath229 where @xmath230 is the current density operator , @xmath231 is the spin density operator , @xmath232 is a vector potential yielding the electric field @xmath233 and the magnetic field @xmath234 , and @xmath235 is gyromagnetic ratio . in the following calculations ,
we evaluate the expectation values of @xmath230 and @xmath231 in the fourier space , whose components are given by @xmath236 where @xmath237 is the conventional velocity , @xmath238 is the anomalous velocity due to the rsoi , which is related to the spin density operator and @xmath239 .
fourier components of the electromagnetic field are also given by using @xmath240 as @xmath241 and @xmath242 , respectively .
the expectation values of @xmath243 and @xmath244 are written on the linear response of @xmath245 and @xmath246 as @xcite @xmath247e_{j}({{\bm q}},\omega ) + e\hbar\gamma \chi_{js}^{ij}({{\bm q}},\omega;{{\bm m } } ) b_{j}({{\bm q}},\omega ) , \label{j}\\ \langle \hat{s}_{{{\bm q}},i } \rangle_{\omega } & = \frac{ie}{\omega+i0^{+}}\chi_{sj}^{ij}({{\bm q}},\omega;{{\bm m } } ) e_{j}({{\bm q}},\omega ) -\hbar\gamma
\chi_{ss}^{ij}({{\bm q}},\omega;{{\bm m}})b_{j}({{\bm q}},\omega ) , \label{s}\end{aligned}\ ] ] where @xmath248 is the elementary charge and @xmath249 is an equilibrium electron density and @xmath250 , @xmath251 , @xmath252 and @xmath253 are current@xmath1current , current@xmath1spin , spin@xmath1current and spin@xmath1spin correlation functions , respectively . here and hereafter , repeated indices generally imply summation @xmath254 .
these are written by using the green s function as @xmath255^ { < } , \label{cjj } \\ & \chi_{js}^{ij}({{\bm q}},\omega;{{\bm m}})= -\sum_{{{\bm k}}}\int_{-\infty}^{\infty } \frac{d\epsilon}{2\pi i } { \rm tr}\left [ \tilde{v}_{i}g_{{{\bm k}}_{+},{{\bm m}}}(\epsilon_{+}){\sigma}_{j}g_{{{\bm k}}_{-},{{\bm m}}}(\epsilon_{- } ) \right]^ { < } , \label{cjs } \\ & \chi_{sj}^{ij}({{\bm q}},\omega;{{\bm m } } ) = -\sum_{{{\bm k}}}\int_{-\infty}^{\infty } \frac{d\epsilon}{2\pi i } { \rm tr}\left [ { \sigma}_{i}g_{{{\bm k}}_{+},{{\bm m}}}(\epsilon_{+ } ) \tilde{v}_{j}g_{{{\bm k}}_{-},{{\bm m}}}(\epsilon_{- } ) \right]^ { < } , \label{csj } \\ & \chi_{ss}^{ij}({{\bm q}},\omega;{{\bm m } } ) = -\sum_{{{\bm k}}}\int_{-\infty}^{\infty } \frac{d\epsilon}{2\pi i } { \rm tr}\left [ { \sigma}_{i}g_{{{\bm k}}_{+},{{\bm m}}}(\epsilon_{+ } ) { \sigma}_{j}g_{{{\bm k}}_{-},{{\bm m}}}(\epsilon_{- } ) \right]^{<}. \label{css}\end{aligned}\ ] ] where @xmath256 , @xmath257 and @xmath258 .
the green s function @xmath259 here is a path - ordered one @xcite
. the lesser component of @xmath259 is given by @xmath260 where @xmath261 is the fermi distribution function with @xmath51 being the temperature .
the key to understanding the optical magnetoelectric properties is to evaluate the optical conductivity , @xmath262 , which can be related to the correlation functions as @xmath263 + \frac{e\hbar\gamma}{\omega+i0^{+}}\left [ \chi_{js}^{il}{\varepsilon}_{lkj } + { \varepsilon}_{ikl}\chi_{sj}^{lj } \right]q_{k } + \frac{i\hbar^2\gamma^2}{\omega+i0^{+}}\chi_{ss}^{mn } { \varepsilon}_{ikm}{\varepsilon}_{jln}q_{k}q_{l } , \label{oc1 } \end{aligned}\ ] ] where @xmath264 is the totally anti - symmetric tensor with @xmath265 . in the main text ,
the optical conductivity @xmath262 is evaluated to the first order in @xmath31 and @xmath123 .
thus @xmath266 may be expanded up to the linear in @xmath31 and @xmath123 . for @xmath267 and @xmath268
, we can put @xmath269 and evaluate these up to the first order in @xmath123 . since the last term in equation ( [ oc1 ] ) is the second order in @xmath31 , this is negligible in the optical conductivity .
in this section , we present calculation results of correlation functions , @xmath270 , @xmath271 , @xmath28 and @xmath272 and the optical conductivity tensor , @xmath273 , and give the explicit form of coefficients , @xmath274 , @xmath162 , @xmath275 , of @xmath276 in the main text .
let us first perform the @xmath179-integral in equations ( [ cjj])-([css ] ) .
following the langreth s method , one can calculate the lesser component of @xmath277 as @xmath278^ { < } & = g^{<}(\epsilon_{+})g^{\rm a}(\epsilon_{- } ) + g^{\rm r}(\epsilon_{+})g^{<}(\epsilon_{-})\nonumber\\ & = f(\epsilon_{+})(g^{\rm a}(\epsilon_{+})-g^{\rm r}(\epsilon_{+}))g^{\rm a}(\epsilon_{- } ) + f(\epsilon_{-})g^{\rm r}(\epsilon_{+})(g^{\rm a}(\epsilon_{-})-g^{\rm r}(\epsilon_{- } ) ) .
\end{aligned}\ ] ] applying this rule to the correlation functions of equations ( [ cjj])-([css ] ) and using equation ( [ r - a ] ) , we can carry out the @xmath179-integral due to the delta function .
the next step is to expand correlation function with respect to @xmath31 and @xmath123 and subsequently to carry out the integration with respect to the rotation angle around the @xmath279-axis .
since these calculations are rather complicated , we here only show the results as @xmath280 \label{chi - jj - f}\\ \chi_{js}^{ij } & \simeq \frac{\hbar}{\alpha}\frac{n_{\rm e}}{m}c(\omega){\varepsilon}_{ijk}\hat{\alpha}_{k } + i\frac{\hbar}{\alpha}\frac{n_{\rm e}}{m}\frac{j_{sd}\hbar\omega}{\epsilon_{\rm f}^2 } \left(n_{i}(\omega)\hat{\alpha}_{j } -\hat{\bm \alpha}\cdot{\bm n}(\omega)\delta_{ij}\right ) , \label{chi - js - f}\\ \chi_{sj}^{ij } & \simeq-\frac{\hbar}{\alpha}\frac{n_{\rm e}}{m}c(\omega){\varepsilon}_{ijk}\hat{\alpha}_{k } -i\frac{\hbar}{\alpha}\frac{n_{\rm e}}{m}\frac{j_{sd}\hbar\omega}{\epsilon_{\rm f}^2 } \left(\hat{\alpha}_{i}n_{j}(\omega ) -\hat{\bm \alpha}\cdot{\bm n}(\omega)\delta_{ij}\right ) , \label{chi - sj - f}\\ \chi_{ss}^{ij}&\simeq \frac{\hbar}{\alpha}\frac{n_{\rm e}}{m}c(\omega ) ( \delta^{\perp}_{ij}-2\delta_{ij})-i\frac{\hbar^2}{\alpha^2 } \frac{n_{\rm e}}{m}\frac{j_{sd}\hbar\omega}{\epsilon_{\rm f}^2 } { \varepsilon}_{ijk}n_{k}(\omega)\end{aligned}\ ] ] where @xmath281 with @xmath217 , @xmath282 and the frequency - dependent coefficients are defined as @xmath283 with @xmath284 being the dimensionless rashba strength parameter , @xmath285 and @xmath286 , @xmath287 , @xmath288 , @xmath289 with @xmath290 and @xmath291 . substituting equations ( [ chi - jj - f ] ) , ( [ chi - js - f ] ) and ( [ chi - sj - f ] ) into the equation ( [ oc1 ] )
, we obtain the optical conductivity as @xmath292 where @xmath293 is the rashba@xmath1edelstein coefficient and @xmath294\nonumber\\ & -i\frac{e^2n_{\rm e}}{m}\frac{1}{k_{\rm f}\epsilon_{\rm f}\omega } \left [ e_{1}(\omega)(\hat{\bm \alpha}\times{{\bm m } } ) \cdot { { \bm q}}\delta^{\perp}_{ij } + \frac{1}{2}e_{2}(\omega ) \left\ { ( \hat{\bm \alpha}\times { { \bm m}})_{i}q^{\perp}_{j } + q^{\perp}_{i}(\hat{\bm \alpha}\times{{\bm m}})_{j } \right\}\right.\nonumber\\ & \left.- \frac{1}{2}e_{3}(\omega ) \left [ m^{\perp}_{i}(\hat{\bm \alpha}\times { { \bm q}})_{j } + ( \hat{\bm \alpha}\times{{\bm q}})_{i}m^{\perp}_{j } \right]\right ] \label{sigma - m}\end{aligned}\ ] ] with @xmath295 and @xmath127 . the first term of equation ( [ op2 - 1 ] ) leads to the softening of plasma frequency in the directions perpendicular to @xmath296 .
the second term represents the direct and inverse edelstein effects , which leads to the natural optical activity ( see main text ) .
the effect of the magnetization is contained in @xmath276 .
the term proportional to @xmath297 of equation ( [ sigma - m ] ) gives an anomalous hall effect , which leads to a magneto - optical response ( kerr effect ) .
other terms including @xmath298 or @xmath299 and having diagonal components linear in the wave vector @xmath31 of equation ( [ sigma - m ] ) induce directional dichroism .
in this section , we consider the wave propagation in the rashba system based on the optical conductivity of equation ( [ op2 - 1 ] ) and lead to the dispersion relation of equation ( 11 ) in the main text .
when we choose @xmath300 , @xmath301 and @xmath302 , the optical conductivity tensor can be written as @xmath303 where @xmath304 and @xmath305 .
thus the dielectric tensor @xmath306 is given by @xmath307 where @xmath308 with @xmath309 .
in general , the electromagntic wave propagation in conducting media is described by the following equation:@xcite @xmath310e_{j}=0 , \end{aligned}\ ] ] where @xmath311 is the light velocity . substituting equation ( [ dt1 ] ) into the equation ( [ weq-0 ] ) ,
we obtain the following characteristic equation yielding a dispersion relation for plane wave as @xmath312 thus for @xmath77 , which represents an ordinary wave , we have the dispersion relation @xmath313 putting the wave vector on @xmath314 , we obtain @xmath315 when @xmath316 , where @xmath317 and @xmath318 with @xmath319 , which leads to equation ( 11 ) in the main text .
thus we see that the second term including the factor @xmath320 of equation ( [ disp - dd ] ) , which depends on the wave propagating direction @xmath321 , leads to the directional dichroism .
the difference in the absorption rate of positive @xmath322 and negative @xmath323 wave propagation is proportional to the imaginary part of @xmath185 ( see main text ) .
in this section , we show the results of @xmath202-integrals of equations ( [ b - c])-([b - e3 ] ) . using @xmath327
we can write equations ( [ b - c])-([b - e3 ] ) as @xmath328 where @xmath329 at zero temperature , we have @xmath330 where @xmath331 at zero temperature , we write @xmath332 in equation ( [ c - s ] ) and @xmath333 in equation ( [ c - n ] ) as @xmath334 carrying out the integrals , we obtain the real and imaginary parts of @xmath332 and @xmath333 as @xmath335 , \label{rs}\\ & { \rm im}s = -\frac{1}{16\pi}\frac{1}{\alpha^3}\frac{1}{\overline{\omega } } \sum_{\lambda=\pm 1}\lambda s_{\lambda}= -\frac{1}{16\pi}\frac{1}{\alpha^3}\frac{1}{\overline{\omega } } \times \begin{cases } s_{+}-s_{- } & 0<\overline{\omega}<\overline{\omega}_{- } \\
-s_{- } & \overline{\omega}_{- } < \overline{\omega } < \overline{\omega}_{+}\\ 0 & \overline{\omega}_{+ } < \overline{\omega } \end{cases } \label{is}\\ & { \rm re}n = - \frac{1}{8\pi^2}\frac{\tilde{\alpha}^2}{\alpha^3}\sum_{\lambda}\frac{\lambda}{\lambda-\overline{\omega } } r_{\lambda } , \label{rn}\\ & { \rm im}n = \frac{1}{16\pi}\frac{\tilde{\alpha}^2}{\alpha^3 } \sum_{\lambda=\pm 1}\frac{1}{s_{\lambda}}= \frac{1}{16\pi}\frac{\tilde{\alpha}^2}{\alpha^3}\times \begin{cases } \dfrac{1}{s_{+}}+\dfrac{1}{s_{- } } & 0<\overline{\omega}<\overline{\omega}_{- } \\ \dfrac{1}{s_{- } } & \overline{\omega}_{- } < \overline{\omega } < \overline{\omega}_{+}\\ 0 & \overline{\omega}_{+ } < \overline{\omega } \end{cases } \label{in}\end{aligned}\ ] ] where @xmath336 ) with @xmath337 is a dimensionless angular frequency normalized by the fermi energy , @xmath338 and @xmath339 are frequencies at the transition edges .
the imaginary part of @xmath56 has cusps at @xmath340 ( see equation ( [ v - ic ] ) bellow ) .
note that when @xmath341 , the following replacement is understood @xmath342 substituting these results into equations ( [ b - c ] ) and ( [ b - d ] ) , we obtain @xmath343 - \frac{\delta_{\tilde{\alpha}}}{1 + 2\delta_{\tilde{\alpha } } } , \label{v - rc } \\ & { \rm im}c = \frac{3\pi}{8 } \frac{\tilde{\alpha}^{-1}}{1 + 2\delta_{\tilde{\alpha } } } \overline{\omega}\sum_{\lambda}\lambda s_{\lambda } , \label{v - ic}\\ & { \rm re}d = \frac{3}{16}\frac{\tilde{\alpha}^{-1}}{1 + 2\delta_{\tilde{\alpha } } } \left[\frac{1}{\overline{\omega}}\sum_{\lambda=\pm1 } ( \lambda-\overline{\omega})l_{\lambda}+2\tan^{-1}\tilde{\alpha}\right ] , \label{v - rd } \\ & { \rm im}d = \frac{3\pi}{32}\frac{\tilde{\alpha}^{-1}}{1 + 2\delta_{\tilde{\alpha } } } \frac{1}{\overline{\omega}}\sum_{\lambda}\lambda s_{\lambda } , \label{v - id}\\ & { \rm re}e_{1}= \frac{2\delta_{\tilde{\alpha}}\tilde{\alpha}^{-1}}{1 + 2\delta_{\tilde{\alpha } } } \frac{\epsilon_{\rm f}}{\hbar\omega } + \frac{3}{8}\frac{\tilde{\alpha}^{-2}}{1 + 2\delta_{\tilde{\alpha } } } \left [ 4\overline{\omega}\tan^{-1}\tilde{\alpha } + 2\sum_{\lambda}(\lambda-\overline{\omega})l_{\lambda } + \frac{9}{8}\tilde{\alpha}^2\sum_{\lambda}\frac{\lambda-2\overline{\omega}}{\lambda-\overline{\omega } } \lambda r_{\lambda}\right .
\nonumber\\ & \qquad~~\left.-\frac{3}{8}\tilde{\alpha}^2\sum_{\lambda } \frac{(\lambda-3\overline{\omega})(1+\tilde{\alpha}^2r_{\lambda } ) } { \overline{\omega}^2 + 2\lambda\tilde{\alpha}^2\overline{\omega}-\tilde{\alpha}^2 } \right ] , \label{v - re1}\\ & { \rm im}e_{1}= \frac{3\pi}{32}\frac{\tilde{\alpha}^{-2}}{1 + 2\delta_{\tilde{\alpha } } } \left[\sum_{\lambda}\lambda s_{\lambda } -\frac{15}{4}\overline{\omega}^2\sum_{\lambda}\frac{\lambda}{s_{\lambda } } + \frac{3}{4}\overline{\omega}^2(2\tilde{\alpha}^4-\tilde{\alpha}^2-\overline{\omega}^2 ) \sum_{\lambda}\frac{\lambda}{s_{\lambda}^3 } \right ] , \label{v - ie1}\\ & { \rm re}e_{2}= \frac{2\delta_{\tilde{\alpha}}\tilde{\alpha}^{-1}}{1 + 2\delta_{\tilde{\alpha } } } \frac{\epsilon_{\rm f}}{\hbar\omega}-\frac{3}{8}\frac{\tilde{\alpha}^{-2}}{1 + 2\delta_{\tilde{\alpha } } } \left [ 2\overline{\omega}\tan^{-1}\tilde{\alpha } + \sum_{\lambda}(\lambda-\overline{\omega})l_{\lambda } + \frac{3}{8}\tilde{\alpha}^2\sum_{\lambda}\frac{\lambda-2\overline{\omega}}{\lambda-\overline{\omega } } \lambda r_{\lambda}\right.\nonumber\\ & \qquad~~\left . -\frac{1}{8}\tilde{\alpha}^2\sum_{\lambda } \frac{(\lambda-3\overline{\omega})(1+\tilde{\alpha}^2r_{\lambda } ) } { \overline{\omega}^2 + 2\lambda\tilde{\alpha}^2\overline{\omega}-\tilde{\alpha}^2 } \right],\label{v - re2}\\ & { \rm im}e_{2}= -\frac{3\pi}{32}\frac{\tilde{\alpha}^{-2}}{1 + 2\delta_{\tilde{\alpha } } } \left[\sum_{\lambda}\lambda s_{\lambda } -\frac{5}{4}\overline{\omega}^2\sum_{\lambda}\frac{\lambda}{s_{\lambda } } + \frac{1}{4}\overline{\omega}^2(2\tilde{\alpha}^4-\tilde{\alpha}^2-\overline{\omega}^2 ) \sum_{\lambda}\frac{\lambda}{s_{\lambda}^3 } \right ] , \label{v - ie2}\\ & { \rm re}e_{3}= \frac{3}{8}\frac{\tilde{\alpha}^{-2}}{1 + 2\delta_{\tilde{\alpha } } } \left [ -2\overline{\omega}\tan^{-1}\tilde{\alpha } -\sum_{\lambda}(\lambda-\overline{\omega})l_{\lambda } -\tilde{\alpha}^2\sum_{\lambda}\frac{\lambda-2\overline{\omega}}{\lambda-\overline{\omega } } \lambda r_{\lambda } \right ] , \label{v - re3}\\ & { \rm im}e_{3}= \frac{3\pi}{32}\frac{\tilde{\alpha}^{-2}}{1 + 2\delta_{\tilde{\alpha } } } 2\overline{\omega}^2\sum_{\lambda}\frac{\lambda}{s_{\lambda } } .\label{v - ie3}\end{aligned}\ ] ] thus , from equations ( [ v - ie1 ] ) and ( [ v - ie3 ] ) , we obtain the explicit form of @xmath344 as @xmath345 .
\label{ie13}\end{aligned}\ ] ] we see that this function diverges towards the transition edges @xmath346 because of the second and last terms .
the dominant term at the edges is the last term .
thus the directional dichroism is strongly enhanced ( see main text ) .
note that similar equations as ( [ is ] ) and ( [ in ] ) are understood in the equations ( [ v - ie1 ] ) , ( [ v - ie2 ] ) , ( [ v - ie3 ] ) and ( [ ie13 ] ) .
in this section , we briefly make an analogy between our microscopic results of @xmath347 in ( [ sigma - m ] ) and the conductivity induced by an effective toroidal and quadrupole moments .
let us consider an effective hamiltonian proposed to describe cross correlation effects : @xcite @xmath348 where @xmath169 and @xmath170 are coefficients , and @xmath349 and @xmath350 are a toroidal and a quadrupole moments , respectively .
the induced current density is obtained from the derivative of the hamiltonian with respect to the vector potential , @xmath351 , as @xmath352 where the conductivity @xmath353 is calculated as @xmath354 + ih \omega\left ( q_{il}{\varepsilon}_{lkj}-{\varepsilon}_{ikl}q_{lj } \right)q_{k}.\end{aligned}\ ] ] in the real space representation , the induced current is written as @xmath355 where @xmath356 and @xmath357 are an effective magnetization and electric polarization induced by the toroidal and quadrupole moments , respectively . if we put respectively @xmath349 and @xmath172 on @xmath358 and @xmath359 , we have @xmath360 \nonumber\\ & + ih\left[m_{i}({\bm \alpha}\times { { \bm q}})_{j}+({\bm \alpha}\times{{\bm q}})_{i}m_{i}\right ] + ih\left[\alpha_{i}({{\bm m}}\times { { \bm q}})_{i}+({{\bm m}}\times{{\bm q}})_{i}\alpha_{j}\right ]
. \label{sigma - q}\end{aligned}\ ] ] thus , except for coefficients , we can have derived the linear terms of @xmath276 with respect to @xmath31 and @xmath123 .
some differences between the two may come from the anisotropic property due to the rashba field @xmath279 . | the rashba interaction induced when inversion symmetry is broken in solids is a key interaction connecting spin and charge for realizing novel magnetoelectric cross - correlation effects . here
, we theoretically explore the optical properties of a bulk rashba conductor by calculating the transport coefficients at finite frequencies .
it is demonstrated that the combination of direct and inverse edelstein effects leads to a softening of the plasma frequency for the electric field perpendicular to the rashba field , resulting in a hyperbolic electromagnetic metamaterial . in the presence of magnetization , a significant enhancement of anisotropic propagation ( directional dichroism )
is predicted because of interband transition edge singularity . based on an effective hamiltonian analysis , the dichroism is demonstrated to be driven by toroidal and quadratic moments of the magnetic rashba system .
the effective theory of the cross - correlation effects has the same mathematical structure as that of insulating multiferroics .
# 1 # 1 _ 0 # 1 # 1 _ i # 1 | # 1|#1 # 1#2 .
|#2 # 1 # 1#2 # 1 # 1 # 1 # 1 # 1#2 # 1#2 # 1#2 # 1 # 1#2 ( c # 1 + # 2 ) # 1#2#3 ( c # 1 + # 2 + # 3 ) # 1#2#3#4 ( cc # 1 & # 2 + # 3 & # 4 ) # 1eq .
( [ # 1 ] ) # 1#2eqs .
( [ # 1])([#2 ] ) # 1([#1]) # 1#2?[#1])([#2 ] ) # 1@xmath0 l # 1#2g_#1^#2 # 1#2j_s#1^#2 # 1#2_s#1^#2 # 1_s#1 # 1#2a_s,#1^#2 # 1e_s,#1 # 1b_s,#1 # 1 * # 1 |
low - energy phenomena in the hadron physics are strongly determined by the nonperturbative ( np ) interactions . at the fundamental ( qcd ) level hadron characteristics
are analyzed in the qcd sum rules with a np vacuum condensates .
hadron structure and interactions can be also considered in the effective lagrangian approach .
starting from the qcd , chiral low - energy expansion of green functions leads to effective lagrangian with phenomenological coupling constants @xcite .
effective chiral hadron model having the nambu - jona - lasinio structure with the quark - meson interactions and incorporating vector mesons as a gauge fields , has been also applied to the hadron reactions studying @xcite .
in fact , quark - meson models contain some input ( np ) parameters : effective couplings , which are defined by the ( diverged ) quark loops .
their regularization fixes some np scale of the theory , @xmath6 .
obviously , these models reproduce both nonperturbative and structure effects of the low - energy interactions @xcite . as it will be demonstrated , dynamical symmetry of hadron interactions can be adequately considered in the gauge scheme for hadron and ( constituent ) quark fields from the very beginning .
so , gauge principles , spontaneous symmetry breaking and the higgs mechanism work well not only at the fundamental quark level , but at the hadron one . in this way
, baryon - meson interactions were analyzed in @xcite .
the suggested renormalizable quantum field model is the @xmath5 model inspired construction , where the fundamental hadron components are reproduced by the constituent quarks and mesons . to analyze strong and electromagnetic processes
it is sufficient to consider @xmath7 gauge group with the spontaneous symmetry breaking .
then , we localize the diagonal sum of chiral @xmath8 subgroups of the global @xmath9 group only . due to extra @xmath10 groups , vector meson dominance
is realized at the gauge level as the mixing of initial gauge fields , @xmath11 , @xmath12 and @xmath13 ( their chiral partners occur after localization of the total chiral group ) . despite of radiative decays of light mesons
, the model describes some details of the scalar meson physics .
namely , after the spontaneous breaking of the local symmetry , residual scalar massive degrees of freedom ( higgs fields ) reproduce physical scalar states : isotriplet @xmath2 and isosinglet @xmath3 .
some properties of the scalar meson spectra are explained in this way .
note , @xmath5 meson is associated with @xmath14 meson @xcite .
mesons @xmath15 and @xmath16 manifest themselves as real external fields , and work also as an effective virtual components of internal hadron vacuum .
their occurence in loops can be understood as an excitation of the quark sea in a hadron .
the initial model lagrangian can be written as follows : @xmath17 here @xmath18 is the first generation quark doublet ( the fundamental representation ) ; @xmath16-triplet is the adjoined representation of the @xmath19 gauge group ; @xmath15-meson is a singlet ; @xmath20 - two doublets of complex scalar fields with the hypercharges @xmath21 , @xmath22 and @xmath23 .
physical states are formed by the primary fields mixing when quadratic mass forms of scalar and vector fields are diagonalized . at the tree level , the mass forms arise as a result of vacuum shifts which generate the gauge fields masses : @xmath24 . the model lagrangian in terms of physical fields can be found in @xcite . due to universality of vector boson couplings , their interactions with quarks and mesons
are described by three gauge parameters : @xmath25 , @xmath26 and @xmath27 .
the lagrangian vertices contains some orthogonal linear combinations of @xmath28 with ( tree ) mixing angles @xmath29 .
these angles are fixed by the experimental data on the vector meson masses and decay widths .
angle @xmath30 is determined by the diagonalization of the vector fields quadratic form .
we have the following relations : @xmath31 in the model , effective `` tree '' couplings on the mass shell of external particles are reproduced by the sum of tree and one - loop quark and meson diagrams , they contain some np scale and depend on the mixing angles , @xmath32 and @xmath30 .
so , the couplings incorporate both short - distance and long - distance contributions ; the `` tree '' couplings values are fixed from the data on meson two - particle decays with a sufficient accuracy .
we suppose , these effective on - shell vertices ( not formfactors ! ) have tensor structure of the bare vertices . from preliminary considerations
it follows that extra terms in off - shell formfactors ( @xmath33 , where @xmath34 is the external momenta , for example ) contribute into physical quantities ( decay widths ) no more than @xmath35 .
thus , firstly , there are loop diagrams , which contribute into on- and off - shell couplings .
approximately , an account of these corrections can be reproduced by some multiplicative parameter in tree amplitudes . secondly , there are loop diagrams , which can not be reduced to contributions into the effective couplings and vertex formfactors .
the obvious difference between effective on - shell constant couplings and off - shell formfactors ( depending on the np cutoff parameter ) prevents from the double - counting in the model .
note , in these formfactors the internal particles masses interpret as the physical ones .
therewith , decays like @xmath36 , @xmath37 , @xmath38 , @xmath39 and @xmath40 are described by finite quark loop diagrams . with a good accuracy
, the vertex couplings in these diagrams can be taken as the effective @xmath41 ( @xmath42 ) tree constants in agreement with the experimental data .
note , these arguments can be applied to the electromagnetic vertices too - the vertices are renormalized by strong interactions , but the corrections from quark and meson loops are small in the case under consideration .
this fact is endorsed by calculations of loop processes with on - shell photons , in particular , @xmath40 ( see below ) .
thus , the consistency and supportability of the approach is verified by the results for all types of possible reactions - both at the tree and the loop level .
to fix the set of the gauge physical couplings and vacuum shifts , we use experimentally observable mass spectrum and widths of some two - particle hadronic decays of vector mesons @xcite - in our model these processes occur at the tree level , fixing on - shell `` tree '' couplings .
constants @xmath43 , @xmath44 ( or @xmath45 ) and @xmath46 are fixed from the experimental values of @xmath47 @xmath48 and @xmath49 .
then , the value of @xmath25 can be extracted from the relation @xmath50 ( see ( [ e:12c ] ) .
thus , from the experimental data and ( [ e:12c ] ) we get the set of main parameters values : @xmath51
due to two - level structure of the model , the meson - meson lagrangian classifies both tree and loop level processes , while the quark - meson interactions occur as loop contributions only .
radiative decays of vector mesons @xmath52 and @xmath53 are some test reactions for the meson - meson sector of the gauge model .
obvious tree diagrams ( see @xcite ) describe these processes with the pair @xmath54 production
. analytical form of the differential width for the process is presented in @xcite . in fig.1 curve ( 1 )
describes the model fit of normalized spectrum of photons , @xmath55 , with a single free multiplicative parameter .
it improves the description at the photon low energy range due to an account of @xmath56 dependence of couplings .
namely , normalizing the spectrum by the theoretical value @xmath57 , the ratio depends on the @xmath58 formfactors at the different energy scales because intermediate pions are off - shell .
totally , this multiplicative parameter increases the ratio no more than @xmath59 , due to the formfactors contribution .
curve ( 2 ) demonstrates comparison of the spectrum with the experimental data @xcite . to describe the spectrum fine structure near @xmath60 ,
it was suggested to consider the loop corrections ( see @xcite ) which do not contribute in vertices formfactors .
integrating of the spectrum from @xmath61 up to @xmath62 , we get the value of @xmath11- meson branching @xmath63 which slightly exceeds the experimental value @xmath64 @xcite
. however , an account of loop contributions ( with the phenomenological couplings ) leads to even more result : @xmath65 @xcite .
an excess of @xmath66 over @xmath67 can be caused by the lack of experimental data at energy range @xmath68 .
replacing @xmath69 ( @xmath70 ) and @xmath71 in the formula for the differential width , we have @xmath72 and @xmath73 for @xmath74 and @xmath75 , respectively ( it agrees with @xmath76 @xcite ) . certainly , loop corrections are important and in analogy with @xcite they can increase @xmath77 up to @xmath78 .
essentially , in the loop level processes all divergencies are summed to zero when all external lines are on the mass shell @xcite . in the quark - meson models
, radiative decays @xmath79 and three - particle decays @xmath80 occur via quark loops only with the gauge ( tree ) vertices .
corresponding one - loop diagrams and analytical expressions for the decays @xmath81 are in @xcite . the goldberger - treiman relation @xmath82 defining the constant of @xmath83 and @xmath84 interaction had been used . from this relation
it also follows that vacuum shift @xmath85 .
constituent quark mass value , as a free parameter , is extracted from the widths fit .
the widths of @xmath79 agree with the experimental data @xcite if we take @xmath86 and keep an imaginary part of the amplitude as nonzero : @xmath87 note , isotopic structure of vertices gives : @xmath88 , and it is confirmed by the results of calculations .
taking the small value @xmath89 , we assume that the ( constituent ) quarks propagate and annihilate in the internal np hadron vacuum , so they are not free asymptotic states .
i.e. , the imaginary parts of amplitudes describe the constituent quark propagation up to the distances , where the np effects are sufficiently large to bind the quarks into a real meson .
attractive forces between the quarks lead to the effective decreasing of their mass . instead , known `` naive '' idea can be used : imaginary part of quark loop equals zero `` by hand '' @xcite .
then , from the above widths we get @xmath90 . a further way to provide the `` infrared confinement '' was suggested in @xcite by the introducing of a special cutoff parameter , @xmath91 , as a common scale in @xmath92 representation of the process amplitude . in particular , from this procedure the decay widths follow in a good agreement with the data , if @xmath90 and @xmath93 .
at the same time , the value @xmath94 leads to light quarks , @xmath95 and nearly the same widths values . in all cases , the ratio @xmath96 .
as it is known , decay @xmath40 is described by finite triangle quark diagrams .
the decay width is : @xmath97 numerically , @xmath98 for @xmath99 . if @xmath100 , the width is @xmath101 . in the exact chiral limit @xcite @xmath102 , and the experimental value is in the interval @xmath103 @xcite .
however , it should be taken into account : a ) a possible violation of the goldberger - treiman relation can be as much as @xmath104 @xcite ( it means a weak @xmath56 dependence of @xmath105 ) ; b ) @xmath106 mixing can be noticeable for the case @xcite , c ) loop corrections to @xmath107 vertex are also significant .
namely , taking @xmath108 and decreasing effective @xmath84 coupling by @xmath109 only , we get @xmath110 . to explain the discrepance with the data , it is sufficient also to decrease the effective electromagnetic coupling @xmath107 no more than @xmath111 ( it can be provided by small variation of np scale , @xmath112 , in the diverged triangles of @xmath113 and @xmath114 types )
. moreover , for the decay amplitude imaginary part equals zero . nevertheless , introducing an infrared cutoff with @xmath115 , for @xmath108 we have @xmath116 .
however , nonuniversality of the parameter @xmath91 arises in the case . if @xmath117 , from exact integration of the amplitude we get a good agreement with the data . from quark
loops an effective couplings can be determined .
then , @xmath118 and the model gives @xmath119 if @xmath120 ; @xmath121 for @xmath122 and @xmath123 when @xmath124 .
analogously , @xmath125 and @xmath126 when @xmath127 ; @xmath128 and @xmath129 . for all these parameters we get the ratio @xmath130 , excepting light quark case with @xmath131 .
scalar isotriplet @xmath2 occurs in the model with an accuracy up to small mixing angles , @xmath32 and @xmath30 .
namely , from initial higgs doublets with @xmath132 after the spontaneous breaking , two charged scalar combinations arise .
they can be unified with the residual neutral components to constitute the isotriplet . in the simple case of
zero mixing ( then exact isotriplet structure emerges and @xmath133 , but this equality should be slightly broken by electromagnetic forces ) , @xmath134 and @xmath135 parts of the lagrangian have the form , where isovectors @xmath136 and @xmath137 are introduced : @xmath138= \frac{i}{2}\ , g_2\,\epsilon_{\alpha\beta\gamma}\rho^{\mu}_{\alpha}a_{\beta}\partial_{\mu}a_{\gamma}.\end{aligned}\ ] ] interaction of scalars with @xmath16-mesons is described by the lagrangian @xmath139 there is an interesting feature of this sector : mixing angle @xmath140 defining the physical fields @xmath141 can be equaled zero .
then , from explicit expressions @xmath142 , @xmath143 we found : @xmath144 , if @xmath145 .
it means @xmath146 , so the term @xmath147 in the initial lagrangian is damped . from experimentally known ( small ) difference of @xmath148 and @xmath149 masses an estimation of @xmath150 can be extracted , because @xmath151 .
also , from @xmath152 it follows @xmath153 ; as a consequence , scalars @xmath154 do not interact with quark current directly . due to free parameters , it is possible to consider dominant decay channels of scalar mesons - @xmath155 and @xmath156 @xcite .
decay @xmath157 is not observed , so from @xmath158 it follows that @xmath159 .
exact symmetry , @xmath133 , is not possible because it forbids the suppressed ( but observed ) decay @xmath160 . at the tree level
two - photon decays of scalar mesons are absent .
considering @xmath39 decay , we have triangle loop diagrams mediated by quarks , pions , @xmath148- and @xmath11-mesons . due to above arguments ,
vector meson loops are absent .
total amplitude of the process is gauge invariant without any divergencies : @xmath161.$ ] experimentally , @xmath162 @xcite . from the `` naive approximation '' we get @xmath163 for @xmath164 and @xmath165 beyond the chiral limit . besides , @xmath166 for @xmath167 ; in this case an agreement can be achieved only if @xmath168 of the chiral limit value and @xmath169 .
this value contradicts to known data @xmath170 .
note , imaginary part of @xmath171 is small for all cases , and @xmath172 .
now , the `` infrared confinement '' procedure results to @xmath173 for @xmath174 and @xmath175 .
all these values provide good agreement with the data for @xmath176 and @xmath155 decays .
as to @xmath177 and @xmath178 , it has been found in @xcite that the decays can be explained only by four-(constituent ) quark components of the mesons . in the gauge model , these components are simulated by vector meson ( due to @xmath179 vertices ) and scalar meson loops ( due to @xmath180 verices ) ; analysis of these decays is in progress . to include axial vector mesons and strange quarks , the model should be generalized to chiral @xmath181 . then dominant decay channel of the scalar mesons ( @xmath182 decay , for example ) will be understood and calculable as a possible admixture of @xmath183 states into scalar structure . note , however , the decay @xmath184 can be described by the symmetry allowed extra term @xmath185 in the model lagrangian .
the gauge quantum field approach is applied to low - energy hadron interactions using the higgs mechanism in a framework of @xmath5 model with the vmd .
residual higgs degrees of freedom allow to reproduce @xmath2 and @xmath3 scalars . due to free parameters , the mass degeneration of these mesons is explained .
decay @xmath186 is damped , as it is known from the expriment , two - pion decays of @xmath187 ans @xmath5 mesons agree with the experimental data well .
we also conclude that the two - quark components of scalar mesons ( @xmath188 ) are small - they occur at the loop level only .
radiative decays of vector mesons ( at the tree and loop level ) and two - gamma decays of @xmath5 and @xmath4 mesons are well described too , the values @xmath189 and @xmath190 are preferred by the analysis with the `` infrared confinement '' procedure . also , the values of effective vertices @xmath191 correspond to known experimental data .
so , the gauge renormalizable generalization of @xmath5 model based on @xmath192 group with the spontaneous symmetry breaking and the vmd is a reasonable and consequtive way to consider radiative and hadronic decays of vector and scalar mesons . some important features of scalar mesons spectrum are explained in the model due to interpretation of these mesons as vacuum fields .
chiral generalization of the model should be done to include axial vector states and other scalar mesons into the gauge scheme .
99 j. gasser , h. leutwyler , ann . of phys .
* 158 * , 142 ( 1984 ) .
j. gasser , h. leutwyler , nucl .
b250 * , 465 ( 1985 ) .
volkov , phys . of part . and nucl .
* 24 * , 35 ( 1993 ) .
volkov and a.e .
radzhabov , sov .
j. phys - usp . *
49 ( 6 ) * , 551 ( 2006 ) .
olean , arxiv : hep - ph/0310005 .
ron - chou hsieh , arxiv : hep - ph/0306143 .
vereshkov and v.i .
kuksa , yad . fiz . * 54 * , 1700 ( 1991 ) .
r. delbourgo and m.d .
scadron , mod .
a10 * , 251 ( 1995 ) ; arxiv : hep - ph/9910242 .
p. minkowski and w. ochs , arxiv : hep - ph/0209225 .
v. beylin , v. kuksa , g. vereshkov , arxiv : hep - ph/1006.3624 .
c. amsler _ et al . _ , [ particle data group ] , phys .
* b667 * , 1 ( 2008 ) .
vasserman _ et al .
_ , sov.j.nucl.phys . * 47 * , 1035 ( 1988 ) .
dolinsky _ et al .
_ , phys . rept . * 202 * , 99 ( 1991 ) .
a. gokalp , a. kucukarslan and o. yilmaz , phys . rev . *
d67 * , 073008 ( 2003 ) ; arxiv : hep - ph/0302240 .
a. bramon , r. escribano , jhep * 0402 * , 048 ( 2004 ) ; arxiv : hep - ph/0305043 .
kuksa , sov j. yad . fiz . * 66 * , 1785 ( 2003 ) . m.d .
scadron , arxiv : hep - ph/0007184 .
m. bystritskiy , e.a .
kuraev , m. secansky and m.k .
volkov , nucl .
suppl . 210
( 2008 ) ; arxiv : hep - ph/0804.3714 .
d. blaschke , g. burau , m.k .
volkov , v.l .
yudichev , arxiv : hep - ph/9812503 . t. branz _ et al .
_ , arxiv : hep - ph/0912.3710 . m. napsuciale , s. rodriguez and e. alvarado - anell , phys . rev . *
d67 * , 036007 ( 2003 ) ; arxiv : hep - ph/0212087 .
r. delbourgo and m.d .
scadron , int .
* a13 * , 657 ( 1998 ) ; arxiv : hep - ph/9807504 .
ioffe and a.g .
oganesian , phys . lett . *
b647 * , 389 ( 2007 ) ; arxiv : hep - ph/0701077 . m. napsuciale , arxiv : hep - ph/020417 . | electromagnetic and strong hadron processes at low energies are considered in the renormalizable model with the spontaneously broken @xmath0 gauge symmetry
. calculated radiative widths of vector mesons and effective couplings @xmath1 agree with the experimental data .
residual higgs degrees of freedom are associated with scalar states @xmath2 and @xmath3 with the degeneration in masses .
two - gamma decays of @xmath4 and @xmath5 mesons are analyzed in detail . to provide an `` infrared confinement ''
a cutoff procedure has been also used in calculations . |
energy losses from heavy elements , such as tungsten , due to radiative emission are one of the crucial problems to be overcome for the successful performance of themonuclear reactors .
nevertheless , tungsten is used as the plasma - facing component in the modern fusion facilities because of its essential properties , like high - energy threshold for sputtering , low sputtering yield , and excellent thermal features .
even small concentration of tungsten ions ( @xmath8 ) relative to the electron density prevents ignition of a deuterium - tritium plasma @xcite .
theoretical modeling provides information about processes in such harsh conditions .
however , fusion plasma modeling requires a significant amount of atomic data .
the ionization and recombination processes determine the charge - state distribution in plasma .
the electron - impact single ionization is the strongest one among the ionization processes .
contribution of the double ionization as a rule is much weaker compared with the single ionization .
furthermore , study of the double ionization process is quite complicated @xcite .
electron - impact ionization for singly charged tungsten ions was previously studied by using the crossed - beam technique @xcite .
these data were supplemented by the electron - impact single , double , and triple ionization measurements of w@xmath9 ions in the charge states @xmath10 , @xmath11 , and @xmath12 , respectively @xcite . the configuration - average distorted - wave ( cadw ) calculations provided good agreement with experimental measurements for higher ionization stages ( @xmath13 ) @xcite . the dw approach succeeded in getting a very good agreement with experiment for the ionization stages up to w@xmath14 ion and
therefore it was applied to all isonuclear sequence @xcite .
further level - to - level studies @xcite included w atom and w@xmath15 ion in the binary - encounter - dipole model as well as w@xmath16 and @xmath17 ions in the dw approach .
for the tungsten isonuclear sequence , cross sections were calculated using semirelativistic and relativistic dw methods for configurations and subconfigurations , respectively @xcite .
influence of the excitation - autoionization ( ea ) process , radiative damping , and relativistic effects were analyzed .
the work of loch et al .
@xcite considered the electron - impact excitations to all shells with @xmath18 and @xmath19 .
the investigation has determined that the ea contribution is relatively small compared to direct ionization ( di ) for the w@xmath20 to w@xmath1 ions .
however , later experimental measurements for the w@xmath21 ion have demonstrated a significant role of indirect process @xcite
. theoretical study of cross sections using the dirac - fock - slater approach confirmed these findings @xcite .
furthermore , the authors found that maxwellian rate coeffcients are larger than the cadw rate coefficients by about 16% .
the discrepancy was attributed to the ea channels originating from the high-@xmath22 shells up to @xmath23 .
therefore , the main aim of our work is to determine the influence of excitations to the high-@xmath22 shells which lead to the ea process . for the study
, we have chosen the w@xmath1 ion which has only one @xmath24 electron in the valence shell . at first
, we will present calculated results of the ea contribution for subconfigurations . later
, level - to - level distorted - wave ( lldw ) data in a single configuration approach are analyzed .
in addition , the influence of the radiative damping is studied in both cases .
the contribution of the direct and indirect processes is compared for the w@xmath1 ion .
direct and indirect processes contribute to the total electron - impact single - ionization cross sections from the level @xmath25 of @xmath26 ion to the level @xmath27 of @xmath28 ion : @xmath29 where @xmath30 is the direct ionization cross section at the electron energy @xmath31 , @xmath32 is the electron - impact excitation cross section to the level @xmath33 of the @xmath26 ion .
autoionization branching ratio @xmath34 is determined by the expression : @xmath35 where @xmath36 and @xmath37 are the auger and radiative transition probabilities , respectively .
therefore , inclusion of the branching ratios in the cross section calculations leads to the radiative damping of the indirect process .
the second term in the numerator presents the transition from the level @xmath33 of the initial ion to the level @xmath27 of the final ion through the intermediate levels @xmath22 of the initial ion reached by the radiative transitions .
this term is not studied in the current work because the amount of calculations drastically increases .
thus , the indirect ionization consists of the two - step process : the excitation with subsequent autoionization .
it is evident that the branching ratios for the levels below the ionization threshold are equal to zero .
furthermore , the branching ratios are equal to zero for the levels above the ionization threshold if they can not decay through the auger transitions directly or through the intermediate states .
the total ionization cross section for the initial level @xmath25 is obtained by performing summation over the all final levels @xmath27 in eq .
( [ sigma ] ) .
we do not consider higher order indirect ionization process such as a resonant - excitation double - autoionization .
previous analysis of this process for the w@xmath21 ion determined a negligible contribution for the ionization cross sections @xcite .
the electron - impact excitation and ionization cross sections are determined in the dw approximation using the flexible atomic code ( fac ) @xcite which implements the dirac - fock - slater method . for moderate and highly charged ions ,
the direct ionization is accurately described by the dw approximation .
the ionization cross sections are calculated in the potential of the ionizing ion .
calculations in the potential of the ionized ion provide approximately 10% smaller cross sections for the direct ionization .
the ground configuration of the w@xmath1 ion is [ kr]@xmath38 which consists of two levels .
our analysis of the ionization cross sections is based on the study of the ground level ( @xmath39 ) .
cross sections from the first excited level ( @xmath40 ) are similar to the ones from the ground level and are not presented here .
the lowest two configurations of the w@xmath1 ion together with the configurations which energy levels straddle the ionization threshold are shown in fig .
[ energies ] .
the presented configurations with the energy levels near the ionization threshold are produced by the one - electron promotions from the ground configuration of the w@xmath1 ion .
they mainly correspond to the excitations from the @xmath41 shell up to the shells with @xmath42 .
all of these configurations , except for the @xmath43 one , have average energies below the ionization threshold .
thus , they do not provide contribution to the ionization cross sections in the cadw approach .
a few configurations formed by promotion from the @xmath44 shell also straddle the ionization threshold .
however , the one - electron excitations reach only @xmath45 shell in this case . the ionization process affecting the @xmath46 , @xmath44 , @xmath41 , and @xmath24 shells contribute to the di process .
on the other hand , study of the w@xmath21 ion has demonstrated that ionization from the @xmath46 shell lies above the double - ionization threshold and contributes to the indirect double - ionization process @xcite .
table i shows calculated energy levels for the lowest configurations of the w@xmath1 , w@xmath47 , and w@xmath48 ions .
in addition , the lowest levels of the configurations for the w@xmath47 ion produced from the ground configuration of the w@xmath1 ion after an electron is ionized from the @xmath49 , @xmath46 , @xmath44 , or @xmath41 shells are shown .
one can see that the calculated ionization threshold energy equals to 878.89 ev .
the ionization energy obtained from the scaled electron binding energies is given as @xmath50 ev @xcite .
that is in quite close agreement with our determined value .
the ionization of the long - living level of the [ kr]@xmath51 configuration corresponds to threshold energy 739.56 ev .
this configuration can decay to the ground configuration through the weak electric octupole transitions ( lifetime @xmath52 s ) .
figure [ ea ] shows contribution of the ea channels originating from the excitations to the high-@xmath22 shells up to @xmath53 .
these data have been produced using the subconfiguration - average distorted - wave ( scadw ) approach .
as it was mentioned eralier , the previous study using the semirelativistic cadw calculations included only the excitations to the outer shells with @xmath4 and @xmath19 @xcite .
it can be seen from fig .
[ ea ] that the additional ea channels originating from the excitations to the higher shells increase cross sections for the indirect part approximately by factor of two .
the convergence for the cross sections is reached when the excitations to the high-@xmath22 shells are included . furthermore , our calculations take into account the excitations to the shells with @xmath54 . fig .
[ ea2nl ] shows that the largest contribution to the ea process comes from the inner shell excitations to the @xmath5 orbital .
the same trend is observed for the @xmath4 and @xmath55 shells .
however , the contribution to the ea process of the excitations to the @xmath5 orbital decreases if compared with other orbitals in the latter case .
it is interesting that none of the excitations from the @xmath46 , @xmath44 , or @xmath41 shells to the @xmath5 orbital give the largest cross sections .
however , the total contribution of the excitations from the @xmath46 , @xmath44 , and @xmath41 shells to the @xmath5 orbital is the largest compared to other orbitals .
it is worth to note that excitations to @xmath56 orbital lead to the strong ea channel for which the relative contribution increases for @xmath55 shells ( fig .
[ ea2nl]b ) . for these shells ,
the largest increase of the relative contribution occurs for the excitations to a @xmath57 orbital compared to the @xmath4 case .
this can be explained by the fact that the strongest excitations to the @xmath57 orbital take place from the @xmath41 shell .
however , the configurations @xmath58 with @xmath59 are below the ionization threshold , and they do not contribute to the ea process .
only some levels of the @xmath60 configuration are above the ionization threshold ( fig .
[ energies ] ) . on the other hand , all levels of @xmath61 configuration are above the ionization threshold .
furthermore , the energy levels of the @xmath62 configuration straddle the ionization threshold .
it explains much larger relative contribution to the ea process of the excitations to the @xmath5 orbital for @xmath4 shells .
other interesting result occurs when the ea channel originating from excitations to the @xmath63 orbital starts to dominate at high electron energies for the @xmath55 shells .
contribution of the ea channel to the @xmath5 orbital is slightly smaller at the higher energies when the excitations to the large principal quantum numbers @xmath22 are considered ( fig .
[ ea2nl]b ) .
the ea channel for the excitations from the @xmath41 shell is predicted to produce the largest contribution compared to the ea channels from the @xmath46 and @xmath44 shells because the excitation cross sections are larger for the @xmath41 shell .
however , as it is mentioned above , many configurations produced by the promotion from the @xmath41 shell to the @xmath0 shell for @xmath4 are below the ionization threshold .
therefore , these configurations do not contribute to the ea process .
figure [ ea4l]a demonstrates that the contribution of the excitations from the @xmath41 shell to the ea process for @xmath4 is more than two times smaller compared to the ea channel for the excitations from the @xmath44 shell .
the similar result was obtained for the w@xmath21 ion when the excitations only up to @xmath4 were considered @xcite .
however , the situation drastically changes for the excitations from the @xmath41 to the higher shells because all arising configurations are above the ionization threshold ( fig .
[ ea4l]b ) .
contribution of this ea channel is much larger compared to the ea channel from the @xmath44 shell .
furthermore , the contribution of the ea channel from the @xmath41 shell for the excitations to @xmath55 shells is comparable to the total cross sections of the ea process for the excitations to @xmath4 shells ( fig .
[ ea4l]a ) .
calculations of the ea process for subconfigurations can differ from those for the detailed level - to - level studies . in our scadw calculations ,
the energy levels are grouped into the subconfigurations . for the energy levels that straddle the ionization threshold ,
the corresponding energy of the subconfiguration can be above or below the ionization threshold .
the subconfiguration with the energy above the ionization threshold contributes to the ea process .
therefore , all levels of the subconfiguration , even those located below the ionization threshold , are included when the cross sections are calculated . on the other hand , for the subconfiguration below the ionization threshold ,
the contributions from the corresponding energy levels above the ionization threshold are neglected .
it is interesting to note that the ea cross sections obtained using scadw and lldw approaches are in good agreement when the autoionization branching ratios are not included in calculations ( fig .
[ damping ] ) .
figure [ damping ] shows that the radiative damping has the large impact on the ionization cross sections .
however , it was found in the previous study that the autoionization branching ratios are close to 1 for this ionization stage @xcite .
the difference can be explained by the fact that the previous study used the configuration - average quantities while the current scadw calculations employ the subconfigurations .
the higher values of the cross sections for the scadw calculations compared to the lldw results suggest that the autoionization branching ratios for the cadw data are closer to 1 .
therefore , the radiative decay paths are restricted for some excited configurations in these two averaged approaches . figure [ damping ] demonstrates that the level - to - level study is crucial for the ea process in the w@xmath1 ion .
figure [ total ] presents the total ionization cross sections for the direct and indirect ionization parts where the indirect part accounts for the radiative damping . in this case , the level - to - level calculations are presented .
the contributions of di from the @xmath46 , @xmath44 , @xmath41 , and @xmath24 shells are highlighted .
the largest contribution comes from the ionization of the @xmath41 electron .
as it is shown above , the ea channel for the excitations from the @xmath41 shell has also the largest impact compared to the @xmath46 and @xmath44 contributions ( fig .
[ ea4l ] ) . for the di process
, the influence of the @xmath46 shell is very small . to ensure
our data can be easily utilized in modeling studies , we present the total ionization cross sections table [ t2 ] . it can be seen that the ea process increases the total cross section at the peak nearly by 40 % ( fig .
[ total ] ) .
this contribution would be about two times smaller if the excitations to the higher shells ( @xmath2 ) were not considered .
the ea channels which include only the excitations with @xmath4 and @xmath19 provide for the ea process approximately 10% of the total cross section at the peak .
the ionization cross sections determined for the metastable [ kr]@xmath51 configuration are approximately by 60 % higher for the ea process compared to the ground configuration when the radiative damping is included for these configurations .
the contribution of the indirect process from the high-@xmath0 shells to the total cross sections increases by a factor of two .
the di from the @xmath64 shell is about two times higher for the metastable configuration compared to the ionization from the @xmath24 shell of the ground configuration .
the contribution of the excitations to high-@xmath0 shells has been studied for the electron - impact excitation - autoionization process .
calculations for the w@xmath1 ion illustrate that the excitations to the high-@xmath0 shells ( @xmath65 ) increase the cross section values by a factor of two for the indirect part of the ionization process . the largest contribution to
the indirect part comes from the excitations to @xmath5 orbital .
surprisingly , large contribution also arises from the excitations to @xmath56 orbital .
the ea cross sections for excitations from the @xmath41 shell to the shells with @xmath4 is approximately two times smaller compared to the excitations from the @xmath44 shell . on the other hand , situation drastically changes for the excitations to the higher shells with @xmath65 where the excitations from the @xmath41 shell dominate . in this case , the ea cross sections for the excitations from the @xmath41 shell are nearly three times higher .
the current calculations for the w@xmath1 ion show that the radiative damping has crucial effect on the cross sections of the indirect process .
the cross section values decrease by two times for the lldw results .
however , the influence of the radiative damping to the subconfiguration - average cross sections is approximately 30 % smaller compared to the influence in the level - to - level calculations .
current results demonstrate that the contribution of the ea channels originating from the excitations to the high-@xmath0 shells has to be estimated in the analysis of the electron - impact ionization process for the highly charged ions .
good agreement with experiment for the low - charge states does not ensure that the same list of the investigated shells for the ea process is enough when the higher - charge states are considered . in order to estimate the contribution from the high-@xmath0 shells ,
the convergence of the cross sections of the ea process must be checked .
this research was funded by european social fund under the global grant measure ( no .
: vp1 - 3.1-mm-07-k-02 - 015 ) .
part of computations was performed on resources at the high performance computing center
hpc saultekis in vilnius university faculty of physics .
14ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) link:\doibase 10.1088/0953 - 4075/40/7/r01 [ * * , ( ) ] link:\doibase 10.1140/epjd / e2012 - 30517 - 2 [ * * , ( ) ] link:\doibase 10.1103/physreva.89.052714 [ * * , ( ) ] http://stacks.iop.org/0022-3700/17/i=13/a=016 [ * * , ( ) ] http://stacks.iop.org/0953-4075/28/i=13/a=021 [ * * , ( ) ] link:\doibase { 10.1088/0953 - 4075/28/22/015 } [ * * , ( ) ] link:\doibase 10.1103/physreva.56.1654 [ * * , ( ) ] link:\doibase { 10.1103/physreva.72.052716 } [ * * , ( ) ] @noop * * , ( ) http://stacks.iop.org/0953-4075/44/i=16/a=165202 [ * * , ( ) ] http://stacks.iop.org/0953-4075/47/i=7/a=075202 [ * * , ( ) ] link:\doibase 10.1139/p07 - 197 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.adt.2006.03.002 [ * * , ( ) ] ( color online ) energy levels of two lowest configurations and configurations which straddle the ionization threshold for the w@xmath1 ion .
red color even configurations , blue color odd configurations . ]
( color online ) total lldw ionization cross section for the w@xmath1 ion .
the contributions of di from different shells are highlighted by different colors .
ea contributions from the excitations to various shells are presented . ]
l c d & & + @xmath67 & @xmath68 & 0 .
+ & @xmath69 & 3.70 + & @xmath70 & 139.33 + @xmath71 & @xmath72 & 878.89 + & @xmath73 & 1068.57 + & @xmath74 & 1266.25 + & @xmath75 & 1458.42 + & @xmath76 & 2743.71 + @xmath77 & @xmath78 & 2006.64 + & @xmath79 & 2023.13 + 880 & & & & 0.010 & & 0.018 & 0.018 & 0.018 & 0.028 + 890 & & & & 0.096 & & 0.701 & 0.701 & 0.701 & 0.797 + 900 & & & & 0.179 & & 1.313 & 1.316 & 1.316 & 1.495 + 930 & & & & 0.409 & & 3.256 & 3.920 & 3.920 & 4.329 + 940 & & & & 0.481 & & 3.372 & 4.478 & 4.478 & 4.959 + 950 & & & & 0.549 & & 3.382 & 4.817 & 4.817 & 5.366 + 960 & & & & 0.615 & & 3.515 & 5.382 & 5.382 & 5.997 + 990 & & & & 0.800 & & 4.529 & 7.392 & 7.392 & 8.192 + 1000 & & & & 0.857 & & 4.558 & 7.798 & 7.798 & 8.655 + 1010 & & & & 0.913 & & 5.140 & 8.667 & 8.667 & 9.580 + 1020 & & & & 0.966 & & 5.240 & 9.024 & 9.024 & 9.990 + 1050 & & & & 1.116 & & 5.591 & 10.146 & 10.146 & 11.262 + 1060 & & & & 1.163 & & 6.059 & 10.821 & 10.853 & 12.016 + 1070 & & & 0.001 & 1.208 & & 6.040 & 10.862 & 11.006 & 12.215 + 1080 & & & 0.063 & 1.252 & & 6.143 & 10.988 & 11.181 & 12.496 + 1090 & & & 0.251 & 1.294 & & 6.410 & 11.224 & 11.463 & 13.009 + 1120 & & & 1.074 & 1.413 & & 7.063 & 12.143 & 12.378 & 14.865 + 1130 & & & 1.346 & 1.450 & & 7.236 & 12.439 & 12.677 & 15.474 + 1140 & & & 1.614 & 1.486 & & 7.490 & 12.964 & 13.199 & 16.300 + 1160 & & & 2.136 & 1.555 & & 8.036 & 13.756 & 13.986 & 17.677 + 1180 & & & 2.630 & 1.619 & & 8.207 & 14.256 & 14.482 & 18.732 + 1200 & & & 3.096 & 1.680 & & 8.184 & 14.668 & 14.890 & 19.666 + 1240 & & & 3.956 & 1.791 & & 8.230 & 15.400 & 15.615 & 21.361 + 1280 & & 0.051 & 4.727 & 1.888 & & 8.186 & 15.619 & 15.911 & 22.577 + 1340 & & 0.374 & 5.742 & 2.014 & & 7.954 & 15.677 & 15.964 & 24.094 + 1410 & & 0.817 & 6.744 & 2.134 & & 7.536 & 15.262 & 15.543 & 25.239 + 1460 & 0.002 & 1.107 & 7.358 & 2.206 & & 7.307 & 14.874 & 15.183 & 25.855 + 1500 & 0.064 & 1.311 & 7.797 & 2.256 & & 7.182 & 14.558 & 14.858 & 26.286 + 1650 & 0.252 & 1.918 & 9.107 & 2.394 & & 6.589 & 13.305 & 13.578 & 27.249 + 1880 & 0.441 & 2.518 & 10.374 & 2.503 & & 5.874 & 11.820 & 12.059 & 27.895 + 2000 & 0.510 & 2.731 & 10.794 & 2.527 & & 5.576 & 11.185 & 11.409 & 27.970 + 2080 & 0.548 & 2.845 & 11.007 & 2.535 & & 5.399 & 10.810 & 11.026 & 27.961 + 2300 & 0.627 & 3.077 & 11.389 & 2.529 & & 4.971 & 9.927 & 10.124 & 27.746 + 2400 & 0.654 & 3.151 & 11.487 & 2.518 & & 4.805 & 9.585 & 9.775 & 27.585 + 2700 & 0.710 & 3.294 & 11.579 & 2.463 & & 4.391 & 8.742 & 8.914 & 26.959 + 3000 & 0.742 & 3.355 & 11.505 & 2.389 & & 4.068 & 8.094 & 8.253 & 26.244 + 3500 & 0.762 & 3.355 & 11.158 & 2.253 & & 3.655 & 7.284 & 7.426 & 24.955 + 3700 & 0.763 & 3.334 & 10.982 & 2.198 & & 3.520 & 7.020 & 7.157 & 24.435 + 4100 & 0.758 & 3.271 & 10.605 & 2.091 & & 3.280 & 6.556 & 6.684 & 23.410 + 4600 & 0.742 & 3.171 & 10.122 & 1.966 & & 3.023 & 6.052 & 6.171 & 22.173 + 6000 & 0.681 & 2.861 & 8.870 & 1.672 & & 2.479 & 4.981 & 5.080 & 19.163 + 7000 & 0.637 & 2.656 & 8.133 & 1.511 & & 2.199 & 4.426 & 4.513 & 17.449 + 8000 & 0.595 & 2.474 & 7.506 & 1.379 & & 1.981 & 3.993 & 4.072 & 16.026 + 9000 & 0.558 & 2.311 & 6.959 & 1.266 & & 1.808 & 3.653 & 3.725 & 14.819 + 10000 & 0.526 & 2.169 & 6.481 & 1.171 & & 1.660 & 3.401 & 3.470 & 13.381 + | contribution to electron - impact ionization cross sections from excitations to high-@xmath0 shells and a consequent autoionization is investigated .
we perform relativistic subconfiguration - average and detailed level - to - level calculations for this process .
ionization cross sections for the w@xmath1 ion are presented to illustrate large influence of the high shells ( @xmath2 ) and orbitals ( @xmath3 ) in the excitation - autoionization process .
the obtained results show that the excitations to the high shells ( @xmath2 ) increase cross sections of indirect ionization process by a factor of two compared to the excitations to the lower shells ( @xmath4 ) .
the excitations to the shells with the orbital quantum number @xmath5 give the largest contribution compared with the other orbital quantum numbers @xmath6 .
radiative damping reduces the cross sections of the indirect process approximately two times in the case of the level - to - level calculations .
determined data show that the excitation - autoionization process contributes approximately @xmath7 to the total ionization cross sections . |
moderate - length high - temperature ( ht ) expansions ( through order 12 ) and low - temperature ( lt ) expansions for the spin - spin correlation function ( sscf ) @xmath1 of the nearest - neighbor ising models with general spin s were first computed@xcite three decades ago on various lattices in 2d and in 3d .
motivations for the study of these models came not only from their direct phenomenological interest , but mainly from the conjecture@xcite that , in a given space dimension , the exponents characterizing the critical behavior are independent both of the lattice structure and of the spin magnitude @xmath0 .
this conjecture was the first step towards the modern notion of universality class .
in the same years also the hypothesis of critical scaling@xcite was put forward .
many studies@xcite of the mentioned ht and lt series were devoted to test the validity and the main consequences of these basic hypotheses@xcite .
although the results sometimes were not as precise as was hoped , or covered only the @xmath2 case , the scaling tests suggested that the critical sscf is a homogeneous function of appropriate variables , while the universality tests indicated that the critical indices and suitable combinations@xcite of critical amplitudes are independent of the spin @xmath0 and lattice structure .
a few years later , the first substantial extension@xcite of ht ising series in 3d ( through order 21 on the body - centered - cubic(bcc ) lattice only ) did not make higher expansion coefficients available for the sscf , but only for its two lowest even moments and therefore various tests could not be repeated and updated . we are now resuming the ht part of those pioneering analyses in order to improve their extent and accuracy by taking advantage of our recent extension@xcite from order @xmath3 through order @xmath4 of the ht expansions for the sscf of the ising model with general spin @xmath0 , in 2d on the square ( sq ) lattice and in 3d on the simple - cubic ( sc ) and the bcc lattices . from these data we have also derived series for related quantities , in particular for a variety of moments of the sscf , which are computed through order 25 , and for the exponential ( or ` true ' ) correlation length defined via the exponential decay of the sscf , which , however , can be extended only through order 19 . for reasons of space we have not tabulated in this paper the series analysed , but have included them into our on - line library@xcite of ht data for the spin-@xmath0 ising model in order to make them more widely available for further study . since this is the largest body of series data so far computed for these systems , we have already been studying other aspects of them in previous papers . in particular , in ref.@xcite we have accurately confirmed that the residual weak spin - dependence observed@xcite in lower order studies of the susceptibility exponent @xmath5 and of the correlation - length exponent @xmath6 in 3d on the bcc lattice , should not be ascribed to small violations of universality , but can be simply explained away as numerical inaccuracies due to expected non - negligible spin - dependent corrections to the leading scale behavior .
moreover we have tested the universality of several amplitude combinations obtaining similar results . in ref.@xcite an analogous survey of universal quantities
was performed in 2d for the sq lattice case .
shorter series ( but only for the @xmath2 case ) had been analysed in refs.@xcite . from the evidence presented here
we can conclude that our ht data for the sscf have by now reached an extension sufficient to make the use of modern series - extrapolation techniques possible and generally reliable .
therefore we are able to exhibit more convincingly both in 2d and in 3d many expected properties related to scaling and universality also in some cases in which the old analyses led to inconclusive or not very precise results .
the rest of the paper is organized as follows . in section ii we shall outline the main features of the model , introduce our notations and conventions and very briefly recall the scaling and universality properties expected for the sscf along with the corresponding tests discussed in full detail by the above cited papers@xcite .
therefore , in section iii we can restrict ourselves to only a few comments on the numerical results .
the spin-@xmath0 ising models with nearest - neighbor interaction are defined by the hamiltonian : @xmath7 where @xmath8 is the exchange coupling , and @xmath9 with @xmath10 a classical spin variable at the lattice site @xmath11 , taking the @xmath12 values @xmath13 .
the sum runs over all nearest - neighbor pairs of sites . for simplicity
, the nearest - neighbor lattice spacing will be set equal to @xmath14 everywhere .
we shall consider expansions in the usual ht variable @xmath15 where @xmath16 is the temperature , @xmath17 the boltzmann constant , and @xmath18 will be called `` inverse temperature '' for brevity . in the critical region
we shall also refer to the standard ` reduced - temperature ' variable @xmath19 .
we shall study the ht expansion of the ( connected ) sscf defined as @xmath20 in order to estimate numerically @xmath1 as @xmath21 , we have allowed for its expected@xcite behavior : in the 2d case @xmath22 and in the 3d case @xmath23 here @xmath24 is the critical amplitude of the leading singular correction , @xmath25@xcite denotes the critical exponent of the specific heat in 3d and the dots indicate higher order corrections .
the correlation - function moment @xmath26 of order @xmath27 is defined as @xmath28 ( for @xmath29 the sum extends to @xmath30 ) .
the expected asymptotic behavior of @xmath31 as @xmath21 is @xmath32 .
\label{asbsm}\ ] ] in 2d the exponent @xmath33 of the leading singular correction is larger than unity , while in 3d a recent simultaneous study@xcite of a set of models in the ising universality class has suggested the very precise estimate @xmath34 .
the scattering function , namely the fourier transform of @xmath35 @xmath36 g(\vec r , t;s)\ ] ] for @xmath37 yields the zero - field reduced susceptibility @xmath38 the second - moment correlation - length is defined in @xmath39 spatial dimensions by @xmath40 for @xmath41 the sscf is exponentially decreasing for large @xmath42 and therefore following ref.@xcite , beside the ` second - moment ' correlation - length we can also define the inverse ` exponential ' ( or ` true ' ) correlation - length in the direction @xmath43 as @xmath44 since the singularity of @xmath45 closest to the real axis in the complex @xmath46 plane is located at @xmath47 , the exponential correlation - length can be obtained by solving recursively@xcite the eq . @xmath48 rather than working directly with @xmath49 which is not an ordinary power series in @xmath18 , it is expedient@xcite to form the quantity @xmath50 } \label{mtcl}\ ] ] which is an ordinary power series in @xmath18 . in eq.([mtcl ] ) @xmath51 is a geometrical factor depending on the unit vector @xmath52 and on the lattice considered .
in particular , if @xmath43 is directed along a lattice axis , we have @xmath53 for the sq and the sc lattices , while @xmath54 for the bcc lattice .
so far , 3d data for this quantity were published exclusively for @xmath2 , and did not extend beyond order 15 in the sc lattice case@xcite or beyond order 10 in the bcc lattice case@xcite . in 2d
the ht expansion can be computed exactly @xcite for @xmath2 , but no data have been published for @xmath55 . in ref.@xcite , we have tabulated the expansion of @xmath56 through order 19 for @xmath52 directed along a lattice axis , in the case of the sq , sc and bcc lattices and with @xmath57 . in order to avoid possible confusion , it should be pointed out that in ref.@xcite our @xmath58 was denoted by @xmath59 , while the symbol @xmath60 was used to denote @xmath61 .
our notation might be more suggestive since our @xmath58 compares very closely with @xmath62 .
indeed , the true and second - moment correlation - lengths are almost identical in magnitude above the critical temperature .
in particular on the sq lattice , when @xmath43 is directed along a lattice axis , the ht expansion coefficients of @xmath63 and @xmath62 coincide through sixth order for @xmath2 , through fourth order for @xmath64 and through second order for higher values of the spin . in 3d , in the sc lattice case
, the expansion coefficients of @xmath63 and @xmath62 coincide through seventh order for @xmath2 , through fifth order for @xmath64 and through third order for higher values of @xmath0 . in the case of the bcc lattice ,
the expansion coefficients coincide through third order for all values of the spin .
moreover , up to the maximum order of our computation , the noncoinciding coefficients differ by less than @xmath65 . the two correlation lengths @xmath66 and @xmath67 are expected to share the same critical exponent @xmath6 so that their asymptotic behavior when @xmath21 can be written as @xmath68 \label{asbxsm}\ ] ] and @xmath69 \label{asbxtrue}\ ] ] here the critical amplitude @xmath70 is independent of @xmath43 , since the sscf becomes spherically symmetric near the critical point .
the ratio @xmath71 is a universal combination of critical amplitudes@xcite i.e. it is expected to depend only on the lattice dimensionality @xmath39 , but not on the spin @xmath0 or the lattice structure . for @xmath72 and in zero magnetic field
, the sscf is expected to exhibit the asymptotic structure @xmath73 when both @xmath42 and @xmath67 are much larger than the lattice spacing ( with arbitrary @xmath74 ) .
eq.([scalfun ] ) together with these assumptions on @xmath42 and @xmath67 , is usually referred to as the `` strong - scaling hypothesis '' ( while it is called the `` weak - scaling hypothesis '' , if its validity is restricted to the @xmath75 limit with fixed @xmath74 ) . in eq.([scalfun ] ) , @xmath76 is the critical exponent describing the decay of the sscf at the critical point , @xmath77 is called the critical scaling function , @xmath78 and @xmath79 are scale factors .
the dots indicate subcritical corrections proportional to a positive power of some irrelevant field .
the scaling function @xmath77 is expected to be universal : its structure does not depend on the particular model under study provided that it belongs to a given universality class . on the contrary the scale factors @xmath80 depend on the spin and the lattice @xmath81 .
the validity of the asymptotic structure eq.([scalfun ] ) was verified analytically@xcite for the spin @xmath2 ising model in 2d . for the scattering function
@xmath45 the analogous scaling form as @xmath21 can be written as @xmath82 if the scale factors @xmath83 are specified adopting the normalization conditions @xmath84 one can write@xcite as @xmath85 @xmath86 where the function @xmath87 is universal and thus the quantities @xmath88 @xmath89 with @xmath90 and @xmath91 for @xmath92 , while @xmath93 and @xmath94 for @xmath95 , have finite universal values as @xmath21 .
all @xmath96 , as well as the difference @xmath97 , vanish@xcite in the mean - field related approximations
. therefore the magnitudes of these quantities at the critical point can be taken as a measure of the deviation of a given system from gaussian behavior , which turns out to be very small on the ht side of the critical point . more generally , it was observed@xcite that the scaling hypothesis eq.([scalfun ] ) implies that , at the critical point , the ratios @xmath98 with @xmath99 , are universal .
these ratios are dominated by the critical singularity also for negative values of the indices @xmath100 provided that each index exceeds @xmath101 , as follows from eq.([asbsm ] ) . finally , the determination of the amplitude @xmath24 of the leading singularity of the sscf ( see eqs.([singpart2d ] ) and ( [ singpart3d ] ) ) gives another opportunity to perform universality and scaling tests . in order that the structure of eqs.([singpart2d ] ) and ( [ singpart3d ] ) be compatible@xcite with the strong - scaling hypothesis eq.([scalfun ] ) , the amplitudes @xmath102 must scale as @xmath103 with @xmath104 , namely @xmath105 for large enough @xmath42 , independently of the spin and the lattice structure . in 2d
the value @xmath106 is expected , while in 3d , adopting our recent estimate@xcite of the values of the correlation length exponent @xmath107 and of the exponent @xmath108 , we should have @xmath109 .
let us first observe that , due to the leading singular corrections in eqs.([singpart2d ] ) and ( [ singpart3d ] ) , whose amplitudes @xmath24 grow with @xmath42 as indicated by eq.([e(r ) ] ) , determining accurately @xmath110 ( as well as @xmath24 itself ) is a rather delicate matter for which it is crucial to rely on sufficiently many expansion coefficients .
we should also consider that the number of non - trivial coefficients in our series decreases with increasing @xmath42 , and correspondingly the precision of our estimates of @xmath111 ( and @xmath112 ) deteriorates .
.our estimates of the critical - point values @xmath110 of the spin - spin correlation function for the nearest - neighbor ising models with spin @xmath113 on the sq , sc and bcc lattices . for comparison with our results ,
the first column labelled @xmath114 $ ] shows the only available previous estimates . in the case of the sq lattice
, we have cited the exact@xcite values . in the case of the nearest - neighbor correlation ( @xmath115 ) on the sc lattice ,
we have reported in the first column the estimate obtained using our numerical procedure with the series @xmath116 of ref.@xcite and have also cited a value@xcite obtained in a recent high - precision montecarlo simulation . in the remaining cases
we have reported the estimates of ref.@xcite obtained from series @xmath117 .
we are not aware of previous calculations for @xmath118 . [ cols="<,^,^,^,^,^,^,^ " , ]
the results of our analysis of the universal ratios @xmath119 are reported in table [ tab4 ] .
they also show independence of the spin and of the lattice structure within a good precision .
our series - extrapolation procedure , based on first- and second - order da s uses only our estimates of @xmath120 and does not need to be biased also with @xmath5 and @xmath6 as it was necessary in the generalized neville procedure@xcite employed with the short series of ref.@xcite .
considering that the values @xmath121 and @xmath122 ( or @xmath123 ) of the exponents accepted at the time of that study are somewhat different from the currently preferred ones and that the extrapolations are very sensitive to those values , a comparison with the numerical results of ref.@xcite has little meaning . in 2d , on the sq lattice , the non - trivial expansion coefficients of the ratio @xmath125 are not sufficiently many and their behavior is not smooth enough to yield very accurate results
. therefore our best estimate of @xmath124 ( by first - order da s biased with @xmath120 ) can not be more precise than @xmath126 , independently of @xmath0 .
our rough estimate is , however , consistent with the more accurate determination @xmath127 obtained in the @xmath2 case in which , as already indicated above , very long series are available@xcite for @xmath128 , while @xmath129 is exactly known@xcite . in 3d we can use both first- and second - order da s biased with @xmath120 .
the very smooth bcc lattice series yield the most accurate results .
our final estimate is @xmath130 , independently of @xmath0 and of the lattice structure .
so far , this ratio could be computed@xcite only for @xmath2 from a 15 term series on the sc lattice , with the result @xmath131 .
a more precise estimate@xcite @xmath132 was obtained indirectly ( and assuming universality ) , from optimized ht series for a continuous - spin model on the sc lattice . within the renormalization group approach@xcite ,
the estimate @xmath133 was obtained in the @xmath134-expansion to third order , while the coupling - constant expansion technique to fourth order gave @xmath135 .
d. m. saul , m. wortis and d. jasnow , phys .
b * 11 * , 2571 ( 1975 ) ; j. p. van dyke and w. j. camp , _ ibid . _
b * 9 * , 3121 ( 1974 ) .
w. j. camp and j. p. van dyke , phys .
b * 11 * , 2579 ( 1975 ) ; w. j. camp , d. m. saul , j. p. van dyke and m. wortis , _ ibid .
_ * 14 * , 3990 ( 1976 ) .
m. e. fisher , phys .
* 16 * , 11 ( 1966 ) ; l. p. kadanoff , physics * 2 * , 263 ( 1966 ) ; d. jasnow and m. wortis , phys . rev . *
176 * , 739 ( 1968 ) ; p. g. watson , j. phys .
c * 2 * , 1883 ( 1969 ) and _ ibid._. c * 2 * , 2158 ( 1969 ) ; r. b. griffiths , phys .
* 24 * , 1479 ( 1970 ) ; r. b. griffiths and j. c. wheeler , phys .
rev . a * 2 * , 1047 ( 1970 ) ; l. p. kadanoff , report at the newport beach conference 1970 ( unpublished ) and in _ proc .
e. fermi 1970 school on critical phenomena _ ,
edited by m. s. green ( academic , london 1971 ) ; c. domb , in _ statistical mechanics at the turn of the decade _ , edited by e. g. d. cohen ( dekker , new york , 1971 ) ; d. d. betts , a. j. guttmann and g. s. joyce , j. phys .
c * 4 * , 1994 ( 1971 ) .
b. g. nickel , in _ phase transitions : cargese 1980 _ , edited by m. levy , j. c. le guillou and j. zinn - justin , ( plenum , new york , 1982 ) ; s. gartenhaus and w. s. mccullough , phys .
b * 38 * , 11688 ( 1988 ) . fig.1 .
estimates of @xmath111 on the sq lattice .
the meaning of the symbols is as follows .
triangles : @xmath2 , squares : @xmath64 , rhombs : @xmath136 , circles : @xmath137 .
the spin @xmath0 points are shifted vertically by the quantity @xmath138 in order to make the figure more legible .
the continuous lines represent fits to the leading asymptotic behaviors @xmath139 expected for large @xmath42 .
we have taken @xmath140 as a fit parameter and fixed @xmath141.,width=384 ] fig.2 .
estimates of @xmath111 on the sc lattice .
the meaning of the symbols is the same as in fig.[gcsq ] the spin @xmath0 points are shifted vertically by the quantity @xmath138 in order to make the figure more legible .
the continuous lines represent fits to the leading asymptotic behaviors @xmath142 expected for large @xmath42 .
we have taken @xmath140 as a fit parameter and fixed @xmath143.,width=384 ] fig.3 .
estimates of @xmath111 on the bcc lattice .
the meaning of the symbols is the same as in fig.[gcsq ] .
the spin @xmath0 points are shifted vertically by the quantity @xmath138 in order to make the figure more legible .
the continuous lines represent fits to the leading asymptotic behaviors @xmath142 expected for large @xmath42 .
we have taken @xmath140 as a fit parameter and fixed @xmath143.,width=384 ] fig.4 .
estimates of @xmath24 on the sq lattice .
the meaning of the symbols is the same as in fig.[gcsq ] .
the spin @xmath0 points are shifted vertically by the quantity @xmath138 in order to make the figure more legible .
the continuous lines represent fits to the leading asymptotic behaviors @xmath144 expected for large @xmath42 .
we have taken @xmath145 as a fit parameter and fixed @xmath146.,width=384 ] fig.5 .
estimates of @xmath24 on the sc lattice .
the meaning of the symbols is the same as in fig.[gcsq ] .
the spin @xmath0 points are shifted vertically by the quantity @xmath138 in order to make the figure more legible .
the continuous lines represent fits to the leading asymptotic behaviors @xmath144 expected for large @xmath42 .
we have taken @xmath145 as a fit parameter and fixed @xmath147.,width=384 ] fig.6 .
estimates of @xmath24 on the bcc lattice .
the meaning of the symbols is the same as in fig.[gcsq ] .
the spin @xmath0 points are shifted vertically by the quantity @xmath138 in order to make the figure more legible .
the continuous lines represent fits to the leading asymptotic behaviors @xmath144 expected for large @xmath42 .
we have taken @xmath145 as a fit parameter and fixed @xmath147.,width=384 ] fig.7 .
the logarithm of the scaling function @xmath148 vs. @xmath149 in the case of the sq lattice .
the data represent the ssfc s with @xmath150 and @xmath151 in the range of temperatures for which @xmath152 .
the meaning of the symbols is the same as in fig.[gcsq].,width=384 ] fig.8 .
the logarithm of the scaling function @xmath148 vs. @xmath149 in the case of the sc and the bcc lattices . for both lattices the data represent the ssfc s with @xmath153 and @xmath151 in the range of temperatures for which @xmath154 .
the meaning of the symbols is the same as in fig.[gcsq ] for the sc lattice case . for the bcc lattice data we have used full symbols of the same shape.,width=384 ] | we have extended , from order 12 through order 25 , the high - temperature series expansions ( in zero magnetic field ) for the spin - spin correlations of the spin-@xmath0 ising models on the square , simple - cubic and body - centered - cubic lattices . on the basis of this large set of data ,
we confirm accurately the validity of the scaling and universality hypotheses by resuming several tests which involve the correlation function , its moments and the exponential or the second - moment correlation - lengths . |
ten years before the advent of qcd , skyrme proposed a model @xcite describing hadronic physics which involves only pion fields and where baryons emerge as topological solitons .
this model is recognized as the simplest chiral realization of qcd at low energies and large @xmath3 @xcite .
the corresponding lagrangian contains , in addition to the well known nonlinear @xmath4 model , an antisymmetric term of fourth order in powers of the derivatives of the pion field ( the so - called skyrme term )
. the latter has been added by skyrme in order to avoid soliton collapse . despite its relative success in describing some properties of baryons ( _ cf . _
@xcite for a review ) , the skyrme model presents several shortcomings such as its prediction of a very small value for the axial coupling constant @xcite and an almost zero nucleon mass when the casimir effects are taken into account @xcite .
but probably the most important crisis of the skyrme model is that it does not allow for the formation of nuclei .
indeed , it predicts a repulsive @xmath0 central potential @xcite . because of these drawbacks ,
several extensions of the standard skyrme model have been proposed in order to improve its prediction power .
the most famous one consists in adding to the skyrme lagrangian a term of order six in powers of the derivatives of the pion field , proportional to the square of the baryon current as first proposed in @xcite .
the extended skyrme model thus obtained contains a second , fourth and sixth - order term , but still is an effective theory of pions .
it has been shown recently @xcite that this extended model is much more realistic when describing low - energy hadronic physics than the standard skyrme model .
in contrast with the velocity - independent part of the @xmath0 potential , the kinetic part and especially the @xmath0 spin - orbit force has not been studied extensively within the framework of skyrme - type models . however , all the authors who worked on this subject , e.g. @xcite , arrived at the disappointing result that the standard skyrme model predicts a repulsive isospin - independent spin - orbit force while , according to the phenomenology @xcite , it should be attractive . a glimpse of hope has emerged from ref .
@xcite where it was claimed that when considering the sixth - order term , an attractive isoscalar spin - orbit interaction could be obtained .
however , we have shown recently @xcite that this is not the case : the sixth - order term contributes with a positive sign as is the case for the skyrme term .
indeed , the authors of ref .
@xcite considered in their calculations only one piece of the force due to the sixth - order term and omitted the piece that stems from the baryon exchange current , which turns out to be not only repulsive , but the dominant one too @xcite .
thus , despite its success in improving the predictions of 1-baryon data , even an extended skyrme model including higher - order terms in powers of the derivatives of the pion field fails to predict an attractive @xmath0 isoscalar spin - orbit interaction as it fails to predict the correct @xmath0 central potential @xcite .
therefore , one has to consider more realistic low - energy models to describe the @xmath0 interaction . the skyrme term as well as the sixth - order term can both be regarded as an _ infinite _ mass limit , the so - called local approximation , of a model with @xmath5-meson @xcite and @xmath6-meson @xcite exchanges respectively .
these findings have provided the ground for improving skyrme - type models by constructing effective lagrangians which include , in addition to the pions , the other low - lying mesons .
the authors of ref .
@xcite have investigated the @xmath0 force by considering such an effective model incorporating the mesons @xmath7 and the scalar meson @xmath1 . unlike previous works
, they have introduced the vector meson fields by gauging the linear @xmath4-model instead of the nonlinear one .
they have found that the different channels of the @xmath0 interaction arising from the potential piece of the lagrangian are very well described , in particular the central potential for which an attraction has been found in a quantitative agreement with the phenomenology @xcite .
these results lead us to think that , similarly to the central potential , it may be that the right way to circumvent the difficulty of finding an attractive isoscalar spin - orbit force with skyrme - type models is to consider effective theories which incorporate low - mass mesons with _
finite _ masses . in order to check these claims
, one has to investigate the kinetic energy of a two - nucleon system within an effective lagrangian such as the one in @xcite .
this is of course not an easy task and goes beyond the scope of this short paper . as a first step , we rather focus here only on the scalar meson contribution to the isoscalar spin - orbit force .
indeed , it has been pointed out in ref .
@xcite that the presence of scalar degrees of freedom plays the crucial role of recovering the missed central attraction , so one may think that this will also be the case for the isoscalar spin - orbit force .
this idea to account for a scalar field in the spin - orbit channel has been investigated in ref .
@xcite . there , a dilaton field
is coupled to skyrmions in order to mimic the scale breaking of qcd .
we will not discuss here if a dilaton field is suitable or not to provide a good description of low - energy hadron physics ; we refer the interested reader to ref .
@xcite for a detailed discussion on this subject .
the fact is that a combination of the sixth - order term and dilaton coupling yields an attractive isoscalar spin - orbit force , as it has been claimed in @xcite .
note though that this result remains questionable since these authors , like those of ref .
@xcite , ignored the baryon exchange current contribution in their calculations @xcite . in any case , here we will follow a more transparent way in order to incorporate scalar degrees of freedom @xcite . it consists in considering the linear @xmath4 model in which the presence of the scalar field is intrinsic . in sec.ii , the model with the scalar - isoscalar @xmath1-meson
is presented .
sec.iii contains the derivation of the isoscalar spin - orbit force .
our results and discussions are presented in sec.iv .
the infinite - mass limit of the effective lagrangian with @xmath8 and @xmath1 mesons @xcite reads : @xmath9 ^ 2 - e_6 ^ 2 b_{\mu } b^{\mu } + \frac { f_{\pi}^2 m_{\pi}^2}{2}\hbox { tr}(u -1 ) } \end{array}\ ] ] where @xmath10 is the baryon current @xcite , and @xmath11 an su(2 ) matrix which characterizes the pion field . the first term in eq .
( [ la ] ) corresponds to the nonlinear @xmath4 model , @xmath12 being the pion decay constant .
the second term , parameterized by the coupling constant @xmath13 , is the so - called skyrme term .
the third term is of order six in the derivatives of the pion field and corresponds to @xmath6-meson exchange in the case of an infinite @xmath6-meson mass @xcite .
the coupling constant @xmath14 is a parameter related to the @xmath15 width .
the last term in eq .
( [ la ] ) which is proportional to the square of the pion mass @xmath16 ( 139 mev ) implements a small explicit breaking of chiral symmetry . as
compared to the standard skyrme model , the extended model ( [ la ] ) , when used with realistic parameters , provides a more accurate description of the 1-baryon properties ( e.g. , the nucleon quantum mass , the @xmath17-@xmath18 mass splitting , the breathing mode energy the axial - vector coupling constant @xmath19 , ... ) @xcite . however , as mentioned in the introduction , it does not describe properly neither the @xmath0 isospin - independent spin - orbit force @xcite nor the central potential @xcite .
the scalar - isoscalar meson @xmath1 is viewed as the responsible for the @xmath20-wave attraction in the @xmath21 interaction and , therefore , should be considered as an essential ingredient in the low - energy hadronic phenomenology .
it is incorporated in the lagrangian ( [ la ] ) by replacing the nonlinear @xmath4 model @xmath22 with the linear one .
namely , @xmath23 where @xmath24 is the isoscalar content of the quaternion field @xmath25 , @xmath4 being the scalar chiral partner of the pion field . the scalar field @xmath1 is defined as @xmath26 .
the coupling constant @xmath27 is related to the @xmath1-meson mass through @xmath28 .
it is straightforward to check that by taking @xmath29 in eq .
( [ ls ] ) , one recovers the nonlinear @xmath4 model @xmath22 . indeed , in that limit , @xmath27 becomes infinite and so the @xmath24-field has to be frozen to its asymptotic value @xmath12 in order to keep the potential energy in eq .
( [ ls ] ) finite .
so far , all the calculations on the @xmath0 spin - orbit force within skyrme - type models have been carried out in the framework of the product ansatz as suggested by skyrme @xcite ( _ cf .
@xcite and references therein ) .
it is worthwhile noticing that the latter is only a simple approximation to the two - baryon configuration and not a self - consistent solution .
we will not discuss here the degree of validity of this approximation .
it is commonly chosen because of its relative simplicity as compared to other two - baryon field configurations which can be found in the literature @xcite .
furthermore , it becomes exact for large @xmath0 separation much larger than 1 fm . ] .
thus , following common practice , we use it here to describe a system of two interacting solitons . furthermore , in order to obtain the appropriate spin and isospin structure , we also introduce rotational dynamics @xcite .
hence , the field configuration of the two - nucleon system separated by a vector @xmath46 reads : @xmath47 where @xmath48 and @xmath49 are su(2 ) matrices , and @xmath50 the hedgehog single soliton ( [ uh ] ) . for the scalar field , we use the configuration suggested in ref .
@xcite @xmath51 which obviously is the form the most compatible with the product - ansatz approximation ( [ pro ] ) .
to carry out a simultaneous quantization of the relative motion of the two nucleons and the rotational motion , we need to treat @xmath52 and @xmath49 as collective coordinates .
therefore , we make all these parameters ( @xmath53 ) time - dependent
. the spin - orbit potential will emerge from a coupling between the relative motion and the spins of the two nucleons so that we have to calculate the kinetic energy corresponding to ( [ ls ] ) . as reported in ref .
@xcite , generally , one has to treat with care the conversion from velocities to canonical momenta before identifying and extracting the spin - orbit potential .
indeed , one has to start from a lagrangian formalism , take its `` classical '' kinetic energy and extract from it the mass matrix and then invert it properly in order to move to a hamiltonian formalism .
however , for a large relative distance @xmath54 , the region of validity of the product ansatz , this procedure @xcite is equivalent to that of refs . @xcite which we will use here . in
this latter one starts , in the case ( [ ls ] ) , directly from @xmath55 where the minus sign in front of the integral is put in explicitly so that account is taken of the change in sign of the off - diagonal terms of the @xmath56 mass matrix under inversion @xcite .
afterwards , one makes the usual identifications @xcite : @xmath57 where @xmath58 and @xmath59 are respectively the radial momentum and the spin of the @xmath60-th nucleon while @xmath61 and @xmath62 are respectively the mass and the moment of inertia of the single soliton .
inserting the product anstze ( [ pro],[xi12 ] ) in the expression ( [ ks ] ) gives @xmath63 where we have omitted to write the second term in eq .
( [ ks ] ) since it does not contain angular velocities and thus will not contribute to the spin - orbit force . in eq .
( [ ks2 ] ) , @xmath64 and @xmath65 are the time components of the right and left currents respectively : @xmath66 in the above equations , the sum from 1 to 3 on repeated indices is understood .
the tensors @xmath67 and @xmath68 depend only on the position @xmath69 of the @xmath60-th nucleon , and their explicit expressions can be found in eq . ( 11 ) of ref .
the @xmath70 s are the matrix elements of the @xmath71 rotation matrix in the adjoint representation : @xmath72 . to obtain the right - hand sides of expressions ( [ currents ] ) , we have used the quantization scheme ( [ subst ] ) .
the next step now is to expand the square in the expression of @xmath73 ( [ ks2 ] ) after inserting expressions ( [ currents ] ) and making use of the @xmath74-matrix properties and the definitions of the tensors @xmath67 and @xmath68 @xcite .
thus , by considering the isoscalar part contribute to the isospin dependent force . ] and keeping only terms proportional to @xmath75 , where @xmath76 is the total spin and @xmath77 the angular momentum ( @xmath78 being the relative momentum , i.e. , @xmath79 ) , we obtain : @xmath80 the total isoscalar spin - orbit potential of the model ( [ lanew ] ) is then : @xmath81 where the linear @xmath4-model contribution @xmath82 is given in eq .
( [ spor ] ) .
the rather lengthy expressions of the skyrme - term contribution @xmath83 and the sixth - order term contribution @xmath84 can be found explicitly in refs .
@xcite and @xcite , respectively , and do not need to be rewritten here .
note that the analytical expressions of @xmath83 and @xmath84 are not affected by the presence of the scalar field and still depend explicitly only on the chiral field @xmath35 .
this is because the @xmath24-field does not couple to @xmath85 nor @xmath86 , as it can be seen from eq .
( [ lanew ] ) .
but obviously their numerical values will change due to the dependence of @xmath35 on the scalar degrees of freedom .
before displaying our numerical results concerning @xmath82 , let us first discuss the analytical properties of the terms @xmath87 and @xmath88 at large distances [ _ cf_. eq . ( [ so ] ) ] . as already shown in ref .
@xcite , these potentials decrease as @xmath89 a large @xmath54 .
indeed , one can see from eq .
( [ eqs ] ) that at large distances the chiral angle @xmath90 falls off as @xmath91 ( note that at large @xmath54 , @xmath92 ) .
then , following the approximation of ref .
@xcite which consists of treating at large distances the field of one soliton as constant in the presence of the other , one obtains from @xcite that @xmath83 decreases as @xmath89 over a power of @xmath54 . for the potential @xmath84
we have shown in ref .
@xcite that it is a sum of two terms . the first term ( the one which has been considered in @xcite ) decreases as @xmath93 .
this behaviour is expected since the sixth - order term @xmath86 in eq .
( [ la ] ) can be regarded as the local approximation of an effective model of @xmath6 meson coupled to three pions @xcite .
however the asymptotic expression of the second part of @xmath84 , which arises from the baryon exchange current @xcite , contains both @xmath89 and @xmath93 terms so that it behaves at large @xmath54 as @xmath89 .
this means that the lagrangian @xmath86 generates , in addition to the expected three - pion exchange piece , a two - pion exchange one ( coming from the exchange current ) similar to that of the the fourth - order skyrme term @xmath85 .
thus we see that for large distance both potentials @xmath83 and @xmath84 are of yukawa type with range @xmath94 .
in fact , the main motivation of including a scalar - isoscalar meson field by hand in skyrme - type models @xcite is to mimic the two - pion exchange since the latter is not well addressed in these models with simple zero - mode quantization . a straightforward manner to calculate numerically the integrals giving the different contributions to the spin - orbit force consists in making the change in variable @xmath95 and taking the @xmath0 separation vector @xmath46 along the @xmath96 axis .
e.g. , with these changes , the expression of the linear @xmath4 model potential @xmath82 in eq .
( [ spor ] ) becomes : @xmath97 where @xmath98 . in fig.2 , we plot @xmath82 , @xmath83 , @xmath84 together with the total sum @xmath99 as functions of the relative distance @xmath54 between the two nucleons . as it was the case for the extended skyrme model ( [ la ] ) , @xmath83 and @xmath84
are still repulsive in the case of the model ( [ lanew ] ) . on the other hand , while the nonlinear @xmath4 model @xmath22 has a zero contribution to the isoscalar spin - orbit force contributes only to the isospin - dependent force @xcite . ]
, the linear @xmath4 model contributes and with a _ negative _ sign as it can be seen from the behaviour of @xmath82 in fig.2 .
this result is in agreement with the phenomenology @xcite .
the total potential ( full line in fig.2 ) is found to be repulsive for intermediate distances ( @xmath100fm ) and attractive for large distances ( @xmath101fm ) . for short distances ( not shown on fig.2 ) ,
neither the product ansatz nor the effective model should be trusted since this region corresponds to processes involving perturbative qcd .
the total potential is attractive at large distances because in this region the scalar meson contribution ( [ spor ] ) is the dominant one as compared to @xmath83 and @xmath84 .
this can be checked straightforwardly by comparing the asymptotic behaviour of each expression in eq ( [ so ] ) .
we have also displayed in fig.2 the potential related to the linear @xmath4 model and the sixth - order term only , @xmath102 . we observe from that curve that the attraction starts at about 1.7 fm instead of the 2 fm in the case of the total potential discussed above .
this result is in some way in agreement with the phenomenological isoscalar spin - orbit potential for which it is well known that scalar - isoscalar and @xmath6 mesons are the mesons which play the most important role @xcite .
similar results have already been found in ref .
@xcite where a different way has been used to introduce a scalar - isoscalar degrees of freedom . in ref .
@xcite , the scalar - isoscalar meson field is a dilaton , and it explicitly couples to the sixth - order term .
however the contribution to the spin - orbit force coming from the sixth - order term was not complete in ref .
@xcite . indeed , these authors ignored the baryon exchange current contribution to the spin - orbit force in their calculations , and we have shown in ref .
@xcite that this contribution is significant , with respect to the direct term @xcite , and should be taken into consideration .
it is then legitimate to question whether the qualitative results of ref . [
17 ] remain valid if the baryon exchange current contribution is taken into account . in this work , by using a different way of including the scalar - isoscalar meson in the skyrme lagrangian [ 14,16 ] , and by considering the _ entire _ contribution of the baryon current to the spin - orbit channel , we show that the main result of ref .
@xcite ( i.e. , an attractive @xmath18-@xmath18 spin - orbit force at large distance from skyrmions with scalar mesons ) is still valid . in this sense
, my results can be viewed as a confirmation of those of klbermann and eisenberg @xcite .
even though we have obtained an attractive force at large distances in the isoscalar spin - orbit channel , attraction is still missing at intermediate distances , and thus the problem remains unsolved within skyrme - type models .
this is not surprising since we have considered here only a finite - mass scalar field without the other vector mesons .
however , the result obtained here , namely , the change from a zero contribution to the isospin - independent spin - orbit interaction to an attractive one when replacing the frozen @xmath1-meson field ( nonlinear @xmath4 model ) with a realistic one ( linear @xmath4 model ) , is very encouraging .
indeed , it suggests that the right way to obtain an attractive isospin - independent @xmath0 spin - orbit force within skyrme - type models in the framework of the product ansatz is to replace the pion theory ( [ la ] ) with a realistic effective model including , in addition to the scalar - isoscalar @xmath1 field , all low - lying vector mesons and taking into account the finiteness of their masses . in a sense ,
the model considered here can be seen as a minimal and modest improvement of the pion skyrme theory .
similarly , it has been shown in ref .
@xcite that the model ( [ lanew ] ) gives rise to attraction in the @xmath0 central potential but the attraction occurs at distances larger than required by phenomenology .
the problem has been finally solved when the other vectors mesons were included @xcite . for this reason , we believe that in order to cure the problem of the isoscalar spin - orbit force , one has to consider effective models which incorporate the first mesonic resonances with finite masses . for instance , when considering a finite - mass @xmath6-meson model , the @xmath6-field couples directly to the nucleon via the baryon current ( defined after eq .
( [ la ] ) ) and generates the common three - pion exchange diagram , contrary to the case of the local approximation @xmath86 ( _ cf .
( [ la ] ) ) . indeed
, in the latter , an unexpected two - pion exchange piece coming from the baryon exchange current arises in addition to the three - pion piece and contributes with a positive sign to the isoscalar spin - orbit force yielding a repulsive interaction @xcite .
this problem with the baryon exchange current is obviously avoided when the sixth order term @xmath86 is replaced with a more realistic @xmath6-meson model . finally , in addition to considering finite - mass mesons lagrangians
, we would like to mention a further way that might lead to the desired attractive isoscalar spin - orbit force within the framework of skyrme - type models .
it is concerned with the approximation of the product ansatz configuration ( [ pro ] ) . in the latter
, the two nucleons are supposed to keep their spherical shape without deformation even after overlapping .
this is certainly not true in reality .
thus , while still using the product ansatz so that we keep benefiting from its simplicity and being able to perform analytical calculations , we may improve on the approximation ( [ pro ] ) by allowing the shape of the nucleon to deform when approaching the other nucleon ( _ cf_. ref .
@xcite and references therein ) .
calculations of the spin - orbit force within this approach are under way @xcite . 99 t. h. r. skyrme , proc .
* a260 * , 127 ( 1961 ) .
e. witten , nucl .
b * 223 * , 422 and 433 ( 1983 ) .
i. zahed and g. e. brown , phys .
rep . * 142 * , 1 ( 1986 ) .
g. s. adkins , c. r. nappi , and e. witten , nucl .
b * 228 * , 552 ( 1983 ) .
b. moussallam , ann .
( n.y . ) *
225 * , 264 ( 1993 ) .
a. jackson , a. d. jackson , and v. pasquier , nucl .
phys . a * 432 * , 567 ( 1985 ) .
a. jackson _ et al _ , phys .
b * 154 * , 101 ( 1985 ) .
a. abada and h. merabet , phys .
d * 48 * , 2337 ( 1993 ) .
e. m. nyman and d. o. riska , phys .
b * 175 * , 392 ( 1986 ) ; + d. o. riska and k. dannbom , phys .
scr . * 37 * , 7 ( 1988 ) ; + t. otofuji _
et al _ , phys .
b * 205 * , 145 ( 1988 ) .
r. d. amado _ et al _ , phys .
b * 314 * , 159 ( 1993 ) ; phys .
b * 324 * , 467 ( 1994 ) ; + b. shao _
et al _ , phys .
c * 48 * , 2498 ( 1993 ) ; phys .
c * 49 * , 3360 ( 1994 ) ; + a. abada , `` _ the isospin independent spin - orbit force in the extended skyrme model _ '' , hep - ph/9401341 , unpublished .
m. lacombe _
et al _ , phys .
c*21 * , 861 ( 1980 ) ; + r. machleidt , k. holinde and ch .
elster , phys . rep . * 149 * , 1 ( 1987 ) .
d. o. riska and b. schwesinger , phys .
b * 229 * , 339 ( 1989 ) .
a. abada , j. phys .
g * 22 * , l57 ( 1996 ) .
d. kalafatis , hep - ph/9406410 , thesis , univ .
k. iketani , kyushu university preprint 84-he-2 ( 1984 ) , unpublished .
d. kalafatis and r. vinh mau , phys .
d * 46 * , 3903 ( 1992 ) and references therein . g. klbermann and j. m. eisenberg , phys .
b * 349 * , 416 ( 1995 ) .
m. c. birse , j. phys .
g * 20 * , 1287 ( 1994 ) .
t. h. r. skyrme , nucl . phys . * 31 * , 556 ( 1962 )
. n. r. walet , nucl .
a * 586 * , 649 ( 1995 ) and references therein .
e. m. nyman , phys .
b * 162 * , 244 ( 1985 ) .
d. o. riska and e. m. nyman , phys .
b * 183 * , 7 ( 1987 ) .
a. rahimov _
et al _ , phys .
b 378 , 12 ( 1996 ) .
a. abada , work in progress . | as a first step toward circumventing the difficulty to obtain an attractive isospin - independent @xmath0 spin - orbit force from skyrme - type models involving only pions , we investigate an improved skyrme lagrangian that incorporates the scalar - isoscalar meson @xmath1 which can be viewed as the cause behind the enhancement of the @xmath2-wave .
we find that at large distances , the main contribution to the spin - orbit potential comes from the scalar lagrangian and it is found to be attractive .
we briefly discuss how to pursue this work to finally obtain a medium - range attractive interaction .
0.5 cm pacs numbers(s ) : 11.10 lm , 12.39 dc , 13.75 cs .
report - no : mc / th 96/17 hep - ph/9606270 |
the two weinberg sum rules ( wsr ) ( at t=0 ) @xcite were first derived in the framework of chiral @xmath7 symmetry and current algebra , and read @xmath8 = 2 \ , f_\pi^2\;,\ ] ] @xmath9 = 0 \;,\ ] ] where @xmath10 @xcite , and @xmath11 & = & ( -g_{\mu\nu}\ , q^2 + q_\mu q_\nu ) \ , \pi_v(q^2 ) \ ; , \end{aligned}\ ] ] @xmath12 & = & -g_{\mu\nu}\ ; \pi_1(q^2 ) -
q_\mu q_\nu\ ; \pi_a(q^2)\,,\end{aligned}\ ] ] with @xmath13 the conserved vector current in the chiral limit , @xmath14 the axial - vector current , and @xmath15 the four - momentum carried by the currents .
the functions @xmath16 are free of kinematical singularities , thus satisfying dispersion relations , and in perturbative qcd ( pqcd ) they are normalized as @xmath17\,.\ ] ] in the framework of qcd eqs .
( 1)-(2 ) become effectively finite energy sum rules ( in the chiral limit : @xmath18 ) @xmath19 = 2 \ , f_\pi^2\ , \delta_{n 0}\;,\ ] ] where @xmath20 is the squared energy beyond which qcd is valid , and both sum rules have been combined .
this result also follows from cauchy s theorem in the complex squared energy plane , together with the assumption of quark - hadron duality , to wit .
given a green function @xmath21 , e.g. @xmath22 , and an analytic function @xmath23 , cauchy s theorem leads to @xmath24 where the discontinuity across the real axis involves the hadronic sector , and the contour integral requires asymptotic information if the integration radius @xmath25 is large enough . according to quark - hadron ( global ) duality @xmath21 is assumed to be given by the qcd operator product expansion ( ope ) . a warning about this assumption
was first raised long ago in @xcite , and is now referred to as potential duality violations associated with the qcd behaviour on or close to the positive real axis .
tests of the wsr were performed @xcite-@xcite using experimental data on @xmath26-decays from the argus collaboration at desy @xcite , and later @xcite from the aleph collaboration @xcite .
both determinations led to a better saturation of the wsr if an integral kernel , e.g. a non - trivial @xmath23 in eq.(7 ) , was introduced , e.g. such that @xmath27 .
this was interpreted in @xcite as a possible signal of duality violations , on account of a considerably improved saturation of the wsr obtained with a _ pinched _
kernel @xcite , @xcite @xmath28 , i.e. @xmath29\nonumber \\ [ .3 cm ] & = & 2\ ; f_\pi^2 \,.\end{aligned}\ ] ] in this paper we discuss the extension of the wsr to finite temperature , which requires reliable information on the thermal behaviour of the hadronic spectral functions .
this information is available from results obtained from finite energy qcd sum rules ( fesr ) at finite @xmath30 in the vector @xcite and independently in the axial - vector channel @xcite .
starting with the hadronic vector spectral function , in the analysis of @xcite it was parametrized by a breit - wigner form @xmath31 where the upper index ( + ) specifies the time - like region ( @xmath32 ) , and all parameters depend on the temperature , and are normalized at @xmath0 as follows @xcite : @xmath33 , @xmath34 gev , and @xmath35 gev .
this parametrization was then used in the first three fesr @xmath36 & - & \frac{s_0^n}{n } \left[1+{\mathcal{o}}(\alpha_s)\right ] \;\ ; ( n=1,2,\cdots ) \;,\label{fesr}\end{aligned}\ ] ] where the leading order vacuum condensates in the chiral limit are the dimension @xmath37 gluon condensate and the dimension @xmath38 four - quark condensate .
the fesr were then used to determine @xmath39 , i.e. the onset of pqcd , and to check that the resulting dimension @xmath37 and @xmath38 vacuum condensates were in agreement with independent determinations from data analyses .
alternative parametrizations were also considered but found to lead to similar results . extending this analysis to finite temperature gave the following results @xmath40 with @xmath41 , @xmath42 , @xmath43 and @xmath44}\;,\ ] ] @xmath45 where @xmath46 , allowed to vary in the range @xmath47 , and finally @xmath48 with @xmath49 and @xmath50 .
+ turning to the axial - vector hadronic resonance parametrization , a fit was made to the aleph data on @xmath26-decay in order to use as the @xmath0 normalization .
as in the vector channel , the first three fesr were employed to determine @xmath25 and to check consistency of the results for the dimension @xmath51 condensates .
the parametrization leading to the best @xmath52 is given by @xcite @xmath53 \nonumber \\ [ .3 cm ] & ( 0 & \leq s \leq 1.2\ , { \mbox{gev}}^2)\;,\end{aligned}\ ] ] @xmath54 \nonumber \\ [ .3 cm ] & & ( 1.2\ , { \mbox{gev}}^2 \leq s \leq 1.45\,{\mbox{gev}}^2)\;,\end{aligned}\ ] ] @xmath53 \nonumber \\ [ .3 cm ] & & ( 1.45\ , { \mbox{gev}}^2 \leq s \leq m_\tau^2)\;,\end{aligned}\ ] ] where @xmath55 , @xmath56 , @xmath57 and @xmath58 ( the latter two parameters were split to facilitate comparison between @xmath59 and @xmath60 , for readers used to zero - width resonance saturation of the wsr ) . regarding eqs.(15)-(17 ) , it is well known that this spectral function can not be fitted with a normal breit - wigner expression , mostly due to the large width of the @xmath5 , and some background contamination . for instance , the early argus data @xcite was fitted with a breit - wigner modulated by a high degree polynomial @xcite . solving the fesr the onset of pqcd turns out to be identical to that in the vector channel , i.e. @xmath39 , so that eq.(17 ) is never needed .
the finite temperature results for @xmath61 , @xmath62 , @xmath63 , and @xmath64 can be written generically as @xmath65 where the various coefficients are given in table i. the @xmath5 mass hardly changes with temperature , so that it was kept constant .
this behaviour can be ascribed to the very large width of the @xmath5 resonance .
it should be noticed the ( different ) way the leptonic decay constants enter their respective spectral functions . however , the full @xmath30-dependence of these functions is the result of the aggregate behaviour of all their parameters .
this @xmath30-dependence is consistent with a quark deconfinement scenario in both channels .
+ the parametrization of the spectral function in the vector channel was later used to predict the dimuon production spectrum in heavy ion collisions at high energies @xcite .
excellent agreement was obtained after comparing with data on in - in collision around the @xmath4-meson peak from the na60 collaboration @xcite .
figure 3 shows the experimental data and the prediction using the vector spectral function from @xcite , as reproduced here in eqs.(9 ) , ( 11)-(14 ) .
+ it should be clear from figs . ( 1 ) and ( 2 ) that the vector spectral function evolves with increasing @xmath30 independently of the axial - vector one .
the latter involves , in addition to the @xmath5 resonance , the pion pole given by @xmath66 .
this contribution vanishes at @xmath6 , signaling chiral - symmetry restoration , i.e. a transition from a nambu - goldstone to a wigner - weyl realization of chiral @xmath7 symmetry
. it should be recalled that the thermal quark condensate from lattice qcd has been used as input to the fesr in the axial - vector channel in @xcite .
as shown in @xcite the @xmath30-dependence of the ratio@xmath67 follows closely that of @xmath68 , vanishing at the same temperature within the accuracy of the method .
this independent and clearly very different approach to @xmath6 exhibited by the thermal spectral functions does not suggest any potential chiral mixing .
ccccc + parameter & @xmath5 & @xmath69 & @xmath70 & @xmath71 + @xmath61 & - 28.5 & -0.6689 & 35.60 & 3.93 + @xmath62 & - 0.2924 & - 0.7557 & 73.43 & 11.08 + @xmath63 & - 19.34 & 14.27 & 7.716 & 6.153 + @xmath64&2.323 & 1.207 & 20.24 & 7.869 + in the qcd sector the pqcd spectral functions for both @xmath32 ( _ annihilation _ ) and @xmath72 ( _ scattering _ ) , as well as the dimension @xmath37 gluon condensate are chiral symmetric , thus not contributing to eq.(6 ) . in the hadronic sector for @xmath72 ( _ scattering _ )
there is an asymmetric contribution from pion loops , which in the vector channel at leading order is @xcite @xmath73 & = & \frac{t^2}{9 } \ ; \delta(s ) \;,\end{aligned}\ ] ] where @xmath74 is the bose thermal function .
the corresponding contribution in the axial - vector channel is a higher order two - loop effect , which is traditionally neglected .
starting at @xmath0 and using the results from the three fesr obtained in @xcite-@xcite , i.e. @xmath75 and the parametrizations eq.(9 ) and eqs.(15)-(17 ) , the wsr , eq.(6 ) , and the pinched wsr , eq.(8 ) , are satisfied to better than 1% .
the corresponding results at finite @xmath30 are shown in fig . 4 for the first wsr , @xmath76 , eq.(6 ) , and in fig . 5 for the pinched wsr , @xmath77 , eq .
the solid ( dash ) lines correspond to the left ( right ) hand sides of the sum rules . clearly , the upper limit of integration in the wsr , i.e. @xmath61 , is not strictly chiral - symmetric , as seen from eqs.(11 ) and ( 18 ) .
the divergence between the left and the right hand sides of the wsr at high @xmath30 is due to the asymmetric hadronic _ scattering _ term present in the vector spectral function , but loop suppressed in the axial - vector channel .
the importance of this term is short lived , as it must disappear at @xmath6 , since there would be no longer pions in the medium . in any case , since the source of this contribution is fully understood , it should not be interpreted as an invalidation of the thermal wsr . + as indicated in figs .
4 - 5 the wsr are very well satisfied up to @xmath78 .
the source of the departure between the left hand side of the wsr and its right hand side , i.e. @xmath79 , or alternatively the thermal quark condensate @xmath80 , is well understood .
it is due to the two - pion loop contribution to the vector spectral function , growing like @xmath3 , but disappearing at @xmath6 .
if this contribution would have been chiral symmetric , then it would have canceled out and the wsr would have been fully satisfied up to @xmath6 .
+ the results for the behaviour of the wsr in the full temperature range can be interpreted as follows .
first of all , at @xmath0 the wsr make the highly non - trivial statement that the integrated difference between the vector and the axial - vector ( including the pion ) spectral functions vanishes .
these functions could not be more different , i.e. the vector correlator has no pole , and the ground state resonance is narrow , while the axial - vector involves a pole and a very broad resonance almost twice as heavy .
next , there is no a - priori reason for the wsr to be satisfied at finite temperature .
their original derivation was done in the context of current algebra and chiral @xmath7 symmetry .
today we reinterpret them in the framework of qcd , which does exhibit this chiral symmetry in the limit of vanishing light quark masses . in principle , an extension to finite @xmath30 could lead to more than one scenario , including one in which the sum rules are not satisfied at all , except at @xmath6 where they must be trivially satisfied .
the results obtained from thermal qcd fesr @xcite-@xcite clearly indicate that the vector spectral function evolves with @xmath30 independently of the axial - vector spectral function .
while both disappear at @xmath6 , they do so following their own distinctive thermal evolution .
in other words , there is no a - priori relation between these two channels , unless one were to demand it .
given the successful saturation of the wsr there seems to be no reason to justify such a demand .
the behaviour of the spectral function parameters points to deconfinement , as the onset of pqcd , @xmath61 , and the resonance couplings decrease , while the widths increase substantially with increasing @xmath30 .
this results in the @xmath4 and @xmath5 resonances becoming progressively broader , with decreasing couplings to their respective currents . at the same time , the pion decay constant / quark condensate decreases with increasing @xmath30 , vanishing at @xmath2 .
the thermal behaviour of the hadron mass is irrelevant in the context of the approach to deconfinement .
in fact , the mass only specifies the position of the real part of the resonance green function in the complex squared energy @xmath81-plane .
its moving upwards or downwards with temperature is no indication of a deconfinement phase transition .
it is the imaginary part of the green function , i.e. the hadronic width , the relevant parameter in this context .
its considerable increase with increasing temperature thus serves as a phenomenological signal for deconfinement .
+ alternative analyses of the thermal wsr have been performed recently @xcite , with somewhat different emphasis , e.g. in connection with potential chiral mixing .
this work has been supported in part by dgapa - unam under grant papiit - in103811 , conacyt - mexico under grant no . 128534 , the national research foundation ( south africa ) , the university of cape town research committee , and fondecyt ( chile ) under grants no .
1130056 and no .
1120770 .
99 s. weinberg , phys .
lett . * 18 * , 507 ( 1967 ) .
j. beringer _
( particle data group ) phys .
d * 86 * , 010001 ( 2012 ) .
r.shankar , phys .
d * 15 * , 755 ( 1977 ) .
r. d. peccei and j. sol , nucl .
b * 281 * , 1 ( 1987 ) . c. a. dominguez and j. sol , phys .
b * 208 * , 131 ( 1988 ) .
h. albrecht _
( argus collaboration ) , z. phys .
c * 33 * , 7 ( 1986 ) . c. a. dominguez and k. schilcher , phys
b * 448 * , 93 ( 1999 ) .
aleph coll .
, r. barate _
j. c * 4 * , 409 ( 1988 ) .
the use of _ pinched _ kernels in other qcd sum rules was independently proposed in k. maltman , phys .
b * 440 * , 367 ( 1998 ) .
a. ayala , c. a. dominguez , m. loewe and y. zhang , phys .
d * 86 * , 114036 ( 2012 ) . notice that there is a misprint in eq.(8 ) of this reference ; the correct formula is given in eqs.(15)-(17 ) of this paper . c. a. dominguez , m. loewe and y. zhang , phys .
d * 86 * , 034030 ( 2012 ) . c. a. dominguez and j. sol , z. phys .
c * 40 * , 63 ( 1988 ) .
a. ayala , c. a. dominguez , l. a. hernandez , m. loewe and a. j. mizher , phys .
d * 88 * , 114028 ( 2013 ) .
r. arnaldi _ et al . _
( na60 collaboration ) , phys . rev . * 96 * , 16302 ( 2006 ) ; ( na60 collaboration ) , phys .
lett . * 100 * , 022302 ( 2008 ) ; s. damjanovic for the na60 colaboration , j. phys .
g * 35 * , 104036 ( 2008 ) ; ( na60 collaboration ) , eur .
j. c * 59 * , 607 ( 2009 ) ; ( na60 collaboration ) , eur .
j. c * 61 * , 711 ( 2009 ) . a. i. bochkarev and m. e. shaposnikov , nucl .
b * 268*. 220 ( 1986 )
. n. p. m. holt , p.m. hohler and r. rapp , phys .
d * 87 * , 076010 ( 2013 ) ; p. m. hohler and r. rapp , phys .
b * 731 * , 103 ( 2014 ) , and references therein . | the saturation of the two weinberg sum rules is studied at finite temperature , using recent independent qcd sum rule results for the thermal behaviour of hadronic parameters in the vector and axial - vector channels .
both sum rules are very well satisfied from @xmath0 up to @xmath1 . at higher temperatures close to @xmath2 a hadronic , pion - loop contribution in the space - like region proportional to @xmath3 , present at leading order in the vector but not in the axial - vector channel
, induces an asymmetry leading to a small deviation . in this region , though , qcd sum rules for the hadronic parameters begin to have no solutions , as the hadronic widths of the @xmath4 and the @xmath5 mesons diverge signaling deconfinement .
close to , and at @xmath6 there are no pions left in the medium and chiral symmetry is restored , so that the sum rules are trivially satisfied . |
in recent years , there has been a growing interest in studying three - body problem with the aim of approximating the behavior of real celestial systems ( @xcite and refs .
therein ) . the results from the study have been found useful in the theory of dynamical systems and in astronomy .
the simplest form of the three - body problem is restricted three - body problem ( r3bp ) .
it describe the motion of an infinitesimal mass moving under the gravitational influence of two massive bodies called the primaries which moves in circular orbits around their center of mass on account of their mutual attraction and the infinitesimal mass not influencing the motion of the primaries .
the classical restricted three - body problem possesses five equilibrium points .
the first three points @xmath5 , denotes the collinear points while the last two points ( @xmath6 and @xmath7 ) denotes the triangular points .
the earth - moon system together with an artificial satellite constitute such problem @xcite .
now , with an interest in four - body problem .
couple r3bp is one of the examples of restricted four - body problem ( r4bp ) .
the smaller body is called as the infinitesimal mas while the remaining massive bodies are called as the primaries . very recently ,
batagiannis and papadakis @xcite studied the families of simple symmetric and non - symmetric periodic orbits in the restricted four body problem .
few among other interesting works are kumari and kushvah @xcite , papadouris and papadakis @xcite , alvarez and vidal @xcite .
the purpose for studying r4bp includes their application in general behavior of the synchronous orbit in presence of moon and the sun . in this paper ,
our aim is to study the motion of an infinitesimal mass by assuming that the primaries of the system are radiating - oblate spheroids surrounded by a circular cluster of material points . by using numerical computation procedure via wolfram mathematica v10 , we locate the libration points and examine their linear stability under the effect of oblateness and gravitational potential from a circular cluster of material points .
let @xmath8 @xmath9 be the masses of the primaries with @xmath10 moving in circular periodic orbits around their center of mass fixed at the origin of the coordinate system .
these masses always lies at the vertices of equilateral triangle with @xmath3 being on the positive x - axis at the origin of time .
the motion of the system is referred to axes rotating with uniform angular velocity @xcite .
the three bodies move in the same plane and their mutual distances remain unchanged with respect to time .
the motion of the primaries consists of circular orbits around their center of gravity .
an equilateral equilibrium configuration of the three - bodies which is a particular solution of the three - body problem given by lagrange is formed by the primaries at any instant of a given time .
we assume that the influence of infinitesimal mass on the motion of primaries moving under their mutual gravitational attraction is negligible . for studying the position of the infinitesimal mass , m , in the plane of motion of the primaries we applied the synodical coordinates .
thus the masses of the primaries are fixed at @xmath11 where the mass parameter @xmath2 is taken as @xmath12 we now take the potential energy of the infinitesimal mass , under the influence of a circular cluster of material points with its center at the origin of the coordinate system , under radiating - oblate primaries as @xmath13.\ ] ] with @xmath14 define as @xmath15 where g is the gravitational constant .
@xmath16 and @xmath17 are the radiation factors of the primaries .
@xmath18 @xmath19 , where @xmath20 is the force caused by radiation pressure force and @xmath21 results due to gravitational force @xcite .
the term @xmath22 is the potential due to the circular cluster of material points ( @xcite and refs therein ) where @xmath23 is the total mass of the circular cluster of material points .
the oblateness coefficients for the more massive primary denoted as @xmath24 @xmath25 and for the less massive primary as @xmath26 @xmath27 with @xmath28 being the zonal harmonic coefficients and @xmath29 denote the mean radii of @xmath30 , supposing that the primaries have their equatorial planes coinciding with the plane of motion . also , @xmath31 is the distance of the infinitesimal mass and which is given by @xmath32 furthermore , @xmath33 , where @xmath34 and @xmath35 are two parameters which determine the density profile of the circular cluster of material points .
@xmath34 , the flatness parameter , control the flatness of the profile while @xmath35 , the core parameter , controls the size of the core of the density .
now , the lagrangian of our problem can be written as @xmath36 where the perturbed mean motion @xmath37 is define as @xcite @xmath38 @xmath39 is the radial distance of the infinitesimal body .
it then follows that the equations of motion of the infinitesimal mass are : @xmath40 @xmath41 where @xmath42 it should be noted here that the suffixes x and y indicate the partial derivatives of @xmath43 with respect to @xmath44 and @xmath45 respectively .
this system admits the well - known jacobi integral : @xmath46 where c is the jacobi constant .
in this section , we attempt to find the equilibrium points for the cases @xmath47 and @xmath48 . from the equations of motion ( [ e9a ] , [ e9b ] ) , it can be deduce that equilibrium solution exists relative to the rotating frame when the partial derivatives of the pseudopotential function @xmath49 is equal zero .
thus , by using the following chain rule , we can find @xmath50 we have splited the differential for the sake of simplicity . now
@xmath51\nonumber\\ & & -\frac{\mu q_1}{8r_2 ^ 6}\left(8r_2 ^ 4 + 12b_1r_2 ^ 2 - 15b_2\right)\left[\frac{\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x}{\sqrt{\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\left(y-\frac{1}{2}\right)^2}}\right]\label{e13}\\ & & -\mu\frac{\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x}{\left[\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\left(y+\frac{1}{2}\right)^2\right]^{3/2}}-\frac{m_bx}{\left(t^2+r^2\right)^{3/2}}=0\nonumber\end{aligned}\ ] ] similarly , we can find @xmath52 as @xmath53\nonumber\\ & & -\frac{\mu q_1}{8r_2 ^ 6}\left(8r_2 ^ 4 + 12b_1r_2 ^ 2 - 15b_2\right)\left[\frac{\left(y-\frac{1}{2}\right)}{\sqrt{\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\left(y-\frac{1}{2}\right)^2}}\right]\label{e14}\\ & & -\mu\frac{\left(y+\frac{1}{2}\right)}{\left[\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\left(y+\frac{1}{2}\right)^2\right]^{3/2}}-\frac{m_by}{\left(t^2+r^2\right)^{3/2}}=0\nonumber\end{aligned}\ ] ] and @xmath1 and @xmath3 .
we choose @xmath54 , @xmath55 , @xmath56 , @xmath57 , @xmath58 , @xmath59 , @xmath60 , @xmath61 , @xmath62 ( b ) same as ( a ) but for @xmath63 and @xmath64.,width=604,height=377 ] and @xmath1 and @xmath3 .
we choose @xmath54 , @xmath55 , @xmath56 , @xmath57 , @xmath58 , @xmath59 , @xmath65 , @xmath66 , @xmath67 ( b ) same as ( a ) but for @xmath63 and @xmath64.,width=604,height=377 ] setting y as zero , we can find the equilibrium points at x - axis by solving equations ( [ e13 ] ) and ( [ e14 ] ) .
this implies that collinear points lie on the line joining the primaries .
we choose @xmath54 and varies the oblateness coefficients .
thus , @xmath68\nonumber\\ & & -\frac{\mu q_1\left(\left|\left(x+\frac{\sqrt{3}}{2}(1 - 2\mu)\right)^2+\frac{1}{4}\right|\left[8\left|\left(x+\frac{\sqrt{3}}{2}(1 - 2\mu)\right)^2+\frac{1}{4}\right|+b_1\right]-15b_2\right)}{8\left[\left(x+\frac{\sqrt{3}}{2}(1 - 2\mu)\right)^2+\frac{1}{4}\right]^{7/2}}\nonumber\\ & \times&\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)-\mu\frac{\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x}{\left[\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\frac{1}{4}\right]^{3/2}}-\frac{m_bx}{\left(t^2+r^2\right)^{3/2}}=0\label{e15}\end{aligned}\ ] ] @xmath69 - 15b_2\right)}{16\left[\left(x+\frac{\sqrt{3}}{2}(1 - 2\mu)\right)^2+\frac{1}{4}\right]^{7/2}}\nonumber\\ & & -\mu\frac{1}{\left[\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\frac{1}{4}\right]^{3/2}}=0\label{e16}\end{aligned}\ ] ] the two collinear points have been represented in figures ( [ fig1]-[fig2 ] ) and presented for better clarification in table ( [ tab1 ] ) the non - collinear points can be found by solving equations ( [ e13 ] ) and ( [ e14 ] ) simultaneously when @xmath70 , i.e. @xmath71 and @xmath72 , where @xmath73\nonumber\\ & & -\frac{\mu q_1}{8r_2 ^ 6}\left(8r_2 ^ 4 + 12b_1r_2 ^ 2 - 15b_2\right)\left[\frac{\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x}{\sqrt{\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\left(y-\frac{1}{2}\right)^2}}\right]\label{e17a}\\ & & -\mu\frac{\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x}{\left[\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\left(y+\frac{1}{2}\right)^2\right]^{3/2}}-\frac{m_bx}{\left(t^2+r^2\right)^{3/2}}=0\nonumber\end{aligned}\ ] ] @xmath74\nonumber\\ & & -\frac{\mu q_1}{8r_2 ^ 6}\left(8r_2 ^ 4 + 12b_1r_2 ^ 2 - 15b_2\right)\left[\frac{\left(y-\frac{1}{2}\right)}{\sqrt{\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\left(y-\frac{1}{2}\right)^2}}\right]\label{e17b}\\ & & -\mu\frac{\left(y+\frac{1}{2}\right)}{\left[\left(\frac{\sqrt{3}}{2}\left(1 - 2\mu\right)+x\right)^2+\left(y+\frac{1}{2}\right)^2\right]^{3/2}}-\frac{m_by}{\left(t^2+r^2\right)^{3/2}}=0\nonumber\end{aligned}\ ] ] we have also choose @xmath54 and varies the oblateness coefficients .
the six non - collinear points have been represented in figures ( [ fig1]-[fig2 ] ) and presented for better clarification in table ( [ tab2 ] )
in this section , we aim to study the stability of a libration point @xmath75 . to achieve this goal ,
firstly , we apply infinitesimal displacement @xmath76 and @xmath77 to the coordinates via : @xmath78 substituting equation ( [ e18 ] ) into equations of the motion in ( [ e9a ] ) and ( [ e9b ] ) , we can find @xmath79 where the superfix ` 0 ' indicates that the partial derivatives have been computed at the triangular libration point by considering @xmath75 . now
, let us assume a solution of the form @xmath80 and @xmath81 , where @xmath82 and @xmath83 is a constant and @xmath84 is a parameter . substituting the assumed solutions in equation ( [ e19 ] )
, we can find the following non - trivial solution for @xmath82 and @xmath83 @xmath85 solving the determinant by expansion , we can find the characteristic equation corresponding to the variational equations of ( [ e19 ] ) as @xmath86 now , the second order derivatives of the potential function w.r.t x and y are as follows @xmath87}{8 { r}_1 ^ 9}\nonumber\\ & & + \frac{\mu { q}_2 \left[4 { r}_2 ^ 2 \left(3 { b}_1 \left(16 { r}_2 ^ 2 - 5 ( 1 - 2 y)^2\right)+2 { r}_2 ^ 2 \left(8 { r}_2 ^ 2 - 3 ( 1 - 2 y)^2\right)\right)-15 { b}_2 \left(24 { r}_2 ^ 2 - 7 ( 1 - 2 y)^2\right)\right]}{32 { r}_2 ^ 9}\nonumber\\ & & + \frac{\mu \left(8 { r}_3 ^ 2 - 3 ( 2 y+1)^2\right)}{4{r}_3 ^ 5 } + n^2-\frac{{m}_b \left(t^2 - 2 x^2+y^2\right)}{\left(t^2+x^2+y^2\right)^{5/2 } } \label{e21a}\end{aligned}\ ] ] @xmath88}{8r_1 ^ 9}\nonumber\\ & & -\frac{\mu q_2(3b_1 + 4r_2 ^ 2)\left(y-\frac{1}{2}\right)^2}{r_2 ^ 7}+\frac{7\mu q_2\left(-15b_2 + 12b_1r_2 ^ 2 + 8r_2 ^ 4\right)\left(y-\frac{1}{2}\right)^2}{8r_2 ^ 9}\label{e21b}\\ & & -\frac{\mu}{r_3 ^ 3}+\frac{3\mu\left(y+\frac{1}{2}\right)^2}{r_3 ^ 5}+n^2-\frac{m_b(t^2+x^2 - 2y^2)}{(t^2+x^2+y^2)^{5/2}}-\frac{\mu q_2\left(-15b_2 + 12b_1r_2 ^ 2 + 8r_2 ^ 4\right)}{8r_2 ^ 7}\nonumber\end{aligned}\ ] ]
@xmath89 for stability of the collinear points @xmath90 , we have @xmath91}{8r_{*1}^9}+\frac{\mu \left(8 r_{*3}^2 - 3\right)}{4r_{*3}^5 } + n^2-\frac{{m}_b \left(t^2 - 2x^2\right)}{\left(t^2+x^2\right)^{5/2}}\nonumber\\ & & + \frac{\mu { q}_2 \left[4 r_{*2}^2 \left(3 { b}_1 \left(16 r_{*2}^2 - 5 \right)+2 r_{*2}^2 \left(8 r_{*2}^2 - 3\right)\right)-15 { b}_2 \left(24 r_{*2}^2 - 7\right)\right]}{32 r_{*2}^9 } \label{e22a}\end{aligned}\ ] ] @xmath92}{8r_{*1}^9}\nonumber\\ & & -\frac{\mu q_2(3b_1 + 4r_{*2}^2)}{4r_{*2}^7}+\frac{7\mu q_2\left(-15b_2 + 12b_1r_{*2}^2 + 8r_{*2}^4\right)}{32r_{*2}^9}\label{e22b}\\ & & -\frac{\mu}{r_{*3}^3}+\frac{3\mu}{4r_{*3}^5}+n^2-\frac{m_b}{(t^2+x^2)^{3/2}}-\frac{\mu q_2\left(-15b_2 + 12b_1r_{*2}^2 + 8r_{*2}^4\right)}{8r_{*2}^7}\nonumber\end{aligned}\ ] ] @xmath93 where we have introduced parameters @xmath94 for mathematical simplicity . substituting these values given by equation ( 22 ) into the characteristic equation
, we have @xmath95 numerical computation of @xmath96 have been presented in tables [ tab3 ] .
we take @xmath97 and @xmath98 .
it can be seen that collinear equilibrium points are unstable for all the variational groups we considered .
again , we solve the characteristic equation for @xmath96 , i.e. @xmath95 and then evaluate them at equilibrium points .
the results have been presented in table [ tab4 ] .
we take @xmath97 and @xmath98 .
an equilibrium point is stable if the characteristic equation evaluated at the equilibrium point , has four complex roots with negative real parts or pure imaginary roots otherwise it is unstable . in this regard ,
the non - collinear equilibrium points are found to be stable . by taking @xmath99 and using @xmath100
, we have the discriminant @xmath101 @xmath102 and by taking @xmath103 with @xmath100 , we have the discriminant @xmath104 @xmath105 .
this implies that @xmath106 is a decreasing function of @xmath2 within an interval @xmath107 and there exist only one value of @xmath2 called as critical mass ( @xmath108 ) in the interval @xmath107 for which the discriminant is zero .
in this research , we examined the motion of an infinitesimal mass by assuming that the primaries of the system are radiating - oblate spheroids surrounded by a circular cluster of material points , within the framework of restricted four - body problem . in our model we assume that the two masses of the primaries @xmath0 and @xmath1 are equal to @xmath2 and the mass @xmath3 is @xmath4
. we have obtained the equilibrium points via numerical computation . in table [ tab1 ] we obtained the two collinear points @xmath109 and @xmath110 on the x - axis .
firstly , we fixed @xmath111 at 0.0015 in the absent of gravitational potential with @xmath112 and then varies @xmath113 as 0.0000:0.0030:0.0090 .
we found that the equilibrium point shifted from left to right .
this is also observed in the presence of gravitational potential with @xmath114 .
however , when we fixed @xmath113 at 0.0015 and then varies @xmath111 , we found that collinear point @xmath115 shifted from left to right whereas that of @xmath110 shifted from right to left .
this is observed in the two cases we considered , i.e. @xmath116 and @xmath117 . in table
[ tab2 ] , we present the non - collinear points for the two cases @xmath116 and @xmath117 . it is shown that @xmath118 .
we found that our results are in excellent agreement with kumari and kushvah @xcite when @xmath116 also with the ones obtained by papadouris and papadakis @xcite when @xmath119 .
we also examined the linear stability of these point and found then to be unstable .
an extension to the effect of potential created by the circular cluster and oblateness coefficients for the more massive primary and the less massive primary , on the existence and linear stability of the libration point have also been presented .
99 j. singh and j. j. taura , astrophys space sci * 350 * ( 2014 ) 127 .
j. singh and j. j. taura , astrophys space sci * 351 * ( 2014 ) 499 .
a. n. baltagiannis and k. e. papadakis , astrophys space sci * 336 * ( 2011 ) 357 .
r. kummari and b. s. kushvah , astrophys .
space sci . * 349 * ( 2014 ) 693 .
j. p. papadouris and k. e. papadakis , astrophys .
space sci .
* 344 * ( 2013 ) 21 .
m. alvarez and c. vidal , math .
eng , doi:10.1155/2009/181360 . v. v. radzievskii , astron . j. * 27 * ( 1950 ) 250 . m. miyamoto , r. nagai , publ . astron
27 , 533 ( 1975 ) i. d. peter , j. j. lissauer , planetary science .
cambridge university press , new york ( 2001 ) | within the framework of restricted four - body problem , we study the motion of an infinitesimal mass by assuming that the primaries of the system are radiating - oblate spheroids surrounded by a circular cluster of material points . in our model , we assume that the two masses of the primaries @xmath0 and @xmath1 are equal to @xmath2 and the mass @xmath3 is @xmath4 . by using numerical approach , we have obtained the equilibrium points and examined their linear stability . the effect of potential created by the circular cluster and oblateness coefficients for the more massive primary and the less massive primary , on the existence and linear stability of the libration point have been critically examine via numerical computation . the stability of these points examined shows that the collinear and the non - collinear equilibrium points are unstable .
the result presented in this paper have practical application in astrophysics .
= 0.20 in * * -1.2cm0.5 cm effect of oblateness , radiation and a circular cluster of material points on the stability of equilibrium points in the restricted four - body problem _ applied theoretical physics division , department of physics , federal university lafia , p. m. b. 146 , lafia , nigeria .
_ few body sys .
( 2014 ) doi : 10.1007/s00601 - 014 - 0922 - 3 * keywords * : restricted four - body problem ; potential from a circular cluster of material points ; stability ; oblateness |
the theoretical and experimental characterization of spin dynamics in ferromagnetic materials due to the interaction with short optical pulses has become an important part of research in magnetism .
@xcite in this connection , spin - dependent hot - electron transport processes in metallic heterostructures have received enormous interest in the past few years .
@xcite in particular , superdiffusive - transport theory has played an increasingly important role in the quantitative interpretation of experimental results . @xcite superdiffusive transport - theory , which was introduced and comprehensively described in refs . , uses spin- and energy - dependent electron lifetimes as input , @xcite and its quantitative results for hot - electron transport on ultrashort timescales in ferromagnetic materials rely heavily , to the best of our knowledge , on the relation between majority and minority electrons for these materials .
the spin - dependent lifetimes that are used for hot - electron transport , both in ferromagnets and normal metals , are the so - called `` inelastic lifetimes . ''
these state ( or energy ) dependent lifetimes result from out - scattering processes due to the coulomb interaction between an excited electron and the inhomogeneous electron gas in the system .
these lifetimes can be measured by tracking optically excited electrons using spin- and time - resolved 2-photon photoemission ( 2ppe ) @xcite and can be calculated as the broadening of the electronic spectral function using many - body green function techniques .
@xcite the problem of the accurate determination of these lifetimes has fueled method development on the experimental and theoretical side , @xcite but has always suffered from the presence of interactions ( electron - phonon , surface effects ) that can not be clearly identified in experiment and are difficult to include in calculations .
qualitative agreement was reached for the spin - integrated lifetimes in simple metals and iron , @xcite but even advanced quasiparticle calculations including many - body @xmath0-matrix contributions , have yielded a ratio between majority and minority lifetimes , which is in qualitative disagreement with experiment for some ferromagnets .
a particularly important material in recent studies has been nickel , @xcite for which the theoretical ratio comes out between 6 and 8 , @xcite while the experimental result @xcite is 2 .
recent experimental results point toward a similar disagreement for cobalt .
@xcite in light normal metals and ferromagnets spin - orbit coupling generally leads to very small corrections to the single - particle energies , i.e. , the _ band structure _ , but it changes the single - particle states qualitatively by introducing a state - dependent spin mixing . with spin - orbit coupling , the average spin of an electron can be changed in transitions due to any spin - diagonal interaction , in particular by electron - phonon momentum scattering .
@xcite this is also true for the two - particle coulomb interaction , @xcite as long as one monitors only the average spin of one of the scattering particles , as is done in lifetime measurements by 2-photon photoemission experiments . while this spin mixing due to spin - orbit coupling
has recently been included in lifetime calculations for lead , @xcite it was not included in dft codes used for existing lifetime calculations for 3d - ferromagnets and aluminum , @xcite whose results are nowadays widely used .
this paper presents results for electron lifetimes in metals and spin - dependent lifetimes in ferromagnets _ including spin - orbit coupling_. we show that spin - orbit coupling can be important for electron lifetimes in metals in general
. moreover , the ratio between the calculated majority and minority lifetimes is , for the first time , in agreement with experiment .
@xcite we believe that our calculated electronic lifetimes should be used as an accurate input for calculations of spin - dependent hot - electron dynamics in ferromagnets .
we first discuss briefly our theoretical approach to calculate the lifetimes .
we start from the dynamical and wave - vector dependent dielectric function @xmath1 in the random phase approximation ( rpa ) .
@xcite our approach , cf . ref .
, evaluates the wave - vector summations in @xmath1 without introducing an additional broadening of the energy - conserving @xmath2 function .
this procedure removes a parameter whose influence on the calculation for small @xmath3 is not easy to control and which would otherwise need to be separately tested over the whole energy range .
the @xmath4- and band - resolved electronic scattering rates , i.e. , the inverse lifetimes , @xmath5 , are calculated using the expression @xcite @xmath6 here , the band indices are denoted by @xmath7 and @xmath8 , and @xmath4 and @xmath9 denote wave - vectors in the first brillouin zone ( 1 .
the energies @xmath10 , occupation numbers @xmath11 and overlap matrix elements @xmath12 are extracted from the elk dft ( density functional theory ) code , @xcite which employs a full - potential linearized augmented plane wave ( fp - lapw ) basis .
last , @xmath13 denotes the fourier transformed coulomb potential and @xmath14 is the energy difference between initial and final state . for negative @xmath15 , the distribution function has to be replaced by @xmath16 . by using the overlap matrix elements
as defined above we neglect corrections due to local field effects . in the language of many - body green functions
, this corresponds to an on - shell @xmath17 calculation , @xcite where the screened coulomb interaction ( @xmath18 ) is obtained from the full rpa dielectric function .
the @xmath4- and band - dependent wave - functions that result from the dft calculations including spin - orbit coupling are of the form @xmath19 , @xcite where @xmath20 are spinors identified by the spin projection @xmath21 , @xmath22 along the magnetization direction .
according to whether @xmath23 or @xmath24 is larger , we relabel each eigenstate by its dominant spin contribution @xmath25 , so that we obtain spin - dependent lifetimes , @xmath26 .
our choice of quantization axis is such that @xmath21 denotes majority carriers states and @xmath27 minority carrier states . due to the existence of several bands ( partially with different symmetries ) in the energy range of interest and the anisotropy of the dft bands @xmath28 , several lifetimes @xmath29
can be associated with the same spin and energy . when we plot these spin and energy dependent lifetimes @xmath30 in the following , in particular figs .
[ fig : lifetimeco ] and [ fig : lifetimeni ] , this leads to a scatter of @xmath30 values .
( color online ) energy - resolved majority ( @xmath31 , blue @xmath32 ) and minority ( @xmath33 , red @xmath34 ) carrier lifetimes for cobalt .
there are in general several different lifetime points at the same energy ( see text ) .
we used @xmath35 @xmath4-points in the full bz.,scaledwidth=50.0% ] ( color online ) same as fig .
[ fig : lifetimeco ] for nickel.,scaledwidth=50.0% ] figures [ fig : lifetimeco ] and [ fig : lifetimeni ] display the calculated energy- and spin - resolved carrier lifetimes @xmath30 around the fermi energy for cobalt and nickel .
the spread of lifetimes at the same energy , which was mentioned above , can serve as an indication for the possible range of results for measurements of energy resolved lifetimes .
these `` raw data '' are important for the interpretation of the theoretical results because they already show two important points .
first , we checked that there is no good fermi - liquid type fit to these lifetimes .
second , even if one fits the lifetimes in a restricted energy range by a smooth @xmath36 curve , this ignores the spread of lifetimes , which can be quite sizable as shown in figs .
[ fig : lifetimeco ] and [ fig : lifetimeni ] .
we believe that such a spread of electronic lifetimes , in particular in the range around 1ev above the fermi energy should be important for the interpretation of photoemission experiments in this energy range , and when these results are used as input in hot - electron transport calculations .
figure [ fig : lifetimeco ] shows the energy- and spin - resolved lifetimes in cobalt .
in addition to the longer lifetimes close to the fermi energy , hole lifetimes in excess of 5fs occur at the top of some d - bands around @xmath37 , @xmath38 , and @xmath39 . for electronic states with energies above 0.5ev longer lifetimes occur at some @xmath4-points .
there are also @xmath40 states with a pronounced spin - asymmetry in the lifetimes ( see discussion below ) .
another important property of cobalt is the existence of two different conduction bands , which intersect the fermi surface with different slope .
this leads to _ two _ rather well - defined lifetime curves , both for electrons and holes .
this can be best seen between @xmath41 and 0ev , where the two curves are shifted by about 0.2ev . the calculated lifetimes in nickel ,
see fig .
[ fig : lifetimeni ] , do not show a pronounced influence of d - bands and/or anisotropy below the fermi energy as in cobalt , which is due to the smaller number of bands in the vicinity of the fermi energy .
however , there is a clear spin - dependence of electronic lifetimes , which is most pronounced around 0.4ev , but persists almost up to 2ev .
in the following , we will mainly be concerned with lifetimes above 0.3ev above the fermi energy , which is the interesting energy range for the interpretation of photoemission experiments and hot - electron transport calculations , because close to the fermi energy the influence of phonons is expected to become more pronounced and lead to significantly shorter lifetimes than those predicted by a calculation that includes only the coulomb interaction . to facilitate comparison with experiment we average the lifetimes in each spin channel in bins of @xmath42 and denote the result by @xmath43 .
the standard deviation of the averaging process then yields `` error bars '' on the @xmath43 values .
note that this procedure does not correspond to a `` random @xmath40 '' approximation .
( color online ) energetically averaged majority ( blue up triangles ) and minority ( red down triangles ) lifetimes for ( a ) fe , ( b ) co and ( c ) ni .
the error bars denote the standard deviation obtained from the scatter of the lifetimes as shown in figs .
[ fig : lifetimeco ] and [ fig : lifetimeni ] .
the insets show the calculated ratio of majority and minority electrons ( `` @xmath34 '' ) , @xmath44 , in comparison to experimental data , where the `` @xmath45 '' ( `` @xmath46 '' ) correspond to values extracted from ref . and ( ) .,scaledwidth=45.0% ]
figure [ fig : lifetimezoom ] displays the averaged electron lifetimes determined from the data shown in figs .
[ fig : lifetimeco ] and [ fig : lifetimeni ] . as insets
we have included the ratio of majority and minority lifetimes , @xmath44 , together with experimental data @xcite for iron , cobalt and nickel .
figure [ fig : lifetimezoom](a ) shows that there is only a very weak spin dependence for iron , and the agreement of the ratio @xmath47 with experiment @xcite and recent investigations @xcite is quite good , but there is a slight disagreement with earlier , semiempirical studies .
@xcite however , even an increase of the ratio around @xmath48 in the experiment @xcite is well reproduced in our results .
the averaged lifetimes of cobalt , which are shown in fig .
[ fig : lifetimezoom](b ) , agree quite well with the experimental lifetimes , @xcite but the large error bars extend to a much wider energy range than in iron .
this can be traced back to the scatter of lifetimes in fig .
[ fig : lifetimeco ] .
the corresponding figure for iron ( not shown ) exhibits a much smaller scatter .
the ratio of majority and minority electron lifetimes , see inset in fig .
[ fig : lifetimezoom](b ) , is around 1 below 0.5ev and increases to @xmath49 for larger energies , a trend that agrees extremely well with measurements .
@xcite to put this result into perspective we note that the experimental data in ref . were compared with a theoretical model based on the random @xmath40 approximation .
@xcite if the random-@xmath40 interaction matrix elements are taken to be spin and energy independent , the majority and minority relaxation times are determined by double convolutions over the spin - dependent density - of - states ( dos ) .
@xcite it was found that the experimental results were not in agreement with the ratio of the dos at the fermi energy , which led the authors of ref . to speculate that the matrix elements for parallel and antiparallel spins should be different due to the pauli exclusion principle . in our calculations , the effective spin - dependence of the matrix elements
is caused exclusively by the spin - mixing due to spin - orbit coupling , but the effect is the same : it makes the ratio of the lifetimes different from the spin - dependent dos at the fermi energy . in fig .
[ fig : lifetimezoom](c ) we turn to nickel .
here , as in the case of iron , the average lifetimes are slightly larger than the measured ones @xcite ( not shown ) , but due to the small anisotropy in the band structure , the lifetimes in nickel show the smallest error bars and thus an extremely well - defined spin dependence . only our calculated
_ majority _ electron lifetimes are similar to earlier ab - initio evaluations , @xcite but there is an important discrepancy in the ratio @xmath50 : the inset of fig .
[ fig : lifetimezoom](c ) shows a ratio of about @xmath51 , which is independent of energy above 0.4ev .
this results compares extremely well with experiment , and should be contrasted with the calculated result of ref . for @xmath52 around 0.5ev .
these gw calculations ( even with a @xmath0-matrix approach ) gave very similar results to those of the random @xmath40 approximation @xcite in the energy range above 0.5ev .
this indicates that the resulting spin asymmetry @xmath538 is solely determined by the spin - dependent dos @xmath54 .
indeed , one has @xmath55 . with the inclusion of spin - orbit coupling , which gives rise to effective spin - flip transitions ,
the spin asymmetry is no longer determined by the spin - dependent dos alone .
this interpretation is again supported by ref . where a strongly enhanced spin - flip matrix element had to be introduced by hand to improve the agreement between a random-@xmath40 calculation and experiment .
( color online ) calculated energy - resolved electronic lifetimes for aluminum ( blue `` @xmath34 '' ) in comparison to earlier investigations without spin - orbit interaction .
the black squares ( stars ) correspond to some data extracted from ref .
( ) . there are in general several different lifetime points at the same energy ( see text ) .
we used @xmath35 @xmath4-points in the full bz.,scaledwidth=45.0% ] to conclude the discussion of the ferromagnets , we comment on the spin - integrated lifetimes which can be obtained from the spin - dependent lifetimes , but are not shown here .
compared with experimental lifetimes of ref .
we generally find an agreement for energies above 0.5ev that is on par with earlier calculations .
@xcite for energies below 0.5ev where the error bars on the averaged lifetimes are largest , the calculated lifetimes are larger than the measured ones , but in this energy range a good agreement with experiments can not be expected because of scattering processes , which appear as elastic due to the energy resolution of the photoemission experiments .
to underscore the importance of spin - orbit coupling for lifetime calculations , we also briefly discuss our calculated results for electronic lifetimes in aluminum in comparison with earlier investigations @xcite without spin - orbit effects .
[ fig : lifetimeal ] shows that smaller electronic lifetimes result for aluminum when spin - orbit coupling is included .
in particular in the energy range between 1 and 3ev the lifetimes differ by almost a factor of two .
thus the inclusion of spin - orbit coupling improves the agreement with experiment ( see , for instance , ref . ) , which was already quite good for the existing calculations .
@xcite it is conceivable that the use of more sophisticated many - body techniques , such as the inclusion of vertex corrections or using a @xmath0-matrix approach , @xcite might lead to further improvements . as in the case of the ferromagnets ,
the electronic band structure is practically unchanged by the inclusion of spin - orbit coupling , but the rather large effect of the spin - orbit coupling on spin relaxation in aluminum through spin hot - spots has already been demonstrated .
@xcite another argument for the importance of the spin - orbit coupling is that the spin - mixing allows transition between the kramers degenerate bands .
these transitions between kramers degenerate bands may have a remarkable influence even on electron - gas properties that are usually assumed to be spin - independent , such as the intraband plasma frequency .
in conclusion , we presented ab - initio results for spin - dependent electronic lifetimes in ferromagnets and aluminum including spin - orbit coupling .
we found that the electronic lifetimes in iron exhibit no visible spin dependence in the range of @xmath56 up to @xmath57 in agreement with earlier results , whereas the ratio @xmath44 between majority and minority lifetimes does not exceed 2 for cobalt and nickel .
our results agree well with experimental data , but differ from earlier calculations , which found that @xmath44 was essentially determined by the spin - dependent density - of - states .
we showed that , by allowing for effectively spin - changing transitions as contributions to the lifetime , spin - orbit coupling is the essential ingredient that can make the spin asymmetry of the electronic lifetimes much smaller than the spin - asymmetry of the density - of - states .
inclusion of our calculated spin dependent lifetimes in transport calculations should make it possible to more accurately characterize the influence of spin - dependent hot - electron transport on magnetization dynamics .
99 b. koopmans , g. malinowski , f. dalla longa , d. steiauf , m. fhnle , t. roth , m. cinchetti , and m. aeschlimann , nature materials * 9 * , 259 ( 2010 ) . g. m. mller , j. walowski , marija djordjevic1 , g .- x .
miao , a. gupta , a. v. ramos , k. gehrke , v. moshnyaga , k. samwer , j. schmalhorst , a. thomas , a. htten , g. reiss , j. s. moodera , and m. mnzenberg , nature materials * 8 * , 56 ( 2009 ) .
b. pfau , s. schaffert , l. mller , c. gutt , a. al - shemmary , f. bttner , r. delaunay , s. dsterer , s. flewett , r. frmter , j. geilhufe , e. guehrs , c.m .
gnther , r. hawaldar , m. hille , n. jaouen , a. kobs , k. li , j. mohanty , h. redlin , w.f .
schlotter , d. stickler , r. treusch , b. vodungbo , m. klui , h.p .
oepen , j. lning , g. grbel , and s. eisebitt , nature comm . * 3 * , 1100 ( 2012 ) .
d. rudolf , c. la - o - vorakiat , m. battiato , r. adam , j. m. shaw , e. turgut , p. maldonado , s. mathias , p. grychtol , h. t. nembach , t. j. silva , m. aeschlimann , h. c. kapteyn , m. m. murnane , c. m. schneider , and p. m. oppeneer , nature comm . * 3 * , 1037 ( 2012 ) .
g. p. zhang , w. hbner , g. lefkidis , y. bai and t. f. george , nature physics * 5 * , 499 ( 2009 ) .
a. eschenlohr , m. battiato , p. maldonado , n. pontius , t. kachel , k. holldack , r. mitzner , a. fhlisch , p. m. oppeneer , and c. stamm , nature mat .
* 12 * , 332 ( 2013 ) .
a. melnikov , i. razdolski , t. o. wehling , e. t. papaioannou , v. roddatis , p. fumagalli , o. aktsipetrov , a. i. lichtenstein , and u. bovensiepen , * 107 * , 076601 ( 2011 ) .
t. kampfrath , m. battiato , p. maldonado , g. eilers , j. ntzold , s. mhrlein , v. zbarsky , f. freimuth , y. mokrousov , s. blgel , m. wolf , i. radu , p.m. oppeneer , and m. mnzenberg , nature nano .
* 8 * , 256 ( 2013 ) .
m. battiato , k. carva , and p. m. oppeneer , * 105 * , 027203 ( 2010 ) .
m. battiato , k. carva , and p. m. oppeneer , * 86 * , 024404 ( 2012 ) . m. aeschlimann , m. bauer , s. pawlik , w. weber , r. burgermeister , d. oberli , and h. c. siegmann , * 79 * , 5158 ( 1997 ) .
a. goris , k. m. dbrich , i. panzer , a. b. schmidt , m. donath , and m. weinelt , * 107 * , 026601 ( 2011 ) .
g. giuliani and g. vignale , _ quantum theory of the electron liquid _ , cambridge university press ( 2005 ) .
g. onida , l. reining , and a. rubio , * 74 * 601 ( 2002 ) .
v. p. zhukov , e. v. chulkov , and p. m. echenique , * 93 * , 096401 ( 2004 ) .
v. p. zhukov , e. v. chulkov , and p. m. echenique , * 73 * , 125105 ( 2006 ) .
r. knorren , k. h. bennemann , r. burgermeister , and m. aeschlimann , * 61 * , 9427 ( 2000 ) .
d. steiauf and m. fhnle , * 79 * , 140401(r ) ( 2009 ) .
b. y. mller , a. baral , s. vollmar , m. cinchetti , m. aeschlimann , h. c. schneider , and b. rethfeld * 111 * , 167204 ( 2013 ) .
m. krau , t. roth , s. alebrand , d. steil . ,
m. cinchetti , m. aeschlimann , and h. c. schneider , * 80 * , 180407(r ) ( 2009 ) .
x. zubizarreta , v. m. silkin , and e. v. chulkov , * 84 * , 115144 ( 2011 ) .
f. ladstdter , u. hohenester , p. puschnig , and c. ambrosch - draxl , * 70 * , 235125 ( 2004 ) .
v. p. zhukov , f. aryasetiawan , e. v. chulkov , i. g. de gurtubay , and p. m. echenique , * 64 * , 195122 ( 2001 ) . s. kaltenborn and h. c. schneider , * 88 * , 045124 ( 2013 ) . the elk fp - lapw code , http://elk.sourceforge.net , version 1.4.18 . v. p. zhukov and e. v. chulkov , physics uspekhi * 52 * , 105 - 136 ( 2009 ) .
d. r. penn , s. p. apell , and s. m. girvin , * 55 * , 518 ( 1985 ) .
d. r. penn , s. p. apell , and s. m. girvin , * 32 * , 7753 ( 1985 ) .
w. f. krolikowski and w. e. spicer , phys .
rev . * 185 * , 882 ( 1969 ) .
v. p. zhukov , e. v. chulkov , and p. m. echenique , * 72 * , 155109 ( 2005 ) .
m. bauer , s. pawlik , and m. aeschlimann , proc .
spie * 3272 * , 201 ( 1998 ) . | for the 3d ferromagnets iron , cobalt and nickel we compute the spin - dependent inelastic electronic lifetimes due to carrier - carrier coulomb interaction including spin - orbit coupling . we find that the spin - dependent density - of - states at the fermi energy does not , in general , determine the spin dependence of the lifetimes because of the effective spin - flip transitions allowed by the spin mixing .
the majority and minority electron lifetimes computed including spin - orbit coupling for these three 3-d ferromagnets do not differ by more than a factor of 2 , and agree with experimental results . |
in the past two decades quantum computation has attracted growing interest , encouraged by the development of tools for the manipulation of single quantum objects as well as by several remarkable theoretical findings @xcite .
different systems have been proposed as candidates for quantum computing ; they are based , for instance , on cavity - laser atoms , bose - einstein condensates , or nmr techniques ( see e.g. @xcite ) . at the same time , a number of quantum algorithms have been designed and some have been shown to be even exponentially faster than their best classical counterparts @xcite . in particular , quantum search algorithms , although able to achieve `` only '' a polynomial speedup , have been proved to be very promising and of widespread use in quantum computation @xcite .
one of the best known quantum search algorithms is attributed to grover @xcite .
the algorithm can find a target within an unsorted database made up of @xmath0 items using @xmath1 queries . due to its broad range of applications and its ability
to be effectively used as a subroutine @xcite , grover s algorithm has been thoroughly investigated and a number of different implementations have been proposed ( see e.g. @xcite ) .
recently , implementations based on continuous - time @xcite as well as on discrete - time quantum walks @xcite have been introduced : given that the application of random walks in classical algorithms provided significant advantages for approximations and optimization , one is strongly motivated to study quantum walks as algorithmic tools . at the current stage
the aim is not only to achieve similar computational improvements , but also to understand the capabilities of quantum computations .
here we focus on the approach pioneered by farhi and gutmann @xcite and further developed by childs and goldstone @xcite , based on continuous - time quantum walks ( ctqws ) .
this implementation provides several important advantages : the algorithm does not need auxiliary storage space and it makes it possible to take into account the geometrical arrangement of the database ( the latter being either a physical position space or an efficiently encoded hilbert space @xcite ) . indeed , while previous studies just considered the cases of the translationally invariant @xmath2 ( named @xmath3-dimensional periodic lattices by childs and goldstone @xcite and called hypercubic lattices in statistical physics , the term which we adopt here ) and of complete graphs @xcite , here we extend the investigations to the case of generic structures and analyze how geometrical parameters [ e.g. , ( fractal ) dimension or ( average ) coordination number ; see later in this article ] affect the dynamics of the ctqw .
in @xcite it was shown that quantum searches based on ctqws recover the optimal quadratic speedup on complete graphs and on high - dimensional hypercubic lattices ( with dimension @xmath4 ) , while for low - dimensional ( @xmath5 ) lattices ctqws can not outperform their classical counterpart .
hence , it may appear that @xmath6 works as a `` critical dimension '' , separating highly performing structures from poorly performing ones .
we will show , both analytically and numerically , that the dimension of the substrate is not sufficient for getting a sharp transition from the ground to the first excited state ; we also take into account the success probability @xmath7 , that is , the probability of finding the quantum walker at the target site @xmath8 at time @xmath9 , given as initial state the equally weighted superposition @xmath10 : an efficient quantum walk gives rise to a success probability close to @xmath11 already at very small times @xmath9 .
in particular , we will take into account several kinds of structures : translationally invariant structures , such as @xmath3-dimensional hypercubic lattices , complete graphs , fractals with low fractal dimension like dual sierpinski gaskets and t - fractals , hierarchical structures as cayley trees , and structures with ( fractal ) dimensions larger than four , such as cartesian products between euclidean lattices and dual sierpinski gaskets ( for the precise definitions of these structures and of the fractal dimension considered here , see later in this article ) . in this way we are able to show that for translationally invariant structures with high dimensions
, @xmath7 displays sharp peaks , while for fractals or low - dimensional structures the peaks are low and broad so that the quantum walk is not particularly effective in the sense that there exists only a low probability @xmath7 for any @xmath9 .
moreover , for any structure , we evidence interference phenomena which give rise to a non - monotonic time dependence of the @xmath7 ; such effects can be significant and must be properly taken into account when considering applications .
interestingly , the ctqw hamiltonian used in the quantum search also describes , in solid - state physics , the dynamics of a tight - binding particle in the presence of static , substitutional impurities . in this context
our results show that in regular , highly - connected geometries ( such as the high - dimensional tori ) the probability of finding the moving particle at the impurity site is ( quasi- ) periodic in time and that the localization can be very effective .
our article is structured as follows . in sec .
[ sec : quantum_search ] we first review basic principles concerning grover s search and we explain how it can be implemented by means of ctqws . in sec .
[ ssec : inho_graphs ] we describe the structures used as substrate for the ctqw .
then , in sec . [ sec : phase_transition ] we present several analytical results , later corroborated and deepened in sec .
[ sec : numerics ] , where our numerical results are shown . in sec .
[ sec : success ] we focus on the success probability .
finally , sec .
[ sec : conclusions ] contains our conclusions and discussion . in appendix
[ appe1 ] , appendix [ appe2 ] and appendix [ appe3 ] we report the details of our analytical calculations .
grover s search algorithm @xcite is meant to solve the unsorted search problem under the assumption that there exists a computational oracle working as a black - box function able to decide whether a candidate solution is the true solution .
hence , the oracle knows which is the target among the @xmath0 entries . the task is to find a target @xmath8 using the fewest calls to the oracle . while the classical algorithm requires exhaustive searches implying @xmath12 queries , grover s algorithm is able to find @xmath8 using @xmath1 queries , giving rise to a quadratic speed up @xcite . a short outline of the idea behind grover s algorithm in the presence of a single marked target is as follows : first of all , one associates each of the @xmath0 index integers with a unique orthonormal vector @xmath13 in an @xmath0-dimensional hilbert space .
then , one chooses as initial state @xmath14 which is delocalized over the entire set of states @xmath15 with equal weights at every site @xmath16 .
this is the least biased initialization one can arrange , given the available information ( since each of the @xmath0 nodes is in principle equally likely to be the target index , the initial state is prepared as an equally weighted superposition of all @xmath0 indices ) .
now , to perform the search , one needs to make the state @xmath17 evolve into a state that has almost all the amplitude concentrated in just the @xmath18 component .
in such a way a single final measurement will find the system in the state @xmath19 , hence revealing the identity of the target index .
more precisely , we need an evolution operator whose repeated application makes the amplitude of @xmath18 grow with the number of iterations @xcite .
this task was originally accomplished within the standard paradigm for quantum computation @xcite , namely using a discrete sequence of unitary logic gates , while in the last years several different implementations have been introduced in which the state of the quantum register evolves continuously under the influence of a driving hamiltonian @xcite . in particular , here we follow the approach developed by childs and goldstone @xcite which relies on ctqw @xcite . as already mentioned , due to their versatility , ctqws allow to model any discrete database .
in fact , a generic discretizable database can be represented by a graph @xmath20 made up of a set of nodes @xmath21 , each corresponding to a different item , and of a set of links @xmath22 joining nodes pairwise in such a way that the topology of the graph mirrors the arrangement of the database .
the graph @xmath23 can be algebraically described by the adjacency matrix @xmath24 whose entry @xmath25 equals one if nodes @xmath26 and @xmath27 are connected , otherwise it is zero ( also for the diagonal elements ) . from @xmath24 one
can directly calculate the degree matrix @xmath28 , which is a diagonal matrix with elements @xmath29 , where @xmath30 is the coordination number ( or degree ) of the @xmath26-th node , that is , the number of its nearest neighbors . ctqws on a graph are defined by the laplacian matrix @xmath31 and obey the following schrdinger equation for the transition amplitude @xmath32 from state @xmath33 to state @xmath34 @xcite : @xmath35 where the hamiltonian is just given by @xmath36 , therefore , the time is dimensionless and given in units of the coupling elements @xmath37 .
the formal solution for @xmath32 can be written as @xmath38 and its magnitude squared provides the quantum mechanical transition probability @xmath39 .
notice that @xmath40 is symmetric and non - negative definite ; its ground state , corresponding to eigenvalue @xmath41 , is given by @xmath42 .
due to the unitary time - evolution generated by @xmath43 , ctqws are symmetric under time - inversion , which precludes @xmath44 from attaining equipartition at long times .
this is different from the behavior of classical continuous - time random walks .
moreover , ctqws keep memory of the initial conditions , as exemplified by the occurrence of ( quasi- ) revivals @xcite .
now , the `` oracle hamiltonian '' is given by @xmath45 whose ground state , with energy @xmath46 , is just the marked item @xmath47 ; all other states have energy zero @xcite .
then , the hamiltonian @xmath43 governing the time evolution of the quantum walk is @xmath48 where @xmath49 is a tunable parameter with units of inverse time , hence dimensionless here . the _
success probability _ @xmath50 is defined as ( see also @xcite ) @xmath51 i.e. , as the transtion probability to be at time @xmath9 at the target @xmath8 when starting in the delocalized state @xmath52 . here
we study its dependence on time @xmath9 and on the parameter @xmath49 and we especially aim to evidence the existence of an optimization parameter @xmath53 , possibly depending on time , which maximizes @xmath50 .
notice that , having fixed @xmath49 , the time @xmath9 can be measured in terms of the number of queries of the discrete grover oracle @xcite .
before proceeding , it is worth introducing the geometrical structures on which we are focusing , first the dual sierpinski gasket ( dsg ) , the t - fractal ( tf ) , and the cayley tree ( ct ) ; examples of these structures are shown in fig .
[ fig : frattali]a , fig .
[ fig : frattali]b , and fig .
[ fig : frattali]c , respectively . notice that these structures differ significantly from those analyzed previously : while hypercubic lattices with periodic boundary conditions ( toroids ) are translationally invariant , the aforementioned structures are not . the dsg , tf and ct can be built iteratively ; at the @xmath54-th iteration we have the fractal of generation @xmath54 ( see e.g. @xcite ) .
the dsg and the tf are examples of exactly decimable fractals for which the fractal dimension , @xmath55 , and the spectral dimension , @xmath56 , are exactly known . while the fractal dimension relates the number of nodes inside a sphere to the radius of the sphere @xcite , the spectral dimension is obtained from the scaling of the eigenmodes of a given structure ( phonons for lattices , fractons for fractal substrates ) @xcite , most simply seen in the probability of a random walker to be ( still or again ) at the original site @xcite . here , we have namely @xmath57 and @xmath58 for the dsg , and @xmath59 and @xmath60 for the tf .
we recall that for the usual , translationally invariant lattices the spectral and fractal dimensions coincide with the euclidean dimension @xmath3 , namely @xmath61 for the chain , @xmath62 for the square lattice and so on . on the other hand , on fractal structures @xmath56 often replaces @xmath3 when dealing with dynamical and thermodynamical properties @xcite . also , for fractals @xmath63 and @xmath55 is smaller than the euclidean dimension of the embedding space , that is , 2 for dsg and tf .
the ct is no fractal in the classical sense , since its growth with increasing generation is exponential .
however , the ct is built in a hierarchical manner , analogous to the tf .
we also notice that both , the tf and the ct , are trees ( and are hence devoid of loops ) and exhibit a large number @xmath64 of end nodes .
examples of fractal structures considered here ; in general the star indicates the position of the trap .
panel @xmath65 : dual sierpinski gasket of generation @xmath66 and volume @xmath67 ; due to symmetry , the three corners are equivalent .
panel @xmath68 : t - fractal of generation @xmath69 and volume @xmath70 ; due to the symmetry all outmost sites ( at distance @xmath71 from the central node ) are equivalent .
panel @xmath72 : cayley tree of generation @xmath73 and volume @xmath74 ; all of the @xmath75 outmost sites are equivalent.,title="fig:",width=207,height=181 ] examples of fractal structures considered here ; in general the star indicates the position of the trap .
panel @xmath65 : dual sierpinski gasket of generation @xmath66 and volume @xmath67 ; due to symmetry , the three corners are equivalent .
panel @xmath68 : t - fractal of generation @xmath69 and volume @xmath70 ; due to the symmetry all outmost sites ( at distance @xmath71 from the central node ) are equivalent .
panel @xmath72 : cayley tree of generation @xmath73 and volume @xmath74 ; all of the @xmath75 outmost sites are equivalent.,title="fig:",width=207,height=181 ] examples of fractal structures considered here ; in general the star indicates the position of the trap .
panel @xmath65 : dual sierpinski gasket of generation @xmath66 and volume @xmath67 ; due to symmetry , the three corners are equivalent .
panel @xmath68 : t - fractal of generation @xmath69 and volume @xmath70 ; due to the symmetry all outmost sites ( at distance @xmath71 from the central node ) are equivalent .
panel @xmath72 : cayley tree of generation @xmath73 and volume @xmath74 ; all of the @xmath75 outmost sites are equivalent.,title="fig:",width=207,height=181 ] as we will show in the following , ctqw on the dsg , the tf and the ct can display a low probability @xmath76 for any @xmath9 and @xmath49 when compared with the case of the translationally invariant lattices , especially when the dimension of the lattice becomes larger than @xmath66 . therefore one can ask whether the probabilities @xmath50 can be improved by adopting ( fractal or hierarchical ) substrates which display a large spectral dimension and a large coordination number . for instance
, we can build up such structures by combining the dsg and euclidean lattices by means of cartesian products .
in general , the cartesian product of two graphs @xmath77 and @xmath78 is a graph @xmath79 with the vertex set @xmath80 , and such that two nodes @xmath81 are adjacent if @xmath82 and @xmath83 , or @xmath84 and @xmath85 .
it has been shown @xcite that the spectral dimension @xmath56 on the product graph @xmath23 is then the sum of the corresponding dimensions of the two initial graphs @xmath86 and @xmath87 .
we combine in this way the dsg with the chain @xmath88 , with the square lattice @xmath89 and with the cubic lattice @xmath90 , to obtain more complex structures displaying spectral dimensions approximately equal to @xmath91 , @xmath92 and @xmath93 , respectively .
finally , it should be underlined that , dealing with such structures , the location of the target , i.e. the node @xmath8 , also ( quantitatively ) matters . in the numerical analysis of sec .
[ sec : numerics ] we place the target on a peripheral site , which means , without loss of generality , the apex for the dsg , the leftmost site for the tf and an outmost site for the ct ( see fig.[fig : frattali ] ) .
we expect that a peripheral position for the target site does not correspond to an optimal situation and , since a priori the target position is unknown , this choice prevents an overestimation of the probability of finding the target .
let us consider the hamiltonian of eq .
( [ eq : hamiltonian ] ) and denote the corresponding set of eigenstates and eigenvalues by @xmath94 and @xmath95 , respectively . now ,
for large @xmath49 the contribution of @xmath96 to @xmath43 is negligible and the ground state @xmath97 is close to @xmath52 . on the other hand , as @xmath98
the ground state is close to @xmath99 since the weight of @xmath40 is small and , from perturbation theory , we expect that @xmath42 is close to @xmath100 , i.e. to the first excited state of @xmath43 @xcite .
as pointed out by childs and goldstone @xcite , there exists an intermediate range of @xmath49 where , for complete graphs and hypercubic lattices with dimension @xmath4 , @xmath17 switches from the ground state @xmath101 to the first excited state @xmath100 ; in the very same region of @xmath49 the target state @xmath18 switches from a state with large overlap with @xmath100 to the ground state @xmath102 .
therefore , by varying @xmath49 , we can find a particular value for @xmath49 for which the hamiltonian @xmath43 can evolve the state @xmath17 into a state close to the target state @xmath19 .
the existence and the narrowness of such a range for @xmath49 are crucial for a large success probability .
as anticipated , the ctqw under study can yield effective results if the hamiltonian @xmath43 is able to rotate the state @xmath42 to a state with a large overlap with @xmath18 . for this to occur a first condition which needs to be fulfilled
is that there exists an intermediate range of @xmath49 over which the state @xmath42 has a substantial overlap with both @xmath102 and @xmath100 . in this subsection
the occurrence of such a condition is investigated analytically for an arbitrary structure , following arguments similar to those exploited in @xcite for @xmath3-dimensional hypercubic lattices ; the details of the calculations are given in appendixes a and b while here we just report the main results .
first of all , we define @xmath103 and @xmath104 , the @xmath105-th eigenstate and eigenvalue of the laplacian @xmath40 , respectively . in the basis of the eigenstates
@xmath103 the target site @xmath19 can be written as @xmath106 where @xmath107 . as derived in appendix [ appe1 ] , once latexmath:[$\xi_j \equiv \sum_{k \neq 0 } @xmath42 with the ground state @xmath102 or with
the first non - degenerate excited state @xmath100 is ( depending on @xmath49 ) limited from below by the same bound .
we find namely @xmath109 , and also @xmath110 for @xmath111 .
now , we define the ( size dependent ) critical value @xmath112 as that value of @xmath49 for which latexmath:[\[\label{eq : intersection } satisfied , i.e. , the @xmath49-value for which the projection of state @xmath10 onto the ground and the first excited state has the same magnitude .
this is particularly interesting in the limit @xmath114 , when a level crossing from state @xmath102 to state @xmath100 can occur .
in fact , we find that if @xmath115 $ ] converges ( at least in the limit @xmath116 ) to @xmath41 as @xmath49 approaches @xmath117 ( from different sides ) , then @xmath118 and @xmath119 both approach @xmath11 ( see eqs .
( [ eq : s_0_short ] ) and ( [ eq : s_l_short ] ) ) , namely a transition occurs at @xmath120 . notice , however , that the condition for this to occur is in general non - trivial as @xmath117 and @xmath121 both depend on @xmath0 . in appendix [ appe1 ]
we find a sufficient condition in the laplacian spectrum and , in particular , in appendix [ appe2 ] we prove that such a condition holds for the dsg for which @xmath104 is exactly known @xcite ; in this case we find that @xmath122 and @xmath123 where @xmath124 is a parameter depending on the given network . in general ,
@xmath125 ; for hypercubic lattices
@xmath126 , regardless of their dimension , while for the dsg we can numerically estimate @xmath124 as being @xmath127 ( see appendix [ appe1 ] ) .
therefore , we expect ( for dsg ) @xmath112 to be , approximately : @xmath128 this result is consistent up to logarithmic corrections with the critical points found in @xcite for the linear chain and the square lattice , namely @xmath129 and @xmath130 , respectively . finally , we point out that the critical parameter @xmath131 provides interesting information in the context of quantum adiabatic computation @xcite : @xmath131 represents a threshold below which we can expect @xmath18 to have a large overlap with the ground state .
we now consider the three examples of low - dimensional inhomogeneous structures described previously , for which the overlaps of the initial state @xmath17 and of the target state @xmath19 with @xmath132 and @xmath133 are shown in figs .
[ fig : ds_ext]-[fig : ct_ext ] for different generations @xmath54 , as a function of the parameter @xmath49 .
these plots evidence that there exists an intermediate range of @xmath49 where the state @xmath17 changes from having a large overlap with the first excited state to having a large overlap with the ground state . in the same region of @xmath49
the overlap @xmath134 is significant for structures of small size ( top panels in figs .
[ fig : ds_ext]-[fig : ct_ext ] ) , while it is still very small when the size is large ( bottom panels in figs .
[ fig : ds_ext]-[fig : ct_ext ] ) .
this is vastly distinct from the situation found for hypercubic lattices @xcite where close to @xmath131 the overlap @xmath134 is significant .
( color online ) overlaps for a dsg of generation @xmath135 ( up ) and @xmath136 ( bottom ) as a function of ( the dimensionless ) @xmath49 , see text for details.,title="fig:",width=302,height=264 ] + ( color online ) overlaps for a dsg of generation @xmath135 ( up ) and @xmath136 ( bottom ) as a function of ( the dimensionless ) @xmath49 , see text for details.,title="fig:",width=302,height=264 ] ( color online ) overlaps for a t - fractal of generation @xmath135 ( up ) and @xmath136 ( bottom ) as a function of @xmath49 .
the symbols are as in fig .
[ fig : ds_ext].,title="fig:",width=302,height=264 ] + ( color online ) overlaps for a t - fractal of generation @xmath135 ( up ) and @xmath136 ( bottom ) as a function of @xmath49 .
the symbols are as in fig .
[ fig : ds_ext].,title="fig:",width=302,height=264 ] ( color online ) overlaps for a ct of generation @xmath135 ( up ) and @xmath137 ( bottom ) as a function of @xmath49 .
the symbols are as in fig .
[ fig : ds_ext].,title="fig:",width=302,height=264 ] + ( color online ) overlaps for a ct of generation @xmath135 ( up ) and @xmath137 ( bottom ) as a function of @xmath49 .
the symbols are as in fig .
[ fig : ds_ext].,title="fig:",width=302,height=264 ] by comparing the plots obtained for the dsg ( fig .
[ fig : ds_ext ] ) , the tf ( fig .
[ fig : t_ext ] ) , and the ct ( fig .
[ fig : ct_ext ] ) , we notice that @xmath131 depends sensitively on the underlying topology . in fact
, going from dsg to ct and then to tf we notice an amplification of the transition region , which is for tf most spread out , requiring relatively large values of @xmath49 in order for @xmath42 to have a large overlap with the ground state .
such effects can be ascribed to the absence of loops and to the existence for tf of a large number of peripheral sites scattered throughout the whole structure which give rise to localization effects @xcite .
we now calculate @xmath131 according to eq .
( [ eq : intersection ] ) and for several values of @xmath54 ; numerical data and relative best fits are depicted in fig .
[ fig : gamma_crit ] . for the dsg ,
numerical points are best fitted by the function @xmath138 , in very good agreement with the analytical findings .
in fact , according to our analytical investigations , @xmath131 scales with the size of the database like @xmath139 ( see eq .
( [ eq : gammacrit_short ] ) ) , where , for the dsg , @xmath140 ( @xmath124 is taken to equal @xmath141 ) .
let us now consider the case of the tf : from numerical data the best fit turns out to be @xmath142 .
interestingly , this result is still in very good agreement with the analytical approximation of eq .
( [ eq : gammacrit_short ] ) .
in fact , for the tf the exponent gets @xmath143 ( where , again , @xmath124 is taken to equal @xmath141 , consistently with the estimates given in appendix [ appe1 ] ) .
such a consistency might suggest that eq .
( [ eq : gammacrit_short ] ) is valid not only for the dsg but for any ( exactly decimable ) fractal with @xmath144 . as for the ct , not being a fractal , eq .
( [ eq : gammacrit_short ] ) does not hold . indeed , we find that the value of @xmath131 corresponding to eq .
( [ eq : intersection ] ) grows linearly with the generation of the fractal , namely logarithmically with @xmath0 . in fig .
[ fig : gamma_crit ] numerical data are fitted by the curve @xmath145 ( notice the semilogarithmic plot ) .
( color online ) estimate of @xmath131 for dsg ( @xmath146 ) , tf ( @xmath147 ) and ct ( o ) in a semi - logarithmic scale .
the continuous lines represent the best fits.,width=302,height=264 ] apart from this , the plots shown in figs . [ fig : ds_ext]-[fig : ct_ext ] look rather similar . in particular , for networks of large sizes @xmath148 decays with @xmath49 more rapidly than @xmath149 . hence , the range of @xmath49 over which the transition occurs is wide , analogously to what happens on low - dimensional hypercubic lattices ( see @xcite ) .
as shown in the next section , this has important effects on the behaviour of the success probability and suggests that a sharp transition is associated with an effective search algorithm . in order to sketch the role of the position of a target we show in the inset of fig .
[ fig : ct_ext ] for the ct of @xmath137 the case of a target placed on a nearest neighbor of the central node .
we see that the transition region is shifted towards lower values of @xmath49 .
this means that a more central placement of the target is improving the probability for the ctqw to reach the target .
analogous results were also found for dsg and for tf .
thus , limiting our focus to peripheral nodes will prevent us from overestimating the success probabilities .
we now consider fractal structures exhibiting large spectral dimension ; in particular , we focus on fractals obtained from cartesian products , such as dsg @xmath150 @xmath88 , dsg @xmath150 @xmath89 and dsg @xmath150 @xmath90 , as introduced in the previous section . again
, we place the target on a `` peripheral site '' , i.e. on a minimally connected site ; this displays the coordination number @xmath73 , @xmath151 and @xmath152 , for dsg @xmath150 @xmath88 , dsg @xmath150 @xmath89 and dsg @xmath150 @xmath90 , respectively .
we calculate for these structures the overlaps of the initial state @xmath17 and of the target state @xmath18 with @xmath133 and @xmath132 ; results for dsg @xmath150 @xmath89 ( @xmath153 ) and for dsg @xmath150 @xmath90 ( @xmath154 ) are shown in fig .
@xmath155 .
[ fig : cartesian ] @xmath89 structure given by the cartesian product of a dsg of generation @xmath156 and a square lattice of size @xmath157 ; bottom panel : overlaps for a dsg @xmath150 @xmath90 obtained from a dsg of generation @xmath156 and a cubic lattice of size @xmath158 .
the lines are as in fig . [
fig : ds_ext].,title="fig:",height=264 ] @xmath89 structure given by the cartesian product of a dsg of generation @xmath156 and a square lattice of size @xmath157 ; bottom panel : overlaps for a dsg @xmath150 @xmath90 obtained from a dsg of generation @xmath156 and a cubic lattice of size @xmath158 .
the lines are as in fig .
[ fig : ds_ext].,title="fig:",height=264 ] as stressed at the beginning of this section , these plots provide some information about the sharpness of the transition from state @xmath42 to the state @xmath133 and from state @xmath159 to the state @xmath133 .
however , around @xmath131 also a significantly large overlap @xmath160 is required . here , the transitions are still rather smooth although , by increasing @xmath56 , the region of @xmath49 values , over which the curves representing @xmath161 , @xmath162 and @xmath163 intersect , is shrinking . on the other hand , the overlaps between @xmath18 and the first excited state @xmath100 are negligible for all values of @xmath49 .
we now turn to the success probability @xmath50 , eq . ( [ eq : success_prob ] ) , and we investigate numerically its dependence on @xmath9 and on @xmath49 . because of its dependence on time , @xmath50 carries more information than the previously discussed overlaps .
we first analyze the case of complete graphs and of @xmath3-dimensional hypercubic lattices ( for which the time dependences of @xmath50 have already been determined for special choices of @xmath49 in ref .
@xcite ) before turning to the dsg , the tf and the ct .
( color online ) contour plot of the success probability @xmath164 as a function of ( the dimensionless ) time @xmath9 and of @xmath49 for the complete graph of size @xmath165 ( top ) and @xmath166 ( bottom ) .
one can notice that for @xmath167 , namely @xmath168 ( top ) and @xmath169 ( bottom ) , @xmath50 has a period @xmath170 and @xmath171 , respectively.,title="fig:",width=321,height=264 ] ( color online ) contour plot of the success probability @xmath164 as a function of ( the dimensionless ) time @xmath9 and of @xmath49 for the complete graph of size @xmath165 ( top ) and @xmath166 ( bottom ) .
one can notice that for @xmath167 , namely @xmath168 ( top ) and @xmath169 ( bottom ) , @xmath50 has a period @xmath170 and @xmath171 , respectively.,title="fig:",width=321,height=264 ] we start our analysis from the complete graph @xmath172 for which , as shown in @xcite , at @xmath173 the ground state changes sharply from @xmath52 to @xmath99 .
this transition takes place at @xmath174 . due to the special topology of @xmath172
, we are able to calculate @xmath50 exactly , obtaining @xmath175,\ ] ] see appendix [ appe3 ] for details . in fig .
[ fig : complete ] we show @xmath164 for the complete graph with @xmath165 and @xmath166 .
we evaluated the figures both numerically , by first diagonalizing @xmath43 in eq .
( [ eq : success_prob ] ) and projecting on the states @xmath18 and @xmath42 and also by making use of eq .
( [ eq : success_completo ] ) .
the results are numerically indistinguishable . from fig .
[ fig : complete ] we see that , around the values @xmath176 and @xmath177 for @xmath165 and @xmath166 respectively , @xmath50 reaches values very close to @xmath11 .
in fact , analyzing eq .
( [ eq : success_completo ] ) ( see appendix [ appe3 ] ) , one finds that @xmath50 attains its maximal value of @xmath11 for @xmath167 and for @xmath178 , where @xmath179 .
therefore , @xmath180 .
furthermore , due to the fact that the period between maxima is @xmath181 for @xmath182 , it follows that the ctqw takes @xmath1 queries to find the target , in agreement with previous results @xcite . on the other hand ,
the exact dependence on @xmath9 and on @xmath49 also allows to highlight the oscillating behaviour of @xmath50 .
this means that , although we properly select @xmath183 , the result for the walk depends sensitively also on @xmath9 .
in particular , for @xmath184 , the success probability is minimal and equals @xmath185 ( which also corresponds to the absolute minimum ) .
for the hypercubic lattices the overlaps of the states @xmath52 and @xmath99 with @xmath133 and @xmath132 have already been analyzed in @xcite ; here we display analogous plots ( but for larger sizes ) in order to compare them with the corresponding success probability @xmath50 .
( color online ) @xmath73-dimensional torus with linear size @xmath186 ; top : overlaps ( symbols are as in fig . [
fig : ds_ext ] ) ; bottom : contour plot of the success probability @xmath164 as a function of @xmath9 and of @xmath49 ; the dashed white line represents @xmath187.,title="fig:",width=302,height=264 ] ( color online ) @xmath73-dimensional torus with linear size @xmath186 ; top : overlaps ( symbols are as in fig .
[ fig : ds_ext ] ) ; bottom : contour plot of the success probability @xmath164 as a function of @xmath9 and of @xmath49 ; the dashed white line represents @xmath187.,title="fig:",width=321,height=264 ] in fig .
[ fig : torus5 ] we consider the case of a @xmath73-dimensional torus ( i.e. a five - dimensional cubic lattice with periodic boundary conditions ) of linear size @xmath188 : the transition at @xmath189 is very clear ( see the top panel ) and the success probability is sharply peaked just at @xmath190 ( see the bottom panel ) .
away of the critical point @xmath131 the success probability quickly decays as a function of @xmath49 : it is just in the region of largest overlap between the initial state @xmath42 and the target state @xmath99 ( namely around @xmath131 ) that one expects an optimal success probability .
again , we notice that @xmath50 oscillates in time ; for @xmath186 and @xmath191 the success probability ranges from about @xmath41 to about @xmath192 .
moreover , for a given time @xmath9 , @xmath50 decays very fast as @xmath193 increases .
for instance , @xmath194 . as a result ,
the computational procedure for this structure can be very efficient , provided that the parameter @xmath49 can be sensitively controlled .
( color online ) @xmath195-dimensional square torus of linear size @xmath196 .
top : overlaps ( symbols are as in fig . [
fig : ds_ext ] ) ; bottom : success probability as a function of time and @xmath49 ; the dashed white line represents @xmath197.,title="fig:",width=302,height=264 ] ( color online ) @xmath195-dimensional square torus of linear size @xmath196 .
top : overlaps ( symbols are as in fig .
[ fig : ds_ext ] ) ; bottom : success probability as a function of time and @xmath49 ; the dashed white line represents @xmath197.,title="fig:",width=321,height=264 ] for hypercubic lattices of dimension @xmath6 , @xmath198 and @xmath199 ( only the latter case is depicted in fig .
[ fig : torus2 ] ) the peaks get more and more broadened and are of smaller magnitude .
notice that the low peaks obtained are in agreement with the analytical results found in @xcite which predict for the @xmath195-dimensional torus a vanishing success probability for @xmath200 .
we now turn to the analysis of the success probability for the dsg , the tf and the ct , represented in figures [ fig : ampl_ds ] , [ fig : ampl_t ] and @xmath201 , respectively ; the upper panels display the situation for small , the lower panels for larger networks .
first of all , we notice that , as previously found for regular lattices @xcite , also for dgs , tf and ct , @xmath50 exhibits peaks which are lower and lower as the size @xmath0 is enlarged .
moreover , for small sizes ( upper panels ) @xmath202 , namely the maxima for the success probability occur for values of @xmath49 which are approximately equal to @xmath131 . on the other hand , for large sizes ( lower panels ) , in the temporal range considered here
, @xmath131 provides an upper bound for @xmath182 .
however , the most striking feature which emerges from the comparison between the contour plot of the success probability for translationally invariant structures ( see figs . [
fig : torus5 ] and [ fig : torus2 ] ) and for fractal / hierarchical structures ( see figs . [
fig : ampl_ds]-@xmath201 ) is that for the latter @xmath7 is much more rough and broadened .
( color online ) contour plot of @xmath50 for the dsg of generation @xmath135 ( top ) and @xmath136 ( bottom ) ; the `` critical '' values @xmath131 are represented by the dashed line : @xmath203 and @xmath204 ( see fig . [
fig : ds_ext]).,title="fig:",width=321,height=264 ] ( color online ) contour plot of @xmath50 for the dsg of generation @xmath135 ( top ) and @xmath136 ( bottom ) ; the `` critical '' values @xmath131 are represented by the dashed line : @xmath203 and @xmath204 ( see fig .
[ fig : ds_ext]).,title="fig:",width=321,height=264 ] ( color online ) contour plot of @xmath50 for the tf of generation @xmath135 ( top ) and @xmath136 ( bottom ) ; notice that @xmath205 and @xmath206 ( see fig .
[ fig : t_ext]).,title="fig:",width=321,height=264 ] ( color online ) contour plot of @xmath50 for the tf of generation @xmath135 ( top ) and @xmath136 ( bottom ) ; notice that @xmath205 and @xmath206 ( see fig . [ fig : t_ext]).,title="fig:",width=321,height=264 ] [ fig : ampl_ct ] for the ct of generation @xmath135 ( top ) and @xmath137 ( bottom ) ; notice that @xmath207 and @xmath208 ( see fig . [ fig : ct_ext]).,title="fig:",width=321,height=264 ] for the ct of generation @xmath135 ( top ) and @xmath137 ( bottom ) ; notice that @xmath207 and @xmath208 ( see fig . [ fig : ct_ext]).,title="fig:",width=321,height=264 ] now , it is worth comparing the @xmath73-dimensional torus of fig .
[ fig : torus5 ] , the square torus of fig .
[ fig : torus2 ] and the ct of fig .
@xmath201 since all three are of comparable size @xmath0 . from the computational point of view
the @xmath73-dimensional torus corresponds to the best situation : @xmath50 is sharp and reaches its maximum value around @xmath192 after approximately @xmath209 unit steps ; the ct displays a success probability around @xmath210 after @xmath211 time steps ; the @xmath195-dimensional torus corresponds to an ineffective candidate situation : after @xmath212 time steps the peak is still less than @xmath213 .
in order to compare also to structures with spectral dimensions larger than four , we show in fig .
@xmath214 the success probabilities @xmath50 for the cartesian products of a dsg of generation @xmath135 and a square lattice of size @xmath157 as well as a cubic lattice of size @xmath157 . although the maxima of the success probabilities are in both cases larger than the ones for the structures with low dimensions , they show still a fairly unregular pattern .
this is in contrast to the highly regular structure of the 5-dimensional torus , see fig .
[ fig : torus5 ] .
[ fig : ampl_cp_dsgl ] for the cartesian product of a dsg of generation @xmath135 and a square lattice of size @xmath157 ( top ) and a cubic lattice of size @xmath157 ( bottom).,title="fig:",width=321,height=264 ] for the cartesian product of a dsg of generation @xmath135 and a square lattice of size @xmath157 ( top ) and a cubic lattice of size @xmath157 ( bottom).,title="fig:",width=321,height=264 ] finally , we stress that , as a result of interference phenomena , @xmath50 oscillates with time .
this has some important consequences : although we can determine and set the optimal @xmath182 , the probability of the ctqw reaching the target depends on the instant of time at which it is calculated .
in particular , oscillations are `` faster '' for systems of smaller size ; for example for the complete graph we find a period @xmath215 ( see eq .
( [ eq : success_completo ] ) ) , namely @xmath216 , while for the @xmath73-dimensional torus the numerical analysis makes it possible to estimate a period @xmath217 which grows exponentially with the lattice size @xmath218 ; for @xmath186 we get @xmath219 ( see fig . [
fig : torus5 ] ) .
in this work we considered ctqws mimicking grover s quantum search problem ; we especially focused on how the topology of the space over which the walk takes place affects the position and sharpness of the transition of the ground state .
previous studies @xcite highlighted that for translationally invariant structures , such as the hypercubic lattices , the quantum walk can be highly efficient for sufficiently high dimensions , i.e. @xmath4 .
however , here we evidence that on generic graphs the dimension does not represent the key geometric parameter ; indeed , both the ( average ) coordination number and the fact that the structure is translationally invariant or not determine the sharpness of the transition .
in fact , on the one hand , a high coordination reduces the distance among nodes and increases the possibility of interference effects , on the other hand , ( in the absence of a target ) translational invariance prevents the emergence of localization effects @xcite . in particular , we considered the success probability @xmath220 ( here @xmath52 and @xmath99 are the initial and the target state , respectively ) as a function of the computation time @xmath9 and of a properly tunable parameter @xmath49 entering the hamiltonian .
we showed that for highly dimensional ( @xmath4 ) translationally invariant structures ( @xmath73-dimensional torus ) there exists a narrow range of @xmath49 around a specific value @xmath131 where the ground state @xmath102 undergoes a transition from having a large overlap with @xmath52 to a state with a large overlap with @xmath99 .
this corresponds to a sharply peaked success probability : @xmath220 displays a set of maxima just at @xmath131 ; the first one is reached after @xmath1 queries .
conversely , for structures with low coordination number and/or fractal or hierarchical topology - such as the cubic ( @xmath90 ) and square ( @xmath89 ) tori , the dsg , the tf , the ct and the cartesian products dsg @xmath150 @xmath89 and dsg @xmath150 @xmath90 - the transition from the initial state to the target state takes place over a wider region of @xmath49 around the value @xmath131 . as a consequence of such a spread - out transition , the success probability displays broadened peaks whose locations depend on time .
therefore , for the non - translationally invariant structures considered here , even with large fractal dimension ( @xmath221 ) , the large success probabilities found for high - dimensional periodic lattices are not recovered .
these results , in agreement with previous findings @xcite , highlight a possible connection between the sharpness of the transition occurring at @xmath131 and the efficiency of the search algorithm . a mathematical , rigorous proof stating whether a sharp transition is a necessary condition for a good algorithm , which is beyond the aim of this article , could provide a very useful tool for further investigations on quantum search algorithms . for the dsg we also proved that @xmath131 scales like @xmath222 ( @xmath125 ) ; interestingly our results suggest that such a scaling might be generalized to all ( exactly decimable ) fractals with spectral dimension @xmath223 . apart from the deterministic fractals considered here
, it will be extremely interesting to also consider disordered structures such as percolation clusters and random graphs characterized by a degree distribution @xmath224 .
these networks display a tunable average degree @xmath225 , which , in principle , can take values ranging from @xmath41 for totally disconnected networks up to @xmath0 for completely connected networks .
according to the results discussed here , we expect that for a sufficiently large @xmath225 and for a sufficiently peaked @xmath224 the transition from the ground state @xmath17 to a state with significant overlap with @xmath19 occurs sharply and that , consequently , the ctqw speeds up .
support from the deutsche forschungsgemeinschaft ( dfg ) and the fonds der chemischen industrie is gratefully acknowledged .
ea thanks the italian foundation `` angelo della riccia '' for financial support .
let us denote by @xmath226 and by @xmath227 the sets of eigenstates and of eigenvalues of the laplacian @xmath40 , respectively .
on the basis of the eigenstates the state @xmath19 localized at the target site can be written @xmath228 now , let us consider the complete hamiltonian @xmath43 and the corresponding eigenvalue equation for the state labeled @xmath65 : @xmath229 as shown in @xcite , when one sets @xmath230 , it is possible to write @xmath231 here we notice that eq .
[ eq : eigenvalue1 ] , and therefore also eq .
[ eq : normalize ] , hold when the eigenvalue @xmath232 is _ non - degenerate _
( for example , this condition is not fulfilled if we place the trap on the central node of the tf and ct ) .
equations ( [ eq : eigenvalue1 ] ) and ( [ eq : normalize ] ) allow then to express the overlap of a given state @xmath233 ( with non - degenerate @xmath232 ) with @xmath234 as @xmath235 where @xmath236 and , in particular , @xmath237 ( see @xcite for more details ) .
it is worth underlining that eq .
( [ eq : overlap ] ) found in @xcite for hypercubic lattices actually holds for all structures .
the topological details are contained in @xmath232 and @xmath238 .
we now proceed with the calculations without making any assumptions on the topology of the database .
first , by using eq .
( [ eq : expansion ] ) and @xmath239 , we rewrite @xmath240 as @xmath241 whose derivative is @xmath242 for a generic index @xmath105 , @xmath243 ; while the lower bound is clear , the upper bound derives from the fact that the eigenstate @xmath99 can not correspond to any laplacian eigenstate . for euclidean lattices bloch s theorem makes it possible to write @xmath244 so that @xmath245 , for any @xmath105 .
on the other hand , for a generic connected structure , a priori , one can only write @xmath246 , as a consequence of the fact that the laplacian eigenstate corresponding to the smallest eigenvalue @xmath247 is just @xmath248 .
this suggests a proper restriction of the previous upper bound : @xmath249 , with @xmath250 and @xmath125 , both depending on the particular topology chosen .
in particular , for euclidean structures , @xmath251 and @xmath126 and one expects that the more inhomogeneous the topology , the larger @xmath124 . by means of numerical calculations
we can estimate @xmath124 : for not too small dsg and tf we find that @xmath252 ( see fig .
@xmath253 ) .
[ fig : alfa ] for the dsg and the tf.,title="fig:",height=226 ] now , before going on , we define the quantity @xmath254 , which will be useful in the following : @xmath255^j } \leq c n^{\alpha } \sum_{k \neq 0 } \frac{1}{[\mathcal{e}(k)]^j } \equiv c n^{\alpha } \zeta_j,\ ] ] where we set @xmath256^j \equiv \zeta_j$ ] and used the upper bound @xmath257 ; for hypercubic lattices @xmath258 which can be approximated by an integral @xcite . using eqs .
( [ eq : overlap ] ) and ( [ eq : f_deriv_gen ] ) the overlap of @xmath259 with the ground state turns out to be @xmath260^{-1 } \\ \nonumber & > & 1 - n e_0 ^ 2 \sum_{k \neq 0 } \frac{|a_k|^{2}}{(\gamma \mathcal{e}(k ) + |e_0|)^2 } \\ &
> & 1 - \frac{n e_0 ^ 2}{\gamma^2 } \sum_{k \neq 0 } \frac{|a_k|^2}{[\mathcal{e}(k)]^2 } , \end{aligned}\ ] ] where in the first inequality we used that the sum is positive , while in second inequality we used that both @xmath104 and @xmath261 are positive . from eq .
( [ eq : s_0_preli ] ) and eq .
( [ eq : xi_j ] ) , we have @xmath262 in general , as @xmath49 is varied , @xmath263 is bounded as @xmath264 and the bounds can be improved by exploiting the following @xmath265 where for the lower bound we used that the first sum appearing in eq .
( [ eq : e_0 ] ) is positive and that @xmath266 .
now , from eq .
( [ eq : s_0 ] ) and eq .
( [ eq : bound_e0 ] ) it follows straightforwardly that @xmath267 following analogous arguments we find for the first non - degenerate excited state labelled as @xmath11 , being @xmath268 non degenerate , @xmath269 and @xmath270 from which we get @xmath271 notice that in this case @xmath111 due to the fact that @xmath272 .
now , by comparing eq .
( [ eq : bound_ground ] ) and eq .
( [ eq : bound_any ] ) we can evidence the existence of a critical value @xmath131 such that when @xmath49 approaches @xmath131 the ground state switches from @xmath273 to @xmath132 , at least in the limit @xmath200 .
in fact , if we take @xmath274 , then in eq .
( [ eq : bound_ground ] ) and eq .
( [ eq : bound_any ] ) we can set @xmath275 , with @xmath276 and @xmath277 respectively , obtaining @xmath278 therefore , recalling that @xmath279 , the condition @xmath280 , as @xmath281 is sufficient for @xmath282 . otherwise stated , as @xmath49 approaches @xmath131 from above ( @xmath283 ) or from below ( @xmath277 ) , the overlap of the ground state with @xmath133 and with @xmath132 , respectively , is close to @xmath11 .
the laplacian spectrum @xmath104 for the dsg is exactly known @xcite and we can therefore calculate exactly the quantities @xmath284 and @xmath285 , obtaining estimates for the critical parameter @xmath131 . at a given generation @xmath54 , the spectrum includes the eigenvalue @xmath69 with degeneracy @xmath286 , the eigenvalue @xmath73 with degeneracy @xmath287 and the eigenvalues @xmath288 and @xmath289 stemming from the eigenvalue @xmath290 , belonging to previous generation and both carrying degeneracy @xmath291 . for each eigenvalue @xmath290 the next - generation eigenvalues @xmath292 are defined according to @xmath293 now , it follows directly from eq .
( [ eq : iteration ] ) that any couple @xmath294 , @xmath295 sum up as @xmath296 and applying this result iteratively to all the couples making up the spectrum we get @xmath297 as for @xmath285 we can implement a similar iterative procedure , by noticing that @xmath298 therefore @xmath285 is made up of two terms stemming from the first and second order contributions , which are , respectively , @xmath299 and @xmath300 subtracting eq .
( [ eq : xi2_dsg ] ) and eq .
( [ eq : xi1_dsg ] ) we finally get @xmath301 ^ 2 } = \frac{1}{900 } ( -13 -14\times 3^g + 21 \times 5^g + 6 \times 25^g).\ ] ] now , recalling that @xmath302 and that the spectral dimension of the dsg is @xmath303 , we can write @xmath304 and obtain expressions for @xmath284 and @xmath285 as a function of just the volume @xmath0 and the spectral dimension @xmath56 of the substrate : @xmath305 @xmath306 where the asymptotic expressions hold for large @xmath0 . we notice that as @xmath200 , @xmath284 and @xmath285 appearing in eqs .
( [ eq : xi1_n ] ) and ( [ eq : xi2_n ] ) satisfy the condition @xmath307 found in appendix [ appe1 ] .
more precisely , @xmath308 and this is sufficient for @xmath309 and @xmath310 to converge to zero as @xmath311 . in particular , from eq .
( [ eq : xi1_n ] ) and eqs .
( [ eq : bound_ground ] ) and ( [ eq : bound_any ] ) , the state @xmath42 is expected to switch from the ground to the first excited state at @xmath312 as explained in sec .
[ sec : phase_transition ] , the expressions found here for @xmath284 and @xmath285 are consistent with @xmath313^j \sim n^{2j / d}$ ] obtained in @xcite for lattices .
therefore , and as suggested by the numerical data discussed in sec .
[ sec : numerics ] , it is plausible that eqs .
( [ eq : xi1_n ] ) , ( [ eq : xi2_n ] ) and , above all , the expression for the critical parameter @xmath131 in eq .
( [ eq : gammacrit ] ) can be extended to arbitrary structures of spectral dimension @xmath144 .
let us start from the definition of success probability given in eq .
( [ eq : success_prob ] ) .
now , the propagator @xmath314 , can be calculated as @xmath315 and the success probability can be rewritten as @xmath316 hence , in order to calculate the success probability @xmath50 we need to find the elements corresponding to the @xmath8-th column of the @xmath317-th power of the hamiltonian . for complete graphs , @xmath318 displays a high degree of symmetry which allows its exact calculation ( see for example @xcite for a similar calculation where the hamiltonian is provided by the adjacency matrix ) . without loss of generality we can fix @xmath319 so that the hamiltonian @xmath43 is @xmath320 and it is easy to see that , regardless of @xmath317 , @xmath321 , with @xmath322 .
therefore , eq .
( [ eq : success_prob_1 ] ) can be rewritten as @xmath323 now , our task is to calculate @xmath324 and @xmath325 for which we need the entries @xmath326 and @xmath327 of the @xmath317-th power of the hamiltonian ; for better readability we set @xmath328 and @xmath329 .
thus , from eq .
( [ eq : hamiltonian_mat ] ) we derive the following recursive relations : @xmath330 and @xmath331 from their combination we get @xmath332 whose solution is @xmath333,\ ] ] with @xmath334 and @xmath335 . from eq .
( [ eq : recursive_1 ] ) and eq .
( [ eq : a1va ] ) , @xmath336 is also explicitly defined : @xmath337,\ ] ] now , according to eq .
( [ eq : success_prob_1 ] ) , we can calculate @xmath324 as @xmath338 \\ \nonumber & = & \frac{\gamma}{b } \left \ { \exp{\left[-it(a+b)/2 \right ] } -\exp{\left[-it(a - b)/2 \right ] } \right\}\\ & = & \frac{\gamma}{b } \exp{-i t a/2 } ( -2 i \sin{\frac{tb}{2 } } ) .
\end{aligned}\ ] ] analogous calculations lead to @xmath339 hence , by inserting eqs .
( [ eq : a1v ] ) and ( [ eq : avv ] ) into eq .
( [ eq : success_prob_2 ] ) , and performing the summation @xmath340,\end{aligned}\ ] ] which provides the exact success probability for the complete graph of size @xmath0 as a function of @xmath49 and of @xmath9 .
we can notice that @xmath50 is strictly larger than zero and that it can be equal to @xmath11 when both @xmath341 , with @xmath342 and @xmath343 are satisfied .
the latter condition holds for @xmath167 , just consistent with @xcite : there it is found that for @xmath167 the walk rotates the state from @xmath52 to @xmath99 and that the gap between the corresponding energies is smallest .
for @xmath167 the first condition reads @xmath344 .
hence , for the exact @xmath49 and at the right time the success probability is unitary , but for larger volumes the right times get sparser .
we also notice that when @xmath0 is large and @xmath345 , eq .
( [ eq : success_prob_3 ] ) can be simply rewritten as @xmath346 from eq .
( [ eq : success_prob_3 ] ) we deduce that for a given time @xmath347 , @xmath50 decreases as @xmath348 gets larger .
chuang and m.a .
nielsen , _ quantum computation and quantum information _ , cambridge university press , cambridge , 2000 l.k .
grover , phys .
. lett . * 79 * , 325 ( 1997 ) c. weiss and m. holthaus , europhys . lett . *
59 * , 486 ( 2002 ) d. daems and s. gurin , phys .
rev . lett . * 99 * , 170503 ( 2007 ) p.w .
shor , _ proceedings of the 35th ieee symposium on foundations of computer science _ ( ieee , los alamos , 1994 ) ,
a. ambainis , _ quantum search algorithms _ , sigact news * 35*(2 ) , 22 ( 2004 ) a.m. childs and j. goldstone , phys . rev .
a * 70 * , 042312 ( 2004 ) l.k .
grover , phys .
lett . * 79 * , 325 ( 1997 ) j. roland and n.j .
cerf , phys .
a * 68 * , 062312 ( 2003 ) s. aaronson and a. ambainis , in _ proceedings of the 44th ieee symposium on foundations of computer science _ ( ieee , los alamos , 2003 ) ,
e. farhi and s. gutmann , phys .
a * 58 * , 915 ( 1998 ) a.m. childs and j. goldstone , phys .
a * 70 * , 022314 ( 2004 ) j. roland and n.j .
cerf , phys .
a * 65 * , 042308 ( 2002 ) a. tulsi , phys .
a * 78 * , 012310 ( 2008 ) j.p .
dowling , nature * 439 * , 919 ( 2006 ) a.m. childs , e. farhi and s. gutmann , quantum inf . process .
* 1 * , 35 ( 2002 ) j. kempe , contemp
* 44 * , 307 ( 2003 ) a. ambainis , j. kempe and a. rivosh , _
soda 05 : proceeding of the sixteenth annual acm - siam symposium of discrete algorithms _ , 1099 ( 2005 ) n. shenvi , j. kempe and k.b .
whaley , phys .
a * 67 * , 052307 ( 2003 ) c.h .
bennett , e. bernstein , g. brassard and u. vazirani , siam j. comput . * 26 * , 1510 ( 1997 ) e. farhi and s. gutmann , phys .
a * 57 * , 2403 ( 1998 ) e. farhi , j. goldstone , s. gutmann and m. sipser , e - print quant - ph/0001106 i. sh . averbukh and n.f .
perelman , phys .
a * 139 * , 449 ( 1989 ) o. mlken and a. blumen , phys .
e * 71 * , 016101 ( 2005 ) d.h .
dunlap and v.m .
kenkre , phys .
b * 34 * , 3625 ( 1986 ) m.g .
cosenza and r. kapral , phys .
a * 46 * , 1850 ( 1992 ) k. hattori , t. hattori and h. watanabe , prog .
. phys.suppl . *
92 * , 108 ( 1987 ) a. blumen and a. jurjiu , j. chem . phys . * 116 * , 2636 ( 2002 ) e. agliari , a. blumen and o. mlken , j. phys .
a * 41 * , 445301 ( 2008 ) b. khang and s. redner , j. phys .
a * 22 * , 887 ( 1989 ) d. ben - avraham and s. havlin , _ diffusion and reactions in fractals and disordered systems _ , cambridge university press , cambridge , 2000 ; s. redner , _ a guide to first - passage processes _ , cambridge university press , cambridge , 2001 . s. alexander , j. bernasconi , w. r. schneider , and r. orbach , rev .
phys . * 53 * , 175 ( 1981 ) .
s. alexander and r. orbach , j. physique lett .
* 43 * , 625 ( 1982 ) .
g.e . santoro and e. tosatti , j. phys .
a * 39 * , r393 ( 2006 ) w. woess , _ random walks on infinite graphs and groups _ , cambridge university press , cambridge , 2000 . | we study continuous - time quantum walks mimicking the quantum search based on grover s procedure .
this allows us to consider structures , that is , databases , with arbitrary topological arrangements of their entries .
we show that the topological structure of the database plays a crucial role by analyzing , both analytically and numerically , the transition from the ground to the first excited state of the hamiltonian associated with different ( fractal ) structures .
additionally , we use the probability of successfully finding a specific target as another indicator of the importance of the topological structure . |
the present situation around @xmath0 oscillation initiated by super - kamiokande hereafter , sk , simply is to be described as follows if we follow majority opinion : + ( 1 ) sk found anomaly in the atmospheric neutrinos ( muon deficit ) @xcite , + ( 2 ) sk found anomaly in the zenith angle distribution @xcite , + ( 3 ) sk discovered atmospheric neutrino oscillation @xcite , + ( 4 ) soudan 2 and macro confirmed sk results @xcite , + ( 5 ) k2k confirmed atmospheric neutrino results based on the baseline experiment @xcite , + ( 6 ) sk found oscillations themselves directly @xcite , + ( 7 ) minos begins precise experiment and confirmed sk results @xcite . +
as one fully understand from the history on @xmath0 oscillation on atmospheric neutrinos , it is said that the experimental results by sk play decisively important role in @xmath1 oscillation .
therefore , to reach the final conclusions on @xmath0 oscillation by sk , it is more desirable that careful examination on the sk conclusion is performed by the minority opinion .
we have been performing computer numerical experiments to re - examine the atmospheric neutrino experiment results , constructing the virtual super - kamiokande apparatus in the computer and analyzing the virtual neutrino events produced both inside and outside the detector . in principle
, we could obtain any physical events which sk obtain and , therefore , could compare our results with sk results directly , which may lead less ambiguous discussion around the atmospheric neutrino oscillation problems . in the discussion on the atmospheric neutrino oscillation , there are two fundamental issues to be clarified from the experimental point of view , namely , the discrimination of the neutrino events ( electron events or muon events ) and the differences between the directions of the incident neutrinos and those of the produced leptons .
sk treats two different types of the neutrino events as for the points of their occurrence with regard to the detector .
ones are the neutrino events occurring inside the detector , namely , _
fully contained events _ and _ partially contained events _ in the sk terminology and the others are neutrino events occurring outside the detector , namely , _
upward through going muon events _ and _ stopping muon events _ in their terminology .
the mutual relation between two fundamental issues mentioned above and different types of the neutrino events is as follows : in the neutrino events occurring inside the detector , one should be careful for the discrimination between muon neutrino events and electron neutrino events , because muon neutrino events produce the different cherenkov light patterns from those of electron neutrino events inside the detector .
furthermore , one should be very sensitive to the differences between the directions of the incident neutrinos and those of the emitted leptons ( muons or electrons ) , because their scattering angles could not be neglected due to their smaller energies .
on the other hand , in the neutrino events occurring outside the detector , one need not worry about the both discrimination among the neutrino events and scattering angles of the neutrino events , because these events should be identified as the muon events due to their larger effective volume and the scattering angle of the produced leptons concerned could be neglected due to their higher energies . in the present paper
, we analyze the neutrino events occurring outside the detector .
the reason why we carry out the analysis of _ upward through going muon events _ and _ stopping muon events _ is that this analysis is more simpler and easier than that of the _ fully contained events _ and _ partially contained events_. our analysis of the neutrino events occurring inside the detector will be presented as a series of the papers elsewhere .
our fundamental approach to examine the experimental results obtained by sk is to carry out `` re - experiment '' of sk . for the purpose
, we have constructed the virtual sk detector in the computer and analyze virtual `` experimental '' events produced in the virtual detector .
the characteristics of our computer numerical experiment are that we could obtain any physical results which sk obtain and we could examine sk results more concretely , comparing sk results with our results directly . here , we give the algorithm for our computer numerical experiment . in figure [ fig:01 ] , we show the concept of our numerical computer experiment schematically .
the procedure of our numerical experiment is as follows : we start a neutrino with some energy on the opposite side to the earth to the super - kamiokande . for the purpose , taking the atmospheric neutrino energy spectrum on the opposite side of the earth to the detector which sk utilizes @xcite
, we choose a neutrino with some energy by the random sampling from this spectrum . .
] . ]
neutrinos thus chosen penetrate into the interior of the earth and interact at some point inside the earth after the traverse through the layer with different densities shown in figure [ fig:02 ] @xcite .
the interaction points of the neutrino events concerned are determined by the random sampling , following the probability function for the given neutrino interaction mean free path where the change of the density of the earth is taken into account , referring to the preliminary reference earth model @xcite .
the interaction mean free paths of the neutrinos for given energies concerned depend on the both density of the location and the zenith angle of the incident neutrino . in figure [ fig:03 ] and figure [ fig:04 ] , we give them in @xmath2 and @xmath3 , where @xmath4 denotes zenith angle of the incident neutrino , as examples .
furthermore , the neutrino concerned produces a daughter muon by deep inelastic scattering @xcite , the energy of which is also determined by the random sampling , following the probability function for the charged current ( deep inelastic scattering ) differential cross section concerned @xcite . as for _ upward going muon events _ and _ upward stopping muon events _ in our computer numerical experiment
, we could get the essential information around them , namely , what nature the neutrinos concerned have ( normal or anti- ) , their location where they are produced and what energies they have , what category they belong to ( _ upward through going muon events _ or _ upward stopping muon events _ ) , while the real experiments by sk give only the discrimination between _ upward through going muon events _ and _ upward stopping muon events_. ) for different energies . ] ) for different energies . ] exactly speaking , the procedure which we really utilize is slightly different from what we just mention above , although it is mathematically equivalent to what we mention above .
our real procedure adopted is as follows : instead of sampling the neutrino concerned directly from the incident neutrino energy spectrum on the surface of the earth and pursing it from its interaction point in stochastic manner , we define the neutrino interaction probability function at @xmath5 which is measured from the surface of the earth , in the following .
@xmath6 , where @xmath7 is the mean free path of the neutrino interaction for normal and anti - neutrino with @xmath8 at the depth @xmath9 , respectively where the density of the earth is @xmath10 and @xmath4 denotes nadir angle . here , we consider the deep inelastic charged current interaction @xcite and density of the earth is taken from the preliminary reference earth model @xcite . here
, we adopt the deep inelastic scattering exclusively as the cause of neutrino interaction , because other neutrino interactions could be neglected for our case . by combining eq.([eq : pintdt ] ) with the atmospheric neutrino energy spectrum at the opposite side of the earth to sk , we construct @xmath11 , neutrino interaction energy spectrum without neutrino oscillations for normal and anti - neutrino which denotes frequency of the neutrino interaction in @xmath12 in the following , @xmath13 , where @xmath14 denotes the atmospheric neutrino spectrum for normal and anti - neutrino at the opposite side of the earth to sk detector @xcite . the neutrino interaction energy spectrum , @xmath15 , , denotes frequency for neutrino interaction in @xmath12for null oscillation . the @xmath16 chosen in relation to the maximum energy of neutrino energy spectrum in the following .
the maximum energies of the emitted muons are equal to energies of the parent neutrino with maximum primary energy .
we utilize the incident neutrino energy spectrum by honda et al .
@xcite which is also utilized by sk , the maximum energy of which is @xmath17 tev .
therefore , we adopt @xmath17 tev as the maximum energy of muon to be pursued in our computer numerical experiment .
tev muon in the standard rock . ]
tev muon in the standard rock .
] now , we decide the maximum range for @xmath17 tev muon which is necessary for our computer numerical experiment . for the purpose ,
we simulate muons with 10 tev in stochastic manner exactly , taking into consideration of the all processes of bremsstrahlung , direct electron pair production , nuclear interaction and ionization loss . in figure [ fig:05 ] ,
we give the range fluctuation of @xmath17 tev muon thus obtained in the standard rock .
it is easily understood from the figure that the range of the muon is widely distributed by the fluctuation due to nature in these elementary processes .
the shorter ranges than their average are caused by catastrophic energy loss due to the bremsstrahlung and nuclear interaction ( see , for example , figure [ fig:09 ] ) , while longer ranges than their average are caused essentially by lager number of the direct electron pair production ( continuous energy loss - like)(see , for example figure [ fig:10 ] ) . from figure [ fig:05 ]
, we could adopt @xmath18 @xmath19 as the maximum range of 10 tev muon .
in other word , when muons with 10 tev are produced at the place far @xmath18 @xmath19 from the detector , such muons never reach the detector .
therefore , it is enough for us to have interest in high energy neutrinos only which are produced within @xmath18 @xmath19 from the detector .
now , we take @xmath18 @xmath19 as @xmath16 which is defined in eq.([eq : nintnulldt ] ) .
the generation points of such muons distribute uniformly within @xmath16 , because @xmath16 is smaller by higher order magnitude of the mean free path of the neutrino for deep inelastic scattering . in figure [ fig:06 ] , we give the range straggling with 1 tev muon obtained in similar manner as in figure [ fig:05 ] . from figures [ fig:05 ] and [ fig:06 ] , it is easily supposed that the effect of the range fluctuation in higher energy muon could not be neglected in spite of the rapid decrease of the higher energy muon in muon energy spectrum .
our computer numerical experiment for definite zenith angle of the incident neutrino is being carried out as follow : we calculate the neutrino interaction probability from eq.([eq : pintdt ] ) .
the value of @xmath5 is defined as the distance from the surface of the earth to @xmath16 .
next , we calculate the interaction neutrino energy spectrum from eq.([eq : nintnulldt ] ) . by using @xmath20
thus obtained from eq.([eq : nintnulldt ] ) , we sample randomly @xmath8 , the energy of the incident ( anti- ) neutrino to be pursued in the following way .
@xmath21 , where @xmath22 , @xmath23 denote the maximum and minimum energy of the neutrino concerned , respectively . and @xmath24 denote uniform random number between ( 0,1 ) .
we determine the interaction point of the ( anti- ) neutrino event with energy determined by procedure b. by using @xmath24 , the uniform random number between ( 0,1 ) , we determine the interaction point @xmath25 .
namely , the neutrino interaction occurs from the detector , @xmath9 is given as , @xmath26 the detailed study on the pursuit of high energy muon , relating to the range energy relation , will appear in a separate paper elsewhere .
the @xmath27 , the energy of the emitted ( anti-)muon which occur at @xmath28 from the detector produced by @xmath29 is determined from the following equation , @xmath30 , where @xmath31 is the differential cross section of the deep inelastic scattering for ( anti- ) neutrino @xcite and @xmath32 equal to the @xmath8 .
the @xmath33 is a uniform random number between ( 0,1 ) . for the ( anti-)muon whose energy and production point is determined from procedures c and d
, we examine the behavior of the trajectory of the ( anti-)muon toward the sk detector in a stochastic manner .
namely , each individual muon is pursued exactly by taking into consideration of the processes of bremsstrahlung , direct electron pair production , nuclear interaction and ionization loss in the stochastic manner without utilizing the average behavior of the muon concerned , the procedure of which is adopted by sk @xcite . as the result of it , we determine finally , which category each individual muon pursued falls into : [ a ] stopping before it reaches the detector , [ b ] stopping inside detector , or [ c ] passing through the detector .
we repeat procedures a to e and obtain the neutrino events without oscillation for a given live time for the real experiment . in our computer numerical experiment ,
we have accumulated the events concerned which correspond to the real live time for sk . for each neutrino event , we know @xmath8 , the energy of the parent neutrino , @xmath34 , the energy of the daughter ( anti-)muon,@xmath35 , the interaction point , @xmath36 ( @xmath37 ) , the direction of both the incident ( anti-)neutrino and the emitted ( anti-)muon , respectively , and @xmath5 , the distance between the interaction point of the neutrino events and the opposite side of the earth . in the case without neutrino oscillation , we repeat from procedures a to e , starting from the interaction energy spectrum(eq.([eq : nintnulldt ] ) ) .
finally , we obtain frequencies for _ upward stopping muon events _ and _ upward through going muon events _ for given zenith angle for null oscillation .
thus , we obtain the zenith angle distributions for these categorically different events without neutrino oscillation after summation over zenith angle . in the case with neutrino oscillation ,
@xmath38 in eq.([eq : nintnulldt ] ) should be replaced by @xmath39 , where @xmath40 , the survival probability of a given @xmath41 . here
, we adopt @xmath42 and @xmath43ev@xmath44 @xcite , as obtained from sk experiment .
other procedures are exactly same as in the case without neutrino oscillation .
thus , eq.([eq : xi_nintnull ] ) should be replaced by @xmath45 , by using eq.([eq : nintoscdt ] ) and other procedure are not changed .
the procedures a to e in both cases without oscillation and with oscillation is illustrated in figure [ fig : chart ] .
in figure [ fig:07 ] , we give the differential neutrino interaction energy spectrum with oscillation ( eq.([eq : nintoscdt ] ) ) for different zenith angle from which we randomly sample the energy of the neutrino event concerned . from the figure
, it is easily understood that the differential neutrino interaction energy spectrum are not so influenced by the characteristics of the nature of oscillatory above @xmath46 gev under sk neutrino oscillation parameters .
the result of the random sampling by eq.([eq : xi_nintosc ] ) is given in figure [ fig:08 ] .
the excellent agreement between the theoretical curve ( solid line ) and the results sampled ( histogram ) guarantees the validity of the sampling procedure by eq.([eq : xi_nintosc ] ) . the nature of oscillatory in the neutrino interaction energy spectrum a play crucial role below @xmath46 gev under the sk neutrino oscillation parameters , on which we will examine in a series of the subsequent papers in detail . the physical origin of _
upward through going muon events _ is the same that of _ stopping muon events _ and the only difference between them lies in the situation that the former may eventually pass through the detector , while the latter may eventually stop the both initiated by the deep inelastic scattering . the generation point of the neutrino event concerned
is given eq.([eq : ti ] ) .
as the generation point of the neutrino event distribute over _ dl _
( @xmath47 g / cm@xmath44 in our case ) randomly shown in figure [ fig : detector_upward ] , the most biggest fluctuation in the muon behavior essentially comes from the generation point of the neutrino events over @xmath16 .
such fluctuation could not be considered in the detector simulation by sk .
the energy of the neutrino is sampled randomly from eq.([eq : xi_nintosc ] ) and the energy of the emitted muon from the neutrino event concerned is sampled randomly from eq.([eq : xi_d ] ) .
the muons whose energies and whose generation points thus determined are simulated in the stochastic manner exactly , taking into account of the all processes of the direct electron pair production , bremsstrahlung , nuclear interaction and ionization loss .
namely , these muons thus sampled lose their energies in stochastic manner in the course of their passages toward sk detector .
in summarized speaking , the fluctuation effect in the muon behavior secondarily comes from the range fluctuation of the muon , although this effect is rather small compared with the effect from the generation point of the neutrino events but could not be neglected . even if the physical natures of the muons are exactly same , some muon may be classified as the _ upward through going muon events _ or may be done as _ stopping muon events _ eventually due to the complicated geometry configuration of @xmath49 for muons concerned , which is schematically shown in figure [ fig : detector_upward ] .
we take into account of the complicated circumstances for the muons concerned as exactly as possible . as the results of it
, we could discriminate finally _ upward through going muon events _ from the _ stopping muon events_. among elementary processes contributed to muon energy losses , the process of the direct electron pair production , bremsstrahlung and nuclear interaction play important roles from the point of fluctuation . in figures
[ fig:09 ] and [ fig:10 ] , how these three elementary processes take place randomly in the stochastic ways are indicated as the functions of the traversed distance in standard rock by a muon with 10 tev .
therefore , we treat the muon energy loss in stochastic manner exactly . in the figures ,
we show graphically where , which interaction ( electron direct pair production , bremsstrahlung or nuclear interaction ) occurs and what energy with respect to the primary energy is lost by the interaction concerned .
the @xmath50 denotes the sustained energy of the muon concerned as the result of all possible energy losses .
the * d * denote the energy loss due to direct electron pair production , * b * is the energy loss due to the bremsstrahlung and * n * denotes nuclear interaction , respectively .
the ratio denotes the ratio of energy loss concerned to the primary energy . in figure
[ fig:09 ] , we give an example of a muon with shorter range , say , @xmath51 meter .
we could understand the muon loses about @xmath52 % catastrophically at the early stage by the bremsstrahlung and bigger energy losses may come from the bremsstrahlung and nuclear interaction and smaller energy losses essentially come from the direct electron pair production . in figure [ fig:10 ] which is contrast to figure [ fig:09 ] , we give an example of a muon with range over @xmath53 meters . in this case ,
bremsstrahlung and nuclear interaction do not contribute to bigger energy losses and the accumulation of large number of the direct electron pair production events with rather smaller energies result in bigger energy loss .
on the contrary to our exact treatment on the individual muon behavior , sk seem to treat them in the sense of as the average theory @xcite .
sk utilize the following conventional formula , revised by lohmann , knop and voss @xcite .
@xmath54 then , this description expresses the average energy loss of the muon which does not give the range straggling of the muon and dose not expresses the real energy loss by individual muon . when the energy of the muon is not so pretty high where we could neglect the fluctuation effect of the muon range in individual case
, we could apply this expression to the analysis of real behavior of the muon .
however , the energy region for muon where we are interested in is now rather high and , therefore , we should pursue the muon behavior in the stochastic manner exactly , but in not average theory , even if the fluctuation effect is taken into account .
it should be emphasized that the average picture of the muons obtained from the average theory like eq.([eq : dedx ] ) does not describe the individual behavior of the muon concerned and the average picture of the muons is the average behavior of the individual muon which is deviated from their average behavior when we could not neglect the fluctuation effect coming from their stochastic nature in the elementary processes .
furthermore , it should be noticed that the stochastic treatment to the individual muon becomes important in the analysis of statistically insufficient number of the events concerned .
in summarized speaking , the fluctuation effect in the muon behavior secondarily comes from the range fluctuation of the muon , although this effect is rather small compared with the effect from the generation point of the neutrino events but could not be neglected . in figure [ fig:11 ]
we compare the experimental results by sk with corresponding our computer numerical results for both with neutrino oscillation and without neutrino oscillation in the case of _ upward through going muon events _ for the same 1645.9 live days @xcite . from the figure
, we could not conclude the experimental data agree with null oscillation or with oscillation .
the 1645.9 live days is not enough for extracting definite conclusion on the neutrino oscillation under sk neutrino oscillation even if it is possible to get sk neutrino oscillation parameters through the analysis of _ upward through going muon events _ , because the statistical fluctuation from the average values is not small .
if we compare our results with oscillation to our results without oscillation in the figure , we could find the event number with oscillation is larger than that without oscillation in the three bins ( 0.0 @xmath55 -0.1 ) , ( -0.1 @xmath55 -0.2 ) and ( -0.9 @xmath55 -1.0 ) .
such phenomena come from the statistical fluctuation due to insufficient sampling ( accumulation ) of the neutrino events . in figure [ fig:12 ] , we give corresponding results from our computer numerical experiment for 164590 days , one hundred times of the sk real live days .
now , we understand from the figure that our results with oscillation are always smaller than that without oscillation which attributes to larger statistics .
this denotes that sampling numbers for 164590 live days is quite enough for extracting the definite conclusion on the neutrino oscillation around the sk neutrino oscillation parameters .
furthermore , it is reasonable to say from the figure that we could not discriminate the difference in both with null oscillation and with oscillation under sk oscillation parameters from our computer numerical experiment . in other words , this is quite understandable , because the definition of the _ upward through going muon events _ is the muon events , the minimum energy of which is only given @xmath55 1.6 gev and nothing more both the incident energy , say , nobody knows both generation points and termination points and energies of the events concerned , but knows their zenith angle merely
. it should be rather curious , if we really could extract definite neutrino oscillation parameters from such extremely poor information around the events concerned . in figure [ fig:12 ] ,
we add , on purpose , the sk experimental data for 1645.9 live days together with our data for 164590 live days 450 live years as one hundred times as sk real live days . from the comparison between sk results for 1645.9 live days and
our result for 164590 live days 450 live years , we could not conclude directly that sk experimental data rather agree with our results with oscillation , because statistics of both results is quite different from each other and we do not allow to compare them directly . comparing the smoothness of the histograms in figure [ fig:11 ] with these in figure [ fig:12 ] , we could conclude that the average values of zenith angle distributions attain sufficiently at the true ones .
in other words , we think sk live days does not give enough statistics to extract clear cut conclusion under sk neutrino oscillation parameters from the analysis of _ upward through going muon events _ , even if sk finally could give definite neutrino oscillation parameters from the analysis of _ fully contained events _ and _ partially contained events_. in figure [ fig:13 ] , we give the correlation diagram between the energies of the emitted muons and the distance from their occurrence points up to the detector for _ upward through going muon events _ for null oscillation , pursuing the muons concerned in stochastic manner .
the average energy of the emitted muons is 390 gev and their average distance 161 meter . in figure [ fig:14 ]
, we give the corresponding quantities for oscillation .
the average energy of the emitted muons is 402 gev and their average distance is 168 meters .
it is easily understood from the figure [ fig:13 ] and figure [ fig:14 ] that the contribution from higher energy part of the muon could not negligible , which is larger beyond the expectation by sk whose analysis is based on the average theory for muon propagation . furthermore , comparing the figure [ fig:13 ] with figure [ fig:14 ] , we could not find clear differences between them , which denote the difficulty in extracting the sk oscillation parameters from the analysis of the _ upward through going muon events_. * figure [ fig:13 ] .
* correlation diagram between the emitted muon energies and their distance from their generation points to the detector for null oscillation in the case of _ upward through going muon events_. + * figure [ fig:14 ] .
* correlation diagram between the emitted muon energies and their distance from their generation points to the detector for the oscillation in the case of _ upward through going muon events_. + * figure [ fig:15 ] . * the distance distribution for
_ upward through going muon events _ in both cases of null oscillation and the oscillation . in figure [ fig:15 ] ,
we give the distance distribution for _ upward through going muon events _ in the cases of both null oscillation and the oscillation for 1645.9 sk live days .
the distance denotes the interval from the generation point of the events concerned to the detector .
it is easily understood from the figure that we could not discriminate the difference in both cases well , which suggests difficulty in extracting the definite sk neutrino oscillation parameters .
now , we return the sk original data for _ upward through going muon events _ which we reproduce it in figure [ fig:16 ] .
now , we point out clear contradiction in figure [ fig:16 ] .
the first to be examined is that frequency of the experimental data is significantly larger than that of their monte carlo simulation without oscillation in the wider region ( -0.6 , 0.0 ) for @xmath56 and the second is rather nice agreement between them in the region ( -1.0,-0.6 ) for @xmath56 .
the first denotes that usually we should expect smaller frequency with oscillation than that without oscillation .
however , here , the situation becomes reverse which contradict with the commonsense .
furthermore , in the region ( 1.0,-0.6 ) for @xmath56 , the effect of the neutrino oscillation should be pronounced due to longer path length of the incident neutrino
. however , the real situation seems to indicate no oscillation . in this case , there are three possible interpretations to such a situation .
namely , [ a ] monte carlo simulation is not unreliable , while the experimental data is reliable including statistical uncertainty , [ b ] monte carlo simulation is reliable , while statistics of the experimental data is too small to give larger frequency of the neutrino events compared with monte carlo events , [ c ] neither experimental data nor monte carlo data is unreliable to exclude the definite oscillation parameters .
furthermore , from the view point of the comparison between experimental data and monte carlo one , we should make two comments : the fist is that sk monte carlo simulation does not consider the real behaviors of the muons in stochastic manner , but their average behaviors only .
the second is that it is not suitable way to compare experimental data for 1645.9 days with monte carlo one for 100 live years , because of big difference in statistics between them .
the normalization of 100 live yeas to 1645.9 days never guarantees the validity of their transformation without examination on the validity concerned . to clarify the contradiction in figure [ fig:16 ] , it is necessary for sk to disclose their monte carlo procedure in detail .
it is natural to think that _ upward stopping muon events _ are much influenced by fluctuation , compared with _
upward through going muon events _ , because they are more directly influenced by the complicated geometry of the detector with regard to the incident neutrinos than _ upward through going muon events _ are done . in figure
[ fig:17 ] , we compare the sk experimental data for 1645.9 live days with our data with oscillation and without oscillation for _ upward stopping muon events _ for the same 1645.9 live days . from the figure , also , we could conclude that sk experimental data neither agree with our data with oscillation nor that without oscillation . in the figure
, it should be noticed that the number of the neutrino events with oscillation is larger than that without oscillation due to fluctuation effect for @xmath57 @xmath58 . in figure
[ fig:18 ] , we give our results with oscillation and without oscillation for 164590 live days together with sk experimental data which , in principle , should not be allowed to be compared with our data due to big difference in statistics as stated in the case of _ upward through going muon events _ ( see , also footnote on page 13 ) . compared figure [ fig:18 ] with figure [ fig:17 ] we understand that the difference between quantity with oscillation and that without oscillation in figure [ fig:18 ] decrease compared with that in figure [ fig:17 ] and both histograms in figure [ fig:18 ] become smooth compared with those in figure [ fig:17 ] , which is evidence that `` average value '' attain at ideal average value . ] ] ] * figure [ fig:19 ] . *
correlation diagram between the emitted muon energies and their distance from their generation points to the detector for null oscillation in the case of _ upward stopping muon events_. + * figure [ fig:20 ] .
* correlation diagram between the emitted muon energies and their distance from their generation points to the detector for the oscillation in the case of _ upward stopping muon events_. + * figure [ fig:21 ] . * the distance distribution for _ upward stopping muon events _ in both cases of null oscillation and the oscillation . in figure [ fig:19 ] ,
we give the correlation diagram between the emitted muon energies and their distances in _ stopping muon events _ for null oscillation .
the average energy of the emitted muons concerned is 55 gev and their average distance is 44 meters . in figure
[ fig:20 ] , we give the similar diagram for sk oscillation parameters as figure [ fig:19 ] .
the average energy of the muons is 56 gev and their average distance is 45 meters .
the situation around figure [ fig:19 ] and figure [ fig:20 ] is essentially same as in that around figure [ fig:13 ] and [ fig:14 ] . in figure [ fig:21 ] ,
we give the distance distribution for _ upward stopping muon events _ in both cases of null oscillation and the oscillation .
it is reasonable to say from the figure that we could not find the significant difference in the both distance distribution , which may lead to the difficulty in extracting the definite oscillation parameters like sk do . ] ] ] * figure [ fig:22 ] . *
energy spectra for parent neutrinos and daughter muons for _ upward through going muon events _ for @xmath59 @xmath60 . +
* figure [ fig:23 ] . *
energy spectra for parent neutrinos and daughter muons for _ upward through going muon events _ for @xmath61 @xmath62 . +
* figure [ fig:24 ] . *
energy spectra for parent neutrinos and daughter muons for _ upward through going muon events _ for @xmath63 @xmath64 . ] ] ] * figure [ fig:25 ] . *
energy spectra for parent neutrinos and daughter muons for _ upward stopping muon events _ for @xmath59 @xmath60 . + * figure [ fig:26 ] . *
energy spectra for parent neutrinos and daughter muons for _ upward stopping muon events _ for @xmath61 @xmath62 . + * figure [ fig:27 ] . *
energy spectra for parent neutrinos and daughter muons for _ upward stopping muon events _ for @xmath63 @xmath64 .
now , we examine energy spectra of the parent neutrinos and their daughter muons for the _ upward through going muon events _ and the _ stopping muon events _ with different @xmath65 . in figures
[ fig:22 ] , [ fig:23 ] and [ fig:24 ] , we give energy spectra of _ upward through going muon events _ for @xmath66 , ( upward ) @xmath67(diagonally ) and @xmath68(horizontally ) , respectively for sk live days , 1645.9 days . in figure
[ fig:25 ] , [ fig:26 ] and [ fig:27 ] , we give the energy spectra of the _ stopping muon events _ for the same for the same sk live days .
it seems to be clear from figures [ fig:22 ] to [ fig:27 ] that all these energy spectra are never smooth which denote insufficient statistics to extract definite conclusions , particularly as shown in figure [ fig:24 ] . in figures
[ fig:22 ] , [ fig:23 ] and [ fig:24 ] , we give the average energies for parent neutrinos and daughter muons ( in parentheses ) in the case of the _ upward through going muon events _ for @xmath66 , @xmath67 and @xmath68 , respectively , namely 529gev ( 343gev ) , 657gev ( 500gev ) and 658 gev ( 407gev ) .
the corresponding average energy for the parent neutrino in sk is @xmath69100 gev @xcite which is pretty smaller than that our values . in figure
[ fig:25 ] , [ fig:26 ] and [ fig:27 ] , we give the average energies of the _ stopping muon events _ for the same @xmath70 , 290gev ( 216gev ) , 134gev ( 54gev ) and 28gev ( 16gev ) , respectively . the corresponding average energy for neutrino by sk is @xmath69 5 gev @xcite which is also too smaller than our corresponding values . as for obtaining the average energy of the _ upward through going muon events _ , sk utilize the average theory which is unsuitable , as already mentioned , for the analysis of real behavior of high energy muon and , therefore , they could not consider the contribution from higher energy muons correctly . as for the estimation of the average energy of _ stopping muon events
_ , it seems to be curious that their average energy by sk is nearly same as or a little higher than that of multi - gev events . because multi - gev events comes from essentially single meson scattering , while _
stopping muon events _ come from deep inelastic scattering , and the effective energy for single meson scattering is too lower than that of deep inelastic scattering ( see , figure [ fig:15 ] in the sk paper @xcite ) .
the difference in the effective energies between _ upward through going muon events _ and _ stopping muon events _ comes from the difference in the effective ranges from the same elementary process , say , deep inelastic scattering . also , sk estimate that generation points up to the detector distributed over 0 meters to several hundred meters @xcite . too small estimation on average energy as well as the generation points for
_ upward stopping muon _ by sk might come from the utilization of the average theory on energy loss of the muon . here
, let us remark the relation between oscillatory nature of the primary neutrino interaction energy spectrum and their influence over _ upward through going muon events _ and _ stopping muon events_. it s easily understood from figures [ fig:22 ] to [ fig:27 ] , referring to figure [ fig:07 ] , that neutrinos contributed to both _
upward through going muon events _ and _ stopping muon events _ for all their zenith angles come from pretty high energies which belong non - oscillatory part of the incident neutrino interaction energy spectrum under the specified neutrino oscillation parameters by sk . in other word , we should say that the analysis of _ upward through going muon events _ and _ stopping muon events _ are not suitable means for proving the existence of neutrino oscillation under sk neutrino oscillation parameters . here , we compare sk monte carlo simulation in the case of null oscillation for enough statistics with our corresponding computer numerical experiment .
now , we compare sk monte carlo results for 100 live years with ours for 450 live years for both _ upward through going muon events _ and _ stopping muon events _ in the case of null oscillation . strictly speaking , sk results for 100 live years should be compared with ours for same live year . or our results for 450 live years
should be compared with the sk results for same live years . however , as sk data for 100 live years are supposed to attain at the enough statistics which is free from fluctuation , compared with that for 1645.9 days , we might be allowed to compare directly sk data for 100 live years with our data for 450 live years . from figure [ fig:28 ]
we understand that general tendency is similar to each other and our results give larger value than sk results .
it is reasonable , if we take into account of the fact that sk neglect the range fluctuation in high energy muon . in figure
[ fig:29 ] , we give the direct comparison between sk results and our results for _ upward stopping muon events_. here , sk results give larger values than our results , on the contrary to figure [ fig:28 ] .
this reversed situation between figure [ fig:28 ] and figure [ fig:29 ] seems to be curious , if we take into account of the range fluctuation of high energy muon .
also , tendencies in the sk histograms in figure [ fig:29 ] are somewhat unnatural in lack in smoothness in spite of huge statistics .
it is desirable for sk to disclose their essential information on their monte carlo method for the case of the null oscillation in order to clarify the difference between sk s and ours .
the techniques for our monte carlo method fully disclose in the present paper . as for our monte carl simulation for the neutrino events inside the detector , we disclose our procedure in detail in the subsequent papers .
thus , the followings should be concluded : + ( a ) the fluctuation effect in the high energy muon behavior from the neutrino event ( due to deep inelastic scattering ) primarily comes from the uniformly randomness of the generation point of the neutrino events concerned .
this effect could not be treated in the correct manner in the detector simulation by sk .
+ ( b ) also , the consideration of the range fluctuation in high energy muon is inevitable , although this effect is pretty smaller than that from the generation point of the neutrino events .
we consider this effect exactly in the stochastic manner . on the other hand
, sk analyzes the behavior of muons in the average theory , namely , in the concept of the average energy loss and average range , not treating them in stochastic manner .
combined of [ a ] with [ b ] , we claim that experimental results for the _ upward through going muon events _ and the _ stopping muon events _ do not show the evidence under sk neutrino oscillation parameters .
+ ( c ) generally speaking , it is difficult to imagine rationally to extract so definite conclusion around neutrino oscillation parameters , say , @xmath71 and @xmath72 through the analysis of the _ upward through going muon events _ and _ stopping muon events_. because , there are too few information around such categorized events to extract the definite conclusion on the neutrino oscillation parameters .
such categorized events provide the direction of the incident neutrino only . _
upward through going muon events _ never provide us the energies of the incident neutrinos , their interaction point and the energies of the emitted muons , their stopping points , giving the information on their penetration through the detector only . _ stopping muon events _
gives the information that they stop inside the detector and other information not obtained by them are same as the _ upward through going muon events_. + ( d ) even if the sk neutrino oscillation parameters exist really , it is difficult to estimate them from the analysis of _ upward through going muon events _ and _ stopping muon events_. because , the effective energies to detect the sk neutrino oscillation parameters , say , smaller than 40 gev , do not lie in the energy region which contribute to the production of _ upward through going muon events _ and the _ stopping muon events_. + ( e ) by the definition of the _ upward through going muon events _ and the _ stopping muon events _ in the sk , the origins of both events are same(deep inelastic scattering)and difference between them is that the former average range is longer than that of the latter so that the former average energy is higher than that of the latter .
however , the latter average is so low which is rather near to the average energy due to multi - gev energy events .
this is difficult to be understood . also , the zenith angle distribution for _ upward through going muon events _ is difficult to be understood in comparison with their monte carlo simulation with 100 live years .
+ ( f ) even if we could obtain the sk neutrino oscillation parameters from the analysis of both _
upward through going muon events _ and the _ stopping muon events _ , present statistics for experimental data obtained by 1645.9 live days is not enough for getting the sk neutrino oscillation parameters .
+ ( g ) as the effective energy to detect oscillatory signature of the neutrino oscillation parameters lies below 40 gev under the assumption of the sk neutrino oscillation parameters , first of all , one should concentrate to analyze the single ring muon event among the fully contained events , the most clear cut event from the quasi elastic scattering in gev energy region . generally speaking
, it is rather difficult to extract so definite physical quantities , such as , neutrino oscillation parameters from the atmospheric neutrino beams which include much uncertainty factors , such as , the incident neutrino energy spectrum .
the throughout and concentrated analysis of the single ring lepton events are inevitable , if we really extract definite parameters from the cosmic ray experiment in spite of the existence of much uncertainty factors involved .
we are now going to complete our works along this line and will present our results from the computer numerical experiment , focusing on the analysis of single - ring muon events as a series of subsequent papers elsewhere .
99 hirata k s _ et al _ 1988 _ phys .
b _ * 205 * 416 hirata k s _ et al _ 1992 _ phys .
b _ * 280 * 146 fukuda y _ et al _ 1994 _ phys .
335 * kajita t 1999 _ nucl . phys .
suppl . _ vol.*77 * 123 fukuda y _ et al _ 1998 _ phys .
_ * 81 * 1562 mann wa 2001 _ nucl .
suppl . _ vol.*19 * 134 ambrosio m _
et al _ 2000 _ phys .
b _ * 478 * 3 ahn m.h _ et al _
2006 _ phys .
d _ * 74 * 072003 ashie y _
et al _ 2004 _ phys .
_ * 93 * 101801 michael d g _
et al _ 2006 _ phys .
_ * 97 * 191801 blake a _ et al _ 2007 _ proceeding of taup 2007 , sendai , japan(to be published ) _ kasuga s _
et al _ 1996 _ phys .
b _ * 374 * 238 anokhina a m and galkin v i 2006 _ physics of atomic nuclei _
vol.*69 * no.1 16 galkin v , _
et al _ 2004 _ hep - ex/0412059 _ galkin v , _
et al _ 2005 _ hep - ex/0501058 _ honda m _
et al _ 1996 _ phys .
d _ * 52 * 4985 dziewonski a and anderson 1981 _ phys .
earth planet .
inter . _ * 25 * 297 gandhi r h _ et al _ 1996 _ astropart
. phys . _ * 5 * 81 gandhi r h _ et al _ 1998 _ phys .
d _ * 58 * 093009 oyama y 1989 _ phys .
d _ * 39 * 1481 hatakeyama s 1998 _ phd thesis ( tohoku university ) _ saji c 2002 _ phd thesis ( niigata university ) _ nitta k 2003 _ phd thesis ( osaka university ) _ ashie y _ et al _ 2005 _ phys .
d _ * 71 * 112005 lohman w , kopp r and voss r 1985 _ cern 85 - 03 _ desai s , _
et al _ 2007 _ hep - ex/0711.0053 _ | adopting neutrino oscillation parameters obtained by super - kamiokande , a numerical computer experiment for neutrino events occurring outside the detector , is carried out in the same sk live days , 1645.9 live days , constructing the virtual super - kamiokande detector in the computer .
the numerical results by the computer experiment could be directly compared with the real sk experimental data .
the comparison shows that it is difficult to obtain convincing conclusion on the existence of the neutrino oscillation with specified neutrino oscillation parameters claimed by sk through analysis for neutrino events occurring outside the detector .
_ keywords _ : neutrino oscillation , exact monte carlo methods , cosmic rays + |
the brightest of the `` unidentified '' high - galactic - latitude egret sources @xcite , at @xmath9 has long been associated with the x - ray emitting neutron - star , which remains its most plausible counterpart even though pulsations have not been detected at any wavelength . unlike blazars , which are highly variable and have steeper @xmath3-ray spectra , shows no evidence for long - term variability , and its spectrum
can be fitted by a relatively flat power law of photon index @xmath10 from 70 mev to 4 gev , with a turn - down above 4 gev @xcite , similar to known @xmath3-ray pulsars .
is the only optically undetected x - ray source in the @xmath3-ray error box , all the others being classified as unlikely @xmath3-ray emitters @xcite .
the detection of as a weak , ultrasoft source in the all - sky survey @xcite suggested that it is a thermally emitting neutron star that is either older or more distant than the @xmath3-ray pulsar geminga @xcite .
observations using the advanced ccd imaging camera ( acis ) and the ( _ hst _ ) further supported this interpretation @xcite .
two components were required to fit the x - ray spectrum , a blackbody of @xmath11 k and a power law of photon index @xmath12 .
the non - thermal x - ray component is important evidence of magnetospheric activity that is found in all @xmath3-ray pulsars , but not in all cooling neutron stars .
an optical upper limit of @xmath13 from the _ hst _ verified the neutron - star interpretation of from its extreme x - ray - to - optical flux ratio , @xmath14 .
we also used the thermal x - ray spectrum and optical limit to bound the distance to in the range @xmath15 pc.y the energetics of are plausible for a pulsar at @xmath16 pc since its @xmath3-ray luminosity ( assumed isotropic ) is @xmath17 ergs s@xmath6 , comparable to the spin - down power @xmath18 of geminga ( @xmath19 ergs s@xmath6 ) . efficiencies approaching 100% are achieved by those @xmath3-ray pulsars having the smallest spin - down power
. @xcite also searched for radio pulsations from using the jodrell bank lovell telescope at a frequency of 1.4 ghz , achieving an upper limit of @xmath20 mjy . adopting an upper limit of @xmath21 pc ,
the pulsed pseudo - luminosity limit of is @xmath22 mjy kpc@xmath23 .
there are only four pulsars known with @xmath24 mjy kpc@xmath23 @xcite in addition to geminga ( * ? ? ?
* and references therein ) . in this paper
, we report an even more sensitive search for radio pulsations from using the nrao green bank telescope ( gbt ) , as well as the first x - ray pulsation search with the high resolution camera ( hrc - s ) .
these investigations approach the sensitivity limits of existing instrumentation to search for the pulsar in .
is circumpolar at the gbt , where on 2002 december 6 we observed it for 24hr using the bcpm spectrometer @xcite . due to the intermediate latitude of the target , with a relatively small predicted dispersion measure ( dm ) and low galactic background temperature , we observed at a central frequency of 820mhz , with a bandwidth of 48mhz in each of two polarizations .
the signals from corresponding polarization channels were summed in hardware , and the total - power samples from each of 96 frequency channels were sampled every @xmath25s and recorded to disk for off - line analysis .
we analyzed the data using standard pulsar search techniques , implemented in the presto software package @xcite .
first we identified and excised radio - frequency interference .
we then dedispersed the data in the dm range 0110 @xmath26 pc , with resolution of 0.3 @xmath26 pc .
this is twice the maximum galactic dm predicted for this line of sight by the @xcite electron density model ; for the distance range of 250800 pc , within which the neutron star likely lies ( 1 ) , the predicted dm range is 29 @xmath26 pc .
each time series contained 1.2 billion points , and would have been needlessly challenging to analyze .
we therefore down - sampled the data to a resolution of 0.288 ms , resulting in much more reasonable time series of 300 million points . for a dm of 40 @xmath26 pc ,
the smearing caused by dispersion across one of the frequency channels is 0.3 ms , so that for smaller dms the effective time resolution of the search was about 0.3 ms .
we performed the periodicity search using fast - fourier techniques , and were sensitive to periods @xmath27 ms .
in order to improve sensitivity for a range of pulse duty cycles , we incoherently summed up to 16 harmonics of the data ; for @xmath28 ms , we maintained good sensitivity for duty cycles down to about @xmath29 .
the sensitivity at long periods decreases gradually due to red noise , but this is not a significant concern for @xmath30 s. unfortunately , we found no convincing pulsar candidates in this search . using the standard modification to the radiometer equation , assuming a pulse shape with 10% duty cycle , we were sensitive to pulsars with @xmath28ms having a period - averaged flux density at 820mhz @xmath31jy , with gradually decreasing sensitivity for smaller periods .
converting this to the more usual 1.4ghz pulsar search frequency with a typical spectral index of 1.6 @xcite gives @xmath32jy . for a distance of 800 pc , this corresponds @xmath33jy kpc@xmath23 .
all known radio pulsars have a greater @xmath34 @xcite .
based on these results , it is therefore unlikely that is a radio pulsar beaming toward the earth .
more likely , it is either beaming away , or is not an active radio pulsar at all .
llrc 6182 & 2005 feb 28 19:27 & 45,018 & 284 + 4606 & 2005 mar 9 15:30 & 28,030 & 189 + 6183 & 2005 mar 11 08:19 & 44,999 & 256 + total & & 118,047 & 729 [ logtable ] three observations of totaling 118 ks of live time were performed between 2005 february 28 and 2005 march 11 using the hrc - s in timing mode , which has time resolution of @xmath35s , and provides absolute accuracy of @xmath36s as determined from observations of the crab pulsar .
table [ logtable ] lists the basic parameters of the observations , including the number of photons extracted from a @xmath37 radius aperture .
the live time is @xmath38 of the elapsed time , limited by event processing .
the total number of photons detected from is 729 , of which 20 are estimated to be background ( using nearby , source - free regions of the image ) .
the count rates from the three observations are within @xmath39 or less of the average count rate , @xmath40 s@xmath6 , consistent with no variation . the spatial distribution of photons is consistent with a point source .
originally conceived as a contiguous , 135 ks pointing , the observation could not be executed as planned because of new operational restrictions that came into effect .
the resulting large gaps in time diminish the sensitivity to pulsations because the increased number of independent frequencies and frequency derivatives that have to be searched yield more false candidates . the total time span of the observations from beginning to end is @xmath41 s , which determines the independent intervals for frequency and frequency derivative , @xmath42 and @xmath43 , respectively .
we chose a frequency step size @xmath44 hz to oversample the independent frequency interval by a factor of @xmath45 , and @xmath46 hz s@xmath6 to achieve a similar oversampling factor for frequency derivative . ) searched for x - ray pulsations from is shaded . the step size @xmath46 hz s@xmath6 , and the search includes @xmath47 . lines of constant @xmath48 and @xmath49 are shown .
the star indicates the parameters of geminga .
, width=316 ] we transformed the photon arrival times to barycentric dynamical time , and did coherent searches of the full data set using the discrete fourier transform technique , also known as the rayleigh or @xmath50 test @xcite .
we searched the entire range of ( @xmath51 ) parameter space for a pulsar that has @xmath52 hz , characteristic age @xmath53 kyr , as well as @xmath54 ergs s@xmath6 . in practice
, we found it simplest to confine the search to the region shaded in figure [ search ] , bounded by @xmath55 hz s@xmath6 for @xmath56 hz , and @xmath57 hz s@xmath6 for @xmath58 hz .
the total number of trials thus defined is @xmath59 , an oversampling of the independent trials by a factor of @xmath60 .
this is a liberal search range that is justified as follows : would have a blackbody temperature greater than its observed @xmath61 k if it were younger than 20 kyr ; an independent argument for an older age is given in 4 .
it would have a wind nebula or stronger nonthermal x - ray component if @xmath62 ergs s@xmath6 . in view of the distance limit
@xmath21 pc , the ratio @xmath63 if @xmath64 ergs s@xmath6 , a smaller ratio than all other pulsars , which typically have @xmath65 .
so it is unlikely that @xmath48 is as large as @xmath66 ergs s@xmath6 .
the range of parameters searched also includes all of the known millisecond ( recycled ) pulsars , and is effective as long as the pulsar does not have a neutron star binary companion .
the deep _ hst _ limit , corresponding to absolute magnitude @xmath67 at @xmath68 pc , rules out the more common white dwarf companions .
no significant pulsed signal was detected .
the largest values of @xmath50 found were @xmath69 .
the theoretical distribution of @xmath70 follows that of @xmath71 with @xmath72 degrees of freedom . for @xmath73 ,
the distribution is an exponential with a mean of 2 , so the single - trial probability that @xmath74 by chance is @xmath75 .
this is expected to arise randomly in @xmath76 independent trials .
@xcite showed that to detect sinusoidal pulsation with amplitude ( pulsed fraction ) @xmath77 at a power level @xmath78 with 50% probability , the number of photons needed is @xmath79 . in our case ,
@xmath80 ( 709 corrected for background ) , so we find @xmath81 corresponding to @xmath82 .
we verified this analytic expression by examining folded light curves at the periods associated with the maximum values of @xmath50 .
it is not surprising that the pulsed fraction of a @xmath3-ray pulsar in x - rays should be less than @xmath83 .
this is not a very sensitive limit consdering that other @xmath3-ray pulsars whose soft x - rays are predominantly thermal ( geminga , psr b105552 , possibly psr b0656 + 14 ) have even smaller pulsed fractions ( 30% , 21% , and 12% , respectively , * ? ? ? * ) . in these cases ,
the pulse profiles are quasi - sinusoidal .
the @xmath50 test is therefore a good one for , whose x - rays are also dominated by a soft thermal source within the response of the hrc , although the result is not very restrictive in this case .
we also note that the vela pulsar has a more complicated pulse shape , while its pulsed fraction is only 9% @xcite .
in order to test for proper motion of the neutron star , we compared the position of on the hrc with the acis position obtained 3 years earlier @xcite .
first , we updated the acis position to the usno - b1.0 system using optically identified sources in the acis image .
the result is r.a .
= @xmath84 , decl .
= @xmath85 ( j2000.0 ) , and differs from the previously used usno - a2.0 system by @xmath86 .
then we registered the hrc image to the acis image using seven x - ray sources in the vicinity of , which required a zero - point shift of the hrc image of @xmath87 in r.a . and
@xmath88 in decl .. the resulting hrc position of is r.a .
= @xmath89 , decl .
= @xmath90 ( j2000.0 ) .
the difference between the positions of in 2002 and 2005 is therefore only @xmath91 , which is comparable to their combined statistical errors .
in order to bound the possible proper motion , we adopt an upper limit that is twice this difference , or @xmath92 , corresponding to @xmath93 yr@xmath6 .
this allows the neutron star to have traveled from a birth in the galactic plane to its present position in @xmath94 yr or longer , which is not an unreasonable age for its x - ray temperature . at the maximum distance of 800 pc ,
the limit on @xmath0 corresponds to a tangential velocity limit @xmath7 km s@xmath6 , which is typical for radio pulsars .
even in the absence of detectable pulsations , remains the leading ( and only ) candidate for identification with .
its soft x - ray spectrum , and absence of optical and radio emission , support the hypothesis that it is an older and possibly more distant cousin of the geminga pulsar .
the upper limit on proper motion that we derived is a significant new observational constraint , allowing an age of at least @xmath94 yr for if it was born in the galactic plane .
this compares favorably with the characteristic age @xmath95 yr of geminga , and is consistent with a surface temperature of only @xmath61 k , compared with @xmath96 k for geminga @xcite . assuming that the age and distance of are both larger than those of geminga , it is likely that is a maximally efficient @xmath3-ray pulsar close to its death line @xcite , following the trend of @xmath3-ray efficiency increasing with decreasing spin - down power @xcite .
alternatively , we can not rule out that is a millisecond pulsar with similar spin - down power and magnetospheric gap voltage as geminga .
if so , its thermal x - ray emission may be due to surface reheating by a magnetospheric accelerator .
it is not now possible to make a significantly more sensitive search for x - ray pulsations from , as the operational limitations of no longer allow long , contiguous pointings .
while _ xmm - newton _ is in principle more sensitive , it has not been used to observe , which lies near the earth - avoidance zone of the satellite orbit , with maximum allowed pointing durations of @xmath97 ks .
finally , the upcoming _ gamma - ray large area space telescope _
( ) may be the best hope of detecting pulsations from .
_ glast _ will certainly reduce the positional uncertainty of , which will be an independent test of association with , and additional motivation for a dedicated pulsar search .
for this , a sufficiently long , pointed observation is best , as argued by @xcite .
we estimate the expected count rate from using its photon flux of @xmath98 @xmath99 s@xmath6 between 100 mev and 4 gev @xcite . assuming an average on - axis effective area of 5000 cm@xmath23 for the large area telescope ( lat ) in this energy range , the predicted count rate is 0.003 s@xmath6 . a more detailed calculation ,
convolving the spectrum of from figure 3 of @xcite with the lat on - axis effective area curve , predicts a similar rate of 0.0027 s@xmath6 in the 1004000 mev range .
this is almost half the count rate from the hrc , and holds great promise for a pulsar discovery .
it is likely that will succeed where failed because , unlike thermal x - rays , @xmath3-rays have sharp peaks with pulsed amplitude approaching 100% . in 1 week of elapsed time , including earth blockage , can collect as many photons from as we have obtained using . on - source time in a pointed observation of can be higher than average because the source is close to the ecliptic pole .
also , galactic background is negligible at these coordinates @xmath9 .
if is successful , the photons analyzed here may yet be used to find a coincident period and period derivative with confidence , thus establishing the identity of with , and determining its physical parameters .
support for this work was provided by nasa through _ chandra _ award sao go4 - 5058x issued by the _
chandra _ x - ray observatory center , which is operated by the smithsonian astrophysical observatory for and on behalf of nasa under contract nas8 - 03060 .
the gbt is operated by the national radio astronomy observatory , a facility of the national science foundation operated under cooperative agreement by associated universities , inc . | we report unsuccessful searches for pulsations from the neutron star identified with the egret source . a 24-hr observation with the nrao green bank telescope at 820 mhz placed an upper limit on flux density of 17 @xmath0jy for @xmath1 ms , and gradually increasing limits for @xmath2 ms .
the equivalent luminosity is lower than that of any known pulsar with the possible exception of the radio - quiet @xmath3-ray pulsar geminga .
a set of observations with the _ chandra x - ray observatory _ hrc totaling 118 ks revealed no pulsar with 1 ms @xmath4 s. the upper limit on its pulsed fraction is 35% assuming a sinusoidal pulse shape .
the position of in observations separated by 3 years is unchanged within errors , leading to an upper limit on its proper motion of @xmath5 yr@xmath6 , or @xmath7 km s@xmath6 at @xmath8 pc , a maximum distance estimated from its thermal x - ray spectrum . with these null results ,
the properties of and its x - ray counterpart are consistent with a more distant or older version of geminga , or perhaps a recycled pulsar . having nearly exhausted the capabilities of current instrumentation at all wavelengths , it will likely fall to the _ gamma - ray large area space telescope _ to discover pulsations from . |
input media for automatic summarization has varied from text @xcite to speech @xcite and video @xcite , but the application domain has been , in general , restricted to informative sources : news @xcite , meetings @xcite , or lectures @xcite .
nevertheless , application areas within the entertainment industry are gaining attention : e.g. summarization of literary short stories @xcite , music summarization @xcite , summarization of books @xcite , or inclusion of character analyses in movie summaries @xcite .
we follow this direction , creating extractive , text - driven video summaries for films and documentaries .
documentaries started as cinematic portrayals of reality @xcite .
today , they continue to portray historical events , argumentation , and research . they are commonly understood as capturing reality and therefore , seen as inherently non - fictional .
films , in contrast , are usually associated with fiction .
however , films and documentaries do not fundamentally differ : many of the strategies and narrative structures employed in films are also used in documentaries @xcite . in the context of our work , films ( fictional ) tell stories based on fictive events , whereas documentaries ( non - fictional ) address , mostly , scientific subjects .
we study the parallelism between the information carried in subtitles and scripts of both films and documentaries .
extractive summarization methods have been extensively explored for news documents @xcite .
our main goal is to understand the quality of automatic summaries , produced for films and documentaries , using the well - known behavior of news articles as reference .
generated summaries are evaluated against manual abstracts using rouge metrics , which correlate with human judgements @xcite .
this article is organized as follows : section [ sec : generic - summarization ] presents the summarization algorithms ; section [ sec : datasets ] presents the collected datasets ; section [ sec : setup ] presents the evaluation setup ; section [ sec : results ] discusses our results ; section [ sec : conclusions ] presents conclusions and directions for future work .
six text - based summarization approaches were used to summarize newspaper articles , subtitles , and scripts . they are described in the following sections . [
[ sub : mmr ] ] is a query - based summarization method @xcite .
it iteratively selects sentences via equation [ eq : e1 ] ( @xmath0 is a query ; @xmath1 and @xmath2 are similarity metrics ; @xmath3 and @xmath4 are non - selected and previously selected sentences , respectively ) .
@xmath5 balances relevance and novelty . can generate generic summaries by considering the input sentences centroid as a query @xcite .
@xmath6\ ] ] lexrank @xcite is a centrality - based method based on google s pagerank @xcite .
a graph is built using sentences , represented by tf - idf vectors , as vertexes .
edges are created when the cosine similarity exceeds a threshold .
equation [ eq : lexrank ] is computed at each vertex until the error rate between two successive iterations is lower than a certain value . in this equation
, @xmath7 is a damping factor to ensure the method s convergence , @xmath8 is the number of vertexes , and @xmath9 is the score of the @xmath10th vertex .
@xmath11}\frac{\operatorname{sim}\left(v_{i},v_{j}\right)}{\sum_{v_{k}\in \operatorname{adj}\left[v_{j}\right]}\operatorname{sim}\left(v_{j},v_{k}\right)}s\left(v_{j}\right)\ ] ] [ [ sub : lsa ] ] infers contextual usage of text based on word co - occurrence @xcite .
important topics are determined without the need for external lexical resources @xcite : each word s occurrence context provides information concerning its meaning , producing relations between words and sentences that correlate with the way humans make associations .
is applied to each document , represented by a @xmath12 term - by - sentences matrix @xmath13 , resulting in its decomposition @xmath14 .
summarization consists of choosing the @xmath15 highest singular values from @xmath16 , giving @xmath17 . @xmath18 and @xmath19 are reduced to @xmath20 and @xmath21 , respectively , approximating @xmath13 by @xmath22 .
the most important sentences are selected from @xmath21 .
documents are typically composed by a mixture of subjects , involving a main and various minor themes .
support sets are defined based on this observation @xcite .
important content is determined by creating a support set for each passage , by comparing it with all others .
the most semantically - related passages , determined via geometric proximity , are included in the support set .
summaries are composed by selecting the most relevant passages , i.e. , the ones present in the largest number of support sets . for a segmented information source @xmath23 , support sets @xmath24 for each passage @xmath25
are defined by equation [ eq : e4 ] , where @xmath26 is a similarity function , and @xmath27 is a threshold .
the most important passages are selected by equation [ eq : e5 ] .
@xmath28 @xmath29 [ [ sub : support - sets_luis ] ] @xcite proposed an extension of the centrality algorithm described in section [ sub : support - sets ] , which uses a two - stage important passage retrieval method .
the first stage consists of a feature - rich supervised key phrase extraction step , using the maui toolkit with additional semantic features : the detection of rhetorical signals , the number of named entities , tags , and 4 n - gram domain model probabilities @xcite .
the second stage consists of the extraction of the most important passages , where key phrases are considered regular passages .
[ [ sub : grasshopper ] ] @xcite is a re - ranking algorithm that maximizes diversity and minimizes redundancy .
it takes a weighted graph @xmath30 ( @xmath31 : @xmath32 vertexes representing sentences ; weights are defined by a similarity measure ) , a probability distribution @xmath33 ( representing a prior ranking ) , and @xmath34 $ ] , that balances the relative importance of @xmath30 and @xmath33 .
if there is no prior ranking , a uniform distribution can be used .
sentences are ranked by applying the teleporting random walks method in an absorbing markov chain , based on the @xmath31 transition matrix @xmath35 ( calculated by normalizing the rows of @xmath30 ) , i.e. , @xmath36 .
the first sentence to be scored is the one with the highest stationary probability @xmath37 according to the stationary distribution of @xmath38 : @xmath39 .
already selected sentences may never be visited again , by defining @xmath40 and @xmath41 .
the expected number of visits is given by matrix @xmath42 ( where @xmath43 is the expected number of visits to the sentence @xmath44 , if the random walker began at sentence @xmath10 ) .
we obtain the average of all possible starting sentences to get the expected number of visits to the @xmath44th sentence , @xmath45 .
the sentence to be selected is the one that satisfies @xmath46 .
we use three datasets : newspaper articles ( baseline data ) , films , and documentaries .
film data consists of subtitles and scripts , containing scene descriptions and dialog .
documentary data consists of subtitles containing mostly monologue .
reference data consists of manual abstracts ( for newspaper articles ) , plot summaries ( for films and documentaries ) , and synopses ( for films ) .
plot summaries are concise descriptions , sufficient for the reader to get a sense of what happens in the film or documentary .
synopses are much longer and may contain important details concerning the turn of events in the story . all datasets were normalized by removing punctuation inside sentences and timestamps from subtitles .
temrio @xcite is composed by 100 newspaper articles in brazilian portuguese ( table [ tab : news_corpus ] ) , covering domains such as world " , politics " , and foreign affairs " .
each article has a human - made reference summary ( abstract ) .
.temrio corpus properties . [ cols="<,<,^,^,^",options="header " , ] ; support sets used cosine distance and threshold=@xmath47 ; kp - centrality used 50 key phrases . ] news articles intend to answer basic questions about a particular event : who , what , when , where , why , and often , how . their structure is sometimes referred to as inverted pyramid " , where the most essential information comes first . typically , the first sentences provide a good overview of the entire article and are more likely to be chosen when composing the final summary . although documentaries follow a narrative structure similar to films , they can be seen as more closely related to news than films , especially regarding their intrinsic informative nature . in spite of their different natures , however ,
summaries created by humans produce similar scores for all of them .
it is possible to observe this behavior in figure [ fig : originais ] .
note that documentaries achieve higher scores than news articles or films , when using the original subtitles documents against the corresponding manual plot summaries .
figure [ fig : graphconcl1 ] presents an overview of the performance of each summarization algorithm across all domains .
the results concerning news articles were the best out of all three datasets for all experiments .
however , summaries for this dataset preserve , approximately , 31% of the original articles , in terms of sentences , which is significantly higher than for films and documentaries ( which preserve less than @xmath48 ) , necessarily leading to higher scores .
nonetheless , we can observe the differences in behavior between these domains .
notably , documentaries achieve the best results for plot summaries , in comparison with films , using scripts , subtitles , or the combination of both .
the relative scores on the films dataset are influenced by two major aspects : the short sentences found in the films dialogs ; and , since the generated summaries are extracts from subtitles and scripts , they are not able to represent the film as a whole , in contrast with what happens with plot summaries or synopses . additionally , the experiments conducted for script+subtitles for films , in general , do not improve scores above those of scripts alone , except for support sets for r-1 .
overall , lsa performed consistently better for news articles and documentaries .
similar relatively good behavior had already been observed for meeting recordings , where the best summarizer was also lsa @xcite .
one possible reason for these results is that lsa tries to capture the relation between words in sentences . by inferring contextual usage of text based on these relations ,
high scores , apart from r-1 , are produced for r-2 and r - su4 . for films ,
lexrank was the best performing algorithm for subtitles , scripts and the combination of both , using plot synopses , followed by lsa and support sets for plot summaries .
mmr has the lowest scores for all metrics and all datasets .
we observed that sentences closer to the centroid typically contain very few words , thus leading to shorter summaries and the corresponding low scores .
interestingly , by observing the average of r-1 , r-2 , and r - su4 , it is possible to notice that it follows very closely the values of r - su4 .
these results suggest that r - su4 adequately reflects the scores of both r-1 and r-2 , capturing the concepts derived from both unigrams and bigrams .
overall , considering plot summaries , documentaries achieved higher results in comparison with films .
however , in general , the highest score for these two domains is achieved using films scripts against plot synopses .
note that synopses have a significant difference in terms of sentences in comparison with plot summaries .
the average synopsis has 120 sentences , while plot summaries have , on average , 5 sentences for films , and 4 for documentaries .
this gives synopses a clear advantage in terms of rouge ( recall - based ) scores , due to the high count of words .
we analyzed the impact of the six summarization algorithms on three datasets .
the newspaper articles dataset was used as a reference .
the other two datasets , consisting of films and documentaries , were evaluated against plot summaries , for films and documentaries , and synopses , for films . despite the different nature of these domains , the abstractive summaries created by humans , used for evaluation , share similar scores across metrics .
the best performing algorithms are lsa , for news and documentaries , and lexrank for films .
moreover , we conducted experiments combining scripts and subtitles for films , in order to assess the performance of generic algorithms by inclusion of redundant content .
our results suggest that this combination is unfavorable . additionally , it is possible to observe that all algorithms behave similarly for both subtitles and scripts .
as previously mentioned , the average of the scores follows closely the values of r - su4 , suggesting that r - su4 is able to capture concepts derived from both unigrams and bigrams .
we plan to use subtitles as a starting point to perform video summaries of films and documentaries . for films ,
the results from our experiments using plot summaries show that the summarization of scripts only marginally improved performance , in comparison with subtitles .
this suggests that subtitles are a viable approach for text - driven film and documentary summarization .
this positive aspect is compounded by their being broadly available , as opposed to scripts .
this work was supported by national funds through fundao para a cincia e a tecnologia ( fct ) with reference uid / cec/50021/2013 .
40 natexlab#1#1[1]`#1 ` [ 2]#2 [ 1]#1 [ 1]http://dx.doi.org/#1 [ ] [ 1]pmid:#1 [ ] [ 2]#2 , , , , , . , in : . , pp . .
, , , . . ,
, , . , in : , pp . .
, , . , in : , pp . . , .
, , , . , in : , pp . .
, , , , . , in : , pp . . , , . , in : , pp .
, , , , , , . , in : , pp . .
, , , . . ,
, in : , pp . . , , . , in : , pp . .
, , . , . ,
. . , . , , , , , . , in : , , , , , , , , ( eds . ) , , . , , , . , in : , .
, in : , pp . . , , , , . , in : , pp . .
, , , , , , , , , , . , in : , pp . .
, , . , in : , pp . . , , , a. , in : , pp . .
, , , b. , in : , pp . .
ncleo interinstitucional de lingustica computacional ( nilc ) .
, in : , pp . . , , , , . . , . , , ,
. . , . , , , , , ,
, , . , in : , pp . .
, , . , in : , pp . .
, , . , in : , pp .
. , . . , . , , . , in : , pp . . , , ,
, , , , . , in : , pp . . | we assess the performance of generic text summarization algorithms applied to films and documentaries , using extracts from news articles produced by reference models of extractive summarization .
we use three datasets : ( i ) news articles , ( ii ) film scripts and subtitles , and ( iii ) documentary subtitles .
standard rouge metrics are used for comparing generated summaries against news abstracts , plot summaries , and synopses .
we show that the best performing algorithms are lsa , for news articles and documentaries , and lexrank and support sets , for films . despite the different nature of films and documentaries ,
their relative behavior is in accordance with that obtained for news articles .
automatic text summarization , generic summarization , summarization of films , summarization of documentaries |
the physical mechanisms describing stellar birth are fairly well understood for low- and intermediate - mass stars but still under debate for their high - mass analogues @xcite . on larger scales , observations show that star forming mechanisms are of relatively poor efficiency , as only a fraction of the gas reservoir in the universe is turning into stars . typical star formation efficiencies ( sfe ) are of the order 3 - 6% in the galaxy @xcite , and 5% or less in other galaxies @xcite . still , in extreme environments like in the starburst galaxy arp 220
, the sfe can reach 50% @xcite suggesting that the star forming mechanism at work in such objects could be of a different nature .
for instance the feedback into the interstellar medium ( ism ) from short - lived massive stars seems to influence the yield of star formation in their local environment .
as they evolve off the main sequence , high - mass stars produce a copious amount of energy while still embedded in their native cocoon ; such disruption of a molecular cloud leads to gravitational instabilities and possibly to the onset of a new episode of star formation .
@xcite showed that under certain conditions runaway triggering can take place around massive ob stars .
in other cases however , turbulence and magnetic fields can have a negative feedback on the local ism and lead to the suppression of the star forming activity @xcite .
@xcite provides an exhaustive review of the theoretical framework to study the dynamical triggers of star formation . in this context , we use the data from the _ spitzer _ legacy surveys mipsgal @xcite and glimpse @xcite to study the star forming activity currently taking place in the vulpecula ob association ( hereafter vul@xmath0ob1 ) .
vul@xmath0ob1 hosts nearly one hundred ob stars and three bright hii regions known as sharpless objects 86 , 87 and 88 @xcite . according to @xcite and @xcite ,
the star forming activity occurring in vul@xmath0ob1 might have been triggered by a common external source , and star formation might be propagating from one hii region to another through the expansion of a supernova shock front .
the aim of this paper is to study the triggered star formation on scales as large as @xmath3100 pc in vul@xmath0ob1 , which could help us understand star formation mechanisms within other giant molecular clouds in the galaxy and beyond .
we use the method developed by @xcite to obtain a census of young stellar objects ( hereafter ysos ) in vul@xmath0ob1 based on infrared color and magnitude criteria , and we rely on the properties and nature of these ysos to trace episodes of star formation .
the mipsgal and glimpse sensitivity makes our search for ysos in vul@xmath0ob1 biased towards massive objects that reveal the most recent star forming activity .
in section [ sec : obs_cat ] we present the dataset used in the paper . in section [ sec : vulob1 ] we give a comprehensive description of vul@xmath0ob1 encompassing the three hii regions as well as the dozen pillar - like structures we discovered in this region .
section [ sec : yso ] describes the identification process of young stellar objects based on their ir - excess emission .
we present and discuss our results in section [ sec : result ] , and we give our conclusions in section [ sec : conclu ] .
the _ spitzer space telescope _
@xcite observed over 270 square degrees of the inner galactic plane in six wavelength bands as part of two legacy programs : the galactic legacy infrared mid - plane survey extraordinaire ( glimpse ; e. churchwell pi ; * ? ? ?
* ) and the mips galactic plane survey ( mipsgal ; s. carey pi ; * ? ? ?
the vulpecula region was covered by glimpse@xmath0i ( pid 188 ) and mipsgal@xmath0i ( pid 20597 ) .
it was imaged with the irac camera @xcite at 3.6 , 4.5 , 5.8 and 8.0@xmath4 m , and with the mips camera @xcite at 24 and 70@xmath4 m .
the angular resolution of _ spitzer _ is 2 , 6 and 18 at 8 , 24 and 70@xmath4 m , respectively .
the mopex package @xcite was used to build mosaics of the sky about 1 square degree wide from individual basic calibrated data ( bcd ) frames .
details of the post - processing and bcd pipeline modifications are described in @xcite for glimpse and in @xcite for mipsgal .
the large vulpecula mosaics presented in this paper are also built with mopex using the 1 square degree plates available through the infrared science archive ( irsa ) .
the search for yso candidates in vulpecula relies on the point source catalogs ( psc ) generated from the glimpse and mipsgal surveys . for glimpse
, point sources are extracted using a modified version of the point - spread function ( psf ) fitting program daophot @xcite .
the glimpse psc we used is the enhanced data product v2.0 available on the irsa website . for mipsgal
, sources are extracted from the 24@xmath4 m maps using a pixel - response function fitting method ( shenoy et al . in preparation ) and the apex software developed at the _ _ spitzer science center__. this catalog is then merged with the glimpse psc to include irac and 2mass fluxes if they are available .
70@xmath4 m sources are extracted from reprocessed 70@xmath4 m maps ( paladini et al . in preparation ) using the _ starfinder _ software @xcite , and matched to the mipsgal psc with a search radius of 9 .
we complement our dataset with visible , sub - millimeter , millimeter and radio data to obtain a broader view of the vulpecula complex , and to investigate the morphology , star formation history and possible relationship between the hii regions .
we make use of the vla galactic plane survey ( vgps ) data , which consists of hi line and 21-cm continuum emission data with a resolution of 1@xmath51@xmath6 km s@xmath7 and 1 respectively @xcite .
these data are analysed in details in section [ subsec : h2 ] .
we also use the virginia tech spectral - line survey ( vtss ) which provides arcminute - resolution images of the 6563 h@xmath8 recombination line of atomic hydrogen in the vulpecula region @xcite .
s. bontemps ( private communication , see @xcite for details ) also provided our team with a 1.2-resolution extinction map of vul@xmath0ob1 generated from stellar counts and colors derived from the 2mass point source catalog ( see figure [ fig : av_co ] ) .
the visual extinction in vulpecula ranges from 3 to 20 with a distribution peaking at av@xmath35 .
we also use the co survey of the galaxy published in @xcite .
however the relatively low spatial resolution of @xmath37.5 over vul@xmath0ob1 prevented a relevant detailed study of the co emission .
we find nonetheless a very good correlation between the co emission and the visual extinction that both trace the cold and dusty molecular material ( cf figure [ fig : av_co ] ) .
in addition , we used the sub - millimeter data obtained during the second flight of the balloon - borne large aperture submillimeter telescope ( blast , * ? ? ?
* ) , where a 4 square degree region was mapped around sh2 - 86 at 250 , 350 and 500@xmath4 m @xcite .
the angular resolution of the blast maps are 40 , 50 and 60 at 250 , 350 and 500@xmath4 m , respectively . and
finally , we exploit the recently released 1.1@xmath0 mm data of the galactic plane survey carried out with caltech submillimeter observatory / bolocam ( aguirre et al . in preparation ) and providing an angular resolution of 31 .
vul@xmath0ob1 is an active star forming region located in the galactic plane at a longitude l@xmath360 . in this article
we focus on the region centered at @xmath9 and covering about 6.6 square degrees ( see figures [ fig : av_co ] and [ fig : vulob1 ] ) .
describe vul@xmath0ob1 as an oval region about 3.5 by 1.5 hosting the star cluster ngc@xmath06823 .
nearly one hundred hot massive stars are found in the direction of vulpecula from the catalog of ob stars compiled by @xcite .
some of these stars have already been associated with and are responsible for the ionization of three bright hii regions in vul@xmath0ob1 , namely sh2 - 86 , 87 and 88 @xcite .
figure [ fig : vulob1 ] presents the mid - infrared morphology of vul@xmath0ob1 and the location of the ob stars .
the vgps 21-cm continuum emission covering vul@xmath0ob1 reveals two bright circular compact sources ( @xmath31.3 in diameter ) coincident with sh2 - 87 and sh2 - 88 , and an extended dimmer region ( 27@xmath512 ) reminiscent of the infrared morphology of sh2 - 86 .
the supernova remnant snr g59.5 + 0.1 @xcite is also clearly visible in the radio continuum data at @xmath10 , its measured diameter is 15 ( cf section [ subsec : snr ] ) .
our analysis of the vgps hi line data indicates that the three hii regions are neighbors ( cf section [ subsec : h2 ] ) , as was previously noted by @xcite , @xcite and @xcite .
distance determinations for the three sharpless objects and their exciting stars range from 1.5 to 3.2 kpc @xcite . following the arguments from @xcite
, we adopt a common distance of 2.3 kpc for the three hii regions . in the catalog of @xcite we find about 30 ob stars in the direction of sh2 - 86 and sh2 - 87 , in what we identify as the _ pillars region _ ( these stars might be responsible for the existence of pillar - structures , see section [ subsubsec : pillars ] ) .
we check their possible association with vul@xmath0ob1 by computing their distance moduli based on the @xmath11-spectral type relation and the intrinsic colors given by @xcite .
we assume a ratio of total - to - selective absorption @xmath12 typical of the diffuse ism , and we associate a visual extinction to each star based on our extinction map of vul@xmath0ob1 . out of the 16 stars for which spectral type and magnitude information are available in the @xcite catalog , 12 fall between 1.3 and 2.3 kpc with a mean distance from the sun of @xmath13 kpc .
we find similar distances using intrinsic colors and absolute magnitudes given by @xcite and @xcite respectively .
although the uncertainties are significant , which is mostly due to the large uncertainty on the values of the absolute magnitudes , our distance estimate for these stars is consistent with an association with the vulpecula complex .
sh2 - 86 is an extended hii region about 40 wide @xcite located in the southern part of vulpecula ( figures [ fig : av_co ] , [ fig : vulob1 ] and [ fig : sh86_i3i4m1 ] in the appendices ) .
it is excited by the open cluster ngc 6823 , which is part of the ob association vul@xmath0ob1 @xcite .
@xcite estimate the age of the star cluster at @xmath14 myr old . in the atomic gas distribution derived from the vgps data
, sh2 - 86 appears as an oblong hole at velocities between 26 and 31 km s@xmath7 .
we find that the dimension of the shell is approximately 60 along the galactic plane and 25 across it . in the optical , the south - eastern part of sh2 - 86
is very contrasted .
it exhibits silhouetted pillar - like structures pointing towards ngc 6823 , emission and reflection nebulae as well as filamentary structures .
the infrared emission is also contrasted in this region tracing the complex interplay between the uv radiation and the surrounding dusty interstellar medium .
+ @xcite also report the detection of 49 compact sub - millimeter sources associated with sh2 - 86 using the balloon - borne large aperture submillimeter telescope ( blast ) .
the presence of these clumps , with masses ranging from 14 to 700 m@xmath15 , indicates the capability of sh2 - 86 to form massive stars . sh2 - 87 and sh2 - 88 are also hii regions initially discovered by @xcite .
these regions are active sites of star formation as indicated by the presence of h@xmath16o maser line emissions and bipolar outflows @xcite .
bolocam data show bright sources at the location of sh2 - 87 and sh2 - 88 indicating the presence of cold core candidates .
our analysis of the vgps data reveals that the morphology of the hi shell around sh2 - 87 at velocities between 22 and 26 km s@xmath7 is very similar to that of the 24@xmath4 m emission .
the hii region appears as a hole in the hi distribution as the hydrogen is ionized by an embedded b0 star @xcite .
a compact source seen in absorption sits at the center of sh2 - 87 , its position is coincident with the location of the peak emission at 24@xmath4 m @xmath17 . @xcite and
@xcite give a detailed description of this hii region .
sh2 - 88 is a diffuse nebula of diameter @xmath320 located at @xmath18 , which is excited by the o8 star bd + 253952 @xcite .
the star formation activity taking place in sh2 - 88 occurs in a couple of nebular knots identified by @xcite , sh2 - 88a and sh2 - 88b , located 15 southeast of sh2 - 88 .
these knots are responsible for the bright 24@xmath4 m emission visible in figure [ fig : vulob1 ] at @xmath19 .
sh2 - 88b is actually the brightest of the three sharpless objects at mid - ir wavelengths .
figure [ fig : sh88b_i1i3i4 ] in the appendices shows a composite image of sh-88b in irac bands .
it consists of a compact cometary hii region and an ultracompact ( uc ) hii region .
@xcite present a detailed analysis of sh2 - 88b and its stellar content . at radio wavelength ,
the association of sh2 - 88 with a hi shell is somewhat more difficult than in the other two cases as the edges of the shell are not as well defined .
we detect however a faint hole at velocities between 22 and 26@xmath0 km s@xmath7 as well as a compact source seen in absorption in the center of sh2 - 88b .
[ cols= " > " , ]
we have presented a thorough description of the vulpecula ob association .
we have complemented the existing observations of isolated objects in vul@xmath0ob1 with a global view of the whole star forming complex from an infrared perspective .
we exploited _
legacy surveys mipsgal and glimpse data to identify 856 young stellar objects with ir - excess emission .
we rely on the nature and properties of these objects to highlight the recent activity of star formation in the complex , and we look for evidences of triggered star formation . we find two populations of yso candidates : one population of distributed objects that likely contains ir - bright evolved stars and some genuine ysos born in isolation or ejected from their parental cloud ; and another population of clustered yso candidates whose spatial distribution correlates very well with the mid - ir morphology of the complex .
yso candidates surface density peaks locally around the three sharpless objects , the hyper compact hii regions , and other embedded star clusters like cr404 .
the vigorous star forming activity around these energetic sources is consistent with scenarios of triggered star formation mechanisms .
still we can not ascertain that these stars were born as a direct consequence of their extreme environment , nor that they would have formed in the absence of what we consider the triggering agents .
nevertheless our analysis allowed to rule out the scenario of @xcite according to which star formation was propagating from sh2 - 88 to sh2 - 86 .
we rather find that the evolutionary stage of the yso population across vul@xmath0ob1 is homogeneous , and thus consistent with the scenario of @xcite .
we have further reported the discovery of a dozen pillar - like structures in vul@xmath0ob1 , and we comment on their morphology from the near - ir to the radio regime .
we were not able to identify the energetic source(s ) responsible for the molding of the pillars , but we argue that these objects are indeed associated with the ob association .
our finding of yso candidates at the tip of most of the pillars is consistent with mechanisms of triggered star formation on small scales .
the authors would like to thank s. bontemps for providing the extinction map of vul@xmath0ob1 .
this work is based on observations made with the _
spitzer space telescope _ , which is operated by the jet propulsion laboratory , california institute of technology under a contract with nasa .
this research used the facilities of the canadian astronomy data centre operated by the national research council of canada with the support of the canadian space agency .
the national radio astronomy observatory is a facility of the national science foundation operated under cooperative agreement by associated universities , inc .
the virginia tech spectral - line survey ( vtss ) is supported by the national science foundation .
this research has made use of the simbad database , operated at cds , strasbourg , france .
this research has made use of the nasa/ ipac infrared science archive , which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
_ facilities : _
l@c@c@c@c@r@@ @xmath20l@c _ vulp1 _ & 60.18 & -0.31 & 296.36 & 23.88 & 5 - 8 & 0.5 & 250 + _
vulp2 _ & 60.03 & -0.35 & 296.32 & 23.74 & 1 & 0.5 & 230 + _ vulp3 _ & 59.97 & -0.31 & 296.25 & 23.70 & 8 & 2 & 260 + _ vulp4 _ & 59.81 & -0.29 & 296.15 & 23.58 & 3 & 1 & 268 + _ vulp5 _ & 60.06 & 0.10 & 295.91 & 23.99 & 4 & 2 & 335 + _ vulp6 _ & 60.11 & 0.21 & 295.83 & 24.10 & 5 & 2 - 4 & 350 + _ vulp7 _ & 60.38 & 0.21 & 295.98 & 24.33 & 9 & 1 - 4 & 17 + _ vulp8 _ & 60.90 & 0.22 & 296.26 & 24.78 & 11 & 4 & 55 + _ vulp9 _ & 61.10 & -0.45 & 297.00 & 24.61 & 4 & 1 & 158 + _ vulp10 _ & 60.77 & -0.63 & 297.00 & 24.24 & 4 & 2 & 190 + _ vulp11 _ & 60.21 & -0.45 & 296.52 & 23.84 & 4 & 2 & 280 + _ vulp12 _ & 59.60 & -0.14 & 295.90 & 23.47 & 3 & 1 & 133 + _ vulp13 _ & 59.51 & -0.22 & 295.92 & 23.36 & 3 & 0.5 & 160 + _ vulp14 _ & 59.49 & -0.19 & 295.88 & 23.35 & 2 & 0.5 & 150 + l@c@c@c@c@c@c@c & bright pillar & & & & & & + [ -1ex ] & compact source & & & & & & + [ 1ex ] _
vulp2 _ & faint & faint & faint & none & none & none & n + [ 1ex ] & bright pillar & bright & bright & & & & + [ -1ex ] & compact source & structured & structured & & & & + [ 1ex ] _
vulp4 _ & faint tip & faint & faint & none & none & none & y + [ 1ex ] & faint pillar & bright pillar & bright pillar & no pillar & no pillar & no pillar & + [ -1ex ] & pedestal & pedestal & pedestal & faint pedestal & faint pedestal & faint pedestal & + [ 1ex ] _
vulp6 _ & faint & faint & faint & none & none & none & y + [ 1ex ] & faint & bright & faint & & & & + [ -1ex ] & structured & structured & structured & & & & + [ 1ex ] & & faint & faint & & & & + [ -1ex ] & & structured & structured & & & & + [ 1ex ] _
vulp9 _ & faint & bright tip & bow shock & faint tip & none & faint & n + [ 1ex ] & & bright tip & & & & & + [ -1ex ] & & possible jets & & & & & + [ 1ex ] _
vulp11 _ & bright tip & bright tip & bright tip & faint tip & faint tip & faint tip & n + [ 1ex ] & bright tip & bright tip & bright tip & & & & + [ -1ex ] & pedestal & pedestal & pedestal & & & & + [ 1ex ] _
vulp13 _ & faint & bright edges & bright edges & none & none & none & y + [ 1ex ] & & & bow shock & & & & + [ -1ex ] & & & bright tip & & & & + [ 1ex ] l@c@r@.@l@r@.@l@r@.@l@r@.@l@r@.@l@[email protected]@r@.@l@r@.@l@r@.@l@r@.@l@c sstglmc g059.6291 - 00.8301 & 59.6291 & -0&8301 & 12&97 & 12&59 & 12&35 & 11&96 & 11&63 & 11&41 & 9&84 & .&. & .&. & galaxy + sstglmc g059.7319 + 00.2144 & 59.7320 & 0&2144 & 12&95 & 12&47 & 12&23 & 11&98 & 11&71 & 11&68 & 10&23 & 7&77 & .&. & galaxy + sstglmc g059.9431 + 00.1783 & 59.9432 & 0&1783 & .&.
& 15&06 & 14&19 & 13&02 & 12&40 & 11&68 & 9&76 & 6&19 & .&. & galaxy + sstglmc g060.9219 - 00.1041 & 60.9220 & -0&1041 & 14&37 & 13&44 & 12&98 & 12&15 & 11&87 & 11&83 & 10&53 & .&. & .&. & galaxy + sstglmc g061.1282 - 00.7549 & 61.1282 & -0&7549 & 13&60 & 12&59 & 12&32 & 12&11 & 12&14 & 11&77 & 10&63 & .&. & .&. & galaxy + sstglmc g061.6453 - 00.7439 & 61.6454 & -0&7439 & .&. & .&. & .&. & 12&19 & 11&91 & 11&79 & 9&53 & .&. & .&. & galaxy + sstglmc g060.0154 + 00.1111 & 60.0154 & 0&1111 & .&. & .&. & .&.
& 12&88 & 10&83 & 10&01 & 9&56 & .&. & -4&96 & shock + sstglmc g059.6940 + 00.1840 & 59.6940 & 0&1840 & .&. & .&. & .&. & 12&80 & 11&62 & 11&14 & 10&69 & .&. & -4&82 & shock + sstglmc g059.7893 + 00.6298 & 59.7893 & 0&6298 & .&. & .&. & .&. & 12&61 & 10&75 & 10&19 & 9&27 & .&. & -4&81 & shock + sstglmc g059.3780 - 00.2438 & 59.3781 & -0&2438 & .&. & .&. & 14&84 & 12&76 & 11&23 & 10&63 & 10&30 & 3&14 & -4&77 & shock + sstglmc g059.7973 + 00.0750 & 59.7973 & 0&0750 & .&. & .&. & .&. & 13&22 & 11&04 & 11&11 & 10&75 & .&. & .&. & shock + sstglmc g059.6366 - 00.1864 & 59.6366 & -0&1864 & .&. & .&. & 14&73 & 11&48 & 9&56 & 8&79 & 8&33 & 3&13 & .&. & shock + sstglmc g058.6970 + 00.6316 & 58.6971 & 0&6316 & .&. & .&. & .&.
& 12&88 & 11&41 & 10&80 & 9&70 & 3&17 & -4&15 & shock + sstglmc g059.0385 - 00.2512 & 59.0386 & -0&2512 & .&. & 14&47 & 13&19 & 11&32 & 10&15 & 9&89 & 8&85 & 3&67 & .&. & shock + adams , f. c. , lada , c. j. , & shu , f. h. , 1987 , , 312 , 788 allen , l. e. , et al . , 2004 , , 154 , 363 anantharamaiah , k. r. , viallefond , f. , mohan , n. r. , goss , w. m. , & zhao , j. h. , 2000 , , 537 , 613 andr , p. , ward - thompson , d. , & barsony , m. , 1993 , , 406 , 122 andr , p. , motte , f. , & neri , r. , 2000 , asp .
conf . , 217 , 152 baraffe , i. , chabrier , g. , allard , f. , & hauschildt , p. h. , 1998 , , 337 , 403 barsony , m. , 1989 , , 345 , 268 benjamin , r. a. et al . , 2003 , , 115 , 953 bertoldi , f. , 1989 , , 346 , 735 bica , e. , bonatto , c. , & dutra , c. m. , 2008 , , 489 , 1129 bonnell , i. a. , bate , m. r. , price , n. m. , 1996 , , 279 , 121 bowler , b. p. , waller , w. h. , megeath , s. t. , patten , b. m. , & tamura , m. , 2009 , , 137 , 3685 brand , j. & blitz , l. , 1993 , , 275 , 67 calvet , n. , hartmann , l. , kenyon , s. j. , & whitney , b. a. , 1994 , , 434 , 330 calvet , n. , daleesio , p. , hartmann , l. , wilner , d. , walsh , a. , & sitko , m. , 2002 , , 568 , 1008 cappa , c. , pineault , s. , arnal , e. m. , & cichowolski , s. , 2002 , , 395 , 955 carey , s. j. et al . , 2009 , , 121 , 76 carlqvist , p. , gahm , g. f. , kristen , h. , 2003 , , 403 , 399 chapin , e. l. et al .
, 2008 , , 681 , 428 chavarra , l. a. , allen , l. e. , hora , j. l. , brunt , c. m. , & fazio , g. g. , 2008 , , 682 , 445 cohen , m. , wheaton , w. a. & megeath , s. t. , 2003 , , 126 , 1090 dalessio , p. , et al . , 2005 , , 621 , 461 dame , t. m. , hartmann , d. & thaddeus , p. , 2001
, , 547 , 792 deharveng , l. , nadeau , d. , zavagno , a. , & caplan , j. , 2000 , , 360 , 1107 dennison , b. , simonetti , j. h. , & topasna , g. a. , 1998 , publications of the astronomical society of australia , 15 , 147 diolaiti , e. , bendinelli , o. , bonaccini , d. , close , l. , currie , d. , & parmeggiani , g. , 2000 , , 147 , 335 efremov , y. n. & elmegreen , b. g. , 1998 , , 299 , 588 ehlerov , s. , palou , j. & huchtmeier , w. k. , 2001 , , 374 , 682 elmegreen , b. g. , 1998 , astronomical society of the pacific conference series , 148 , 150 evans , n. j. , et al . , 2009 , , 181 , 321 fazio , g. g. , et al .
, 2004 , , 154 , 10 felli , m. & harten , r. h. , 1981 , , 100 , 42 fich , m. & blitz , l. , 1984 , , 279 , 125 flaherty , k. m. , 2007 , , 663 , 1069 garmany , c. d. & stencel , r. e. , 1992 , , 94,211 green , d. a. , 2006 , ` a catalogue of galactic supernova remnants ( 2006 april version ) ' , astrophysics group , cavendish laboratory , cambridge , uk ( available at `` http://www.mrao.cam.ac.uk/surveys/snrs/ '' ) greene , t. p. , wilking , b. a. , andre , p. , young , e. t. , & lada , c. j. , 1994 , , 434 , 614 gritschneder , m. , naab , t. , walch , s. , burkert , a. , & heitsch , f. , 2009 , , 694 , 26 guetter , h. h. , 1992 , , 103 , 197 guieu , s. et al . , 2009 , , 697 , 787 guseinov , o. h. , ankay , a. , sezer , a. , & tagieva , s. o. , 2003 , astronomical and astrophysical transactions , 22 , 273 gutermuth , r. a. et al . , 2008 , , 674 , 336 hartmann , l. , megeath , s. t. , allen , l. , luhman , k. , calvet , n. , dalessio , p. , franco - hernandez , r. , & fazio , g. , 2005 , , 629 , 681 harvey , p. , mern , b. , huard , t. l. , rebull , l. m. , chapman , n. , evans , n. j. , & myers , p. c. , 2007 , , 623 , 1149 hester , j. j. et al . , 1996 , , 111 , 2349 hosokawa , t. & inutsuka , s .-
i . , 2006 , , 648 , 131 hora , j. l. , et al . , 2008 , , 135 , 726 hoyle , f. , shanks , t. , & tanvir , n. r. , 2003 , , 345 , 269 indebetouw , r. , et al . , 2005 , , 619 , 931 indebetouw , r. , robitaille , t. p. , whitney , b. a. , churchwell , e. , babler , b. , meade , m. , watson , c. , & wolfire , m. , 2007 , , 666 , 321 johnson , h. l. , 1958 , lowell observatory bulletin , 4 , 37 jrgensen , j. k. , schier , f. l. & van dishoeck , e. f. , 2002 , , 389 , 908 karr , j. l. & martin , p. g. , 2003 , , 595 , 900 kharchenko , n. v. , piskunov , a. e. , rser , s. , schilbach , e. , & scholz , r. d. , 2005 , , 438 , 1163 koenig , x. p. , allen , l. e. , gutermuth , r. a. , hora , j. l. , brunt , c. m. , & muzerolle , j. , 2008 , , 688 , 1142 kroupa , p. , 2001
, , 322 , 231 lada , c. j. , 1987 , iau symposium , 115 , 1 lada , c. j. , et al . , 2006 , , 131 , 1574 landolt - bornstein , 1982 , bulletin of the astronomical institutes of czechoslovakia , 33 , v.2,subvol .
b lefloch , b. & lazareff , b. , 1994 , , 289 , 559 lortet - zuckermann , m. c. , 1974 , , 30 , 67 lutz , d. , 1999 , in the universe as seen by iso , ed .
p. cox & m. f. kessler , esa sp-427 , 63 makovoz , d. , khan , i. , & masci , f. , 2006 , proc .
spie , 6065 , 330 marleau , f. , et al .
, 2008 , , 136 , 662 massey , p. , johnson , k. e. & degioia - eastwood , k. , 1995 , , 454 , 151 mckee , c. f. , & ostriker , e. c. , 2007 , , 45 , 565 meade , m. r. et al . , 2007 , glimpse i v2.0 data release - http://data.spitzer.caltech.edu/popular/ glimpse/20070416_enhanced_v2/documents/ glimpse1_dataprod_v2.0.pdf megeath , s. t. , allgaier , e. , young , e. , allen , t. , pipher , j. l. , & wilson , t. l. , 2009 , , 137 , 4072 melioli , c. , de gouveia dal pino , e. m. , de la reza , r. & raga , a. , 2006 , , 373 , 811 mern , b. et al . , 2008 , , 177 , 151 miao , j. , white , g. j. , nelson , r. , thompson , m. , & morgan , l. , 2006 , , 369 , 143 mizuno , d. r. et al .
, 2008 , , 120 , 1028 mizuta , a. , kane , j. o. , pound , m. w. , remington , b. a. , ryutov , d. d. , & takabe , h. , 2006 , , 647 , 1151 molinari , s. , pezzuto , s. , cesaroni , r. , brand , j. , faustini , f. & testi , l. , 2008 , , 481 , 345 molinari , s. , 2008 , 37th cospar scientific meeting , 37 , 2085 nomura , h. & kamaya , h. , 2001 , , 121 , 1024 pascale , e. , et al . , 2008 , , 681 , 400 pigulski , a. , kolaczkowski , z. , kopacki , g. , 2000 , acta astronomica , 50 , 113 price , d. j. , & bate , m. r. , 2009 , arxiv e - prints , 0904.4071 reach , w. t. et al . , 2004 , , 154 , 385 reach , w. t. et al
, 2005 , , 117 , 978 reach , w. t. et al . , 2006 , , 131 , 1479 reach , w. t. et al . , 2009 , , 690 , 683 reed , b. c. , 2003 , , 125 , 380 rieke , g. h. et al . , 2004 , , 154 , 25 rieke , g. h. et al . , 2008 , , 135 , 2245 robitaille , t. p. , whitney , b. a. , indebetouw , r. , & wood , k. , 2007 , , 169 , 328 robitaille , t. p. , et al . , 2008 , , 136 , 2413 rownd , b. k. & young , j. s. , 1999 , , 118 , 670 schneider , n. , bontemps , s. , simon , r. , jakob , h. , motte , f. , miller , m. , kramer , c. , & stutzki , j. , 2006 , , 458 , 855 sharpless , s. , 1959 , , 4 , 257 spitzer , l. j. , 1954 , , 120 , 1 srinivasan , s. et al .
, 2009 , , 137 , 4810 stern , d. et al . , 2005 , , 631 , 163 stetson , p. b. , 1987 , , 99 , 191 stil , j. m. et al .
, 2006 , , 132 , 1158 stone , m. e. , 1970 , , 159 , 293 sugitani , k. et al . , 2002 , , 565 , 25 taylor , a. r. , wallace , b. j. , & goss , w. m. , 1992 , , 103 , 931 terebey , s. , shu , f. h. , & cassen , p. , 1984
, , 286 , 529 turner , d. g. , 1986 , , 167 , 157 urquhart , j. s. , white , g. j. , pilbratt , g. l. , & fridlund , c. v. m. , 2003 , , 409 , 193 vicente , s. m. & alves , j. , 2005 , , 441 , 195 wegner , w. , 2006 , , 371 , 185 werner , m. w. , et al . , 2004 , , 154 , 1 xu , j. w. , zhang , x. z. & han , j. l. , 2005,chinese journal of astronomy and astrophysics , 5 , 165 xue , r. & wu , y. , 2008 , , 680 , 446 zavagno , a. , deharveng , l , comern , f. , bran , j. , massi , f. , caplan , j. , & russeil , d. , 2006 , , 446 , 171 | the vulpecula ob association , vul@xmath0ob1 , is a region of active star formation located in the galactic plane at 2.3@xmath1kpc from the sun .
previous studies suggest that sequential star formation is propagating along this 100@xmath1pc long molecular complex . in this paper
, we use _ spitzer _ mipsgal and glimpse data to reconstruct the star formation history of vul@xmath0ob1 , and search for signatures of past triggering events .
we make a census of young stellar objects ( yso ) in vul@xmath0ob1 based on ir color and magnitude criteria , and we rely on the properties and nature of these ysos to trace recent episodes of massive star formation .
we find 856 yso candidates , and show that the evolutionary stage of the yso population in vul@xmath0ob1 is rather homogeneous - ruling out the scenario of propagating star formation .
we estimate the current star formation efficiency to be @xmath2% .
we also report the discovery of a dozen pillar - like structures , which are confirmed to be sites of small scale triggered star formation . |
the @xmath0 is a new hadronic resonance discovered in 2003 by the belle collaboration @xcite and subsequently confirmed by the cdf , babar , and d0 collaborations @xcite .
all the properties of the @xmath0 that have been measured thus far are compatible with its identification as a weakly - bound molecule whose constituents are a superposition of the charm meson pairs @xmath1 and @xmath2 @xcite . because it is so weakly bound
, the @xmath0 has many features in common with the deuteron , which is a weakly - bound baryonic molecule consisting of a proton and a neutron .
their binding energies are both small compared to the natural energy scale associated with pion exchange : @xmath3 , where @xmath4 is the reduced mass of the two constituents .
the binding energy 2.2 mev of the deuteron is small compared to the natural scale of about 20 mev . after taking into account a recent precision measurement of the @xmath5 mass by the cleo collaboration @xcite , the difference between the measured mass of the @xmath0 and the @xmath1 threshold is @xmath6 this is small compared to the natural energy scale of about 10 mev .
the belle collaboration has set an upper bound on the width of the @xmath7 @xcite : @xmath8 this is also small compared to the natural energy scale .
the small width is most easily understood if the @xmath7 is below the @xmath1 threshold .
because the binding energies of the deuteron and the @xmath0 are small compared to the natural energy scales , they have universal properties that are determined by the large scattering length @xmath9 of the constituents @xcite .
the universality of few - body systems with a large scattering length has many applications in atomic , nuclear , and particle physics @xcite .
the universal features of the @xmath0 were first exploited by voloshin to describe its decays into @xmath10 and @xmath11 , which can proceed through decay of the constituent @xmath12 or @xmath13 @xcite .
universality has also been applied to the production process @xmath14 @xcite , to the line shape of the @xmath7 @xcite , and to decays of @xmath7 into @xmath15 and pions @xcite .
these applications rely on factorization formulas that separate the length scale @xmath9 from all the shorter distance scales of qcd @xcite .
the factorization formulas can be derived using the operator product expansion for a low - energy effective field theory @xcite .
there is one feature of the @xmath0 that is very different from the deuteron .
the mass difference between the constituents @xmath12 and @xmath16 , @xmath17 mev , is very close to the @xmath18 mass : @xmath19 mev .
thus the @xmath12 mass is very close to the @xmath20 threshold : @xmath21 mev .
this is important because the pion is the lightest meson that can be exchanged between the @xmath12 and @xmath16 .
the small energy gap between the @xmath12 mass and the @xmath20 threshold makes it easy for the @xmath12 to make a transition to @xmath20 .
an important consequence is that the @xmath7 could have a substantial 3-body component @xmath10 .
suzuki has argued that the near equality of @xmath22 and @xmath23 implies that the pion - exchange interaction is too weak to bind the @xmath7 @xcite , but this conclusion has been criticized @xcite .
an analysis of the effect of the near equality @xmath24 in the case of the @xmath0 is complicated by other effects that may be comparable in importance .
there are other nearby two - body and three - body thresholds .
the thresholds for the charged mesons @xmath25 and @xmath26 are higher than the @xmath1 threshold by @xmath27 mev .
the @xmath28 and @xmath29 thresholds are higher than the @xmath1 threshold by only @xmath30 mev .
there are several other features that complicate the analysis of the effects of pions on the @xmath7 .
the constituent @xmath31 s are spin-1 particles .
the charge conjugation @xmath32 requires the molecule to be a superposition of charm mesons : @xmath33 .
the interaction of the charm mesons with pions depends on the pion momentum .
a quantitative analysis of the effects of pions on the @xmath7 might require all these effects to be taken into account . since the new feature of the @xmath0 is the near equality between the mass difference @xmath22 of the constituents and the mass @xmath23 of a meson that can be exchanged between them , it is worthwhile to analyze this feature in a simple model that avoids all the other complications of the @xmath7 .
the simplest such model is one in which the primary constituents @xmath34 and @xmath35 are spin-0 mesons with a large positive scattering length and a momentum - independent coupling to a lighter spin-0 meson @xmath36 .
we will calculate the effects of the exchanged meson in this model to next - to - leading order in the @xmath37 coupling constant @xmath38 . in sec .
[ sec : model ] , we write down the lagrangian for the model and discuss its parameters . in sec . [ sec : zeroth ] , we summarize the results of the model at 0@xmath39 order in @xmath38 that were derived previously in refs .
the results include the short - distance production rate for @xmath7 and the line shape of @xmath7 in a short - distance decay channel .
the ultraviolet divergences can be removed by renormalization of the coupling constant for the @xmath40 contact interaction and by renormalization of short - distance coefficients in the operator product expansion . in sec .
[ sec : second ] , we calculate the observables in sec . [
sec : zeroth ] to next - to - leading order in @xmath38 .
we also calculate the short - distance production rate for @xmath41 and the decay rate of @xmath7 into @xmath41 .
there are new ultraviolet divergences that can be removed by renormalization of the @xmath35 mass .
we show that after summing a geometric series of next - to - leading order corrections , all remaining ultraviolet divergences can again be removed by renormalization of the coupling constant for the @xmath40 contact interaction and by renormalization of short - distance coefficients in the operator product expansion .
a further resummation is required to eliminate an infrared divergence at the @xmath40 threshold in the next - to - leading order corrections . in sec .
[ sec : summary ] , we summarize our results .
the two - loop feynman diagrams that arise at next - to - leading order in @xmath38 are calculated in an appendix .
we consider the simplest possible model with a bound state @xmath7 whose mass is close to a 3-body threshold .
the primary constituents @xmath34 and @xmath35 of the bound state are scalar mesons whose masses @xmath42 and @xmath43 satisfy @xmath44 . there is also a scalar meson @xmath36 whose mass @xmath23 is close to the mass difference @xmath45
. we will refer to this model as the _ scalar meson model_. it is defined by a nonrelativistic quantum field theory with three complex scalar fields @xmath34 , @xmath35 , and @xmath36 . the free terms in the lagrangian on a field indicates its complex conjugate .
] are @xmath46 the interaction terms for the scalar meson model are @xmath47 the interaction vertices are illustrated in fig .
[ fig : vertex ] .
the coupling constants @xmath48 and @xmath38 have mass dimensions @xmath49 and @xmath50 , respectively .
we assume that there is a fine - tuning that makes the @xmath40 scattering length large , which implies that the @xmath40 contact interaction with coupling constant @xmath48 must be treated nonperturbatively .
the subscript on @xmath48 emphasizes that it is a bare coupling constant that requires renormalization .
we assume that the @xmath37 interaction can be treated perturbatively and we calculate its effects through order @xmath51 . at this order ,
no renormalization of @xmath38 is required .
however , we will see that nonperturbative renormalization of @xmath48 requires summing a geometric series to all orders in @xmath51 .
there are also ultraviolet divergent corrections to the mass of the @xmath35 which are cancelled by the mass counterterm @xmath52 in eq .
( [ lint ] ) .
contact interaction , ( b ) the @xmath53 interaction , and
( c ) the @xmath35 mass counterterm . [
fig : vertex],width=604 ] it is convenient to introduce concise notations for various combinations of the masses . the sum and difference of the @xmath34 and @xmath35 masses are @xmath54 the reduced masses of @xmath40 and of @xmath55 are @xmath56 [ mred ] if @xmath23 is much smaller than @xmath42 , the natural energy and 3-momentum scales associated with exchange of the meson @xmath36 between the mesons @xmath34 and @xmath35 are @xmath57 and @xmath23 , respectively .
we assume that the energy gap between the @xmath35 mass and the @xmath55 threshold is small compared to the natural energy scale : @xmath58 the total number of heavy mesons @xmath34 and @xmath35 is conserved by the interactions in eq .
( [ lint ] ) .
we are interested in the sector of the theory with two heavy mesons and , more specifically , in the threshold region where the invariant mass @xmath59 of all the particles is very close to @xmath60 : @xmath61 this restricts the possible scattering states to @xmath40 and @xmath41 . for numerical illustrations , we will use masses that correspond to the @xmath62 system .
the meson masses are @xmath63 mev , @xmath64 mev , and @xmath65 mev .
thus @xmath66 mev and @xmath67 mev .
the natural energy scale is @xmath68 mev .
the condition in eq .
( [ dm - small ] ) is only marginally satisfied : @xmath69 .
we take the binding energy of @xmath7 to be @xmath70 mev , which is near the central value of the measurement of @xmath71 in eq .
( [ mx - cleo ] ) .
if we express this binding energy as @xmath72 , the binding momentum is @xmath73 mev .
we impose an ultraviolet cutoff on the momenta @xmath74 of particles in the center - of - momentum frame that restricts them to the region @xmath75 .
this guarantees that all the particles remain nonrelativistic .
it was shown in ref .
@xcite that at leading order @xmath76 , all ultraviolet divergences can be absorbed into @xmath48 and into short - distance coefficients in the operator product expansion . at order
@xmath51 , there are new ultraviolet divergences that can be absorbed into the @xmath35 mass .
we will find that the remaining ultraviolet divergences can again be absorbed into @xmath48 and into short - distance coefficients in the operator product expansion .
the coupling constant @xmath48 must be tuned to near the critical value at which the @xmath40 scattering length diverges in order for the bound state @xmath7 to have a small binding energy @xmath77 . in order to compare results at order @xmath76 and order @xmath51 , we must have a well - defined renormalization prescription for @xmath48 or , equivalently , a clear definition of the binding energy @xmath78 .
we will find that all the amplitudes for processes near the @xmath40 threshold have a resonant term with a common factor @xmath79 that depends on the energy @xmath80 in the rest frame .
for example , the line shape in a short - distance decay channel of @xmath7 produced by a short - distance process has a long - distance factor @xmath81 , where @xmath59 is the invariant mass of the decay products .
the square of the resonant factor can be expressed in the form @xmath82 ^ 2 + [ { \rm i m } \ , \mathcal{a}^{-1}(e)]^2 } . \ ] ] as a function of the real energy @xmath80 , this resonance factor has a peak when @xmath80 is close to @xmath83 . the peak arises because @xmath84 vanishes near that value of @xmath80 and @xmath85 is much smaller than @xmath84 except near that value of @xmath80 . as our renormalization prescription for @xmath48 , we demand that @xmath84 vanishes at an energy determined by a real variable @xmath86 : @xmath87 the dependence of an observable on the bare parameter @xmath48 can be eliminated in favor of the renormalization parameter @xmath86 .
we assume that @xmath86 is small compared to the natural momentum scale : @xmath88 .
the amplitude @xmath79 has a pole at a complex energy @xmath89 near the real energy at which @xmath90 has its maximum .
we choose to define the mass @xmath71 and the width @xmath91 of @xmath7 by expressing that complex energy in the form @xmath92 we also choose to define the binding energy @xmath78 by @xmath93 we assume that the binding energy and the width of @xmath7 are both small compared to the natural energy scale : @xmath94 . in ref .
@xcite , the authors proposed a renormalization scheme in which @xmath48 was tuned to give a prescribed value for the complex parameter @xmath89 .
the binding energy @xmath78 and the width @xmath91 were taken as the input parameters that determine @xmath89 .
this renormalization scheme requires @xmath48 to be a complex coupling constant .
the imaginary part of @xmath48 takes into account inelastic scattering channels for @xmath95 that are outside the energy range described by the effective field theory . in this paper
, we assume that any such inelastic channels do not give a significant contribution to the width of @xmath7 .
the coupling constant @xmath48 is therefore real valued .
we could use the binding energy @xmath78 as the input parameter and eliminate @xmath48 in favor of @xmath78 , and then @xmath91 would be calculable .
we find it more convenient to eliminate @xmath48 in favor of the variable @xmath86 introduced in eq .
( [ reainv : zero ] ) . in this renormalization scheme , @xmath78 and @xmath91
are both calculable in terms of @xmath86 , @xmath38 , and the masses @xmath42 , @xmath43 , and @xmath23 .
the most convenient input for determining the coupling constant @xmath38 is the width @xmath96 of the heavy meson @xmath35 . since @xmath97 , the @xmath53 vertex in fig .
[ fig : vertex](b ) allows @xmath35 to decay into @xmath55 . calculating the width using nonrelativistic phase space , we obtain @xmath98^{1/2 } , \label{gam2-nr}\ ] ] where @xmath4 and @xmath99 are the reduced masses defined in eqs .
( [ mred ] ) .
the value of the coupling constant @xmath38 can be determined from @xmath96 and the masses .
we assume that the width of @xmath35 is small compared to the natural energy scale : @xmath100 .
we also assume @xmath96 is small enough to allow perturbation theory in the coupling constant @xmath38 . for numerical illustrations , we will use values of the parameters that correspond to the @xmath62 system .
we take the width of @xmath35 to be @xmath101 mev , which is approximately the total width for @xmath12 @xcite .
this is much smaller than the natural energy scale @xmath57 and also rather small compared to @xmath102 : @xmath103 . using eq .
( [ gam2-nr ] ) , we determine the numerical value of the @xmath53 coupling constant to be @xmath104 mev@xmath105 . when considering the dependence of observables on @xmath38 , it will be more convenient to regard them as functions of @xmath96 .
we now list all the energies that are assumed to be small compared to the natural scale @xmath57 : * the energy @xmath106 of a @xmath40 or @xmath41 system relative to the @xmath40 threshold , * the energy difference @xmath107 between the @xmath35 mass and the @xmath55 threshold , * the decay width @xmath96 of @xmath35 , which is given in eq .
( [ gam2-nr ] ) , * the binding energy @xmath78 of the molecule , which is approximately equal to @xmath108 , where @xmath86 is a renormalization parameter . we will carry out our calculations under the assumption that these energies are all comparable and that the masses @xmath43 , @xmath42 , and @xmath23 are also comparable .
the only approximations we will make are those that can be justified by a nonrelativistic approximation , such as @xmath109 . at order
@xmath51 , renormalization of the @xmath35 mass is necessary . in the _ on - shell renormalization scheme _ , the parameter @xmath43 in the free lagrangian in eq .
( [ lfree ] ) is chosen to be the physical @xmath35 mass .
the physical mass @xmath43 and the width @xmath96 of @xmath35 can be defined by specifying that the exact @xmath35 propagator at 0 momentum has a pole in the energy at @xmath111 .
the pole in the energy @xmath112 at 0 momentum is the solution to the equation @xmath113 where @xmath114 is the @xmath35 self - energy .
the exact @xmath35 self - energy is the sum of the one - loop diagram in fig .
[ fig : self](a ) and the mass counterterm in fig .
[ fig : self](b ) : @xmath115 there are no other diagrams for the self - energy at higher order in @xmath38 .
the one - loop diagram is calculated in section [ sec : self - energy ] of the appendix .
the analytic expression for the self - energy diagram using dimensional regularization is given in eq .
( [ sigma2-dimreg0 ] ) .
the result in a general regularization scheme is @xmath116 \right)^{1/2 } .
\label{d2self}\end{aligned}\ ] ] the term @xmath117 is real valued and includes a linear ultraviolet divergence .
the divergence is cancelled by a divergent term in the mass counterterm @xmath52 in eq .
( [ sigma : exact ] ) . in the on - shell renormalization scheme , the finite terms in @xmath52
are chosen so that the solution to eq .
( [ pole - eq ] ) satisfies @xmath118 .
the wavefunction normalization factor for a @xmath35 with momentum @xmath119 is @xmath120 self - energy .
[ fig : self],width=529 ] since eq .
( [ pole - eq ] ) can not be solved analytically , the on - shell renormalization scheme is not the most convenient choice .
a simpler renormalization scheme for the @xmath35 mass is to choose the mass counterterm so the self - energy @xmath121 vanishes at the @xmath40 threshold : @xmath122 .
the self - energy is then @xmath123 \right)^{1/2 } .
\label{d2self : ren}\end{aligned}\ ] ] with this renormalization prescription , the parameter @xmath43 differs at order @xmath124 from the physical @xmath35 mass , which is given by the real part of the solution to eq .
( [ pole - eq ] ) . the width @xmath96 , which is given by the imaginary part of the solution to eq .
( [ pole - eq ] ) , reduces at leading order in @xmath38 to @xmath125 .
this agrees with the explicit expression for the decay rate of @xmath35 at order @xmath51 obtained in eq .
( [ gam2-nr ] ) . to order @xmath51 , the wavefunction normalization factor in eq .
( [ z2-def ] ) is @xmath126 \big)^{-1/2 } .\ ] ] since the momentum @xmath119 is restricted to the nonrelativistic region , the term @xmath127 can always be neglected compared to 1 .
the resulting wavefunction normalization factor is independent of @xmath119 : @xmath128^{-1/2}. \label{z2-nr}\ ] ] the correction term is pure imaginary .
the @xmath37 interaction allows a virtual @xmath35 whose energy is greater than @xmath129 to decay into @xmath130 . in the amplitude for such a process at leading order in @xmath38 , the propagator of the virtual @xmath35 diverges .
near such a point in momentum space , the corrections to the @xmath35 propagator from higher orders in @xmath38 are not suppressed .
this problem can be solved by summing the @xmath35 propagator corrections to all orders .
the effect is to add @xmath131 to the denominator @xmath132 of the @xmath35 propagator .
the imaginary part of the self - energy @xmath114 eliminates any divergence in the @xmath35 propagator for real values of the energy .
one drawback of this solution is that it leads to unnecessarily complicated expressions for the integrals in loop diagrams with @xmath35 propagators .
summing the @xmath35 self - energy corrections to all orders is essential only near the pole in the propagator . in other regions of the energy @xmath112
, @xmath114 is a perturbative correction to the denominator of the propagator that is suppressed by a power of @xmath51 . in those regions , summing the @xmath35 propagator corrections to all orders
is optional . at the pole in @xmath112 ,
the self - energy @xmath114 reduces to @xmath133 .
an alternative resummation that eliminates the divergence in the propagator for real values of the energy without changing the order in @xmath38 of the truncation error is to sum only the @xmath133 term in the self - energy to all orders and treat the remainder of the self - energy as a perturbation .
an advantage of this partial resummation is that it leads to simpler expressions for the integrals in loop diagrams with @xmath35 propagators .
a systematic method for implementing this partial resummation is the _ complex mass scheme _ @xcite .
this method has been used in standard model calculations at next - to - leading order @xcite .
the application of this method to the scalar meson model involves adding cancelling terms to the free and interaction terms in the lagrangian : @xmath134 where @xmath96 is the @xmath35 width and @xmath135 is the @xmath35 wavefunction normalization factor . to order @xmath51 , @xmath96 and @xmath135
are given by eqs .
( [ gam2-nr ] ) and ( [ z2-nr ] ) . since @xmath136 is pure imaginary and @xmath135 is complex , the free lagrangian consisting of the sum of eqs .
( [ lfree ] ) and ( [ dlfree ] ) corresponds to a nonhermitian hamiltonian .
the effects of the nonhermiticity are cancelled exactly if the interaction terms in eq .
( [ dlint ] ) are calculated to all orders . at any finite order in perturbation theory ,
the complex mass scheme gives amplitudes that are uniformly accurate to the appropriate order in @xmath38 , even if there is a virtual @xmath35 that decays into @xmath55 .
the effect of adding eq .
( [ dlfree ] ) to the free lagrangian in eq .
( [ lfree ] ) is to change the @xmath35 propagator : @xmath137 the effect of adding eq .
( [ dlint ] ) to the interaction lagrangian in eq .
( [ lint ] ) is to change the counterterm vertex in fig .
[ fig : vertex ] : @xmath138 . \label{newcounterterm}\ ] ] it is convenient to also rescale the fields @xmath139 and @xmath140 by the same complex factor @xmath141 .
this eliminates the factor of @xmath135 from the numerator of the @xmath35 propagator in eq .
( [ newpropagator ] ) .
it also multiplies all the interaction terms in eqs .
( [ lint ] ) and ( [ dlint ] ) by @xmath135 or @xmath141 . since we will use nonperturbative renormalization for the coupling constant @xmath142 for the @xmath40 contact interaction , the factor of @xmath135 only affects the value of the unphysical bare coupling constant @xmath48 . since @xmath143 , the other factors of @xmath135 or @xmath141 in the interaction vertices contribute only at order @xmath124 and higher .
we will be calculating only to order @xmath51 , so we will ignore these factors of @xmath135 and @xmath141 . the perturbation expansion in the complex mass scheme corresponds to ordinary perturbation theory in @xmath38 followed by a resummation to all orders of a subset of terms that are higher order in @xmath38 .
we will refer to calculations to order @xmath76 in this perturbation expansion as _ leading order ( lo ) _ in the complex mass scheme .
we will refer to calculations through order @xmath51 in this perturbation expansion as _ next - to - leading order ( nlo ) _ in the complex mass scheme .
we can use the scalar meson model to describe processes with asymptotic states that include not only the particles @xmath34 and @xmath36 , but also the bound state @xmath7 .
the local composite operator @xmath145 has a nonzero amplitude to create @xmath7 from the vacuum .
thus @xmath146 can be used as an interpolating field for @xmath7 .
it is more convenient to use the operator @xmath147 , because this is a renormalized operator whose matrix elements do not depend on the ultraviolet cutoff @xcite .
the corresponding propagator for @xmath7 is @xmath148 the momentum 4-vector in the fourier transform is @xmath149 , where @xmath80 and @xmath150 are the energy and momentum of @xmath7 . at @xmath151 ,
this propagator has a pole at the complex energy @xmath89 .
near the pole in the energy , the behavior of the propagator at @xmath151 is @xmath152 the energy @xmath89 determines the binding energy @xmath78 and the width @xmath91 of @xmath7 through eq .
( [ epole ] ) .
the residue of the pole defines the wavefunction normalization factor @xmath153 , which is complex valued . because the propagator @xmath154 has mass dimension @xmath49 , @xmath153 has mass dimension @xmath155 . strictly speaking ,
since @xmath7 has a nonzero width @xmath91 , there are no t - matrix elements for processes involving @xmath7 in the initial or final state , because the decay of the @xmath7 prevents it from being a truly asymptotic state .
however if @xmath91 is sufficiently small , the @xmath7 can propagate over a long time interval before decaying , and it can therefore be treated as a quasi - asymptotic state .
we can use the lehmann - symanzik - zimmermann ( lsz ) formalism to define t - matrix elements for the @xmath7 that are closely related to t - matrix elements for decay products of the @xmath7 whose total energy is tuned to the peak of the @xmath7 resonance . to define the t - matrix element @xmath156 for a process with @xmath7 in the initial or final state ,
we start with a green s functions for the operator @xmath157 or @xmath158 . the connected green s function at 4-momentum @xmath159 is amputated by multiplying by the inverse propagator @xmath160 , it is normalized by multiplying by @xmath161 , and then it is evaluated at the real energy @xmath162 .
this gives the t - matrix element multiplied by @xmath163 for a state @xmath7 with the standard nonrelativistic normalization .
the t - matrix element @xmath156 for a state @xmath7 with the standard relativistic normalization is obtained by multiplying by @xmath164 . to justify this prescription
, we consider a process @xmath165 , where @xmath166 and @xmath167 both represent one or more particles and @xmath168 denotes a set of particles that can be decay products of @xmath7 . the t - matrix for this process will have a resonant enhancement when the invariant mass @xmath59 of the particles in @xmath168 is near the mass @xmath71 .
the resonant contribution to the t - matrix element for @xmath165 in the rest frame of @xmath168 can be approximated by @xmath169 \approx i { \cal t}[a \to b + x ] \ , \frac{i}{2 m_x(m - m_x + i \gamma_x/2 ) } \ , i { \cal t}[x \to c ] .
\label{tabc : res}\ ] ] our prescription for t - matrix elements such as @xmath170 $ ] and @xmath171 $ ] involves evaluating the amputated connected green s function at the real energy @xmath172 .
this ensures that the expression in eq .
( [ tabc : res ] ) is as good an approximation as possible to the t - matrix element @xmath173 $ ] near the peak of the resonance . as a simple illustration of the prescription for t - matrix elements for processes involving @xmath7
, we consider the forward scattering process @xmath174 . the relevant green s function in the rest frame of @xmath7 is the negative of the propagator @xmath175 .
this must be multiplied by two factors of @xmath176 , one for the @xmath7 in the inital state and one for the @xmath7 in the final state .
the t - matrix element @xmath177 $ ] is then obtained by evaluating this at @xmath178 and multiplying it by @xmath179 : @xmath180 = - \frac{2 m_x z_x}{\delta_x(m_x,0 ) } .
\label{txx}\ ] ] the scalar meson model defined by the lagrangian in eqs .
( [ lfree ] ) and ( [ lint ] ) can be a low - energy approximation to a more fundamental quantum field theory .
the fundamental theory may include high energy processes that can create @xmath40 or @xmath41 with invariant mass near @xmath60 .
if _ long - distance _ effects involving momenta much smaller than @xmath23 can be separated from _ short - distance _ effects involving momenta of order @xmath23 and larger , we can use the scalar meson model to calculate the long - distance effects .
the basic tool required to separate long - distance effects from short - distance effects is the _ operator product expansion _
applications of the operator product expansion to the scalar meson model at 0@xmath39 order in @xmath38 were described in ref .
below , we give the leading terms in the operator product expansions of the t - matrix elements for some of the processes considered in ref . @xcite .
the general production process for @xmath7 has the form @xmath181 , where @xmath166 and @xmath167 each represent one or more particles .
there can be analogous production processes for @xmath41 . in the case of @xmath41 , we assume that the invariant mass @xmath59 satisfies the condition in eq .
( [ e - small ] ) .
this implies that the relative momenta of the mesons , which we denote generically by @xmath182 , are small compared to the natural momentum scale @xmath23 .
we call the production process a short - distance process if all the particles in @xmath166 and @xmath167 have momenta in the rest frame of @xmath7 or @xmath41 that are of order @xmath23 or larger .
the t - matrix element for a short - distance production process can be expanded in powers of the small energy differences @xmath106 and @xmath107 divided by @xmath57 and higher energy scales and in powers of the small relative momenta @xmath74 divided by @xmath23 and higher momentum scales .
the operator product expansion can be used to organize the expansions of the t - matrix elements into sums of products of short - distance coefficients and matrix elements of local operators between the vacuum state @xmath183 and the appropriate final state .
the leading terms in the expansions are those with the lowest dimension operator , which is @xmath145 .
the corresponding renormalized operator is @xmath157 . in feynman diagrams , the local operator @xmath157 is represented by a dot from which a @xmath34 line and a @xmath35 line emerge , as illustrated in fig .
[ fig : operators ] .
if we keep only the leading terms , the factorization formulas for the t - matrix elements reduce to and @xmath184 differ from those in ref .
@xcite by a factor of @xmath48 .
this choice simplifies many equations . ]
@xmath185 & = & \sqrt{2m_x } \ ; { \cal c}_a^{b,12 } \ , \langle x| \lambda_0
d_1^\dagger d_2^\dagger(0 ) | \emptyset \rangle , \label{taxpb } \\
{ \cal t}[a \to b + d_1 d_1 \phi ] & = & \sqrt{8m_1 ^ 2 m } \ ; { \cal c}_a^{b,12 } \ , \langle d_1 d_1
\phi | \lambda_0 d_1^\dagger d_2^\dagger(0 ) | \emptyset \rangle . \label{taddphib0}\end{aligned}\ ] ] [ taddx ] the wilson coefficient @xmath186 is a function of high energy scales , such as @xmath23 and @xmath4 , and of the total momentum of the @xmath7 or @xmath41 . the only dependence on whether the final state contains @xmath7 or @xmath41 is in the matrix elements of the local operator @xmath187 . short - distance and long - distance effects are separated in eqs .
( [ taddx ] ) . and
@xmath188 operators .
[ fig : operators],width=94 ] the decay of @xmath7 into @xmath41 can be described within the scalar meson model .
the fundamental theory may also allow the decay of @xmath7 into other final states @xmath168 .
we call such a process a short - distance decay if all the particles in @xmath168 have momenta in the rest frame that are of order @xmath23 or larger .
the t - matrix element for such a process can be expanded in powers of the small energy differences @xmath189 and @xmath107 divided by @xmath57 and higher energy scales .
the operator product expansion can be used to organize the expansion of the t - matrix element into sums of products of short - distance coefficients and matrix elements of local operators between the state @xmath190 and the vacuum state .
the leading term in the expansion is the one with the lowest dimension operator , which is @xmath146 .
the corresponding renormalized operator is @xmath147 .
if we keep only this term , the factorization formula for the t - matrix element reduces to@xmath191 = \sqrt{2m_x } \ ; { \cal c}_{12}^{c } \ , \langle \emptyset | \lambda_0
d_1 d_2(0 ) | x \rangle .
\label{txcm}\ ] ] the wilson coefficient @xmath184 is a function of high energy scales , such as @xmath23 and @xmath4 . short - distance and long - distance effects are separated in eqs .
( [ txcm ] ) .
if the fundamental theory includes processes that allow the production of @xmath7 via @xmath181 and the decay of @xmath7 via @xmath192 , it also allows the process @xmath165 , where @xmath168 represents the same particles but with a variable invariant mass @xmath59 instead of @xmath71 .
this process has a resonant enhancement when @xmath59 is near the @xmath40 threshold as specified by eq .
( [ e - small ] ) .
we call it a _ short - distance _ process if each of the particles in @xmath166 and @xmath167 has momentum large compared to @xmath23 in the rest frame of @xmath168 and if the relative momentum between each particle in @xmath168 and each particle in @xmath166 or @xmath167 is large compared to @xmath23 .
the t - matrix element for such a process can be described within the scalar meson model by a double operator product expansion .
the leading terms in this expansion are @xmath193 & = & { \cal c}_a^{b,\,c } + { \cal c}_a^{b,12 } { \cal c}_{12}^c \int d^4x \ , e^{i p \cdot x } \langle \emptyset | \lambda_0
d_1 d_2(x ) \lambda_0 d_1^\dagger d_2^\dagger(0 ) | \emptyset \rangle , \label{t : abcm}\end{aligned}\ ] ] where the 4-vector is @xmath194 .
the wilson coefficients @xmath186 and @xmath195 are the same ones that appear in the operator product expansions in eqs .
( [ taddx ] ) and ( [ txcm ] ) .
the first term @xmath196 on the right side of eq .
( [ t : abcm ] ) takes into account the direct production of @xmath168 at short distances .
this term can be expanded in powers of the small energy differences @xmath106 and @xmath102 divided by energy scales of order @xmath57 or higher .
the leading term in this expansion is simply a constant . according to eq .
( [ x - prop ] ) , the fourier transform in eq .
( [ t : abcm ] ) is just the @xmath7 propagator evaluated at @xmath197 .
short - distance and long - distance effects are not yet separated in eq .
( [ t : abcm ] ) , because the @xmath7 propagator requires an additive renormalization @xcite .
at @xmath199 order in @xmath38 , the only interaction in the scalar meson model is the contact interaction between @xmath34 and @xmath35 with coupling constant @xmath48 .
the @xmath41 states remain noninteracting at this order in @xmath38 . in this section ,
we summarize some of the results at @xmath199 order in @xmath38 that were obtained in ref .
@xcite and we generalize those results to the complex mass scheme . for the amplitude for the propagation of @xmath95 between contact interactions .
the open dots indicate that the vertex factors @xmath201 are omitted .
[ fig : l - lo],width=188 ] at 0@xmath39 order in @xmath38 , all the observables for processes near the @xmath40 threshold are related in a simple way to the connected green s function for @xmath202 .
the basic building block for this green s function is the amplitude @xmath203 for the propagation of the @xmath40 pair between successive contact interactions , which is given by the feynman diagram in fig .
[ fig : l - lo ] .
the amplitude @xmath204 is calculated in section [ sec : propcon0 ] of the appendix . the analytic expression using dimensional regularization
is given in eq .
( [ l0 - 3d ] ) .
the result in a general regularization scheme is @xmath205 where @xmath206 is the energy variable defined by @xmath207 this variable vanishes at the @xmath95 threshold @xmath208 .
it is real and positive if @xmath209 and it is pure imaginary with a negative imaginary part if @xmath210 .
the term @xmath211 in eq .
( [ l0-sub ] ) is real valued and includes a linear ultraviolet divergence .
the amputated connected green s function @xmath212 for @xmath202 can be calculated nonperturbatively by summing the geometric series represented by fig .
[ fig : alo ] to all orders in @xmath48 : @xmath213 the renormalization prescription for @xmath48 specified by eq .
( [ reainv : zero ] ) is that @xmath214 must vanish at @xmath215 .
this requires the inverse bare coupling constant to be @xmath216 using this expression for @xmath217 and the expression for @xmath204 in eq .
( [ l0 - 3d ] ) , the amplitude in eq .
( [ a - bare ] ) reduces to @xmath218 if we were to use the complex mass scheme , the amplitude @xmath204 would be given by eq .
( [ l0-sub ] ) with @xmath60 replaced by @xmath219 .
if we again demand that @xmath214 must vanish at @xmath215 , the renormalized expression for the amplitude would be @xmath220 where @xmath221 is a complex energy variable defined by @xmath222 and @xmath223 is its value at @xmath224 : @xmath225 the variable @xmath221 vanishes at the complex threshold @xmath226 and it has a negative imaginary part for all real values of @xmath80 . for @xmath202 at @xmath227 order in @xmath38
can be obtained by summing a geometric series of one - loop diagrams .
[ fig : alo],width=604 ] if the local composite operator @xmath147 is used as an interpolating field for @xmath7 , the propagator for @xmath7 is given in eq .
( [ x - prop ] ) .
the diagrams for the propagator of @xmath7 at 0@xmath39 order in @xmath38 are shown in fig .
[ fig : xprop ] .
these diagrams form a geometric series whose sum in the rest frame @xmath228 is @xmath229 using the expression for @xmath230 in eq .
( [ a - bare ] ) , we can express the propagator for @xmath7 as @xmath231 .
\label{deltax : ren}\ ] ] to obtain a renormalized propagator for @xmath7 that does not depend on the ultraviolet cutoff , an additive renormalization is necessary . after adding the constant @xmath232
, we obtain the renormalized propagator @xmath233 .
propagator at 0@xmath39 order in @xmath38 .
the interpolating field for the @xmath7 is @xmath234 .
[ fig : xprop],width=604 ] the term @xmath235 in the propagator for @xmath7 in eq .
( [ deltax : ren ] ) has a pole at the complex energy @xmath89 given in eq .
( [ epole ] ) .
using the renormalized expression for @xmath230 in eq .
( [ a0-ren ] ) , we find that the binding energy and width of @xmath7 at 0@xmath39 order in @xmath38 are @xmath236 [ egamx:00 ] the wavefunction normalization factor for @xmath7 is @xmath237 if we were to use the complex mass scheme , the amplitude @xmath230 would be given by eq .
( [ a0-cms ] ) .
this amplitude has a pole at a complex energy @xmath89 that determines the binding energy and width of @xmath7 .
the complex parameter @xmath89 satisfies @xmath238 where @xmath221 and @xmath223 are defined in eqs .
( [ gamma - e ] ) and ( [ gammax ] ) . comparing with eq .
( [ epole ] ) , we find that at lo in the complex mass scheme , the binding energy and width of @xmath7 are @xmath239 [ egamx:0 ] if we treat @xmath96 as being of order @xmath51 , the binding energy of @xmath7 in eq .
( [ ex:0 ] ) differs from @xmath108 only at order @xmath124 .
the wavefunction normalization constant for @xmath7 at lo in the complex mass scheme is @xmath240 the operator product expansion of the t - matrix element for a short - distance production process @xmath181 is given in eq .
( [ taxpb ] ) . the vacuum
to@xmath7 matrix element can be calculated via the lsz formalism using @xmath158 as an interpolating field for @xmath7 @xcite : @xmath241 at 0@xmath39 order in @xmath38 , the normalization constant @xmath153 is given in eq .
( [ zx ] ) . at lo in the complex mass scheme , @xmath153
is given in eq .
( [ zx : cms ] ) . the t - matrix element in eq .
( [ taxpb ] ) can be expressed as the product of a short - distance factor and a long - distance factor : @xmath242 = \sqrt{2m_x } \ , { \cal c}_a^{b,12 } \ , z_x^{1/2}. \label{taxb - fact}\ ] ] the rate for producing @xmath7 is obtained by squaring the t - matrix element in eq .
( [ taxb - fact ] ) and integrating over the appropriate phase space .
if @xmath166 consists of a single particle , its decay rate into @xmath243 can be expressed in the factored form @xmath244 = \gamma_a^b \ , m_{12 } |z_x| .
\label{gamabx}\ ] ] we have followed ref .
@xcite in choosing the long - distance factor in eq .
( [ gamabx ] ) to be @xmath245 , which is dimensionless .
the short - distance factor @xmath246 is @xmath247 where the 4-vector @xmath248 satisfies @xmath249 .
the operator product expansion of the t - matrix element for a short - distance decay process @xmath192 is given in eq .
( [ txcm ] ) .
the @xmath7to vacuum matrix element can be calculated via the lsz formalism using @xmath157 as an interpolating field for @xmath7 @xcite : ) is the complex conjugate of the matrix element in eq .
( [ xdd0 ] ) .
however @xmath190 in eq .
( [ 0ddx ] ) is an in state , while @xmath250 in eq .
( [ xdd0 ] ) is the hermitian conjugate of an out state .
these two states are related by the s - matrix : @xmath251 .
] @xmath252 at 0@xmath39 order in @xmath38 , the normalization constant @xmath153 is given in eq .
( [ zx ] ) . at lo in the complex mass scheme , @xmath153
is given by eq .
( [ zx : cms ] )
. the t - matrix element in eq . ( [ txcm ] ) can be expressed as the product of a short - distance factor and a long - distance factor : @xmath191 = \sqrt{2m_x } \ , { \cal c}_{12}^{c } \ , z_x^{1/2 } .
\label{txcm - fact}\ ] ] the decay rate of @xmath7 into the particles represented by @xmath168 is obtained by squaring the t - matrix element in eq .
( [ txcm - fact ] ) and integrating over the phase space of those particles .
it can be expressed in the factored form @xmath253 = \gamma^c \ , m_{12 } | z_x | .
\label{gamxcm - fact}\ ] ] we have followed ref .
@xcite in choosing the long - distance factor to be @xmath254 , which is dimensionless .
the short - distance factor @xmath255 is @xmath256 where the 4-vector @xmath248 satisfies @xmath249 . if @xmath7 can be produced via the short - distance process @xmath181 and
if it can decay via the short - distance process @xmath192 , then there can be resonant enhancement of the process @xmath165 , where @xmath168 represents the same particles but with a variable invariant mass @xmath59 instead of @xmath71 .
the t - matrix element for this process can be expressed as the double operator product expansion in eq .
( [ t : abcm ] ) .
the fourier transform of the matrix element is just the @xmath7 propagator evaluated at @xmath257 .
the t - matrix element therefore reduces to @xmath193 & = & { \cal c}_a^{b , c } + { \cal c}_a^{b,12 } { \cal c}_{12}^c \frac{i \lambda_0 ^ 2 l_0(m)}{1-\lambda_0 l_0(m)}. \label{t : abc - bare}\end{aligned}\ ] ] the t - matrix element can be expressed in a form in which the short - distance effects and long - distance effects are separated : @xmath193 & = & -i{\cal c}_a^{b,12 } { \cal c}_{12}^c \left [ \mathcal{a}_0(m ) - ( 2 \pi / m_{12 } ) c_a^{b , c } \right ] , \label{t : abc - simple}\end{aligned}\ ] ] where @xmath230 is given in eq .
( [ a0-ren ] ) and @xmath258 is a complex constant with dimensions of length that is completely determined by short - distance factors : @xmath259 the nonresonant term @xmath258 in eq .
( [ t : abc - simple ] ) is a combination of short - distance factors , so the natural scale for @xmath258 is @xmath260 .
the condition @xmath88 and the condition on @xmath59 in eq .
( [ e - small ] ) imply that the nonresonant term in eq .
( [ t : abc - simple ] ) is small compared to the resonant term in the threhsold region secified by eq .
( [ e - small ] ) .
we therefore set @xmath261 .
the invariant mass distribution of the particles in @xmath168 is obtained by squaring the t - matrix element and integrating over the momenta of all the particles in the final state .
if @xmath166 consists of a single heavy particle , the invariant - mass distribution of the particles in @xmath168 has the form @xmath262 & = & \big ( \gamma_a^b \gamma^c \big ) \ , \frac{m_{12}^2}{2 \pi } \left| \mathcal{a}_0(m ) \right|^2 . \label{dgamabc}\end{aligned}\ ] ] the short - distance factors @xmath246 and @xmath255 are the same as in eqs .
( [ gamabx ] ) and ( [ gamxcm - fact ] ) . at 0@xmath39 order in @xmath38 ,
@xmath230 is given in eq .
( [ a0-ren ] ) . at lo in the complex mass scheme , @xmath230
is given in eq .
( [ a0-cms ] ) .
note that if the invariant mass distribution in eq .
( [ dgamabc ] ) is divided by the product of the decay rates @xmath263 $ ] and @xmath264 $ ] in eqs .
( [ gamabx ] ) and ( [ gamxcm - fact ] ) , the short - distance factors cancel .
this combination of observables has only a long - distance factor given by @xmath265 . of the @xmath7 resonance in a short - distance decay channel at leading order in @xmath38 .
the dotted line is at order @xmath76 in ordinary perturbation theory ( using @xmath230 in eq .
( [ a0-ren ] ) ) .
the dashed lines are at lo in the complex mass scheme ( using @xmath230 in eq .
( [ a0-cms ] ) ) .
the line shapes are shown for @xmath73 mev and for both @xmath101 mev ( upper solid line ) and @xmath266 mev ( lower solid line ) .
the units on the vertical axis are @xmath267 .
[ fig : lineshapelo],width=453 ] the line shape of @xmath7 as a function of the invariant mass @xmath80 of the particles in @xmath168 is given by the factor @xmath268 in eq .
( [ dgamabc ] ) . in fig .
[ fig : lineshapelo ] , we compare the line shapes at 0@xmath39 order in @xmath38 and at lo in the complex mass scheme . at 0@xmath39 order in @xmath38 ,
the line shape diverges at @xmath269 . in the complex mass scheme ,
the line shape is a resonance whose maximum is near @xmath269 and whose width is determined by @xmath96 . if @xmath270 , the line shape for @xmath271 is a nonrelativistic breit - wigner resonance whose full width at half - maximum is @xmath96 : @xmath272 ^ 2 + \gamma_2 ^ 2/4}.\end{aligned}\ ] ] however , a breit - wigner resonance has tails that fall off like @xmath273 and it is integrable .
in contrast , the line shape @xmath268 with @xmath230 given by eq .
( [ a0-cms ] ) has tails that fall off only like @xmath274 and it is therefore not integrable .
thus the area under the resonance is not as well - defined as for a breit - wigner resonance .
in this section , we calculate the results in section [ sec : zeroth ] to next - to - leading order in @xmath38 . since the @xmath275 interaction allows transitions to @xmath41 states , we also calculate the rates for processes whose final state includes @xmath41 .
one of the basic building blocks for the amplitudes for processes near the @xmath95 threshold is the amplitude for the propagation of @xmath95 between contact interactions .
the term in this amplitude of order @xmath76 , @xmath203 , is given in eq .
( [ l0-sub ] ) .
the term of order @xmath51 , @xmath276 , is the sum of the three feynman diagrams in fig .
[ fig : l - nlo ] .
these diagrams are calculated in section [ sec : propcon2 ] of the appendix .
the final expression for @xmath277 is the sum of the amplitudes @xmath278 and @xmath279 given by eqs .
( [ l3-integral ] ) and ( [ l2-analytic ] ) : @xmath280 where @xmath206 is the energy variable defined in eq .
( [ kappa - e ] ) , the function @xmath281 is @xmath282 , \label{f - kappa}\end{aligned}\ ] ] and @xmath283 is the value of @xmath206 at the @xmath41 threshold @xmath284 : @xmath285^{1/2}. \label{kappa1}\end{aligned}\ ] ] the function @xmath286 in the integrand in eq .
( [ f - kappa ] ) is a rational function of @xmath287 that increases from 0 to 1 as @xmath287 increases from 0 to 1 : @xmath288 the term @xmath289 in eq .
( [ l2-bare ] ) is real valued and includes logarithmic ultraviolet divergences .
the function @xmath281 vanishes at the @xmath41 threshold @xmath290 by definition .
the first few terms in the laurent expansion of @xmath281 around the @xmath40 threshold at @xmath291 can be obtained from eqs .
( [ l2a - laurent ] ) and ( [ l2b - laurent ] ) .
the leading term is @xmath292 for the amplitude for the propagation of @xmath95 between contact interactions .
the open dots indicate that the vertex factors @xmath201 are omitted .
[ fig : l - nlo],width=453 ] at 0@xmath39 order in @xmath38 , the ultraviolet divergence in @xmath204 was eliminated by summing a geometric series in @xmath293 and then renormalizing the coupling constant @xmath48 . the ultraviolet divergence in @xmath277
can be eliminated in a similar way .
the geometric series of diagrams that must be summed are those that can be obtained by replacing each one - loop subdiagram in fig .
[ fig : alo ] by the sum of the one - loop subdiagram and the three diagrams for @xmath277 in fig .
[ fig : l - nlo ] .
the sum of that geometric series gives an amplitude @xmath294 that can be obtained by making the substitution @xmath295 in eq .
( [ a - bare ] ) : @xmath296}. \label{a2-bare}\ ] ] our renormalization prescription for @xmath48 in eq .
( [ reainv : zero ] ) requires @xmath297 to vanish at @xmath269 . using this prescription , the amplitude in eq .
( [ a2-bare ] ) can be expressed as @xmath298 } .
\label{a2-nlo}\end{aligned}\ ] ] of the @xmath7 resonance in a short - distance decay channel .
the dotted lines are at order @xmath51 in ordinary perturbation theory ( using @xmath299 in eq .
( [ a2-nlo ] ) ) .
the dashed lines are at lo in the complex mass scheme ( using @xmath230 in eq .
( [ a0-cms ] ) ) .
the line shapes are shown for @xmath73 mev and for both @xmath101 mev ( the two narrower resonances that are almost indistinguishable ) and for @xmath266 mev ( the two wider resonances ) .
the line shapes are normalized so their maximum values are 1 .
the inset shows their behavior near the @xmath40 threshold .
[ fig : lineshapenlocmslo],width=453 ] the quantity @xmath300 with @xmath299 given in eq .
( [ a2-nlo ] ) is the line shape of the @xmath7 resonance in a short - distance decay channel to order @xmath51 in ordinary perturbation theory . in fig .
[ fig : lineshapenlocmslo ] , we compare this line shape with that at lo in the complex mass scheme , which is @xmath301 with @xmath230 given by eq .
( [ a0-cms ] ) .
they are shown for two values of the coupling constant @xmath38 that correspond to the width of @xmath35 being @xmath101 mev and @xmath302 mev .
the two line shapes have the same qualitative behavior below the resonance and near the peak of the resonance .
they have qualitatively different behavior near the @xmath40 threshold @xmath303 .
unlike @xmath301 , the function @xmath300 vanishes at the @xmath40 threshold and has a sharp peak just above the threshold .
the inset in fig .
[ fig : lineshapenlocmslo ] shows the behavior near the @xmath40 threshold , which is clearly unphysical .
the reason @xmath300 vanishes at the @xmath40 threshold can be seen from the laurent expansion of @xmath281 around @xmath291 .
the leading term , which is given in eq .
( [ f - laurent ] ) , diverges like @xmath304 as @xmath305 .
no matter how small the coupling constant @xmath38 , the term proportional to @xmath306 can not be treated as a perturbation in the region near @xmath291 .
the resolution of the problem can be found by noting that the diverging term in eq .
( [ f - laurent ] ) can be expressed as @xmath307 where @xmath96 is the width of the @xmath35 from its decay into @xmath55 , which is given in eq .
( [ gam2-nr ] ) .
thus the divergence at @xmath291 implies that near the @xmath40 threshold , the imaginary part of the @xmath35 self - energy must be resummed to all orders . that resummation shifts the branch point in the leading order amplitude from the real threshold @xmath208 to the complex threshold @xmath308 .
making this change and then imposing our renormalization scheme , the amplitude in eq .
( [ a2-nlo ] ) becomes @xmath309 } , \label{a2-resum}\end{aligned}\ ] ] where @xmath221 and @xmath223 are defined in eqs .
( [ gamma - e ] ) and ( [ gammax ] ) .
this amplitude is a smooth function of @xmath80 at @xmath291 . a convenient way to implement the resummation in eq .
( [ a2-resum ] ) that can be extended staightforwardly to higher orders in @xmath38 is to use the complex mass scheme .
this automatically gives an amplitude @xmath299 that has smooth behavior at the @xmath95 threshold and is correct through relative order @xmath51 .
in the complex mass scheme , the contribution to @xmath277 from the diagrams in figs . [
fig : l - nlo](a ) and [ fig : l - nlo](b ) are obtained by replacing the energy variable @xmath206 by @xmath221 defined in eq .
( [ gamma - e ] ) and by replacing @xmath283 by @xmath310 , which is the value of @xmath221 at the @xmath41 threshold @xmath311 : @xmath312^{1/2 } . \label{gamma1}\end{aligned}\ ] ] in the complex mass scheme , there are additional contributions to the function @xmath277 coming from the one - loop diagram in fig . [ fig : l - nlo](c ) with the counterterm vertex in eq .
( [ dlint ] ) .
the additional contributions are given in eq .
( [ l2c ] ) . adding those terms is equivalent to making the substitution @xmath313 using the expressions to order @xmath51 for @xmath96 in eq .
( [ gam2-nr ] ) and @xmath314 in eq .
( [ z2-nr ] ) , we can see that the additional terms in eq .
( [ fsub ] ) cancel the @xmath315 and @xmath316 terms coming from the laurent expansions of @xmath279 in eq .
( [ l2b - laurent ] ) .
our renormalization prescription for @xmath48 in eq .
( [ reainv : zero ] ) requires @xmath297 to have a zero at @xmath269 . using this prescription
, the amplitude can be expressed as @xmath317 } , \label{a2-ren : resum}\end{aligned}\ ] ] where @xmath221 and @xmath223 are given by eqs .
( [ gamma - e ] ) and ( [ gammax ] ) and the function @xmath318 is @xmath319 .
\label{fcms}\end{aligned}\ ] ] the function @xmath318 can be obtained from @xmath281 in eq .
( [ f - kappa ] ) by subtracting several terms in the laurent expansion in @xmath316 and then replacing @xmath206 and @xmath283 by @xmath221 and @xmath310 .
the terms that are subtracted are the imaginary term of order @xmath315 , the real terms of order @xmath320 , and the imaginary term of order @xmath316 in the second of the two terms on the right side of eq .
( [ f - kappa ] ) .
the real terms of order @xmath320 are subtracted because they would cancel in the denominator in eq .
( [ a2-ren : resum ] ) anyway .
the leading terms in the laurent expansion of @xmath318 around the complex @xmath40 threshold at @xmath321 are @xmath322.\end{aligned}\ ] ] for real values of @xmath80 , there is an ambiguity in the function @xmath323 in eq .
( [ a2-ren : resum ] ) for @xmath324 because of the branch point at @xmath325 in eq .
( [ fcms ] ) .
this ambiguity can be resolved by the substitution @xmath326 . of the @xmath7 resonance in a short - distance decay channel using the complex mass scheme .
the dashed lines are at lo in the complex mass scheme ( using @xmath230 in eq .
( [ a0-cms ] ) ) .
the solid lines are at nlo in the complex mass scheme ( using @xmath299 in eq .
( [ a2-ren : resum ] ) ) .
the line shapes are shown for @xmath73 mev and for both @xmath101 mev ( the two narrower resonances that are almost indistiguishable ) and for @xmath266 mev ( the two wider resonances ) .
the line shapes are normalized so their maximum values are 1 .
[ fig : lineshapenlocmsnlo],width=453 ] in fig .
[ fig : lineshapenlocmsnlo ] , we compare the line shapes of @xmath7 in a short - distance decay channel at nlo in the complex mass scheme , which is @xmath300 with @xmath299 given by eq .
( [ a2-ren : resum ] ) , and that at lo in the complex mass scheme , which is @xmath301 with @xmath230 given by eq .
( [ a0-cms ] ) .
they are shown for two values of the coupling constant @xmath38 that correspond to the width of @xmath35 being @xmath101 mev and @xmath302 mev .
the curves are normalized so their maximum values are 1 .
the line shapes at lo and nlo have the same qualitative behavior . as @xmath38 decreases , the quantitative differences between the line shapes at lo and nlo decrease .
thus the complex mass scheme seems to provide a good solution to the infrared problem near the @xmath40 threshold that arises in ordinary perturbation theory .
we will therefore use the complex mass scheme in most of the calculations in the remainder of this section .
if the local composite operator @xmath147 is used as the interpolating field for @xmath7 , the propagator for @xmath7 is defined in eq .
( [ x - prop ] ) .
the expression for the propagator at 0@xmath39 in @xmath38 is given in eq .
( [ deltax-0 ] ) .
a renormalizable expression for the @xmath7 propagator that is accurate to order @xmath51 can be obtained simply by making the substitution @xmath327 in eq .
( [ deltax-0 ] ) : @xmath328 } { 1 - \lambda_0 [ l_0(e ) + l_2(e ) ] } .
\label{deltax-2}\ ] ] comparing with the expression for the amplitude @xmath299 in eq .
( [ a2-bare ] ) , we see that the propagator in eq .
( [ deltax-2 ] ) can be expressed as @xmath329 .
\label{deltax-2ren}\ ] ] the expression for @xmath299 at nlo in the complex mass scheme is given in eq .
( [ a2-ren : resum ] ) .
this amplitude has a pole at a complex energy @xmath89 that determines the binding energy and the width of @xmath7 .
the energy @xmath89 is the solution to @xmath330 = 0 , \label{gamma - pole : nlo}\ ] ] where @xmath221 and @xmath223 are defined in eqs .
( [ gamma - e ] ) and ( [ gammax ] ) .
the wavefunction normalization constant for @xmath7 is @xmath331 if we treat @xmath96 as order @xmath51 , the solution to eq .
( [ gamma - pole : nlo ] ) for the complex variable @xmath89 through order @xmath51 is @xmath332 . \ ] ] using the expression for @xmath318 in eq .
( [ fcms ] ) , we find that the binding energy and width of the @xmath7 to order @xmath51 are @xmath333 , \label{gamx : nlo}\end{aligned}\ ] ] [ egamx : nlo ] where the upper endpoint @xmath334 of the integral is @xmath335 for @xmath73 mev , the second term on the right side of eq . ( [ gamx : nlo ] ) is @xmath336 .
thus the width of the constituent @xmath35 accounts for most of the width of @xmath7 at order @xmath51 .
we can obtain an analytic expression for the width @xmath91 if @xmath337 . in this limit , the upper limit @xmath334 on the integral approaches 1 and the width reduces to @xmath338 .
\label{gamma - x : nlolim}\ ] ] as @xmath339 varies from 0 to @xmath340 , the factor in square brackets ranges from @xmath341 to @xmath342 .
note that the expression for @xmath91 in eq .
( [ gamma - x : nlolim ] ) can be smaller than @xmath96 for some values of the parameters .
the operator product expansion of the t - matrix element for the short - distance production process @xmath181 is given in eq .
( [ taxpb ] ) . at nlo in the complex mass scheme ,
the operator matrix element is given by eq .
( [ xdd0 ] ) , where @xmath153 is now the wavefunction normalization constant for @xmath7 in eq .
( [ zx - nlo ] ) . the factored expression for the t - matrix element
is given by eq .
( [ taxb - fact ] ) .
if @xmath166 consists of a single particle , its decay rate into @xmath343 can be expressed in the factored form in eq .
( [ gamabx ] ) , where @xmath153 is given in eq .
( [ zx - nlo ] ) .
the operator product expansion for the t - matrix element for the short - distance decay process @xmath192 is given in eq .
( [ txcm ] ) . at nlo in the complex mass scheme ,
the operator matrix element is given in eq .
( [ 0ddx ] ) , where @xmath153 is the wavefunction normalization constant for @xmath7 in eq .
( [ zx - nlo ] ) . the factored expression for the t - matrix element
is given by eq .
( [ txcm - fact ] ) . the decay rate of @xmath7 into the particles represented by @xmath168 can be expressed in the factored form in eq .
( [ gamxcm - fact ] ) , where @xmath153 is given in eq .
( [ zx - nlo ] ) .
the t - matrix element for the short - distance process @xmath165 , where @xmath168 is a set of particles with invariant mass @xmath59 close to the @xmath40 threshold , can be expressed as the double operator product expansion in eq .
( [ t : abcm ] ) . the t - matrix element at order @xmath76
is given in eq .
( [ t : abc - bare ] ) .
an expression for this t - matrix element that is accurate through order @xmath51 can be obtained by substituting @xmath327 : @xmath193 & = & { \cal c}_a^{b , c } + { \cal c}_a^{b,12 } { \cal c}_{12}^c \frac{i \lambda_0 ^ 2 [ l_0(m ) + l_2(m ) ] } { 1 - \lambda_0 [ l_0(m ) + l_2(m ) ] } .
\label{t : abc - bare - nlo}\end{aligned}\ ] ] the expression for the t - matrix element in which short - distance and long - distance factors are separated is given by eq .
( [ t : abc - simple ] ) with @xmath344 replaced by @xmath345 .
the invariant mass distribution for the short - distance decay channel is @xmath262 & = & \big ( \gamma_a^b \gamma^c \big ) \ , \frac{m_{12}^2}{2 \pi } \left| \mathcal{a}_2(m ) \right|^2 , \label{dgamabc : nlo}\end{aligned}\ ] ] where @xmath246 and @xmath255 are the short - distance factors defined in eqs .
( [ gamaddb ] ) and ( [ gam - c ] ) . the expression for @xmath345 at nlo in the complex mass scheme
is given in eq .
( [ a2-ren : resum ] ) .
the line shapes at lo and nlo in the complex mass scheme are compared in fig .
[ fig : lineshapenlocmsnlo ] . if the invariant mass distribution in eq .
( [ dgamabc : nlo ] ) is divided by the product of the decay rates @xmath263 $ ] and @xmath264 $ ] in eqs .
( [ gamabx ] ) and ( [ gamxcm - fact ] ) , the short - distance factors cancel , leaving only the long - distance factors @xmath346 , where @xmath299 and @xmath153 are given in eqs .
( [ a2-ren : resum ] ) and ( [ zx - nlo ] ) .
the corresponding quantity at lo in the complex mass scheme is @xmath347 , where @xmath230 and @xmath153 are given in eqs .
( [ a0-cms ] ) and ( [ zx : cms ] ) .
the dependence of this long - distance factor on the invariant mass @xmath59 is illustrated in fig .
[ fig : lineshapenlocmsnlo ] , where the curves are all normalized to maximum value 1 .
the peak value of @xmath348 is also completely determined by long distances . for @xmath73 mev and @xmath101 mev ,
the peak value decreases from 130.1 mev@xmath349 at lo to 119.2 mev@xmath349 at nlo . for @xmath73 mev and @xmath266 mev ,
the peak value decreases from 8.34 mev@xmath349 at lo to 6.58 mev@xmath349 at nlo .
the decrease in the difference between the peak values at nlo and lo from about 27% at @xmath266 mev to about 9% at @xmath101 mev supports the assumption that the perturbative expansion is well - behaved in the complex mass scheme .
the fundamental theory may have short - distance production processes @xmath351 , where @xmath166 and @xmath167 both represent one or more particles whose momenta in the @xmath41 rest frame are of order @xmath23 or larger .
the @xmath41 invariant mass @xmath59 is assumed to be close to @xmath60 , as specified by the condition in eq .
( [ e - small ] ) .
the operator product expansion of the t - matrix element for this process is given in eq .
( [ taddphib0 ] ) . for the vacuum
to@xmath352 matrix element of the operator @xmath188 .
[ fig : ddphi],width=453 ] we first calculate the rate for @xmath351 to order @xmath51 using ordinary perturbation theory . at order @xmath38 , the operator matrix element in eq .
( [ taddphib0 ] ) is given by the series of feynman diagrams in fig .
[ fig : ddphi ] summed over the two permutations of the external @xmath34 lines .
the sum of the geometric series of diagrams is @xmath353 multiplied by the first diagram in the series . according to the optical theorem , the square of the first diagram integrated over the 3-body phase space for @xmath41 is given by the sum of the 3-body cuts through the feynman diagrams for @xmath277 in fig . [
fig : l - nlo ] .
if @xmath354 , the sum of the 3-body cuts gives the imaginary part of @xmath277 .
thus the square of the t - matrix element in eq .
( [ taddphib0 ] ) integrated over the 3-body phase space of @xmath41 is @xmath355 \right|^2 \nonumber \\ & & \hspace{7 cm } = |{\cal c}_a^{b,12 } |^2 |\mathcal{a}_0(m)|^2 ( -2 ) \ , { \rm i m } \ , l_2(m ) , \label{tddphi : g2}\end{aligned}\ ] ] where @xmath356 , @xmath357 , and @xmath358 are 4-momenta for @xmath34 , @xmath34 , and @xmath36 , respectively , and @xmath359 . if @xmath166 consists of a single particle , the invariant mass distribution for @xmath360 at order @xmath51 can be expressed as @xmath361 & = & \gamma_a^b \ ,
\frac{m_{12}}{2\pi } \left| \mathcal{a}_0 ( m)\right|^2 ( -2 ) \ , { \rm i m } \ , l_2(m ) , \label{ddphiimd : g2}\end{aligned}\ ] ] where @xmath246 is the short - distance factor defined in eq .
( [ gamaddb ] ) and we have replaced a multiplicative factor of @xmath59 by @xmath60 . if there is a short - distance process @xmath351 , then the process @xmath362 is also allowed . the invariant mass distribution for @xmath362 at order @xmath76 is given by the expression on the right side of eq . ( [ ddphiimd : g2 ] ) with @xmath363 replaced by @xmath364 .
the threshold for this process is @xmath365 . since @xmath35 ultimately decays into @xmath55 , a decay into @xmath40
can also be regarded as a contribution to the inclusive decay into @xmath366 .
the result in eq .
( [ ddphiimd : g2 ] ) is an order @xmath51 correction to the inclusive invariant mass distribution for @xmath41 .
it is nonzero for all @xmath367 . for @xmath368 , there are additional corrections of order @xmath51 that can be obtained by replacing any of the one - loop subdiagrams in fig .
[ fig : ddphi ] by one of the two - loop diagrams for @xmath277 in fig .
[ fig : l - nlo ] .
nonperturbative renormalization of the coupling constant @xmath48 then requires that a geometric series of these corrections be summed to all orders .
the net effect is to replace @xmath344 in eq .
( [ ddphiimd : g2 ] ) by @xmath345 .
thus the complete result for the inclusive invariant mass distribution for @xmath41 through order @xmath51 is @xmath361 & = & \gamma_a^b \ , \frac{m_{12}}{2\pi } \left| \mathcal{a}_2 ( m)\right|^2 ( -2 ) \ , { \rm i m } \ , [ l_0(m ) + l_2(m ) ] , \label{dgamddphi : g2}\end{aligned}\ ] ] where the functions @xmath369 , @xmath370 , and @xmath371 are given in eqs .
( [ a2-nlo ] ) , ( [ l0-sub ] ) , and ( [ l2-bare ] ) . note that taking the imaginary parts of @xmath370 and @xmath371 eliminates the additive ultraviolet - divergent terms . a more compact expression for the distribution in eq .
( [ dgamddphi : g2 ] ) is @xmath361 & = & \gamma_a^b \ ,
\frac{m_{12}}{\pi } \ , { \rm i m } \ , \mathcal{a}_2(m ) .
\label{ddphiimd : nlocms}\end{aligned}\ ] ] the inclusive invariant mass distribution in eq .
( [ dgamddphi : g2 ] ) when calculated using ordinary perturbation theory has unphysical behavior at the threshold @xmath372 .
the factor @xmath373 has a zero at @xmath372 and a sharp peak just above that threshold as illustrated in fig .
[ fig : lineshapenlocmslo ] . the term @xmath374 also has unphysical behavior at this threshold .
the imaginary part of @xmath277 in the region @xmath375 is given by the sum of eqs .
( [ iml2a ] ) and ( [ iml2b ] ) .
the resulting expression for @xmath376 is @xmath377 , \label{iml2ab}\end{aligned}\ ] ] where the upper limit @xmath334 of the integral is @xmath378 the function in eq .
( [ iml2ab ] ) diverges like @xmath379 as @xmath305 . as a consequence , near the @xmath40 threshold
, corrections of higher order in @xmath51 that come from the @xmath35 self energy are not suppressed .
this is the same problem that prompted us to introduce the complex mass scheme in section [ sec : xamp2 ] .
we next consider the inclusive rate for @xmath351 in the complex mass scheme . at lo in the complex mass scheme ,
the invariant mass distribution for @xmath41 can be obtained from eq .
( [ ddphiimd : nlocms ] ) by replacing @xmath345 by @xmath344 in eq .
( [ a0-cms ] ) .
this can be written more explicitly as @xmath361 & = & \gamma_a^b \ , \frac{-2 \ ,
{ \rm i m } \ , \gamma(m)}{|\gamma(m ) - { \rm re } \ , \gamma_x|^2 } , \label{ddphiimd : locms}\end{aligned}\ ] ] where @xmath221 and @xmath223 are defined in eqs .
( [ gamma - e ] ) and ( [ gammax ] ) .
this expression has a nonzero imaginary part at all energies @xmath80 , even below the @xmath41 theshold . for @xmath380 ,
the imaginary part is small and it is cancelled by higher orders in the complex mass scheme . at nlo in the complex mass scheme , the invariant mass distribution for @xmath41 is obtained from eq .
( [ ddphiimd : nlocms ] ) by using the expression for @xmath345 in eq .
( [ a2-ren : resum ] ) .
resonance at nlo in the complex mass scheme .
the solid lines are for a short - distance decay channel ( @xmath300 using @xmath299 in eq .
( [ a2-ren : resum ] ) ) .
the dash - dotted lines are for the @xmath41 channel ( @xmath381 using @xmath299 in eq .
( [ a2-ren : resum ] ) ) .
the line shapes are shown for @xmath73 mev and for both @xmath101 mev ( narrower resonances ) and @xmath266 mev ( wider resonances ) .
the line shapes are normalized so their maximum values are 1 . [
fig : lineshapeaw],width=453 ] in fig .
[ fig : lineshapeaw ] , we compare the line shape of @xmath7 in the @xmath41 channel , which is proportional to @xmath381 , with the line shape of @xmath7 in a short - distance decay channel , which is proportional to @xmath300 .
they are shown for two values of the coupling constant @xmath38 that correspond to the width of @xmath35 being @xmath101 mev and @xmath302 mev .
the curves are normalized so their maximum values are 1 .
the line shapes have the same qualitative behavior .
the line shape for @xmath41 is shifted towards higher energy by an amount that decreases as @xmath38 decreases .
the belle collaboration has observed an enhancement in the production of @xmath10 near threshold in the decay @xmath382 @xcite .
the enhancement peaks at the invariant mass @xmath383 mev , where we have combined the errors in quadrature .
if we use the recent precision measurement of the @xmath5 mass by the cleo collaboration @xcite , the peak is above the @xmath1 threshold by @xmath384 mev .
in contrast , the peak observed in the short - distance decay channel @xmath385 is below the @xmath1 threshold by @xmath386 mev .
the difference is @xmath387 mev , which differs from 0 by two standard deviations .
part of the discrepancy may be due to a shift in the invariant mass distribution analogous to the one illustrated in fig .
[ fig : lineshapeaw ] .
the width @xmath91 of the @xmath7 resonance was defined in eq .
( [ epole ] ) in terms of the energy @xmath89 of the pole in the @xmath7 propagator : @xmath389 . at lo in the complex mass scheme ,
@xmath91 is simply @xmath96 . at nlo in the complex mass scheme , @xmath91
is obtained by solving eq .
( [ gamma - pole : nlo ] ) for @xmath89 .
if the width of the @xmath7 resonance is sufficiently small , its width can also be interpreted as its decay rate .
a direct calculation of that decay rate will give a result that is closely related to but not identical to @xmath91 .
we will denote the result of the direct calculation of the inclusive decay rate by @xmath390 $ ] .
we can use the optical theorem to calculate the inclusive decay rate for @xmath391 : @xmath392 = \frac{1}{m_x } \ , { \rm i m } \ , { \cal t}[x \to x ] .
\label{opticalthm}\end{aligned}\ ] ] the forward t - matrix element for @xmath174 can be defined using the lsz formalism and it is given in eq .
( [ txx ] ) . at nlo in the complex mass scheme ,
the normalization constant @xmath153 and the propagator @xmath175 are given in eqs .
( [ zx - nlo ] ) and ( [ deltax-2 ] ) . the resulting expression for the decay rate of @xmath7 in eq .
( [ opticalthm ] ) is @xmath392 = -2 \ , { \rm i m } \frac{z_x{\cal a}_2^{-1}(m_x)}{\lambda_0 [ l_0(m_x ) + l_2(m_x)]}. \label{imtxx}\end{aligned}\ ] ] we have used eq .
( [ a2-bare ] ) to express this in a form with a factor of @xmath393 , because @xmath394 does not depend on the ultraviolet cutoff @xmath395 .
the decay rate in eq .
( [ imtxx ] ) depends on @xmath395 through the bare coupling constant @xmath48 and through the ultraviolet - divergent additive constants in @xmath396 and @xmath397 . from the renormalized expression for @xmath394 in eq .
( [ a2-bare ] ) , we can see that the divergent terms in @xmath398 must be cancelled by @xmath217 as @xmath399 .
the leading divergence is a linear divergence in @xmath211 that is linear in @xmath395 .
thus @xmath217 must include a cancelling linear divergence .
this implies that @xmath400 $ ] must approach 1 as @xmath399 . upon taking the limit @xmath399 , eq .
( [ imtxx ] ) reduces to @xmath392 = -2 \ , { \rm i m } \left ( z_x \mathcal{a}_2^{-1}(m_x ) \right ) .
\label{imtxx - finite}\end{aligned}\ ] ] an explicit expression for the inclusive decay rate of @xmath7 can be obtained from eq .
( [ imtxx - finite ] ) by inserting the renormalized expression for @xmath394 in eq .
( [ a2-ren : resum ] ) : @xmath392 = \frac{m_{12}}{2 \pi } \ , ( -2 ) \ , { \rm i m } \left [ z_x \big ( \gamma(m_x ) + f_{\rm cms}(\gamma(m_x ) ) - { \rm re } [ \gamma_x + f_{\rm cms}(\gamma_x ) ] \big ) \right ] , \nonumber \\ \label{gamddphi}\end{aligned}\ ] ] where @xmath153 , @xmath221 , @xmath223 , and @xmath318 are given in eqs .
( [ zx - nlo ] ) , ( [ gamma - e ] ) , ( [ gammax ] ) , and ( [ fcms ] ) . if we treat @xmath96 as order @xmath51 , the decay rate in eq .
( [ gamddphi ] ) reduces at order @xmath51 to the corresponding result for the width @xmath91 in eq .
( [ gamx : nlo ] ) . of the @xmath7 resonance ( solid line ) and the decay rate @xmath401 $ ] ( dashed line ) at nlo in the complex mass scheme as functions of @xmath96 .
the vertical axis is in units of @xmath96 .
the intercept on the vertical axis is approximately @xmath402 .
[ fig : widths],width=453 ] in fig .
[ fig : widths ] , we compare the width @xmath91 obtained by solving eq .
( [ gamma - pole : nlo ] ) and the decay rate @xmath401 $ ] in eq .
( [ gamddphi ] ) as functions of @xmath96 .
they agree for small @xmath96 , because the difference between @xmath91 and @xmath401 $ ] is of order @xmath124 .
they differ by less than 1% if @xmath403 mev .
we have examined the effects of a weakly - bound hadronic molecule on a nearby 3-body threshold , where the three bodies consist of a primary constituent of the molecule and the decay products of the other constituent .
we studied these effects in the simplest possible model : the scalar meson model , with spin-0 constituents @xmath34 , @xmath35 , and @xmath36 and momentum - independent interactions .
the primary constitutents of the molecule @xmath7 are @xmath40 and the 3-body threshold is that for @xmath41 .
the @xmath40 contact interaction with coupling constant @xmath48 was treated nonperturbatively .
the @xmath37 interaction with coupling constant @xmath38 was treated perturbatively .
several observables were calculated to all orders in @xmath48 and to next - to - leading order in @xmath38 .
both ultraviolet problems and infrared problems were encountered in these calculations .
we found that at next - to - leading order in @xmath38 , the ultraviolet divergences can be removed by a perturbative renormalization of the @xmath35 mass , a nonperturbative renormalization of @xmath48 , and a nonperturbative renormalization of short - distance coefficients in the operator product expansion .
the nonperturbative renormalizations required the summing of a geometric series of order-@xmath51 subdiagrams to all orders .
the subdiagrams represent order-@xmath51 contributions to the amplitude for the propagation of @xmath40 between contact interactions .
thus the renormalized amplitudes at next - to - leading order include all terms of order @xmath51 together with subsets of terms from all higher orders in @xmath38 .
we found that the next - to - leading order amplitudes suffer from an infrared problem at the @xmath40 threshold .
the problem is illustrated dramatically in fig .
[ fig : lineshapenlocmslo ] , which shows the line shape @xmath90 in a short - distance decay channel as a function of the energy @xmath80 .
the pathological behavior at @xmath303 arises because the @xmath37 interaction shifts the @xmath40 threshold from the real value @xmath60 to the complex value @xmath219 .
since @xmath96 is of order @xmath51 , a strict expansion in powers of @xmath38 leads to singularities at @xmath303 .
one possible solution to this problem is to sum the geometric series of self - energy corrections to the @xmath35 propagator to all orders , but this leads to very complicated loop integrals .
a simpler solution is to use the complex mass scheme , in which the width @xmath96 is included in the feynman rule for the @xmath35 propagator and its effects are systematically compensated at higher orders in @xmath38 through counterterms .
this partial resummation of the @xmath35 propagator corrections leads to smooth behavior of the line shape @xmath90 at the @xmath40 threshold as illustrated in fig .
[ fig : lineshapenlocmsnlo ] .
having solved the infrared problems , we calculated several observables to nlo in the complex mass scheme .
they include the line shapes of the @xmath7 resonance in both a short - distance decay channel and in the @xmath41 channel .
the @xmath0 is a weakly - bound hadronic molecule whose primary constitutents are a superposition of charm mesons : @xmath33 .
its mass is near the 3-body thresholds for @xmath10 , @xmath28 , and @xmath29 .
the methods we applied to the scalar meson model can be extended straightforwardly to this multichannel problem .
a conventional perturbation expansion in the @xmath404 coupling constant will have infrared problems at the @xmath1 threshold which are associated with the decay @xmath405 .
these infrared problems can be avoided by using the complex mass scheme .
the ultraviolet problems can be expected to be more severe in this system , because the @xmath406 interaction is proportional to the 3-momentum of the pion .
thus renormalization may not be as simple as in the scalar meson model .
if the renormalization problem can be solved , it will be straightforward to calculate observables to nlo in the complex mass scheme .
one of the most interesting applications will be to predict the difference between the line shapes of the @xmath0 resonance in a short - distance decay channel , such as @xmath385 , and in the @xmath10 channel .
eb thanks s. fleming , m. kusunoki , and t. mehen for valuable discussions .
jl thanks the physics department at the ohio state university for its hospitality while some of this work was carried out .
this research was supported in part by the department of energy under grant de - fg02 - 91-er4069 and by the korea research foundation under moehrd basic research promotion grant krf-2006 - 311-c00020 .
in this appendix , we calculate the loop integrals that appear at order @xmath76 and at order @xmath51 in the scalar meson model .
many of the loop integrals we need to evaluate are ultraviolet divergent .
they can be regularized by using dimensional regularization of the integrals over the 3-momenta .
we analytically continue the 3-dimensional integrals to @xmath407 dimensions using the prescription @xmath408 where @xmath409 is the renormalization scale and @xmath410 is the pochhammer symbol : @xmath411 the function of @xmath412 in the prefactor is designed to simplify the analytic expressions for dimensionally regularized integrals by cancelling the effects of the analytic continuation of angular integrals .
thus the integral of a scalar function of @xmath413 is @xmath414 loop integrals are evaluated by first inserting the appropriate nonrelativistic propagators : @xmath415 where @xmath416 and @xmath417 .
integrals over the loop energy are then evaluated using the residue theorem . after using the feynman parameter trick to combine denominators , the dimensionally regularized integrals over the 3-momenta
can be evaluated .
the final steps are the evaluation of the feynman parameter integrals and the analytic continuation to @xmath418
. it will be convenient to express the results in terms of the variables the one - loop diagram for the self - energy @xmath114 of @xmath35 is shown in fig .
[ fig : self](a ) .
the expression for the diagram is @xmath420 the residue theorem can be used to evaluate the energy integral : @xmath421 if we implement dimensional regularization using the prescription in eq .
( [ dimreg ] ) , the analytic result is @xmath422 - i \varepsilon
\right)^{\frac{1}{2 } - \epsilon } .
\label{sigma2-dimreg}\ ] ] the self - energy in dimensional regularization is given by the analytic continuation of eq .
( [ sigma2-dimreg ] ) to @xmath418 : @xmath423 - i \varepsilon \right)^{\frac{1}{2 } } .
\label{sigma2-dimreg0}\ ] ] the momentum integral in eq .
( [ sigma2a ] ) has a linear ultraviolet divergence that is set to zero by dimensional regularization .
the self - energy in a general regularization scheme is given in eq .
( [ d2self ] ) . the linear ultraviolet divergence is contained in the extra term @xmath117 , which is real valued .
the amplitude for the propagation of @xmath40 between contact interactions is given at order @xmath76 by the one - loop diagram in fig .
[ fig : l - lo ] . in the center - of - momentum frame where the 4-momentum is @xmath424 , this amplitude is @xmath425 the residue theorem can be used to evaluate the energy integral : @xmath426 if we implement dimensional regularization using the prescription in eq .
( [ dimreg ] ) , the analytic result is @xmath427 where @xmath316 is defined in eq .
( [ kappa - def ] ) .
the amplitude in dimensional regularization is given by analytic continuation to @xmath418 : @xmath428 the momentum integral in eq .
( [ l0-e ] ) has a linear ultraviolet divergence that is set to zero by dimensional regularization .
the amplitude in a general regularization scheme is given in eq .
( [ l0-sub ] ) . the linear ultraviolet divergence is contained in the extra term @xmath211 , which is real valued .
the amplitude for the propagation of @xmath40 between contact interactions has contribution of order @xmath51 from the two - loop diagrams in fig .
[ fig : l - nlo ] . in the center - of - momentum frame where the 4-momentum is @xmath429 ,
the order-@xmath51 amplitude @xmath276 is a function of @xmath80 only . we proceed to calculate the contributions from each of the three diagrams .
the contribution to the amplitude @xmath276 from the two - loop diagram in fig .
[ fig : l - nlo](a ) in which @xmath36 is exchanged between the @xmath34 and @xmath35 is @xmath431 the integrals over the loop energies @xmath432 and @xmath433 can be evaluated by closing the contours in the upper half - planes . after introducing feynman parameters and evaluating the integrals over the loop momenta
, the amplitude reduces to @xmath434^{-3/2 + \epsilon } \nonumber \\ & & \hspace{2 cm } \times \left [ ( x+y ) \kappa^2 + ( 1-x - y)(m_1/m_{12})(\kappa^2-\kappa_1 ^ 2)\right]^{-2 \epsilon } , \end{aligned}\ ] ] where @xmath316 and @xmath283 are defined in eqs . ( [ kappakappa1 ] ) .
one of the feynman parameter integrals can be evaluated analytically by changing variables to @xmath435 and @xmath436 .
the integral over @xmath437 gives a hypergeometric function : @xmath438^{-2 \epsilon } , \label{l3-eps}\end{aligned}\ ] ] where @xmath286 is the rational function of @xmath287 given in eq .
( [ t - z ] ) .
the pole at @xmath418 in eq .
( [ l3-eps ] ) is a logarithmic ultraviolet divergence . the amplitude @xmath278 in dimensional regularization is defined by the laurent expansion of the expression in eq .
( [ l3-eps ] ) to order @xmath439 .
the expansion in powers of @xmath412 is facilitated by using a transformation formula to replace the hypergeometric function in eq .
( [ l3-eps ] ) by one of the form @xmath440 : @xmath441 the expansion of @xmath440 to order @xmath412 is simple : @xmath442 expanding the amplitude @xmath278 in eq .
( [ l3-eps ] ) to 0@xmath39 order in @xmath412 , it reduces to @xmath443}{t(z ) } + \frac{2t(z)}{3 } \ , { } _ 2f_1 \left ( 1,1,\mbox{$\frac{5}{2}$};t(z ) \right ) \nonumber \\ & & \hspace{5 cm } -2 \ln \frac{z\kappa^2+(1-z)(m_1/m_{12})(\kappa^2-\kappa_1 ^ 2)}{\mu^2 } \bigg ] \bigg ) .
\label{l3-dimreg}\end{aligned}\ ] ] we have used the integral @xmath444 this can be evaluated by changing variables to @xmath445 $ ] .
most of the terms in eq .
( [ l3-dimreg ] ) are independent of @xmath80 .
the expression can be simplified by expressing it in the form @xmath446 the amplitude @xmath447 at the @xmath41 threshold is real - valued and includes the logarithmic divergence .
the amplitude @xmath278 has an imaginary part for @xmath448 . for energies in the range @xmath375 ,
the imaginary part is @xmath449 where the upper limit @xmath334 of the integral is @xmath450 the first few terms in the laurent expansion of @xmath278 in powers of @xmath316 are @xmath451 .
\label{l2a - laurent}\end{aligned}\ ] ] the contribution to @xmath276 from the two - loop diagram in fig .
[ fig : l - nlo](b ) with a self - energy correction to the @xmath35 line is @xmath452 where @xmath453 is the contribution to the @xmath35 self - energy from the one - loop diagram in fig .
[ fig : self](a ) . the analytic result for @xmath453 in dimensional regularization
is given in eq .
( [ sigma2-dimreg ] ) .
since @xmath453 has a branch cut for @xmath432 in the lower half - plane , it is convenient to close the @xmath432 contour in the upper half - plane . after introducing feynman parameters , the integral over @xmath454 can be evaluated . if we implement dimensional regularization using the prescription in eq .
( [ dimreg ] ) , the result is @xmath455^{-2\epsilon } , \label{l2-eps}\end{aligned}\ ] ] where @xmath316 and @xmath283 are defined in eqs .
( [ kappakappa1 ] ) and @xmath456 is given by @xmath457 the pole at @xmath418 in eq .
( [ l2-eps ] ) is a logarithmic ultraviolet divergence .
the amplitude @xmath279 in dimensional regularization is defined by the laurent expansion of the expression in eq .
( [ l2-eps ] ) to order @xmath439 .
expanding to 0@xmath39 order in @xmath412 , the expression reduces to @xmath458 .
\label{l2-dimreg}\end{aligned}\ ] ] the integral over @xmath459 can be evaluated analytically .
we can simplify the expression by replacing @xmath460 by 1 , because the difference is suppressed by a factor of @xmath461 .
most of the terms in eq .
( [ l2-dimreg ] ) are independent of @xmath80 .
the expression can be simplified by expressing it in the form @xmath462 the amplitude @xmath463 at the @xmath41 threshold is real - valued and includes the logarithmic divergence .
the amplitude @xmath279 has an imaginary part for @xmath448 . for energies in the range @xmath375 ,
the imaginary part of @xmath279 is @xmath464 note that this diverges at the @xmath40 threshold @xmath291 .
the first few terms in the laurent expansion of @xmath278 in powers of @xmath316 are @xmath465 .
\label{l2b - laurent}\end{aligned}\ ] ] in a general regularization scheme , the one - loop self - energy subdiagram @xmath453 in eq .
( [ l2b ] ) has a linear ultraviolet divergence , which is included in the term @xmath117 in eq .
( [ d2self ] ) . the resulting divergence in @xmath279
is cancelled by the one - loop diagram with a @xmath35 mass counterterm in fig .
[ fig : l - nlo](c ) .
the counterterm vertex is @xmath466 , with @xmath467 . in the complex mass scheme , the @xmath35 propagator counterterm has the more complicated form given in eq .
( [ newcounterterm ] ) . with this counterterm , the contribution to @xmath277 of order @xmath51 from the one - loop diagram in fig .
[ fig : l - nlo](c ) is @xmath468 \big).\end{aligned}\ ] ] in the term with the factor @xmath469 , the integral is proportional to @xmath204 .
this amplitude in a general regularization scheme is given in eq .
( [ l0-sub ] ) . in the term with the factor @xmath470 , the integral is convergent .
the complete result for the diagram is @xmath471 | the @xmath0 seems to be a loosely - bound hadronic molecule whose constituents are two charm mesons .
a novel feature of this molecule is that the mass difference of the constituents is close to the mass of a lighter meson that can be exchanged between them , namely the pion .
we analyze this feature in a simple model with spin-0 mesons only .
various observables are calculated to next - to - leading order in the interaction strength of the exchanged meson .
renormalization requires summing a geometric series of next - to - leading order corrections .
the dependence of observables on the ultraviolet cutoff can be removed by renormalizations of the mass of the heaviest meson , the coupling constant for the contact interaction between the heavy mesons , and short - distance coefficients in the operator product expansion .
the next - to - leading order correction has an unphysical infrared divergence at the threshold of the two heavier mesons that can be eliminated by a further resummation that takes into account the nonzero width of the heaviest meson . |
within the famous total variation ( tv ) regularization framework , image reconstruction in low - dose ct is formulated as minimizing the cost function expressed as @xmath0 subject to @xmath1 , and that in few - view ct is formulated as @xmath2 , where @xmath3 is the tv norm which aims at regularizing the ill - conditioned reconstruction problem .
a variety of iterative algorithms have been proposed for solving the above problems in ct reconstruction fields based on modifying classical iterative algorithms such as the gradient method , art ( algebraic reconstruction technique ) , sirt ( simultaneous iterative reconstruction technique ) , and sart ( simultaneous algebraic reconstruction technique ) , often combined with the ordered - subsets ( os ) technique [ 1]-[10 ] ( and many others ) .
however , it is fair to say that there exist very few iterative algorithms in engineering literatures , which not only converge fast but also can exactly solve the above tv - regularized problems .
it is known that , from mathematical point of view , this difficulty partly comes from the non - differentiability of tv penalty . however , in applied mathematics fields , the research progress on this topic is very fast so that many nice algorithms to handle the tv regularization have been developed since 2010 .
therefore , we believe that it would be time for ct engineers like us to develop and use more rigorous iterative algorithms for the tv - regularized image reconstructions . in this paper , we propose a very fast iterative algorithm which can be applied to the above two typical tv - regularized image reconstructions in a unified way .
the algorithm derivation is outlined as follows .
first , the original minimization problem is reformulated into the standard saddle point ( primal - dual ) problem by using the lagrangian duality , to which we apply the alternating projection proximal ( app ) algorithm which belongs to a class of first - order primal - dual methods including chambolle - pock algorithm , generalized iterative soft - thresholding ( gist ) algorithm , and alternating extragradient ( ae ) algorithm [ 11]-[14 ] .
however , the resulting algorithm converges very slowly , mainly because its overall structure is same as the simultaneous iterative reconstruction methods such as the classical sirt and sart algorithms . to overcome this drawback ,
we precondition the iteration formula using the ramp filter of filtered backprojection ( fbp ) reconstruction algorithm in such a way that the solution to the preconditioned iteration perfectly coincides with the solution to the original problem .
the final algorithm can be interpreted as the first - order primal - dual method accelerated by the fbp - type preconditioning using the ramp filter .
unlike the famous known fbp - based acceleration technique called the iterative fbp algorithm [ 15],[16 ] , for both the above two different formulations ( low - dose ct and few - view ct ) , the proposed algorithm converges to the solution exactly minimizing the cost function .
we evaluate this algorithm for both the low - dose ct case and the few - view ct case with 32 projection data . in the low - dose ct experiment ,
the proposed algorithm converged with approximately 10 iterations to the almost same image as the gist algorithm with 1,000 iterations . in the few - view ct experiment , the proposed algorithm required only 3 iterations to reach to the almost same image as the chambolle - pock algorithm with 1,000 iterations .
these accelerations were achieved by the introduction of fbp - type preconditioning in the primal - dual space ( in the saddle point formulation ) , which has not been investigated yet and is an original contribution of this paper .
in this section , we formulate image reconstruction in the low - dose ct and review typical existing iterative algorithms .
we denote an image by a @xmath4-dimensional vector @xmath5 and denote a corresponding projection data ( sinogram ) by an @xmath6-dimensional vector @xmath7 .
we denote the system matrix which relates @xmath8 to @xmath9 by @xmath10 , where we assume that @xmath11 and @xmath9 is contaminated with statistical noise . throughout this paper , we assume that the projection data is acquired by the standard parallel - beam geometry so that @xmath9 consists of uniform samples of continuous projection data @xmath12 .
normally , image reconstruction in this setup can be formulated as the problem of minimizing the penalized weighted least - squares ( p - wls ) cost function expressed as @xmath13 where @xmath14 denotes @xmath15 diagonal matrix in which each diagonal element @xmath16 is the inverse of noise variance @xmath17 , and @xmath18 is the penalty function to smooth the image [ 17],[18 ] .
although there exist several choices in @xmath18 , throughout this paper , we assume that @xmath18 is the tv penalty function defined by @xmath19 where @xmath20 and @xmath21 are inner product representations of finite difference operations around the @xmath22-th pixel along the horizontal and vertical directions , respectively .
see fig . 1 for the detailed definitions of @xmath23 and @xmath24 .
since 2008 when ge healthcare developed the asir ( advanced statistical iterative reconstruction ) software , iterative low - dose ct reconstruction has been progressed according to the following three stages .
[ image space denoising ( 1-st generation ) ] this class of reconstruction methods first perform an fbp reconstruction followed by a smoothing by using an iterative edge - preserving denoising algorithm such as the tv - denoising and the mrf - based denoising . [ iterative fbp algorithm ( 2-nd generation ) ] typically , this class of reconstruction methods are based on the following iteration formula @xmath25,\end{aligned}\ ] ] where @xmath26 $ ] denotes the projection operator onto the positive orthant @xmath1 and @xmath27 is the 1-d ramp filter of fbp reconstruction algorithm having the frequency response @xmath28 . thanks to the introduction of @xmath27 during the iteration , eq .
( 3 ) converges very fast .
however , it is well - known that is can not exactly minimize the p - wls cost function of eq .
another limitation is that the penalty function @xmath29 needs to be differentiable so that the tv penalty , which is non - differentiable , can not be used .
[ true iterative reconstruction ( ir ) algorithm ( 3-rd generation ) ] this class of reconstruction methods try to exactly minimize the p - wls cost function of eq .
( 1 ) . for example , if we use the gradient method , its iteration formula can be expressed as @xmath30.\end{aligned}\ ] ] the drawback of this method is that its convergence is rather slow compared with the iterative fbp algorithm when no acceleration techniques are incorporated .
another limitation is that the penalty function @xmath29 needs to be differentiable so that the tv penalty can not be used again .
[ preconditioned gradient method ] an alternative approach which is known in image reconstruction community is to use the iteration formula as @xmath31,\end{aligned}\ ] ] where @xmath32 denotes the 2-d ramp filter in image space having the frequency response @xmath33 [ 19 ] . unlike eq .
( 3 ) , this method converges to an exact minimizer of eq .
( 1 ) , but it is rarely used mainly because the 2-d filter is computationally intensive and its design is troublesome .
we note that the proposed algorithm derived in the following sections also uses the fbp - type acceleration as in eq .
( 3 ) , but it exactly converges to a minimizer of eq .
( 1 ) and also allows to use the tv penalty .
therefore , we believe that the proposed algorithm overcomes the most major drawbacks of existing low - dose ct iterative reconstruction algorithms . in fig .
2 , we show a comparison between the true ir reconstruction and the iterative fbp reconstruction for a numerical chest phantom without a regularization term in the low - dose ct setup . as expected from the theory , it can be observed that the noise property of iterative fbp reconstruction is worse compared with the true ir reconstruction .
this result motivates us to develop an alternative fbp - type acceleration technique which is able to exactly minimize the wls and p - wls cost functions . and the vertical difference @xmath34 used in the tv penalty function.,height=264 ] from this section , we derive the proposed algorithm , which is based on reformulating the minimization of eq .
( 1 ) into a saddle point ( primal - dual ) problem , precondition it , and applying the alternating projection proximal ( app ) algorithm which belongs to a class of first - order primal - dual methods [ 11 ] .
first , eq .
( 1 ) can be reformulated into the following equality constrained minimization by introducing an additional variable @xmath35 .
@xmath36 where we note that @xmath35 can be interpreted as forward - projected projection data computed from @xmath8 .
next , we perform a preconditioning on the constraint @xmath37 . when we solve eq .
( 6 ) using an iterative algorithm , its convergence speed strongly depends on the condition of linear constraint @xmath37 .
so , we perform the preconditioning by multiplying @xmath38 by some non - singular @xmath15 matrix @xmath39 . then , eq . ( 6 ) can be converted into @xmath40 the detailed form of @xmath39 is not shown here , but we discuss later which choice of @xmath39 is best to accelerate the convergence . intuitively , if we choose @xmath39 in such a way that @xmath41 approximates the ramp filter in the fbp reconstruction , the final structure of iteration formula resembles the iterative fbp algorithm leading to a fast convergence .
furthermore , we remark that the rewriting from eq .
( 6 ) to eq .
( 7 ) does not alter the problem solution @xmath42 .
next , we reformulate eq .
( 7 ) into a form of saddle point problem [ 11]-[14 ] .
the procedure is as follows .
the lagrangian function @xmath43 corresponding to eq .
( 7 ) with respect to the constraint @xmath44 is defined by @xmath45 where @xmath46 is the lagrange multiplier vector , which is also called the dual variable .
therefore , eq . ( 7 ) can be converted into the saddle point problem@xmath47 furthermore , the variable @xmath35 can be eliminated in eq .
( 9 ) by solving the minimization with respect to @xmath35 . by performing this computation , we obtain @xmath48 where @xmath49 denotes the inner product . in some literatures ,
( 10 ) is called the standard form of saddle point problem [ 12 ] .
it is also called the primal - dual ( pd ) problem whereas the original problem of eq .
( 1 ) is called the primal ( p ) problem [ 20]-[22 ] . by using the so - called strong duality theorem in optimization literatures
, we can prove the following theorem [ 20]-[22 ] .
[ theorem ] assume that the penalty function @xmath18 is a possibly non - differentiable convex function .
we denote the solution to the primal problem ( eq .
( 1 ) ) by @xmath50 .
we denote the solution to the primal - dual problem ( eq .
( 10 ) ) by @xmath51 . then
, the following two properties hold .
\(1 ) @xmath52 \(2 ) @xmath53 from this theorem , if we succeed in exactly solving the saddle point problem of eq .
( 10 ) , its solution vector @xmath54 is also a solution of the original problem of eq .
( 1 ) , _ i.e. _ @xmath50 .
in summary of the above discussion , we need to solve the following saddle point problem
. @xmath55 where the non - negativity constraint @xmath1 was put into the cost function @xmath56 as the indicator function @xmath57 defined by @xmath58
the proposed algorithm is based on solving eq . ( 11 ) instead of eq .
there exist a variety of iterative methods to solve eq .
( 11 ) in optimization literatures such as the classical multiplier method , augmented lagrangian method , alternating direction method of multiplier ( admm ) method , etc [ 20]-[23 ] . among them
, we use a class of iterative methods called the first - order primal - dual methods , because they require relatively simple computations per iteration [ 11]-[14 ] .
the basic overall structure of this iterative method is to repeat an update of primal variable @xmath8 along the descent direction of lagrangian @xmath59 , _
i.e. _ the descent direction of @xmath60 , and an update of dual variable @xmath46 along the ascent direction of lagrangian @xmath61 , _
i.e. _ the ascent direction of @xmath62 , alternately with some additional extrapolation step .
the role of extrapolation step is to guarantee that each update from @xmath63 to @xmath64 becomes a non - expansive mapping leading to a convergence to the saddle point . in each update along the descent or ascent direction , either the gradient step or the proximal step
can be used dependent on differentiabilities of @xmath56 and @xmath65 .
typically , there exist four variations in the first - order primal - dual methods , which are the chambolle - pock algorithm , generalized iterative soft - thresholding ( gist ) algorithm , alternating projection proximal ( app ) algorithm , and alternating extragradient ( ae ) algorithm [ 11]-[14 ] .
the differences in used operations among the four algorithms are briefly summarized in table 1 .
.differences among the four first - order primal - dual algorithms . [ cols="<,<,<,<,<",options="header " , ] in our problem , as is clear from eq .
( 11 ) , @xmath56 is non - differentiable because it corresponds to the tv penalty ( combined with the non - negativity constraint ) whereas @xmath65 is differentiable ( quadratic form ) .
thus , we decided to employ the app algorithm [ 11 ] .
for general ( non - differentiable ) @xmath56 and ( differentiable ) @xmath65 , the iteration formula of app algorithm in its original form is summarized as follows .
@xmath66 } \nonumber \\ & & \ \ \ \ \ \ \ \ \overline{\vec{\mu}}^{(k+1)}=\vec{\mu}^{(k)}+\sigma(d^{1/2}a\vec{x}^{(k)}-\nabla
h(\vec{\mu}^{(k ) } ) ) \\ & & { \rm [ primal\ update\ ( proximal ) ] } \nonumber \\ & & \ \ \ \ \ \ \ \ \vec{x}^{(k+1)}={\rm prox}_{\tau g}(\vec{x}^{(k)}-\tau a^td^{t/2}\overline{\vec{\mu}}^{(k+1)})\\ & & { \rm [ dual\ update\ ( gradient ) ] } \nonumber \\ & & \ \ \ \ \ \ \ \
\vec{\mu}^{(k+1)}=\vec{\mu}^{(k)}+\sigma(d^{1/2}a\vec{x}^{(k+1)}-\nabla h(\vec{\mu}^{(k)})),\end{aligned}\ ] ] where @xmath67 is the iteration number , and @xmath68 , @xmath69 are stepsize parameters corresponding to the primal update and the dual update , respectively . to guarantee the convergence , we assume that @xmath70 and @xmath71 are selected such that @xmath72 where @xmath73 is lipschitz constant of the gradient @xmath74 [ 11 ] .
furthermore , the operator @xmath75 appearing in eq .
( 14 ) denotes the proximity ( prox ) operator defined by @xmath76 where @xmath70 is called the stepsize parameter [ 24 ] .
we will describe later on how to compute the prox operator when @xmath18 is the tv penalty .
we note that eq .
( 13 ) represents the extrapolation step to mathematically guarantee the convergence , eq .
( 14 ) is the primal update using the prox operator , and eq .
( 15 ) is the dual update using the gradient step . the iteration formula in our special case of eq .
( 11 ) is obtained by applying the general algorithm of eqs .
( 13)-(15 ) followed by the variable change @xmath77 and rewriting the extrapolation step into a much simpler form .
the resulting iteration formula is expressed as @xmath66 } \nonumber \\ & & \ \ \ \ \ \ \ \
\overline{\vec{\mu}}^{(k+1)}=\vec{\mu}^{(k)}+\sigma d(a\vec{x}^{(k)}-\vec{b}-w^{-1}\vec{\mu}^{(k)})=2\vec{\mu}^{(k)}-\vec{\mu}^{(k-1)}-\sigma dw^{-1}(\vec{\mu}^{(k)}-\vec{\mu}^{(k-1 ) } ) \\ & & { \rm [ primal\ update ] } \nonumber \\ & & \ \ \ \ \ \ \ \ \vec{x}^{(k+1)}={\rm prox}_{\tau g}(\vec{x}^{(k)}-\tau a^t\overline{\vec{\mu}}^{(k+1)})\\ & & { \rm [ dual\ update ] } \nonumber \\ & & \ \ \ \ \ \ \ \ \vec{\mu}^{(k+1)}=\vec{\mu}^{(k)}+\sigma d(a\vec{x}^{(k+1)}-\vec{b}-w^{-1}\vec{\mu}^{(k)}),\end{aligned}\ ] ] where @xmath78 and we simplified the extrapolation step using the fact that the @xmath79 term in the extrapolation step is same as @xmath80 appearing in the previous dual update .
the algorithm derivation was finished .
with respect to computational requirements , the dominant computations in implementing eqs .
( 18)-(20 ) are a forward projection @xmath81 , a backprojection @xmath82 , and a prox operator associated with the tv penalty in each iteration .
therefore , we expect that the computational costs of the proposed algorithm are similar to those of the other iterative reconstruction algorithms which use the tv penalty . in implementing eqs .
( 18)-(20 ) , there still exist two unclear points .
the first point is how to compute @xmath75 in the primal update , which amounts to solving the minimization of eq .
when @xmath18 is the tv penalty of eq .
( 2 ) , eq . (
17 ) is same as the so - called tv denoising problem ( rof problem ) investigated in [ 25],[26 ] , to which a variety of efficient algorithms are available . in our implementations , we use chambolle s projection algorithm for this purpose modified in such a way that the non - negativity constraint @xmath1 is incorporated [ 26 ] . the second unclear point concerns how to design the preconditioning matrix @xmath83 to achieve a fast convergence . intuitively , from eqs .
( 18)-(20 ) , it can be inspired that a nice choice of @xmath83 is the ramp filter in the fbp reconstruction , because , in this case , the first iteration initialized with @xmath84 coincides with the fbp reconstruction followed by the tv denoising .
however , this is not the best choice when the projection data @xmath9 contains statistical noise .
we discuss this issue below .
first , we note that @xmath83 needs to be a positive definite matrix , because @xmath83 is required to be expressed as @xmath78 . by forgetting the extrapolation step ( for some moment ) and combining eqs .
( 19 ) and ( 20 ) , the approximate iteration formula with respect to the dual variable @xmath46 is expressed as @xmath85 furthermore , neglecting the prox operator , we have @xmath86 equation ( 22 ) implies that the meaning of @xmath46 is the noise compoment in the projection data and the app method is implicitly solving the equation @xmath87 at each iteration @xmath67 by using the preconditioned iteration of eq .
( 22 ) . from this observation , the best choice of @xmath83
is clearly given by @xmath88 however , eq .
( 23 ) is not practical because its computation requires a large matrix inverse .
so , we use a shift - invariant approximation @xmath89 where @xmath90 is the average value of diagonal elements of @xmath91 .
then , noting that @xmath92 is the blurring operator @xmath93 in projection data space ( @xmath94 is the number of view angles over @xmath95 in the parallel - beam projection data ) , @xmath83 becomes the frequency domain filter expressed as @xmath96 where @xmath97 and @xmath98 denote the 1-d fourier transform with respect to the radial variable in the projection data space and its inverse , respectively .
we note that eq .
( 24 ) is the smoothed ramp filter .
we show the frequency response @xmath99 with varying parameter @xmath70 in fig .
3 . finally , we summarize that the smoothed ramp filter is the best preconditioning matrix @xmath83 in the case of low - dose ct reconstruction .
we experimentally confirmed that there exists an essential difference between the ordinary ( non - smoothed ) ramp filter and the smoothed ramp filter in the convergence speed of cost function value @xmath100 .
used in the proposed algorithm for the low - dose ct.,height=188 ] we summarize the proposed algorithm for the low - dose ct in algorithm 1 . as a special case of @xmath101 , _
i.e. _ absence of the tv penalty term , we obtain a fast fbp - preconditioned algorithm to exactly minimize the wls cost function , which is also summarized in algorithm 1. p40em algorithm 1 : low - dose ct tv - regularized wls reconstruction + prepare the preconditioning matrix @xmath83
. give initial primal and dual vectors by @xmath102 and @xmath103 .
set stepsize parameters @xmath104 such that @xmath105 and @xmath106 .
execute the following steps for @xmath107 .
[ step 1 ] extrapolation step + @xmath108 + [ step 2 ] primal update + @xmath109 + @xmath110 + [ step 3 ] dual update + @xmath111 + p40em algorithm 1 : low - dose ct wls reconstruction + prepare the preconditioning matrix @xmath83
. give initial primal and dual vectors by @xmath102 and @xmath103 .
set stepsize parameters @xmath104 such that @xmath105 and @xmath106 .
execute the following steps for @xmath107 .
[ step 1 ] extrapolation step + @xmath108 + [ step 2 ] primal update + @xmath112 $ ] + [ step 3 ] dual update + @xmath111 +
the algorithm derivation used in section 2 based on the first - order primal - dual methods combined with the preconditioning is a rather general framework so that it can be applied to develop a fast convergent reconstruction algorithm in the few - view ct formulation .
we denote an image by a @xmath4-dimensional vector @xmath5 and denote a corresponding projection data ( sinogram ) by an @xmath6-dimensional vector @xmath7 .
we denote the system matrix which relates @xmath8 to @xmath9 by @xmath10 , where we assume that @xmath113 and @xmath9 does not contain statistical noise .
normally , image reconstruction in this setup can be formulated as the problem of minimizing a penalty function @xmath18 under the linear constraint @xmath114 as @xmath115 where we assume that @xmath29 is the tv penalty defined by eq .
( 2 ) throughout this section . very often , in engineering literatures , the problem of eq . ( 25 ) is converted into the unconstrained minimization @xmath116 followed by solving it using a technique of unconstrained optimizations .
however , the solution obtained in this way can be quite different from that of the original constrained problem of eq .
the proposed algorithm below allows us to exactly solve eq .
( 25 ) .
we derive the proposed algorithm to solve eq .
( 25 ) using the same mathematical framework as in section 2 below .
first , we introduce the @xmath15 non - singular preconditioning matrix @xmath39 into eq .
( 25 ) as @xmath117 we note that the solution to eq . ( 25 ) is same as the solution to eq .
. the lagrangian function @xmath118 corresponding to eq .
( 26 ) with respect to the constraint @xmath119 is defined by @xmath120 where @xmath46 is the lagrange multiplier vector ( the dual variable ) .
therefore , eq . ( 26 ) can be converted into the saddle point problem @xmath121 where @xmath57 is the indicator function defined by eq .
( 12 ) . since the structure of eq .
( 28 ) is same as that of eq .
( 11 ) ( there exists only a difference in the form of @xmath65 ) , we can use the same approach as in the low - dose ct case to develop an iterative algorithm . applying the app algorithm given by eqs .
( 13)-(15 ) to eq .
( 28 ) yields the following iterative algorithm . @xmath66 } \nonumber \\ & & \ \ \ \ \ \ \ \ \overline{\vec{\mu}}^{(k+1)}=\vec{\mu}^{(k)}+\sigma
d(a\vec{x}^{(k)}-\vec{b})=2\vec{\mu}^{(k)}-\vec{\mu}^{(k-1 ) } \\ & & { \rm [ primal\ update ] } \nonumber \\ & & \ \ \ \ \ \ \ \ \vec{x}^{(k+1)}={\rm prox}_{\tau g}(\vec{x}^{(k)}-\tau a^t\overline{\vec{\mu}}^{(k+1)})\\ & & { \rm [ dual\ update ] } \nonumber \\ & & \ \ \ \ \ \ \ \ \vec{\mu}^{(k+1)}=\vec{\mu}^{(k)}+\sigma d(a\vec{x}^{(k+1)}-\vec{b}).\end{aligned}\ ] ] the algorithm derivation was finished .
we note that the only difference between the low - dose ct case and the few - view ct case lies in the absence of @xmath122 term in the latter from the comparison between eqs .
( 18)-(20 ) and eqs .
( 29)-(31 ) .
next , we discuss about the preconditioning matrix @xmath83 to achieve a fast convergence . by following the same discussion as in the low - dose ct case ,
thanks to the absence of @xmath122 term , the best preconditioning matrix @xmath83 in eqs .
( 29)-(31 ) can be shown to be the ramp filter in the fbp reconstruction , which is expressed as @xmath123 where we note again that @xmath94 is the number of view angles over @xmath95 in the parallel - beam projection data , and @xmath97 and @xmath98 denote the 1-d fourier transform with respect to the radial variable in the projection data space and its inverse , respectively . with this choice of @xmath83 , the first iteration initialized with @xmath84 coincides with the fbp reconstruction followed by the tv smoothing ( denoising ) .
we summarize the proposed algorithm for the few - view ct in algorithm 2 .
p40em algorithm 2 : few - view ct tv - regularized reconstruction + prepare the preconditioning matrix @xmath83
. give initial primal and dual vectors by @xmath102 and @xmath103 .
set stepsize parameters @xmath104 such that @xmath124 .
execute the following steps for @xmath107 .
[ step 1 ] extrapolation step + @xmath125 + [ step 2 ] primal update + @xmath109 + @xmath126 + [ step 3 ] dual update + @xmath127 +
we have performed two simulation studies to demonstrate performances of algorithm 1 in the low - dose ct reconstruction and algorithm 2 in the few - view ct reconstruction .
the details of simulation studies are summarized below .
[ low - dose ct simulation ] we used a numerical phantom called the spot phantom consisting of @xmath128 ( pixels ) and a single slice of chest - scan ct image consisting of 320@xmath129320 ( pixels ) .
the simulated projection data was computed with 1,200 ( angles ) by the parallel - beam geometry followed by adding noise which follows transmission poisson statistics , from which image reconstructions were performed .
we compared the proposed algorithm with the following three competitive algorithms .
\(a ) gradient descent algorithm : the standard gradient descent algorithm to minimize the tv - regularized wls cost function of eq .
( 1 ) was implemented . to deal with the non - differentiability of tv penalty ,
the tv norm was approximated by the standard smoothing technique as @xmath130 where @xmath131 is a small positive number to control the degree of smoothing .
\(b ) iterative fbp algorithm [ 15],[16 ] : the iterative fbp algorithm was implemented mainly to demonstrate that it can not accurately minimize the tv - regularized wls cost function of eq .
( 1 ) and there exists an image degradation although it converges very fast .
\(c ) generalized iterative soft - thresholding ( gist ) algorithm [ 13 ] : we also implemented one of newest iterative algorithms developed by applied mathematicians , which is able to exactly minimize the tv - regularized wls cost function of eq .
since this algorithm is of simultaneous update type such as the sirt and sart algorithms , we expect that its convergence is rather slow .
[ few - view ct simulation ] we used the spot phantom same as in the low - dose ct simulation .
the simulated projection data was computed with 32 ( angles ) by the parallel - beam geometry , from which image reconstructions were performed .
we compared the proposed algorithm with the following two competitive algorithms .
\(a ) chambolle - pock algorithm [ 4],[12 ] : we implemented one of newest iterative algorithms developed by applied mathematicians , which is able to exactly solve the minimization of tv penalty under the constraints @xmath132 .
since this algorithm is of simultaneous update type as in the gist algorithm , we expect that its convergence is rather slow .
\(b ) art algorithm [ 27 ] : we also implemented the standard art algorithm without the tv penalty , which is known to converge fast thanks to its row - action ( sequential update ) structure . in figs .
4 - 5 , we show reconstructed images in the low - dose ct case together with corresponding root mean squares ( rmse ) reconstruction errors defined by @xmath133 where @xmath134 denotes the reconstructed pixel value and @xmath135 denotes the true pixel value . also , in fig .
6 , we show the corresponding convergence properties of each reconstruction algorithm .
these figures clearly demonstrate that algorithm 1 converges to the exact minimizer of the tv - regularized wls cost function very fast .
in particular , it can be observed that the iterative fbp algorithm converges to an image with worse rmse value although its convergence is very fast , whereas algorithm 1 converges to the exact minimizer with a much faster speed compared with the other slow algorithms such as the gradient descent and gist algorithms thanks to the fbp - type preconditioning . in fig .
7 , we show reconstructed images in the few - view ct case together with corresponding rmse reconstruction errors . also , in fig .
8 , we show the corresponding convergence properties of each reconstruction algorithm .
these figures clearly demonstrate that algorithm 2 converges to an image with much smaller rmse value compared with the art algorithm , and its convergence is much faster than the chambolle - pock algorithm again thanks to the fbp - type preconditioning .
was plotted.,height=566 ]
in this paper , we proposed new image reconstruction algorithms for two typical tv - regularized reconstruction problems in tomography .
the first algorithm ( algorithm 1 ) minimizes the tv - regularized wls cost function for the low - dose ct , and the second algorithm ( algorithm 2 ) minimizes the tv penalty under the data fidelity linear constraint for the few - view ct .
the both algorithms were derived from the saddle point reformulation of each optimization problem followed by the preconditioning using the ramp filter of fbp reconstruction algorithm and applying the alternating projection proximal ( app ) iterative algorithm .
we demonstrated that both algorithm 1 and algorithm 2 converge very fast while its convergence to the exact solution of each optimization problem is guaranteed . on the low - dose ct reconstruction ,
very recently , we began to use a shift - variant preconditioner instead of the smoothed ramp filter , which approximates @xmath136 better , and confirmed an essential improvement .
on the few - view ct reconstruction , we began to extend the formulation to @xmath137 by introducing the error tolerance @xmath138 .
these newest results will be published elsewhere .
this work was partially supported by jsps kakenhi grant number 15k06103 .
1 e.y.sidky and x.pan `` image reconstruction in circular cone - beam computed tomography by constrained , total - variation minimization , '' _
phys.med.biol._ , vol.53 , pp.4777 - 4807 , 2008 .
h.yu and g.wang `` a soft - thresholding filtering approach for reconstruction from a limited number of projections , '' _
phy.med.biol._ , vol.55 , pp.3905 - 3916 , 2010 .
s.ramani and j.a.fessler `` a splitting - based iterative algorithm for accelerated statistical x - ray ct reconstruction , '' _
ieee trans.med.imaging_ , vol.31 , pp.677 - 688 , 2012 .
e.y.sidky , j.h.jorgensen , and x.pan `` convex optimization problem prototyping for image reconstruction in computed tomography with the chambolle - pock algorithm , '' _ phys.med.biol._ , vol.57 , pp.3065 - 3091 , 2012 .
l.ritschl , f.bergner , c.fleischmann , m.kachelriess `` improved total variation - based ct image reconstruction applied to clinical data , '' _
phys.med.biol._ , vol.56 , pp.1545 - 1561 , 2011 .
j.song , q.h.liu , g.a.johnson , and c.t.bade `` sparseness prior based iterative image reconstruction for retrospectively gated cardiac micro - ct , '' _
med.phys._ , vol.34 , pp.4476 - 4483 , 2007 .
z.chen , x.jin , l.li , and g.wang `` a limited - angle ct reconstruction method based on anisotropic tv minimization , '' _
phys.med.biol._ , vol.58 , pp.2119 - 2141 , 2013 .
z.tian , x.jia , k.yuan , t.pan , and s.b.jiang `` low dose ct reconstruction via edge - preserving total variation regularization , '' _
phys.med.biol._ , vol.56 , pp.5949 - 5967 , 2011 .
m.defrise , c.vanhove , and x.liu `` an algorithm for total variation regularization in high - dimensional linear problems , '' _ inverse problems _ , vol.27 , paper no . 065002 , 2011 .
v.p.gopia , p.palanisamya , k.a.wahidb , p.babync , and d.cooperd `` micro - ct image reconstruction based on alternating direction augmented lagrangian method and total variation , '' _ comput.med.imaging graph .
_ , vol.37 , pp.419 - 429 , 2013 .
p.tseng `` alternating projection - proximal methods for convex programming and variational inequalities , '' _
siam j.opt._ , vol.7 , pp.951 - 965 , 1997 .
a.chambolle and t.pock `` a first - order primal - dual algorithm for convex problems with applications to imaging , '' _ j.math.imaging vis .
_ , vol.40 , pp.120 - 145 , 2011 .
i.loris and c.verhoeven `` on a generalization of the iterative soft - thresholding algorithm for the case of non - separable penalty , '' _ inverse problems _ , vol.27 , paper no.125007 , 2011 .
s.bonettini and v.ruggiero `` an alternating extragradient method for total variation - based image restoration from poisson data , '' _ inverse problems _ , vol.27 , paper no.095001 , 2011 .
k.zeng , z.chen , l.zhang , and g.wang `` an error - reduction - based algorithm for cone - beam computed tomography , '' _
med.phys._ , vol.31 , pp.3206 - 3212 , 2004 .
j.sunnegardh , p-e.danielsson `` regularized iterative weighted filtered backprojection for helical cone - beam ct , '' _ med phys _ ,
vol.35 , pp.4173 - 4185 , 2008 .
j.a.fessler `` penalized weighted least - squares image reconstruction for positron emission tomography , '' _
ieee trans.med.imaging_ , vol.13 , pp.290 - 300 , 1994 .
e.u.mumcuoglu , r.leahy , s.r.cherry , and z.zhou `` fast gradient - based methods for bayesian reconstruction of transmission and emission pet images , '' _
ieee trans.med.imaging_ , vol.13 , pp.687 - 701 , 1994 . g.chinn and s-c.huang `` a general class of preconditioners for statistical iterative reconstruction of emission computed tomography , '' _ ieee trans.med.imaging_ , vol.16 , pp.1 - 10 , 1997 .
d.p.bertsekas `` nonlinear programming , '' athena scientific , 1999 .
d.p.bertsekas `` convex optimization algorithms , '' athena scientific , 2015 .
r.glowinski and p.letallec `` augmented lagrangian and operator - splitting methods in nonlinear mechanics , '' siam , 1987 .
s.boyd , n.parikh , and
e.chu `` distributed optimization and statistical learning via the alternating direction method of multipliers , '' foundations and trends(r ) in machine learning , now publishers inc . , 2011 .
n.parikh and s.boyd `` proximal algorithms , '' foundations and trends(r ) in optimization , now publishers inc . , 2013 .
l.i.rudin , s.osher , and e.fatemi `` nonlinear total variation noise removal algorithm , '' _ physica d _ , vol.60 , pp.259 - 268 , 1992 .
a.chambolle `` an algorithm for total variation minimization and applications , '' _ j.math.imaging vis .
_ , vol.20 , pp.89 - 97 , 2004 .
g.t.herman and l.b.meyer `` algebraic reconstruction techniques can be made computationally efficient , '' _
ieee trans.med.imaging_ , vol.12 , pp.600 - 609 , 1993 . | this paper concerns iterative reconstruction for low - dose and few - view ct by minimizing a data - fidelity term regularized with the total variation ( tv ) penalty .
we propose a very fast iterative algorithm to solve this problem .
the algorithm derivation is outlined as follows .
first , the original minimization problem is reformulated into the saddle point ( primal - dual ) problem by using the lagrangian duality , to which we apply the first - order primal - dual iterative methods .
second , we precondition the iteration formula using the ramp filter of filtered backprojection ( fbp ) reconstruction algorithm in such a way that the problem solution is not altered .
the resulting algorithm resembles the structure of so - called iterative fbp algorithm , and it converges to the exact minimizer of cost function very fast . |
the solar corona supports a rich variety of wave phenomena whose main features are determined by the magnetic field and plasma structuring . for our purposes
it is important to distinguish between different wave manifestations such as driven and impulsively excited perturbations .
driven waves are often caused by a continuous , periodic exciter at the photospheric or chromospheric level and propagate away from the place where they are generated .
this is the case , for example , of waves propagating upwards along spicules @xcite or along the coronal magnetic field @xcite .
the disturbances propagating in filament threads observed by @xcite are probably representative of driven waves too , although the exciting agent is not clear in this case . rapidly propagating quasi - periodic coronal disturbances
@xcite also seem to be caused by a continuous driver that in this case is associated to a flare or the initiation of a cme . for a modelling of these events
see @xcite .
when a magnetic flux tube ( e.g. , spicule , coronal loop , or prominence thread ) is subject to a periodic driving it shows a distinct response . an axisymmetric driver that periodically compresses and expands the tube at one position leads to the propagation of a sausage - type perturbation ( see figure [ fig_cylinder_sausage](a ) ) .
a periodic transverse shaking of the tube results in a kink - type perturbation by which the tube boundary acquires the shape of a snake ( see figure 1(a ) of * ? ? ?
* hereafter paper i ) .
a torsional driver leads to alfvn wave propagation along the tube with no apparent modification of its boundary .
( a)(b ) an impulsive energy release can lead to both standing and propagating disturbances .
the transverse coronal loop oscillations studied by @xcite constitute a good example of the first type of phenomenon , for which there are many observational studies in the literature . for recent observations using the atmospheric imaging assembly ( aia ) on board
the _ solar dynamics observatory _ , see @xcite .
our interest in this work is in the second type of event , i.e. , a magnetic flux tube perturbed by a sudden , localized expansion or compression such as that of figure [ fig_cylinder_sausage](b ) ( see also figure 1(b ) of paper i for a localized transverse excitation ) . in a dispersive medium ,
such initial pulse disperses and transforms into a wave train that induces quasi - periodic plasma oscillations with a range of periods ( rather than the single - period oscillations of driven propagating waves or standing oscillations ) .
hence , the time signal ( e.g. , doppler shift or spectral line intensity ) collected at a given position shows a characteristic `` tadpole '' shape in its wavelet diagram ( e.g. , * ? ? ?
* ; * ? ? ?
the important difference between driven and impulsively excited waves , then , is that the first ones have the period of the driving agent , whereas the second ones display a continuous variation of the period that depends on the size of the initial perturbation and the physical properties of the medium .
several observations of quasi - periodic oscillations have been ascribed to the dispersive behavior of propagating wave packets .
@xcite used the solar eclipse corona imaging system ( secis ) to record white light and coronal green line images of the solar corona during a solar eclipse .
they studied the intensity variations in the vicinity of an active region and detected periodic variations with period in the range 47 s that propagate through the loop apex with a speed of 2100 km s@xmath0 .
the wavelet spectra of signals at different points along the wave path display the `` tadpole '' shape mentioned above .
this wavelet feature is also present in solar decimetric radio bursts @xcite and in radio sources lying in a magnetic fan structure @xcite .
the impulsive excitation of fast magnetohydrodynamic waves , and their subsequent evolution as an oscillatory wave train , has been addressed for different magnetic structures : fast sausage wave trains in a coronal slab @xcite , in a current sheet @xcite , and in a magnetic funnel @xcite .
impulsively generated fast sausage and kink wave packets in coronal holes have also been studied by @xcite .
our paper is concerned with the temporal evolution of a concentrated axisymmetric compression or expansion of a uniform , cylindrical magnetic tube ; see figure [ fig_cylinder_sausage](b ) .
this initial disturbance leads to the propagation of a fast sausage wave train along the tube .
the propagation and dispersion of wave packets in magnetic slabs has already been investigated numerically in slab geometry , but not in cylindrical geometry .
this is the outline of our paper .
section [ sect_gov_eqns ] contains the governing equations of linear fast perturbations in a plasma cylinder in the @xmath1 limit . in sections [ sect_proper ] and [ sect_improper ]
we give a description of fast sausage proper and improper modes , whose spectra are discrete and continuous , respectively .
the initial perturbation is used in section [ sect_fourier ] to obtain analytical expressions for the time evolution of the perturbed variables and the results are presented in section [ sect_results ] .
we consider a homogeneous cylindrical magnetic tube embedded in a uniform environment . in this work we use cylindrical coordinates @xmath2 , with the @xmath3-axis pointing along the cylinder axis .
the magnetic field is uniform in the whole space and parallel to the @xmath3-axis , and so can be written as @xmath4 .
the equilibrium density is @xmath5 where the subscripts ` @xmath6 ' and ` @xmath7 ' denote internal and external values , respectively , and @xmath8 is the tube radius .
next we consider small perturbations about this equilibrium , governed by the linear , ideal magnetohydrodynamic equations , in the cold plasma approximation ( equations ( 1)(3 ) of * ? ? ?
* ) . in our work
@xmath9 is the velocity perturbation , @xmath10 the lagrangian plasma displacement , @xmath11 the magnetic field perturbation , @xmath12 the magnetic pressure perturbation ( with @xmath13 the magnetic permeability of free space ) , and @xmath14 the density perturbation . using equations ( 1)(3 ) of @xcite it is straightforward to see that , if @xmath15 and @xmath16 are initially zero , then they vanish for all times .
in addition , perturbations are governed by @xmath17 @xmath18 @xmath19 @xmath20 @xmath21 where we have written @xmath22 and @xmath23 moreover , @xmath24 is the alfvn speed , whose values inside and outside the magnetic cylinder are respectively given by @xmath25 in this paper we consider a magnetic tube that is denser than the environment , i.e. , @xmath26 and @xmath27 .
for sausage perturbations , @xmath28 , @xmath14 , @xmath10 , @xmath11 , and @xmath9 have no azimuthal dependence and as a consequence the @xmath29-components of equations ( [ mom_eqn ] ) , ( [ b_eqn ] ) , and ( [ v_eqn ] ) dictate that @xmath30 , @xmath31 , and @xmath32 vanish ( provided that they are initially zero ) .
in addition , the variables are of the form @xmath33,\ ] ] and similarly for @xmath14 , @xmath34 , @xmath35 , and @xmath36 .
the eigenfunctions of fast magnetohydrodynamic ( mhd ) modes in a magnetic cylinder have the azimuthal dependence @xmath37 , with @xmath38 for sausage modes .
equation ( [ conti_eqn ] ) gives the radial dependence of the density perturbation @xmath39 equation ( [ mom_eqn ] ) leads to @xmath40 where the radial wavenumber , @xmath41 , is given by @xmath42 equations ( [ b_eqn ] ) and ( [ v_eqn ] ) give the radial components of the magnetic field and velocity perturbations in terms of the radial plasma displacement , @xmath43 @xmath44 finally , from equations ( [ p_eqn ] ) and ( [ xi_pert ] ) we obtain @xmath45 to determine the solution to equation ( [ equ_for_hatp ] ) we need to take into account the possible signs of @xmath46 inside and outside the magnetic cylinder ( the internal and external radial wavenumbers are here denoted as @xmath47 and @xmath48 , respectively ) . for proper fast sausage modes
@xmath49 and this implies that @xmath50 , @xmath51 .
then , the solution to equation ( [ equ_for_hatp ] ) regular at @xmath52 and that does not diverge as @xmath53 is @xmath54 where @xmath55 and @xmath56 are two constants , @xmath57 is the bessel function of the first kind and order @xmath58 , and @xmath59 is the modified bessel function of the second kind and order @xmath58 .
moreover , @xmath60 and @xmath61 is @xmath62 it is important to mention that when using these equations to determine the radial wavenumbers , the positive value is taken , i.e. , @xmath63 , @xmath64 .
an algebraic expression for @xmath55 and @xmath56 is obtained by imposing the continuity of @xmath65 at @xmath66 .
the remaining constant can be imposed arbitrarily .
here we use expressions analogous to those of paper i , that is @xmath67 now that the function @xmath65 is determined , all other perturbations can be obtained from equations ( [ rho_pert ] ) , ( [ xi_pert ] ) , ( [ b_pert ] ) , and ( [ v_pert ] ) .
we write the expression of the radial plasma displacement because it will be used later , @xmath68 where a prime indicates a derivative . from now on
we use the relations @xcite
@xmath69 and write equation ( [ xir_of_r_primes ] ) as @xmath70 imposing the continuity of the radial displacement at @xmath66 leads to the dispersion relation for proper fast sausage modes , @xmath71 in this work the dispersion relation is used to obtain the frequency of a proper mode with a given longitudinal wavenumber , @xmath72 .
therefore , we often write the frequency as @xmath73 . moreover ,
the phase speed and group velocity are defined in the usual manner , @xmath74 although the phase speed is not used in our calculations nor in the interpretation of results , it is also given here because it is an important parameter of proper modes .
the main features of @xmath75 , @xmath76 , and @xmath77 of proper sausage modes in a magnetic cylinder have been discussed in previous works ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
they are summarised below .
first , we discuss some properties of the eigensolutions for the possible sign combinations of @xmath72 and @xmath75 ; they will be useful when the method of fourier integrals is applied to the wave propagation problem .
it is clear that the two pairs of values @xmath78 and @xmath79 correspond to the same eigenmode , propagating in a given direction along the @xmath3-axis .
it is obvious that the solutions presented above have this property : if the signs of both @xmath72 and @xmath75 are simultaneously changed , then equations ( [ ki ] ) and ( [ kappae ] ) show that the internal and external radial wavenumbers are not modified . as a result , these new values of the longitudinal wavenumber and frequency also satisfy the dispersion relation . in other words , if @xmath73 is the frequency of a proper mode , then @xmath80 is also a frequency of a proper mode .
using equations ( [ rho_pert ] ) , ( [ b_pert ] ) , ( [ v_pert ] ) , ( [ hatp ] ) , and ( [ xir_of_r ] ) , we see that the perturbations of both modes are identical .
finally , we also have @xmath81 and @xmath82 .
next let us consider a fixed @xmath72 : if @xmath73 is a solution to the dispersion relation , then @xmath83 corresponds to the eigenmode propagating in the opposite direction along the magnetic tube .
again , the two eigenmodes have the same eigenfunctions , but opposite phase speed and group velocity .
thus , in the rest of this section it is enough to concentrate in @xmath84 , @xmath85 .
both positive and negative values of @xmath72 and @xmath75 will be taken into account when deriving the solution to the initial value problem by the method of fourier integrals ( section [ sect_fourier ] ) .
the magnetic cylinder supports an infinite number of fast sausage radial harmonics , characterised by the number of nodes of @xmath86 .
all radial harmonics have a node at the tube axis , @xmath87 .
this is the only node for the fundamental sausage eigenmode .
the first overtone has two nodes in the radial direction , etc . , and all these nodes are inside the magnetic cylinder .
an additional feature of proper fast sausage modes is that they only exist for longitudinal wavenumbers exceeding the cut - off value @xcite , @xmath88 where @xmath89 is the @xmath90th zero ( in increasing order ) of @xmath91 . here
@xmath92 correspond to the fundamental fast sausage mode , its first radial overtone , etc .
+ + in figure [ fig_dr_over ] the frequency , phase speed , and group velocity of the fundamental fast sausage eigenmode and its first overtones are shown for two values of the density ratio , @xmath93 .
the dispersion diagram ( figure [ fig_dr_over](a ) ) is rarely found in the literature .
it shows that all proper modes , including the fundamental one , have a wavenumber cut - off and that their frequency satisfies @xmath94 .
one can see that increasing the density ratio raises the upper frequency limit @xmath95 and so more overtones are present in this diagram .
it is worth to mention that linear waves with @xmath96 also exist , but instead of discrete values of @xmath75 these solutions can have any frequency ( see section [ sect_improper ] ) .
figure [ fig_dr_over](b ) has been presented many times in other works and so is not discussed here .
it is enough to note that the phase speed lies between the internal and external alfvn speeds .
finally , figure [ fig_dr_over](c ) shows the group velocity as a function of @xmath72 .
when the wavenumber approaches its cut - off value from the right , @xmath77 rapidly increases and tends to @xmath97 , i.e. , it tends to 3 and 2 for the red and black lines , respectively . for larger wavenumbers ,
the group velocity decreases to its minimum and finally increases slowly towards @xmath98 . in appendix [ app_large_k ]
we show that the phase speed tends to @xmath98 from above as @xmath99
. moreover , the group velocity tends to @xmath98 from below for very large wavenumbers .
this behavior of @xmath76 and @xmath77 is in agreement with figures [ fig_dr_over](b ) and ( c ) .
when a fixed @xmath72 is considered , proper modes have discrete frequencies in the range @xmath100 ( see purple dashed lines of figure [ fig_dr_over](a ) ) .
improper mode eigenfunctions can also be expressed in the form of equation ( [ rzt_dependence ] ) , but they can have any frequency in the ranges @xmath101 $ ] and @xmath102 .
paper i contains a detailed discussion of kink improper modes properties .
all the arguments presented in paper i are valid for improper sausage modes and so are not repeated here . the total pressure and radial displacement perturbations
are given by @xmath103 , & r \geq a , \end{cases}\ ] ] @xmath104 , & r > a , \end{cases}\ ] ] where @xmath105 is the bessel function of the second kind and order @xmath58 , and @xmath106 and @xmath107 are two constants , @xmath108 , \vspace{2mm}\\ \displaystyle c_y = \frac{\pi a{k_e}}{2{k_i}}\left[-{k_e}j_1({k_i}a ) j_0({k_e}a ) + { k_i}j_0({k_i}a ) j_1({k_e}a)\right ] .
\end{array}\ ] ] all other perturbations can be derived from the expressions of section [ sect_proper_eigen ] .
in this section we derive the expressions governing the propagation of an axisymmetric impulsive excitation such as that of figure [ fig_cylinder_sausage](a ) .
the procedure to obtain the evolution of the perturbations after the magnetic tube suffers a localized transverse displacement is described in detail in paper i , where all the relevant information can be found ; here only the main equations are given .
the initial value problem is specified by two initial conditions on the radial displacement , @xmath109 \dfrac{\partial\xi_r}{\partial t } ( t=0,r , z)= g(r , z ) .
\end{array}\ ] ] to impose these initial conditions we write the fourier transform of @xmath110 with respect to @xmath3 , @xmath111 for a given longitudinal wavenumber , @xmath112 can be expressed as a linear combination of proper and improper eigenfunctions , @xmath113\hat{\xi}_j(r , k ) \nonumber \\ & + \int_{|k|{v_{\rm ae}}}^\infty\left[a_\omega^+(k)e^{-i\omega t } + a_\omega^-(k)e^{i\omega t}\right]\hat{\xi}_\omega(r , k)\,d\omega , \end{aligned}\ ] ] where @xmath114 and @xmath115 are assumed with no loss of generality . in this expression ,
the fundamental fast sausage mode ( @xmath116 ) and its radial overtones ( @xmath117 ) have frequencies @xmath118 , which are the solutions to the dispersion relation ( equation ( [ dr_kisq_pos ] ) ) , and their radial displacement , @xmath119 , is given by equation ( [ xir_of_r ] ) .
the contribution of each of these proper modes to equation ( [ expand_eigen - main ] ) depends on the amplitude @xmath120 , where the @xmath121 and @xmath122 signs correspond to propagation in opposite directions along the magnetic tube .
moreover , @xmath123 is the number of proper sausage modes for the selected longitudinal wavenumber .
for @xmath124 no proper eigenmodes exist and so @xmath125 , which means that the sum is not present in equation ( [ expand_eigen - main ] ) ; for @xmath126 only the fundamental fast sausage mode exists and then @xmath127 ; and so on . regarding improper modes , their contribution to equation ( [ expand_eigen - main ] ) has the form of an integral because they have a continuous variation of frequency for a given @xmath72 .
the improper modes eigenfunction @xmath128 has been presented in equation ( [ xir_of_r_im ] ) .
we next define the hilbert space of functions on the interval @xmath129 by the scalar product @xmath130 with this definition , the eigenfunctions satisfy the orthogonality conditions @xmath131 where @xmath132 and @xmath133 see appendix [ app_q_of_omega ] . to determine the functions @xmath120 ( @xmath134 ) and @xmath135 we use the initial conditions .
taking @xmath136 in equation ( [ expand_eigen - main ] ) we obtain @xmath137\hat{\xi}_j(r , k ) \nonumber \\ & + \int_{|k|{v_{\rm ae}}}^\infty\left[a_\omega^+(k ) + a_\omega^-(k)\right]\hat{\xi}_\omega(r , k)\,d\omega , \end{aligned}\ ] ] @xmath138 being the fourier transform with respect to @xmath3 of @xmath139 .
now , differentiating equation ( [ expand_eigen - main ] ) with respect to @xmath140 and substituting @xmath141 yields @xmath142\hat{\xi}_j(r , k ) \nonumber \\ & - i\int_{|k|{v_{\rm ae}}}^\infty \omega\left[a_\omega^+(k ) - a_\omega^-(k)\right]\hat{\xi}_\omega(r , k)\,d\omega , \end{aligned}\ ] ] with @xmath143 the fourier transform of @xmath144 with respect to @xmath3 .
the next step is to take the scalar product of equations ( [ init_expand_xi ] ) and ( [ init_expand_der ] ) with @xmath119 . using equation ( [ scalar ] ) , this results in two algebraic expressions for @xmath120 that lead to @xmath145 with @xmath146\,dr \nonumber \\ & + \frac{{\rho_e}j_0({k_i}a)}{{\kappa_e}^2 } \int_a^\infty \left[(\omega_j \tilde{f}\pm i\tilde{g } ) { \kappa_e}r k_1 ( { \kappa_e}r ) \right]\,dr , \end{aligned}\ ] ] @xmath147 analytical expressions for the integrals in this formula can be obtained . using equation ( [ j0pk0p ] ) and the relations @xcite
@xmath148 xk_1'(x)=-xk_0(x)-k_1(x ) , \end{array}\ ] ] we obtain @xmath149\right\}=2xj_0 ^ 2(x ) , \\ [ 2ex ] \dfrac{d}{dx}\left\{x^2\left[k_0
^ 2(x)-k_1 ^ 2(x)\right]\right\}=2xk_0 ^ 2(x ) . \end{array}\ ] ] with the aid of these formulae we get @xmath150 \nonumber \\ & -{k_i}a j_0({k_i}a ) j_1({k_i}a),\end{aligned}\ ] ] @xmath151 \nonumber \\ & + { \kappa_e}a k_0({\kappa_e}a ) k_1({\kappa_e}a).\end{aligned}\ ] ] inserting these expressions in equation ( [ a_denom - main_int ] ) @xmath152 can be cast as @xmath153 \nonumber \\ & \hspace{15ex } -{k_i}a
j_0({k_i}a ) j_1({k_i}a)\bigg\ } \nonumber \\ & + \frac{{\rho_e}j_0 ^ 2({k_i}a)}{{\kappa_e}^4 } \bigg\{\frac{{\kappa_e}^2a^2}{2}\left[k_0 ^ 2({\kappa_e}a)-k_1 ^ 2({\kappa_e}a)\right ] \nonumber \\ & \hspace{15ex } + { \kappa_e}a k_0({\kappa_e}a ) k_1({\kappa_e}a)\bigg\}.\end{aligned}\ ] ] furthermore , the amplitudes of improper modes , @xmath135 , are obtained in a similar fashion by taking the scalar product of equations ( [ init_expand_xi ] ) and ( [ init_expand_der ] ) with @xmath128 .
then we have @xmath154 where @xmath155 \,dr .\end{aligned}\ ] ] we can finally present the main expression of this paper .
the inverse fourier transform of @xmath112 with respect to @xmath72 yields the radial displacement as a function of @xmath156 and @xmath3 , @xmath157 where @xmath112 is given in equation ( [ expand_eigen - main ] ) .
all other perturbations ( i.e. , magnetic field , velocity , and density ) are described by analogous formulas .
the expressions presented in section [ sect_amplitudes ] are valid for any axisymmetric initial conditions described by the functions @xmath139 and @xmath144 in equation ( [ init_cond ] ) . here
we consider @xmath158 ( i.e. , the perturbed system is initially at rest ) and @xmath139 given by @xmath159 with @xmath160 the parameter @xmath161 determines the length of the initial tube disturbance .
in addition , @xmath162 is the maximum radial displacement of the tube boundary at @xmath141 . in our calculations we consider various values of @xmath161 .
the fourier transforms of @xmath139 and @xmath144 with respect to @xmath3 are @xmath163 \tilde g(r , k)=0 , \end{array}\end{aligned}\ ] ] and so the following formulae for @xmath164 and @xmath165 can be derived @xmath166 @xmath167\bigg\}.\end{aligned}\ ] ] therefore , analytical expressions for all the integrals in the amplitudes @xmath120 and @xmath135 have been obtained .
note that if @xmath158 then @xmath168 and @xmath169 .
the procedure to obtain the perturbations with the help of equations ( [ expand_eigen - main ] ) and ( [ four_trans_inv - main ] ) is described in section 5.4 of paper i and thus is not repeated here . in this work we concentrate on the radial displacement at a point on the magnetic tube boundary ( @xmath66 )
. the dimensionless results only depend on two parameters , namely the density ratio @xmath93 ( or equivalently the alfvn speed ratio @xmath170 ) and @xmath161 .
this parameter determines the length of the initial perturbation , which is of the order of @xmath171 ; see equation ( [ f_r ] ) .
we use the parameter values @xmath172 and @xmath173 .
if dimensional results are needed , then two more parameters are required : the tube radius , @xmath8 , and alfvn speed , @xmath98 .
+ + we first consider the radial displacement at the magnetic tube boundary for @xmath141 .
this is obtained from equations ( [ four_trans_inv - main ] ) and ( [ expand_eigen - main ] ) , where the contribution of proper and improper modes are computed separately .
the black line of figure [ fig_t=0 ] shows that this numerically computed @xmath34 reproduces the gaussian initial condition of equations ( [ init_cond ] ) and ( [ f_r ] ) .
in addition , figure [ fig_t=0 ] also reveals that improper modes ( red line ) have a large contribution to the initial wave form . for the considered parameter values , this contribution is larger than that of proper modes ( green line ) .
the reason for the large amplitude of improper modes can be found in the wavenumber cut - off of the fundamental fast sausage eigenmode , that does not exist for @xmath124 .
its overtones have even larger wavenumber cut - offs . then , given the initial disturbance of equation ( [ f_r ] ) and its fourier transform of equation ( [ f_r_k ] ) , that peaks at @xmath174 , the contributions to @xmath175 for @xmath124 can not be accounted for by proper modes .
this implies that improper modes must have an important contribution to the initial amplitude of the perturbation and , therefore , to the posterior evolution of the wave train propagation .
another significant feature of figure [ fig_t=0 ] is that the contribution of both proper and improper modes extends well beyond the length of the tube affected by the initial disturbance .
the two contributions display spatial variations that added together yield the gaussian shape of equation ( [ f_r_k ] ) .
it is worth noticing that improper eigenmodes are more relevant for the present axisymmetric disturbance than for the transverse disturbance : compare figure [ fig_t=0 ] and figures 4(a ) and ( b ) of paper i. the main difference between fast kink and sausage proper modes is that the fundamental proper kink mode exists for all wavenumbers , so that it can account for the main part of the function @xmath175 .
this fact lowers the importance of improper modes in the case of a transverse disturbance and also has a second effect : contrary to the case studied here , at @xmath141 proper and improper modes are confined to the vicinity of the initial perturbation , such as revealed by figures 4(a ) and ( b ) of paper i. from figure [ fig_t=0 ] we also see that the contribution of improper modes increases when either @xmath161 is increased or @xmath93 is decreased .
we next explain this behavior .
an increase of @xmath161 corresponds to a longer initial perturbation , whose fourier transform becomes more concentrated around @xmath174 .
since the fundamental fast sausage mode only exists for @xmath176 , an increase in @xmath161 results in a decrease of @xmath175 for @xmath176 , so that the contribution of proper modes becomes smaller .
this is corroborated by figures [ fig_amplr](a ) and
( b ) , where the contribution of proper modes to the amplitude of @xmath177 at @xmath66 is shown for two values of @xmath161 .
in addition , the role of the density ratio is to modify the cut - off wavenumbers , in such a way that a decrease in @xmath93 leads to a larger wavenumber cut - off ( see figure [ fig_dr_over ] ) .
this then yields a smaller contribution of proper modes : compare figures [ fig_amplr](a ) and ( c ) .
figure [ fig_amplr ] also provides information about the relative importance of proper mode overtones .
such as expected , the largest amplitude of all proper modes corresponds to the fundamental fast sausage mode , whereas its radial overtones possess smaller amplitudes . by comparing panels ( a ) and ( b )
we see that longer initial perturbations result in a negligible amplitude of all radial overtones .
we must mention that in figures [ fig_amplr](a ) and ( b ) only the fundamental mode and its first overtone are present ( solid and dashed lines ) , although in panel ( b ) the first overtone has negligible amplitude , so its curve is invisible . furthermore , in figure [ fig_amplr](c ) the dashed and dotted curves overlap for @xmath178 and may seem to be just one .
+ + next , we concentrate in the evolution of the pulse of figure [ fig_t=0](a ) .
this initial perturbation splits in two identical wave trains that propagate symmetrically in both directions along the magnetic tube .
the early stages of the propagating wave trains are presented in figure [ fig_vae2_delta1_short_t ] and animation 1 .
figures [ fig_vae2_delta1_short_t](a)(c ) correspond to the separation of the two wave trains , whereas in figure [ fig_vae2_delta1_short_t](d ) these two features are fully formed .
the strong dispersion of the axisymmetric perturbation can be well appreciated in this last panel : both wave packets have developed some 6 maxima and minima and their length is more than 4 times that of the perturbation at @xmath141 .
proper and improper modes contribute with similar amplitudes to the total signal represented in figure [ fig_vae2_delta1_short_t ] and animation 1 .
a secondary effect of wave dispersion is the strong reduction of the wave packets amplitude . in the absence of dispersion
, the two wave packets of figure [ fig_vae2_delta1_short_t](d ) would have amplitude 0.5 , whereas their actual peak - to - peak amplitude is under 0.3 .
when discussing figure [ fig_t=0 ] we explained that , when considered separately , the contributions of proper and improper modes affect a section of the magnetic tube much longer than that of the initial perturbation . at @xmath141
these two contributions simply cancel out in the parts of the magnetic tube where the initial excitation is negligible .
this situation also takes place for @xmath179 : although proper and improper mode contributions disturb a large part of the magnetic tube , their sum is concentrated in a smaller section .
thus , ahead of the propagating wave fronts , where the initial perturbation has not yet arrived , these contributions are spatially out of phase and cancel out .
this is also found at @xmath180 of figure [ fig_vae2_delta1_short_t](d ) because the two wave packets have moved beyond this section of the magnetic tube , which is barely disturbed .
+ + + the green and red curves of figure [ fig_vae2_delta1_short_t ] reveal the separate behavior of proper and improper modes .
the first ones propagate with group velocity ranging from @xmath181 ( see the minimum of the black line in figure [ fig_dr_over](c ) ) to @xmath182 . because of this variation of group velocity the green curve displays some dispersion , that becomes more clear for later times ( this issue will be discussed below with the help of figure [ fig_longer_t ] ) . on the other hand ,
improper modes propagate at their own phase speed and so are subject to phase mixing , that leads to stronger dispersion .
this dispersive behavior in turn causes a severe amplitude decrease that is clearly visible when comparing the red lines of figures [ fig_vae2_delta1_short_t](a ) and ( d ) .
the asymptotic behavior of improper modes for large @xmath140 was studied in section 5.3 of paper i , that deals with a transverse initial perturbation .
it was found that the contribution of improper modes decays as @xmath183 .
the same applies to the present axisymmetric perturbation .
+ + at @xmath184 the perturbation of figure [ fig_t=0](a ) , whose early development is presented in figure [ fig_vae2_delta1_short_t ] , is spread over a long section of the tube of length @xmath185 ( compare this value with the initial length of the perturbation @xmath186 ) ; see figure [ fig_longer_t](a ) . here
the total perturbation of the radial displacement and the contributions of proper and improper modes are represented with the usual colors ( black , green , and red , respectively ) .
the vertical lines denote the position of a point initially at @xmath187 and moving at specific velocities : the dash - dotted , dashed , and dotted lines correspond to the maximum group velocity of proper modes ( @xmath188 ) , the minimum group velocity of the fundamental fast sausage mode ( @xmath189 ) , and the minimum group velocity of its first overtone ( @xmath190 ) .
this range of group velocities determines the length of the wave train at a given time .
the wave train shape is controlled by two more features : the dependence @xmath191 of proper modes ( black lines in figure [ fig_dr_over](c ) ) and the amplitude of the perturbation ( figure [ fig_amplr](a ) ) .
wavenumbers slightly above the cut - off occupy the leading part of the wave packet and have the largest amplitudes , but their @xmath77 varies strongly with @xmath72 .
thus , these wavenumbers are strongly dispersed and contribute moderately to the wave train amplitude . on the other hand ,
wavenumbers @xmath192 have group velocities in a small range around the minimum of @xmath77 and for this reason the maximum amplitude of the wave train is found near the dashed line of figure [ fig_longer_t](a ) . a side effect of the wave dispersion
is the decrease of the wave packet amplitude ( compare figures [ fig_vae2_delta1_short_t](d ) and [ fig_longer_t](a ) ) .
again , note the cancellation of proper and improper mode contributions ahead and behind the propagating wave train .
we now investigate the effect of the initial disturbance length and the density ratio on the wave train dispersion .
if the internal to external density ratio is kept fixed and the length of the initial disturbance is doubled ( figure [ fig_longer_t](b ) ) the dispersion of the wave train in its constituent frequencies does not change because the group velocity of proper modes is the same .
then , the positions of maxima and minima in figures [ fig_longer_t](a ) and ( b ) are the same .
nevertheless , according to figures [ fig_t=0](a ) and ( b ) and figures [ fig_amplr](a ) and ( b ) , a longer initial disturbance corresponds to smaller amplitudes of proper modes and for this reason the amplitude of the wave train is much smaller for @xmath193 than for @xmath194 .
finally , we consider the case of a magnetic tube with a larger internal to external density contrast ( @xmath195 ; figure [ fig_longer_t](c ) ) . in this case
the minimum group velocity of the fundamental fast sausage mode and its first overtone are similar to those for @xmath196 and this implies that the position of the vertical dashed and dotted lines in figures [ fig_longer_t](a ) and ( c ) are similar .
but by increasing the density ratio from 4 to 9 , the maximum group velocity of proper modes increases from @xmath188 to @xmath197 and so the front of the wave train travels considerably faster .
hence , the wave packet extends over a much longer section of the magnetic tube , with length @xmath198 .
the leading part of the wave train has a small amplitude because of the strong gradient of @xmath191 near the cut - off wavenumber .
in addition , the largest amplitude is found at the back of the wave packet , were , again , a wide range of wavenumbers with values of @xmath77 near the minimum group velocity concentrate their contribution .
here we compare the response of the magnetic tube boundary under initial axisymmetric and transverse disturbances , respectively described in this paper and in paper i. to make this comparison we consider the wave train passage at a fixed position ( @xmath199 ) and plot the radial displacement as a function of time ( see figure [ fig_comp_kink_sausage ] , in which the left and right columns correspond to axisymmetric and transverse , which gives the radial profile of the initial perturbation , was different from that used in the present work ( equation ( [ init_psi ] ) ) . in order to make a comparison with the results of paper i ,
the proper and improper mode amplitudes of transverse disturbances are presented in appendix [ app_kink ] for the same @xmath200 . ]
disturbances ) .
+ + + + + inspection of figure [ fig_comp_kink_sausage ] reveals the very different dispersive properties of fast sausage and kink waves .
the first ones display rapid oscillations with period @xmath201 , whereas the latter have only one full oscillation with larger amplitude and period @xmath202 .
this disparate behavior occurs because the fundamental fast kink mode does not have a cut - off wavenumber and its group velocity has a more moderate range of variation than that of its sausage counterpart .
the length of the initial perturbation plays a different role in the propagation of axisymmetric and transverse disturbances . in the first case
( compare figures [ fig_comp_kink_sausage](a ) and ( c ) ) a longer disturbance imparts less energy to proper modes and the wave train disperses more quickly , with a stronger decrease of its amplitude . on the other hand ,
a longer transverse disturbance ( compare figures [ fig_comp_kink_sausage](b ) and ( d ) ) imparts less energy to the first overtone ( and so the trailing oscillations are missing in panel ( d ) ) and the fundamental mode receives more energy in the long - wavelength regime , for which @xmath77 has a small spread .
this then results in less dispersion of the wave train .
it is also worth noting that the first fast kink mode overtone leaves its imprint in figures [ fig_comp_kink_sausage](b ) and ( f ) as small amplitude , short period oscillations that arrive at the detection point after the main pulse .
finally , the small amplitude oscillations in panels ( a ) and ( c ) before and after the wave train passage are numerical artifacts . the improper modes are closely related to leaky quasi - modes studied by , e.g. , @xcite ; see also the discussion by @xcite . the interference of improper modes will create a wave front propagating from the tube in the radial direction .
the leaky quasi - modes describe the so - called intermediate asymptotic of this wave front ( see * ? ? ?
* for a discussion of the role of leaky quasi - modes in the case of kink waves ) .
wave radiation after the initial excitation can be well appreciated by plotting snapshots of the dependence of @xmath34 with respect to @xmath156 for several times ( figure [ fig_leakage ] ) .
panel ( a ) shows the imposed radial profile of the radial displacement at @xmath141 , given by equation ( [ init_psi ] ) , with improper modes ( red curve ) having a more important contribution than proper modes ( green curve ) .
the subsequent temporal evolution of @xmath34 shows oscillations inside the magnetic cylinder ( @xmath203 ) and in its vicinity , together with a wave pulse propagating outwards in the radial direction , that is specially visible in panel ( d ) as a depression in @xmath34 in the range @xmath204 .
this propagating feature is clearly caused by improper modes because the amplitude of proper modes at its position is negligible .
the wave emission process has been presented here by doing the cuts of figure [ fig_leakage ] at the center of the initial perturbation ( @xmath187 ) , but as far as the amplitude of improper modes is non zero , it also takes place at other positions along the magnetic tube when they are affected by the propagating disturbance .
+ + + + +
in this paper we have used a method based on fourier integrals to solve the initial value problem of impulsively excited axisymmetric ( i.e. , sausage ) perturbations in a uniform , dense , cylindrical magnetic tube .
the linear and @xmath1 approximations have been assumed .
this method gives analytical expressions for the perturbations ( plasma displacement , density , velocity , and magnetic field ) that require the numerical computation of the two integrals in equations ( [ expand_eigen - main ] ) and ( [ four_trans_inv - main ] ) .
the accuracy of the results depends exclusively on the precision of the numerical computation of these integrals .
an advantage of this method is that one can select which perturbations will be computed and at which spatial position(s ) and time(s ) . the impulsive generation of fast sausage wave trains in a dense magnetic tube has also been studied before by numerically integrating the linearized mhd equations .
most of these works use a cartesian slab to represent the magnetic tube @xcite . in the solar context
, only @xcite have considered this problem in cylindrical geometry , although they did not take into account the dispersive behavior of the propagating wave packet .
one of the key features of the present paper is the importance of the group velocity in the dispersion of a concentrated perturbation .
concerning fast sausage wave trains , this was already acknowledged by @xcite and later by other authors ( e.g. , * ? ? ?
* ; * ? ? ?
in addition to this , we have emphasized the role of the length of the initial perturbation and the internal to external density ratio in the wave packet dispersion ( figures [ fig_longer_t ] and [ fig_comp_kink_sausage ] ) .
we have found that longer initial perturbations have much lower amplitudes because the initial excitation imparts less energy to proper modes .
in addition , larger values of the density ratio lead to a wider wave packet that nevertheless has a larger amplitude because most of the oscillatory energy is concentrated in its trailing part . some of these results have also been found in a non - uniform dense slab by @xcite we are now concerned with the observational implications of our work .
one of the main conclusions of paper i is that a spatially concentrated , transverse disturbance of a magnetic tube gives rise to non - negligible density perturbations that take their maximum value at the tube boundary .
the reason for these density perturbations is that fast kink waves are compressive for longitudinal wavelengths of the order of the tube radius .
the axisymmetric wave trains studied in the present work also generate density perturbations caused by the compressive character of fast sausage waves . to determine whether the wave trains propagating along coronal loops studied by @xcite ( and similar events detected in solar radio bursts , see * ?
* ; * ? ? ?
* ; * ? ? ?
* ) are caused by fast sausage or kink excitations it is necessary to carry out a forward modelling of the magnetohydrodynamic perturbations to determine how they would appear when observed with available instruments .
@xcite assumed that the observed emission intensity in optically thin coronal lines is proportional to the line integral along the line - of - sight ( los ) of the electron number density squared .
they considered fast sausage and kink waves with fixed wavelength and different orientations of the magnetic tube axis with respect to the los .
they concluded that the two wave modes generate similar amplitudes of integrated density variations .
nevertheless , the integrated intensity alone does not govern the line intensity and formation of emission lines .
@xcite took into account that it is actually governed by a complex , non - linear interaction between the density squared , the ionisation balance and emissivity , and the instrument response function .
the important point here is that the ionisation equilibrium timescale of and ions in a coronal environment is of the order of a few to tens of seconds .
the effect of the ionisation balance can be ignored only if the wave period is much longer than this timescale .
otherwise , observed intensity perturbations do not necessarily behave like the theoretical temperature and density .
therefore , to make a theoretical interpretation of the @xmath205 s oscillations described by @xcite a careful forward modelling of impulsively excited fast sausage and kink perturbations is required .
@xcite ( see also * ? ? ?
* ) studied the influence of los angle and spatial , temporal , and spectral resolution of observing instruments on fast sausage wave observables . under the assumption of equilibrium ionisation
, these authors concluded that the intensity variation can be very low ( @xmath206% for 171 ) or significant ( @xmath207% for 193 ) .
most of the intensity variation , however , disappears when considering the radiative relaxation times of the ions .
another important conclusion of this work is that the intensity modulation has strong dependence on the los angle and on the spatial and temporal resolutions .
our future aim is to repeat the calculations of @xcite for impulsive initial excitations of the fast sausage and kink types .
j.t . acknowledges support from the spanish ministerio de educacin y ciencia through a ramn y cajal grant . r.o . and j.t .
also acknowledge the funding provided under the project aya2011 - 22846 by the spanish micinn and feder funds .
is indebted to d. w. fanning for making available the coyote library of idl programs ( http://www.idlcoyote.com/ ) and to w. rosenheinrich for his table of bessel functions integrals ( http://www.fh-jena.de/@xmath208rsh/forschung/stoer/besint.pdf ) .
m.s.r . also acknowledges the support by the stfc grant .
our purpose here is to study the solutions to the dispersion relation @xmath209 in the limit of very large longitudinal wavenumber , i.e. , for @xmath210 .
we make use of the asymptotic formulae for large arguments @xcite @xmath211 figure [ fig_dr_over ] shows that , for @xmath212 , @xmath213 is not close to @xmath97 and so @xmath214 .
then , the second term in the dispersion relation is @xmath215 again using figure [ fig_dr_over ] we see that , for @xmath212 , @xmath213 is close to @xmath98 and so @xmath47 is not large .
since the second term in equation ( [ appa : eq:1 ] ) is small , the first one also has to be small .
this is only possible when @xmath216 is close to one of the zeros of the function @xmath217 , which are @xmath218 , @xmath219 as a result , we have @xmath220 from this expression , the phase speed and group velocity for large longitudinal wavenumber are @xmath221 we conclude that as @xmath222 tends to infinity , @xmath76 tends to @xmath98 from above and @xmath77 tends to @xmath98 from below .
using equation ( [ xir_of_r_im ] ) we write the expression for the scalar product of two improper eigenfunctions as @xmath223 where @xmath224 now we use the formulae @xmath225 , \nonumber \\ & y_1(z ) = \sqrt{\frac2{\pi z}}\sin\left(z - \mbox{$\frac34$}\pi\right ) \left[1 + y_1(z)\right ] , \label{eq:8}\end{aligned}\ ] ] where @xmath226 and @xmath227 decay as @xmath228 when @xmath229 . using these formulae , after some algebra , we rewrite equation ( [ eq:6 ] ) as @xmath230 where @xmath231 @xmath232 @xmath233\right .
\nonumber \\ & \hspace{10ex } \times \cos\left(k_e r - \mbox{$\frac34$}\pi\right ) \nonumber \\ & \hspace{8ex } + \left[c_y c'_j j_1(k'_e r ) + c_y c'_y y_1(k'_e r)\right ] \nonumber \\ & \hspace{10ex } \times \sin\left(k_e r - \mbox{$\frac34$}\pi\right ) \nonumber \\ & \hspace{8ex } + \left[c_j c'_j j_1(k_e r ) + c_y c'_j y_1(k_e r)\right ] \nonumber \\ & \hspace{10ex } \times \cos\left(k'_e r - \mbox{$\frac34$}\pi\right ) \nonumber \\ & \hspace{8ex } + \left[c_j c'_y j_1(k_e r ) + c_y c'_y y_1(k_e r)\right ] \nonumber \\ & \hspace{10ex } \times \left . \sin\left(k'_e r - \mbox{$\frac34$}\pi\right)\right\}\,dr , \label{eq:12 } \end{aligned}\ ] ] @xmath234 dr , \label{eq:13}\end{aligned}\ ] ] @xmath235\right . \nonumber \\ & \hspace{8ex } - ( c_j c'_y + c_y c'_j)\cos\left[r(k_e + k'_e)\right ] \nonumber \\ & \hspace{8ex } - ( c_j c'_y - c_y c'_j)\sin\left[r(k_e - k'_e)\right ] \nonumber \\ & \hspace{8ex } + \left .
( c_j c'_j - c_y c'_y)\sin\left[r(k_e + k'_e)\right]\right\}dr . \label{eq:13b}\end{aligned}\ ] ] as it is explained in paper i , @xmath236 is a regular function of @xmath75 and @xmath237 . repeating the same calculation as in paper i we obtain that @xmath238 where @xmath239 is a regular function and @xmath240 .
\label{eq:15}\ ] ] using the same arguments as in paper i we prove that @xmath241 and @xmath242 .
then we obtain @xmath243
in paper i the evolution of an initial transverse displacement of the magnetic tube was studied .
the radial variation of the initial disturbance outside the magnetic cylinder was taken as @xmath244 , with @xmath245 a prescribed length . in the present paper
this external radial variation is @xmath246 .
we now derive expressions for the amplitudes @xmath247 and @xmath248 for a transverse perturbation whose initial radial distribution is @xmath249 for @xmath250 .
note that for this type of perturbation we still use @xmath251 for @xmath252 . to derive this result the primitive of @xmath259 is required .
we have used equation ( [ derivj1k1 ] ) to express @xmath260 in terms of @xmath261 and @xmath262 .
next , the primitive has been calculated with the help of expressions in @xcite , pp .
143 and 160 , with @xmath263 and @xmath264 replaced by @xmath261 and @xmath262 . with @xmath268 and
@xmath269 the struve functions .
the primitive of @xmath270 is derived in the same way by substituting @xmath91 and @xmath217 by @xmath271 and @xmath272 and @xmath273 by @xmath274 .
this function is defined as | a localized perturbation of a magnetic flux tube produces a pair of wave trains that propagate in opposite directions along the tube .
these wave packets disperse as they propagate , where the extent of dispersion depends on the physical properties of the magnetic structure , on the length of the initial excitation , and on its nature ( e.g. , transverse or axisymmetric ) . in @xcite
we considered a transverse initial perturbation , whereas the temporal evolution of an axisymmetric one is examined here . in both papers
we use a method based on fourier integrals to solve the initial value problem .
previous studies on wave propagation in magnetic wave guides have emphasized that the wave train dispersion is influenced by the particular dependence of the group velocity on the longitudinal wavenumber .
here we also find that long initial perturbations result in low amplitude wave packets and that large values of the magnetic tube to environment density ratio yield longer wave trains . to test the detectability of propagating transverse or axisymmetric wave packets in magnetic tubes of the solar atmosphere ( e.g. , coronal loops , spicules , or prominence threads ) a forward modelling of the perturbations must be carried out .
this is left for a future work . |
low - density parity - check ( ldpc ) codes , invented by gallager @xcite in the @xmath1s , constitute a hot research topic since they are a main competitor to turbo codes @xcite .
recently , a connection between ldpc codes and codes based on graphs was shown ( see , for example , @xcite ) , which opens new perspectives in searching for powerful ldpc codes .
moreover , coding theory methods can be applied in describing and searching for graphs better than previously known .
for example , in @xcite compact representations based on convolutional ldpc codes for famous bipartite graphs such as heawood s , tutte s , and balaban s graphs@xcite are presented .
typically , ldpc codes have a minimum distance which is less than that of the best known linear codes , but due to their structure they are suitable for low - complexity iterative decoding , like for example the believe - propagation algorithm .
an important parameter determining the efficiency of iterative decoding algorithms for ldpc codes is the _ girth _ , which determines the number of independent iterations@xcite and is a parameter of the underlying graph .
the minimum distance seems not to play an important role within iterative decoding algorithms , since the error - correcting capabilities of such a suboptimal procedure are often less than those guaranteed by the minimum distance .
in fact , it was shown in @xcite that the performance of ldpc codes in the high signal - to - noise ( snr ) region is predominantly dictated by the structure of the smallest absorbing sets . however , as the size of these absorbing sets is upper - bounded by the minimum distance , ldpc codes with large minimum distance are of particular interest .
ldpc codes can be characterized as either random / pseudo - random or nonrandom , where nonrandom codes can be subdivided into regular or irregular @xcite , while random / pseudo - random codes are always irregular @xcite .
a @xmath2-regular ( nonrandom ) ldpc code is determined by a parity - check matrix with exactly @xmath3 ones in each column and exactly @xmath4 ones in each row .
the class of quasi - cyclic ( qc ) @xmath5-regular ldpc codes is a subclass of regular ldpc codes with low encoding complexity .
such codes are most suitable for algebraic design and are commonly constructed based on combinatorial approaches using either finite geometries @xcite or steiner triple systems @xcite , having girth @xmath6 . amongst other algebraic constructions leading to qc ldpc codes with larger girth
we would like to mention @xcite , where a class of qc ldpc codes of rate @xmath7 with girth up to @xmath8 based on subgroups of the multiplicative group of the finite field @xmath9 was obtained .
the same method was used for convolutional codes in @xcite .
although qc ldpc codes are not asymptotically optimal they can outperform random or pseudorandom ldpc codes ( from asymptotically optimal ensembles ) for short or moderate block lengths @xcite .
this motivates searching for good short qc ldpc codes .
the problem of finding qc ldpc codes with large girth and large minimum distance for a wide range of code rates was considered in several papers , for example , @xcite .
codes with girth at most @xmath8 are constructed in @xcite , while @xcite gives examples of rather short codes with girth @xmath10 .
codes with girth up to @xmath11 with @xmath12 are presented in @xcite and it is shown that qc ldpc codes with girth @xmath13 and block length between @xmath14 and @xmath15 outperform random codes of the same block length and rate . most of the papers devoted to constructing nonrandom ldpc codes with large girth combine some algebraic techniques and computer search . commonly these procedures start by choosing a proper base matrix ( also called weight or degree matrix ) or the corresponding base graph ( also called seed graph @xcite or protograph@xcite ) .
the references @xcite are focused on all - ones base matrices , while in @xcite base matrices are constructed from steiner triple systems and integer lattices . in both cases ,
a system of inequalities with integer coefficients describing all cycles of a given length is obtained and suitable labels or degrees are derived .
for example , if we replace all nonzero entries in the base matrix by permutation matrices @xcite , circulant matrices @xcite , or sums of circulant matrices @xcite , we obtain the corresponding qc ldpc block codes . on the other hand , if we replace all nonzero entries in the base matrix by either monomials or binomials , we obtain the corresponding ( parent ) ldpc convolutional codes@xcite . notice that both constructing the inequalities and the labeling require significant computational efforts . some methods directed towards reducing the computational complexity of these steps can be found in @xcite .
parameters of the so far shortest qc ldpc block codes with @xmath16 and girth @xmath17 , and @xmath18 found via computer search are presented in @xcite , improving previous results from @xcite . as mentioned earlier ,
it is important that the constructed qc ldpc block codes have large minimum distance for achieving a suitable upper - bound on their error - correcting performance at high snr .
it is proved in @xcite that the minimum distance of qc ldpc codes whose base parity - check matrices are @xmath19 all - one matrices is upper - bounded by @xmath20 .
however , considering base matrices with zeros leads to qc ldpc codes with larger minimum distance .
for example , in @xcite it is shown that replacing all nonzero entries in the base matrix by sums of circulants and all zero entries by all - zero matrices , increases the minimum distance of the resulting code while preserving its regularity .
for ldpc convolutional codes this approach implies that a parity - check matrix contains binomials instead of monomials .
the corresponding upper - bound on the minimum distance of such ldpc codes is presented in @xcite .
a particular case of this upper - bound , valid only for codes with zeros and monomials is derived in @xcite . in section [ sec : 2 ] , we introduce notations for generator and parity - check matrices of convolutional codes and for their corresponding tailbiting block codes .
section [ sec : 3 ] focuses on bipartite graphs , biadjacency matrices , and their relation to parity - check matrices of ldpc block codes .
our constructions of base and voltage matrices , used when searching for ldpc block codes with large girth , are introduced in section [ sec : 4 ] .
bounds on the girth and the minimum distance for qc @xmath5-regular ldpc block codes are discussed in section [ sec : 5 ] .
new search algorithms for qc ldpc block codes constructed from all - one matrices , steiner triple systems , and qc regular matrices are presented in section [ sec : 6 ] . moreover , depending on the desired girth , algorithms of different complexity for constructing the set of inequalities and searching for suitable labelings are described .
a new algorithm for computing the minimum distance of qc @xmath5-regular ldpc codes is described in section [ sec : 7 ] and used to compute the minimum distance of our newly found codes .
moreover , we determined the hitherto unknown minimum distance for some of the shortest known ldpc codes given in @xcite . in section [ sec :
8 ] , we present new examples of @xmath5-regular qc ldpc codes in the form of tailbiting ldpc codes with girth between @xmath18 and @xmath0 .
this representation is compact and it is possible to apply low - complexity encoding , searching , and decoding procedures well developed for convolutional and tailbiting block codes @xcite . in particular , the presented codes with girth @xmath18 and @xmath8 are shorter than the codes presented in @xcite and @xcite , respectively .
moreover , our codes with girth @xmath10 to @xmath11 are shorter than the corresponding codes presented in @xcite . section [ sec : 9 ] concludes the paper with some final remarks .
consider a rate @xmath21 binary convolutional code @xmath22 with the semi - infinite generator matrix rcl [ eq : semi - infinite - generator - matrix ] g & = & ( ccccccc g_0 & g_1 & & & g_m_g + & g_0 & g_1 & & & g_m_g + & & & & & & ) of memory @xmath23 where @xmath24 , @xmath25 , are @xmath26 binary matrices .
its semi - infinite syndrome former rcl h^t & = & ( ccccccc h^t_0 & h^t_1 & & & h^t_m + & h^t_0 & h^t_1 & & & h^t_m + & & & & & & ) of memory @xmath27 , where in general @xmath28 , @xmath29 , @xmath30 , are @xmath31 binary matrices , and @xmath32 denotes transpose . clearly @xmath33 and @xmath34 satisfy rcl [ eq : gh - equal - zero ] g h^t & = & @xmath35 and rcl @xmath36 h^t & = & @xmath35 where rcl @xmath36 & = & @xmath37 g is the code sequence and @xmath38 is the information sequence .
next we tailbite the semi - infinite generator matrix to length @xmath39 @xmath40-tuples , where @xmath41 .
then we obtain the @xmath42 generator matrix of the quasi - cyclic ( qc ) block code @xmath43 as rcl [ eq : tb - generator - matrix ] g _ & = & ( ccccccccc g_0 & g_1 & & & g_m_g + & g_0 & g_1 & & & g_m_g + & & & & & & + & & & g_0 & g_1 & & & g_m_g + g_m_g & & & & g_0 & g_1 & & g_m_g - 1 + g_m_g - 1 & g_m_g & & & & & & + & & & & & & & g_1 + g_1 & g_2 & & g_m_g & & & & g_0 ) every cyclic shift of a codeword of @xmath43 by @xmath40 places modulo @xmath44 is a codeword .
the corresponding tailbiting parity - check matrix is the @xmath45 matrix rcl [ eq : tb - parity - check - matrix ] h _ & = & ( ccccccccc h_0 & & & & h_m & h_m-1 & & h_1 + h_1 & h_0 & & & & h_m & & h_2 + & h_1 & & & & & & + & & & h_0 & & & & h_m + h_m & & & h_1 & h_0 & & & + & h_m & & & h_1 & & & + & & & & & & & + & & & h_m & h_m-1 & & h_1 & h_0 ) it is easily shown that @xmath46 and @xmath47 satisfy rcl g _ h_^t & = & @xmath35 given that is fulfilled .
the parity - check matrix for the convolutional code @xmath22 can also be written as the @xmath31 polynomial matrix rcl h(d ) & = & h_0 + h_1 d + h_2 d^2 + + h_m d^m or , equivalently , as rcl [ eq : convolutional - parity - check - matrix ] h(d ) & = & ( cccc h_11(d ) & h_12(d ) & & h_1c(d ) + h_21(d ) & h_22(d ) & & h_2c(d ) + & & + h_(c - b)1(d ) & h_(c - b)2(d ) & & h_(c - b)c(d ) ) in the sequel we mostly consider parity - check matrices with only monomial entries @xmath48 of degree @xmath49 , where @xmath49 are nonnegative integers .
clearly , such a parity - check matrix @xmath50 can be represented by its _ degree matrix _ @xmath51 , @xmath52 and @xmath53 .
note that starting from section [ sec : 8 ] we will relax the restriction to only monomial entries and also include zero entries .
[ ex : 1 ] consider the rate @xmath54 convolutional code @xmath22 with parity - check matrix c [ eq : example_conv_parity_check ] h(d ) = ( cccc 1 & 1 & 1 & 1 + 1 & 1 & d & d + 1 & d & 1 & d ) whose degree matrix is c w = ( cccc 0 & 0 & 0 & 0 + 0 & 0 & 1 & 1 + 0 & 1 & 0 & 1 ) tailbiting to length @xmath55 , we obtain the tailbitten @xmath56 parity - check matrix of a qc ldpc block code rcl [ eq : example_tb_parity_check ] h_tb & = & in particular , every cyclic shift of a codeword by @xmath57 places modulo @xmath58 is a codeword . due to the restriction to monomial entries in @xmath50
, @xmath47 is @xmath2__-regular _ _ , that is , it has exactly @xmath3 and @xmath4 ones in each column and row , respectively . moreover , to fulfill the _ low _ density criterion , @xmath39 has to be much larger than @xmath3 and @xmath4 , and thus the matrix @xmath47 is sparse .
note that the first @xmath40 columns of @xmath47 are repeated throughout the whole matrix in a cyclicly shifted manner . by reordering the columns as @xmath59 , @xmath60 , _ etc . _ and the rows as @xmath61 , @xmath62 , _ etc .
_ we obtain a parity - check matrix of an equivalent @xmath5-regular ldpc block code constructed from circulant matrices rcl [ eq : circulant - ldpc_parity - check - matrix ] h _ & = & ( llll i_w_11 & i_w_12 & & i_w_1c + i_w_21 & i_w_22 & & i_w_2c + & & & + i_w_(c - b)1 & i_w_(c - b)2 & & i_w_(c - b)c + ) where @xmath49 are the entries of the degree matrix @xmath63 and @xmath64 denotes an @xmath65 _ circulant matrix _ , that is , an identity matrix with its rows shifted cyclically to the left by @xmath49 positions . note , that the @xmath5-regular ldpc block code determined by @xmath66 is not quasi - cyclic , although equivalent to the qc block code determined by @xmath47 . 1 we return to in example [ ex : 1 ] and
reorder the columns as @xmath67 and the rows as @xmath68 .
then we obtain the equivalent rate @xmath69 @xmath70-regular ldpc block code with parity - check matrix rl [ eq : example_tb_qc_parity_check ] c + h _
a graph @xmath71 is determined by a set of _ vertices _ @xmath72 and a set of _ edges _ @xmath73 , where each edge connects exactly two vertices .
the _ degree of a vertex _ denotes the number of edges that are connected to it . if all vertices have the same degree @xmath40 , the _ degree of the graph _ is @xmath40 , or , in other words , the graph is @xmath40-_regular_. consider the set of vertices @xmath74 of a graph partitioned into @xmath75 disjoint subsets @xmath76 , @xmath77 .
such a graph is said to be @xmath75-_partite _ , if no edge connects two vertices from the same set @xmath76 , @xmath77 .
a _ path _ of length @xmath78 in a graph is an alternating sequence of @xmath79 vertices @xmath80 , @xmath81 , and @xmath78 edges @xmath82 , @xmath83 , with @xmath84 .
if the first and the final vertex coincide , that is , if @xmath85 , then we obtain a _ cycle_. a cycle
is called _ simple _ if all its vertices and edges are distinct , except for the first and final vertex which coincide .
the length of the shortest simple cycle is denoted the _ girth _ of the graph . in @xcite
it was proven that the girth of a graph coincide with the minimum distance of the block code , whose parity - check matrix corresponds to the _ incidence matrix _ of the graph . moreover
, the girth determines the number of independent iterations in belief - propagation decoding @xcite .
every parity - check matrix @xmath34 of a rate @xmath86 ldpc block code can be interpreted as the _ _ biadjacency matrix__@xcite of a bipartite graph , the so - called _ tanner graph _ @xcite , having two disjoint subsets @xmath87 and @xmath88 containing @xmath89 and @xmath90 vertices , respectively .
the @xmath89 vertices in @xmath87 are called _ symbol nodes _ , while the @xmath90 vertices in @xmath88 are called _ constraint nodes_.
if the underlying ldpc block code is @xmath5-regular , the symbol and constraint nodes have degree @xmath3 and @xmath4 , respectively .
consider the tanner graph with the biadjacency matrix @xmath47 , corresponding to a qc @xmath5-regular ldpc code , obtained from the parity - check matrix of a tailbiting ldpc block code . clearly , by letting the tailbiting length @xmath39 tend to infinity
, we obtain a convolutional parity - check matrix @xmath50 of the parent convolutional code @xmath22 . in terms of tanner graph representations
, this procedure corresponds to unwrapping the underlying graph and extending it in the time domain towards infinity .
hereinafter , we will call the girth of this infinite tanner graph the _ free girth _ and denote it @xmath91 . in 1,2,3, ... ,8 coordinate ( s ) at ( 1.1 * , 3 ) ; in 1,2,3, ... ,6 coordinate ( c ) at ( 1.3*+0.6 , 0 ) ; ( c1 ) ( s1 ) ; ( c1 ) ( s2 ) ; ( c1 )
( s3 ) ; ( c1 ) ( s4 ) ; ( c2 ) ( s1 ) ; ( c2 ) ( s2 ) ; ( c2 ) ( s7 ) ; ( c2 ) ( s8 ) ; ( c3 ) ( s1 ) ; ( c3 ) ( s3 ) ; ( c3 ) ( s6 ) ; ( c3 ) ( s8 ) ; ( c4 ) ( s5 ) ; ( c4 ) ( s6 ) ; ( c4 ) ( s7 ) ; ( c4 ) ( s8 ) ; ( c5 ) ( s3 ) ; ( c5 ) ( s4 ) ; ( c5 ) ( s5 ) ; ( c5 ) ( s6 ) ; ( c6 ) ( s2 ) ; ( c6 ) ( s4 ) ; ( c6 ) ( s5 ) ; ( c6 )
( s7 ) ; in 1,2,3, ... ,8 ( s ) circle ( 3pt ) ; at ( @xmath92 ) ( 0,0)[b]@xmath93 ; in 1,2,3, ... ,6 ( c ) circle ( 3pt ) ; at ( @xmath94 ) ( 0,0)[t]@xmath95 ; 1 interpreting as a biadjacency matrix , we obtain the corresponding tanner graph @xmath71 as illustrated in fig . [ fig : example_graph ] with @xmath96 symbol nodes and @xmath97 constraint nodes , having girth @xmath98 .
a binary matrix @xmath99 is called _ base matrix _ for a tailbiting ldpc block code if its parent convolutional code with parity - check matrix @xmath50 has only monomial or zero entries and satisfies c b = h(d)|_d=1 that is , all nonzero entries in @xmath50 are replaced by @xmath100 .
different tailbiting ldpc block codes can have the same base matrix @xmath99 . the _ base graph _
@xmath101 follows as the bipartite graph , whose biadjacency matrix is given by the base matrix @xmath99 .
denote the girth of such a base graph @xmath102 .
the terminology `` base graph '' originates from graph theory and is used , for example , in @xcite .
it differs from the terminology used in @xcite , where _ protograph _ or _ seed graph _ are used . consider the additive group @xmath103 , where @xmath104 . from the base graph @xmath105
we obtain the _ voltage graph _
@xcite @xmath106 by assigning a voltage value @xmath107 to the edge @xmath108 connecting the vertices @xmath109 and @xmath110 , satisfying the property @xmath111 .
although the graph is not directed , the voltage of the edge depends on the direction in which the edge is passed .
finally , define the _ voltage of the path _ to be the sum of the voltages of its edges .
let @xmath112 be a _ lifted graph _ obtained from a voltage graph @xmath113 , where @xmath114 and @xmath115 .
two vertices @xmath116 and @xmath117 are connected in the lifted graph by an edge if and only if @xmath109 and @xmath110 are connected in the voltage graph @xmath113 with the voltage value of the corresponding edge given by @xmath118 .
it is easy to see that cycles in the lifted graph correspond to cycles in the voltage graph with zero voltage .
consequently , the girth @xmath119 of a voltage graph follows as the length of its shortest cycle with voltage zero , which is equal to the free girth @xmath91 @xcite .
a voltage assignment corresponds directly to selecting the degrees of the parity - check monomials in @xmath50 . in the following we start from a base graph @xmath101 and
use a voltage assignment based on the monomial degrees @xmath49 of the degree matrix @xmath63 to determine the corresponding voltage graph @xmath113 .
the edge voltage from the constraint node @xmath120 to the symbol node @xmath121 is denoted by @xmath122 while the opposite direction from symbol node @xmath121 to constraint node @xmath120 follows as @xmath123 , @xmath124 and @xmath53 , where c [ eq : voltage_labeling ] \ { rcl _
i j & = & w_ij + .
when searching for ldpc convolutional codes with given free girth @xmath91 , we use integer edge voltages , that is , we deal with an infinite additive group .
however , when searching for qc ldpc block codes with given girth @xmath125 , obtained by tailbiting a parent convolutional code to length @xmath39 , we use a group of modulo @xmath39 residues , that is , is replaced by c [ eq : voltage_labeling_mod_m ] \ { rcll _
i j & = & w_ij & m + .
the definitions of path and cycle in a voltage graph coincide with those in a regular graph , except for the additional restriction that two neighboring edges may not connect the same nodes in reversed order .
the _ voltage _ of a path or cycle within a voltage graph , follows as the sum of all edge voltages involved . in 1,2,3,4
coordinate ( s ) at ( 2.5 * , 2.7 ) ; in 1,2,3 coordinate ( c ) at ( 2.5*+1.25 , 0 ) ; ( c1 ) ( s1 ) [ xshift=-5pt ] node[pos=0.3]@xmath126 ; ( c1 ) ( s2 ) [ xshift=-5pt ] node[pos=0.9]@xmath126 ; ( c1 ) ( s3 ) [ xshift=-8pt ] node[pos=0.9]@xmath126 ; ( c1 ) ( @xmath127 ) [ xshift=-10pt ] node[pos=0.95]@xmath126 ; ( c2 ) ( s1 ) [ xshift=-6pt ] node[pos=0.6]@xmath126 ; ( c2 ) ( s2 ) [ xshift=-4pt ] node[pos=0.7]@xmath126 ; ( c2 ) ( s3 ) [ xshift=+4pt ] node[pos=0.7]@xmath128 ; ( c2 ) ( s4 ) [ xshift=+6pt ] node[pos=0.6]@xmath128 ; ( c3 )
( @xmath129 ) [ xshift=+10pt ] node[pos=0.95]@xmath126 ; ( c3 ) ( s2 ) [ xshift=+8pt ] node[pos=0.9]@xmath128 ; ( c3 ) ( s3 ) [ xshift=+5pt ] node[pos=0.9]@xmath126 ; ( c3 ) ( s4 ) [ xshift=+5pt ] node[pos=0.3]@xmath128 ; in 1,2,3,4 ( s ) circle ( 5pt ) ; at ( @xmath92 ) ( 0,0)[b]@xmath93 ; in 1,2,3 ( c ) circle ( 5pt ) ; at ( @xmath94 ) ( 0,0)[t]@xmath95 ; 1 the bipartite graph whose biadjacency matrix is given by the base matrix @xmath99 of the rate @xmath54 @xmath130-regular ldpc convolutional code @xmath22 is illustrated in fig .
[ fig : voltage_graph ] . as the edges are labeled
according to , fig .
[ fig : voltage_graph ] corresponds to a voltage graph with girth @xmath131 ( for example , @xmath132 @xmath133 @xmath134 @xmath133 @xmath135 @xmath133 @xmath136 @xmath133 @xmath132 )
. the edge from , for example , constraint node @xmath136 to symbol node @xmath137 is labeled according to rclclcl _ 23 & = & -|_32 & = & w_23 & = & 1 as the free girth of the infinite tanner graph , corresponding to the parent convolutional code @xmath22 , determined by the convolutional parity - check matrix @xmath50 is equal to the girth of the voltage graph , we can conclude that @xmath138 .
if we neglect all edge labels , we would obtain the corresponding base graph .
there are a number of approaches which can be applied to construct and search for qc @xmath140-regular ldpc block and convolutional codes @xcite or the bipartite graphs constructed by their incidence matrices . since every ldpc convolutional code can be represented by a bipartite tanner graph using the biadjacency matrix , these techniques can be applied to @xmath141 qc ldpc codes .
moreover , bounds on the girth and the minimum distance of @xmath140 qc ldpc codes @xcite can be generalized to an arbitrary @xmath3 .
[ th : one ] the minimum distance @xmath142 and the girth @xmath125 of an @xmath143 qc ldpc block code @xmath43 obtained from a rate @xmath21 convolutional code @xmath22 with free distance @xmath144 and girth @xmath145 by tailbiting to length @xmath39 are upper - bounded by the inequalities rcl d_min & & d_free + g & & g_free the first statement follows directly from the fact that any codeword @xmath146 of the tailbiting block code @xmath43 , obtained from the parity - check matrix @xmath50 of the parent convolutional code @xmath22 , satisfies c [ eq : proof_theorem_1 ] @xmath36(d ) h^t(d ) = @xmath35 ( d^m-1 ) since the parent convolutional code @xmath22 satisfies without reduction modulo @xmath147 and reducing modulo @xmath147 does not increase the weight of a polynomial , the first statement follows directly . for the second statement
we consider the voltage graph @xmath113 representation of the parent convolutional code @xmath22 with girth @xmath148 together with the tanner graph representation of the qc ldpc block code @xmath43 with girth @xmath125 .
similar to the relations between the free distance @xmath144 and the minimum distance @xmath142 , there exists a relation between each cycle within the voltage graph @xmath113 of the parent convolutional code and the tanner graph @xmath71 of the corresponding block code obtained by tailbiting to length @xmath39 .
the edge voltages for every cycle in @xmath113 have to sum up to zero .
similarly , every cycle in @xmath71 corresponds to a cycle in @xmath113 such that its edge voltages have to sum up to zero modulo @xmath39 . with the same argument as before it follows directly that rclcl g & & g _ & = & g _ in @xcite a lower bound on the girth of a voltage graph @xmath119 was found via the girth of the corresponding base graph @xmath102 for ordinary graphs .
it is straightforward to generalize this bound : consider a base graph of a qc @xmath139-regular ldpc convolutional code with girth @xmath102 and let @xmath149 denote the @xmath150th generalized minimum hamming distance of the linear @xmath151 block code determined by the parity - check matrix which corresponds to the incidence matrix of the tanner graph . in order words
, @xmath149 corresponds to the number of nontrivial ( not identically zero ) positions of an @xmath150-dimensional linear subcode .
[ th : two ] there exist a tailbiting length @xmath39 and a voltage assignment , such that the girth @xmath125 of the tanner graph for the corresponding tb block code of length @xmath152 satisfies the inequality rclcl [ eq : girth_lower_bound ] g & & 2 \ { g _ + g_/2 , d_2 } & & 3 g _ where @xmath153 is the second generalized minimum hamming distance , that is , the minimum support of a subcode of dimension two .
we have equality in , if the underlying base graph consists of two connected cycles , having at least one common vertex .
according to theorem [ th : one ] , any cycle in the tanner graph of a qc ldpc block code corresponds to a cycle of the same length in the voltage graph .
as the labels of the voltage tanner graph can be freely chosen , it is enough to prove that there is no zero cycle shorter than @xmath154 , that is , no such cycle whose voltage is zero regardless of the labeling of the base graph . in particular ,
such a cycle is also known as an _ inevitable cycle _ @xcite or _ balanced cycle _ @xcite .
the number of times each edge in such a cycle of the voltage graph is passed in different directions has to be even .
this cycle can not be simple , since in a simple cycle each edge is passed in one direction only .
hence , the cycle passes through the vertices of a subgraph which contains at least two different cycles , corresponding to two different nonzero codewords .
the minimum distance of the base graph is @xmath102 , and , according to the griesmer bound , the smallest length of a linear code with two nonzero codewords of minimum distance @xmath155 is @xmath156 , and , hence , the first lower bound of inequality follows .
consider the second lower bound .
the definition of the second generalized minimum hamming distance implies that the smallest subgraph with two cycles has to have at least @xmath153 edges .
thus , the second of the two lower bounds gives the precise value of the girth of a subgraph containing two connected cycles , having at least one common vertex .
otherwise , @xmath153 is a lower bound . the bounds are tighter than the @xmath157 bound @xcite but not tight if the shortest non - simple cycle consists of two simple cycles connected by a path
. finally , we want to recall an upper bound on the achievable girth and minimum distance .
we start by reformulating the theorem on the achievable girth by fossorier @xcite and thereby generalize it to include base matrices with zero elements .
[ th : three - a ] consider the parity - check matrix @xmath50 of a rate @xmath21 convolutional code with base matrix @xmath99 .
denote the corresponding base graph @xmath101 and let @xmath158 be the @xmath159 submatrix @xmath160 if the base matrix @xmath99 , after possibly reordering its rows and columns , contains the submatrix @xmath158 , then the girth @xmath119 of the corresponding voltage graph @xmath113 is upper - bounded by @xmath161 regardless of the voltage assignment .
the subgraph determined by the @xmath159 submatrix @xmath158 contains @xmath162 symbol nodes , @xmath163 constraint nodes , and @xmath97 edges .
moreover , there exist @xmath162 shortest cycles of length @xmath164 .
thus , the base graph @xmath101 has girth @xmath165 and its second generalized hamming distance is @xmath166 . applying theorem [ th : two ] , we obtain the precise value of the achievable girth as @xmath167 , which completes the proof . for parity - check matrices with only nonzero monomial entries ,
the inequality was proven in @xcite .
moreover , let @xmath50 be the parity - check matrix of a rate @xmath21 @xmath5-regular ldpc convolutional code with free distance @xmath144 . by tailbiting to length @xmath39
we obtain a qc ldpc block code of block length @xmath44 and minimum distance @xmath142 .
as proven in @xcite for parity - check matrices without zero elements and reformulated in @xcite for parity - check matrices with zero elements , the corresponding minimum distance @xmath142 can be upper - bounded by @xmath168 for parity - check matrices with only nonzero monomials , the inequality simplifies to @xmath20 .
when searching for qc ldpc block codes with large girth , we start from a base graph of a rate @xmath21 @xmath5-regular ldpc convolutional code . using the following algorithm
, we determine a suitable voltage assignment based on the group of nonnegative integers , such that the girth of this voltage graph is greater than or equal to a given girth @xmath125 .
afterwards we replace all edge labels by their corresponding modulo @xmath39 residuals , where we try to minimize @xmath39 while preserving the girth @xmath125 . using the concept of biadjacency matrices leads to the corresponding degree matrix @xmath63 and we obtain the parity - check matrix of a convolutional code whose bipartite graph has girth @xmath169 .
tailbiting to lengths @xmath39 , yields the rate @xmath170 qc ldpc block code whose parity - check matrix is equal to the biadjacency matrix of a bipartite graph with girth @xmath125 . the algorithm for determining a suitable voltage assignment for a base graph consists of the following two main steps : 1 .
construct a list containing all inequalities describing cycles of length smaller than @xmath125 within the base graph .
search for such a voltage assignment of the base graph that all inequalities are satisfied .
the efficiency of the second step , searching for a suitable voltage assignment , depends on the chosen representation for the list of inequalities determined during the first step . in general , when searching for all cycles of length @xmath125 roughly @xmath171 different paths have been considered .
however , by using a similar idea as in @xcite when searching for a path within a trellis , we create a tree of maximum depth @xmath172 and search only for identical nodes within the tree and thereby reduce the complexity to roughly @xmath173 . utilizing the base graph of a rate @xmath21 @xmath5-regular ldpc convolutional code , with @xmath40 symbol and @xmath174 constraint nodes , we construct a separate subtree starting with each of the @xmath40 symbol nodes . before describing the algorithm
, we have to introduce some notations .
a node in the tree will be denoted by @xmath175 and has a unique parent node @xmath176 .
the underlying base graph is bipartite , that is , every node @xmath175 in the tree with @xmath177 is only connected to nodes @xmath178 with @xmath179 , @xmath180 , where @xmath181 and @xmath182 are the sets of symbol and constraint nodes , respectively . in other words ,
a symbol node is only connected to constraint nodes and vice versa .
moreover , every node @xmath175 is characterized by its depth @xmath183 and its number @xmath184 , where @xmath185 follows directly from @xmath186 or @xmath187 depending on whether its depth @xmath183 is even or odd . in particular ,
@xmath188 .
having introduced those basic notations , we can grow @xmath40 separate subtrees , with the root node @xmath189 of the @xmath190th subtree being initialized with @xmath191 and depth @xmath192 .
extend every node @xmath177 at depth @xmath193 with @xmath194 by connecting it with the nodes @xmath195 at depth @xmath196 according to the underlying base graph , except @xmath197 which is already connected to @xmath175 at depth @xmath198 .
finally we label the edges according to and obtain the voltage for node @xmath175 in the @xmath190th subtree as the sum of the edge voltages of the path @xmath199 . in 1,2,3,4
coordinate ( s ) at ( 2.5 * , 2.7 ) ; in 1,2,3 coordinate ( c ) at ( 2.5*+1.25 , 0 ) ; ( c1 ) ( s1 ) [ xshift=-10pt ] node[pos=0.3]@xmath200 ; ( c1 ) ( s2 ) [ xshift=-10pt ] node[pos=0.9]@xmath201 ; ( c1 )
( s3 ) [ xshift=-13pt ] node[pos=0.9]@xmath202 ; ( c1 ) ( @xmath127 ) [ xshift=-15pt ] node[pos=0.95]@xmath203 ; ( c2 ) ( s1 ) [ xshift=-10pt ] node[pos=0.6]@xmath204 ; ( c2 ) ( s2 ) [ xshift=-9pt ] node[pos=0.7]@xmath205 ; ( c2 ) ( s3 ) [ xshift=+11pt ] node[pos=0.7]@xmath206 ; ( c2 ) ( s4 ) [ xshift=+13pt ] node[pos=0.6]@xmath207 ; ( c3 ) ( @xmath129 ) [ xshift=+14pt ] node[pos=0.95]@xmath208 ; ( c3 ) ( s2 ) [ xshift=+12pt ] node[pos=0.9]@xmath209 ; ( c3 )
( s3 ) [ xshift=+11pt ] node[pos=0.9]@xmath210 ; ( c3 ) ( s4 ) [ xshift=+11pt ] node[pos=0.3]@xmath211 ; in 1,2,3,4 ( s ) circle ( 5pt ) ; at ( @xmath92 ) ( 0,0)[b]@xmath93 ; in 1,2,3 ( c ) circle ( 5pt ) ; at ( @xmath94 ) ( 0,0)[t]@xmath95 ;
coordinate ( start ) at ( 0 , 5.6 ) ; ( start ) circle ( 5pt ) ; at ( @xmath212 ) @xmath132 ; in 1,2,3 coordinate ( c ) at ( 2.8 , 2.8 * ) ; in 2,3,4 coordinate ( s ) at ( 5 , 2.8 * -2.4 + 0.8 * ) ; ( start ) ( c1 )
[ xshift=+14pt ] node[pos=0.4]@xmath213 ; ( start ) ( c2 ) [ yshift=+5pt , xshift=+14pt ] node[pos=0.4]@xmath214 ; ( start ) ( c3 ) [ xshift=+14pt ] node[pos=0.4]@xmath215 ; ( c1 ) ( s12 ) [ yshift=-5pt ] node[pos=0.4]@xmath201 ; ( c1 ) ( s13 ) [ yshift=+5pt ] node[pos=0.65]@xmath202 ; ( c1 ) ( s14 ) [ yshift=+9pt ] node[pos=0.4]@xmath203 ; ( c2 ) ( s22 ) [ yshift=-5pt ] node[pos=0.4]@xmath205 ; ( c2 ) ( s23 ) [ yshift=+5pt ] node[pos=0.65]@xmath206 ; ( c2 ) ( s24 ) [ yshift=+9pt ] node[pos=0.4]@xmath207 ; ( c3 )
( s32 ) [ yshift=-5pt ] node[pos=0.4]@xmath209 ; ( c3 ) ( s33 ) [ yshift=+5pt ] node[pos=0.65]@xmath210 ; ( c3 ) ( s34 ) [ yshift=+9pt ] node[pos=0.4]@xmath211 ; in 1,2,3 ( c ) circle ( 5pt ) ; at ( @xmath216 ) @xmath95 ; in 2,3,4 ( s ) circle ( 5pt ) ; at ( @xmath217 ) @xmath218 ; clearly , all subtrees together contain all paths of a given length in the voltage graph .
moreover , taking into account that the girth @xmath125 of bipartite graphs is always even , we can conclude that in order to check all possible cycles of length at most @xmath219 in the voltage graph , it is sufficient to grow the corresponding @xmath40 subtrees up to depth @xmath220 and to construct voltage inequalities for all node pairs @xmath221 in the same subtree @xmath190 with the same number @xmath222 and depth @xmath223 but different parent nodes @xmath224 .
the corresponding voltage inequality for the node pair @xmath221 follows directly as the difference between the voltages for the paths from @xmath189 to @xmath175 and @xmath225 , respectively , that is , @xmath226 .
if there exists a cycle @xmath227 , then at depth @xmath228 there exists at least one pair of nodes @xmath221 , whose corresponding voltage inequality is not satisfied , that is , is equal to zero . if there is no cycle shorter than @xmath125 in the voltage graph , then there are no such pairs , and all voltage inequalities are satisfied .
[ ex : 2 ] consider the rate @xmath54 @xmath130-regular ldpc convolutional code given by . the corresponding base graph , with four symbol nodes @xmath229 , @xmath230 , and three constraint nodes @xmath231 , @xmath232 , is illustrated in fig . [
fig : voltage_graph_variables ] . in the following ,
we shall search for a set of edge labels , that is , monomial degrees in @xmath63 , such that the corresponding voltage graph has at least a given girth @xmath125 .
thus , we label the branches by the general edge voltages according to and obtain a bipartite voltage graph . in order to find a suitable labeling for the edge voltages from the @xmath190th constraint node @xmath120 to the @xmath233th symbol node @xmath234 , that is @xmath122 , @xmath232 , @xmath235 , whose underlying voltage graph has at least girth @xmath236
, we have to grow @xmath164 subtrees up to length @xmath237 , with their root nodes being initialized by @xmath234 , @xmath230 .
the subtree with root node @xmath132 is illustrated in fig .
[ fig : tree ] .
clearly , at depth @xmath238 there are no identical nodes , while at depth @xmath239 there are @xmath240 pairs of identical nodes ( @xmath222 ) , but with different parents .
taking into account that a similar subtree is constructed using the remaining three symbol nodes @xmath135 , @xmath137 and @xmath241 as root nodes , we obtain in total @xmath242 node pairs , which all correspond to a voltage inequality .
for example , the voltage inequality obtained by checking all node pairs @xmath221 with @xmath243 , that is , @xmath244 , at depth @xmath239 in the subtree starting with symbol node @xmath132 , are rcl -_11 + _ 12 - _ 22 + _ 21 & & 0 + -_11 + _ 12 - _ 32 + _ 31 & & 0 + -_21 + _ 22 - _
32 + _ 31 & & 0 note that amongst all @xmath245 voltage inequalities , there are only @xmath11 unique ones .
constructing a tree representation 1 .
grow @xmath40 separate subtrees according to the underlying base graph up to depth @xmath246 , with the root node @xmath189 of the @xmath190th subtree being initialized with @xmath191 and depth @xmath192 .
2 . extend every node @xmath177 at depth @xmath193 with @xmath194 by connecting it with the nodes @xmath195 at depth @xmath196 according to the underlying base graph , except @xmath197 which is already connected to @xmath175 at depth @xmath198 .
denote the set of all nodes within the @xmath190th subtree by @xmath247 . using the obtained subtrees
@xmath247 , @xmath248 , with maximum depth @xmath249 , we have found all cycles of length smaller than or equal to @xmath219 as well as their corresponding voltage inequalities .
the same cycle might be found several times within the @xmath40 subtrees .
moreover , two different cycles can correspond to the same voltage inequality .
we continue by creating a reduced list @xmath250 of node pairs @xmath221 of all @xmath40 subtrees @xmath247 , @xmath248 , containing all unique voltage inequalities .
thereby , we remove all duplicate cycles , as well as different cycles corresponding to the same voltage inequality . using the reduced list @xmath250 we can reduce the obtained @xmath40 subtrees @xmath247 , @xmath248 , in a similar way by removing all nodes , not participating in any of the cycles in @xmath250 , and denote the reduced subtree by @xmath251 .
in other words , we remove all nodes in @xmath247 , @xmath248 , which only participate in already known cycles or new cycles with already known voltage inequalities . in the following we present two different approaches for finding suitable edge labels ( edge voltages ) , which we shall denote as algorithm a and algorithm b. in algorithm a , we label the edges of the reduced subtrees @xmath251 , @xmath248 , with a set of randomly chosen voltages . for every node pair @xmath221 in the list @xmath250
, we calculate the voltage of the corresponding cycle as the difference of the path voltages @xmath199 and @xmath252 .
if none of these cycle voltages is equal to zero , the girth of the underlying base graph with such a voltage assignment is greater than or equal to @xmath125 . in algorithm b , we discard the list @xmath250 and only focus on the @xmath40 reduced subtrees @xmath251 .
after labeling their edges with a set of randomly chosen voltages , we sort the nodes @xmath175 of each subtree according to their path voltage @xmath199 .
if there exists no pair of nodes @xmath221 with the same path voltage , number @xmath253 , and depth @xmath254 , but different parent nodes @xmath255 , the girth of the underlying base graph with such a voltage assignment is greater than or equal to @xmath125 . a formal description of those two algorithms is given below : constructing a system of voltage inequalities and searching for an optimum voltage assignment using a list 1 .
create a reduced list @xmath250 of node pairs @xmath221 for all @xmath40 subtrees @xmath247 , @xmath248 , containing all node pairs @xmath221 with a unique voltage inequality , having the same number @xmath222 , depth @xmath223 , but different parent nodes @xmath224 .
2 . reduce each of the @xmath40 subtrees @xmath247 by removing all nodes , which do not participate in any of the found cycles corresponding to the voltage inequalities in @xmath250 , and denote the reduced subtree structure by @xmath251 .
3 . assign randomly chosen voltages to the edges of all trees and perform the following steps : 1 .
find the voltages for all paths leading from the root node @xmath189 of the @xmath190th reduced subtree @xmath251 to all nodes @xmath256 , @xmath248 .
2 . determine the voltage inequality for all cycles @xmath257 , as the difference of the corresponding path voltages in @xmath251 , @xmath248 , computed previously .
if all voltage inequalities are satisfied , the girth of the underlying base graph with such a voltage assignment is greater than or equal to @xmath125 . constructing a system of voltage inequalities and searching for an optimum voltage assignment using a tree 1 . construct the reduced list @xmath250 and the reduced subtrees @xmath251 , @xmath248 , as in algorithm a without storing the corresponding list @xmath250 .
2 . assign randomly chosen voltages to the edges of all trees and perform the following steps : 1 .
find the voltages for all paths from the root node @xmath189 to all nodes within @xmath251 , @xmath248 , and sort all elements within @xmath251 according to their voltages .
search for a pair of nodes @xmath221 in the sorted list with the same path voltage , number @xmath253 , and depth @xmath254 , but different parent nodes @xmath255 .
if no such pair exists , then the girth of the corresponding voltage graph with such a voltage assignment is greater than or equal to @xmath125 . denote the sum of all nodes in the reduced tree @xmath251 , @xmath248 , and the number of unique inequalities in the list @xmath250 by @xmath258 and @xmath259 , respectively , that is , rcltrcl n _ & = & _ i=1^c | _ i,| & and & n _ & = & | | where @xmath260 denotes the number of entries in the set @xmath261 . algorithm a requires @xmath258 summations for computing the path voltages and @xmath259 comparisons for finding cycles , leading to the complexity estimate @xmath262 .
algorithm b requires the same number of @xmath258 summations for computing the path voltages , roughly @xmath263 operations for sorting the set , and @xmath258 comparisons , leading to a total complexity estimate of @xmath263 .
c!c|c!c|c!c|c & & & + & @xmath258 & @xmath259 & @xmath258 & @xmath259 & @xmath258 & @xmath259 + @xmath164 & @xmath264 & @xmath265 & @xmath266 & @xmath267 & @xmath268 & @xmath269 + @xmath270 & @xmath271 & @xmath272 & @xmath273 & @xmath274 & @xmath275 & @xmath276 + @xmath97 & @xmath277 & @xmath278 & @xmath279 & @xmath280 & @xmath281 & @xmath282 + @xmath283 & @xmath284 & @xmath285 & @xmath286 & @xmath287 & @xmath288 & @xmath289 + @xmath96 & @xmath290 & @xmath291 & @xmath292 & @xmath293 & @xmath294 & @xmath295 + @xmath296 & @xmath297 & @xmath298 & @xmath299 & @xmath300 & @xmath301 & @xmath302 + @xmath18 & @xmath303 & @xmath304 & @xmath305 & @xmath306 & @xmath307 & @xmath308 + @xmath309 & @xmath310 & @xmath311 & @xmath312 & @xmath313 & @xmath314 & @xmath315 + @xmath8 & @xmath316 & @xmath317 & @xmath318 & @xmath319 & @xmath320 & @xmath321 + in table [ tab : complexity ] the values of @xmath258 and @xmath259 are given when searching for suitable voltage assignment for a @xmath5-regular rate @xmath322 qc ldpc convolutional codes with @xmath16 and arbitrary @xmath323 and girth @xmath125 constructed from all - ones base matrices .
in this case , up to @xmath324 , algorithm a is preferable , while when searching for voltage assignment with girth @xmath325 , algorithm b should be used . in the general case we have to consider all node pairs , and as @xmath259 is roughly @xmath326
we conclude that algorithm b performs asymptotically better ( when @xmath327 ) .
usually the girth of the tanner graph of an qc ldpc block code is considered to be the most important parameter that affects the performance of belief - propagation decoding , as it determines the number of independent iterations @xcite .
therefore , most research is focused on finding qc ldpc block codes with large girth , while their corresponding minimum distance is mostly unknown . in @xcite
it was shown , that the performance of belief - propagation decoding algorithms at high snrs depends on the structure and the size of the smallest absorbing sets , which however can be upper - bounded by the minimum distance .
this is the rationale for computing the minimum distance of the shortest known qc ldpc block codes as well as our search for qc ldpc codes with both large girth and large minimum distance .
our method of calculating the minimum distance is based on the well - known fact that the minimum distance of a linear block code @xmath43 with parity - check matrix @xmath34 is equal to the minimum number of columns of @xmath34 which sum up to zero .
c!c!c!c!l @xmath4 & @xmath328 & @xmath142 & @xmath39 & + + & & & & @xmath329 + & & & & @xmath330 + & & & & @xmath331 + & & & & @xmath332 + & & & & @xmath333 + & & & & @xmath334 + & & & & @xmath335 + & & & & @xmath336 + & & & & @xmath337 + & & & & @xmath338 + & & & & @xmath339 + & & & & @xmath340 + & & & & @xmath341 + & & & & @xmath342 + & & & & @xmath343 + & & & & @xmath344 + & & & & @xmath345 + & & & & @xmath346 + + & & & & @xmath329 + & & & & @xmath347 + & & & & @xmath348 + & & & & @xmath349 + & & & & @xmath350 + & & & & @xmath351 + & & & & @xmath352 + & & & & @xmath353 + & & & & @xmath354 + & & & & @xmath355 + consider the @xmath45 parity - check matrix @xmath47 of the @xmath2-regular rate @xmath170 tailbiting block code @xmath43 with block - length @xmath152 . starting with each of the first @xmath40 columns of @xmath47 as a root
, we will construct @xmath40 separate trees , where each node @xmath175 is characterized by its depth @xmath183 and partial syndrome state column - vector @xmath356 .
initialize the partial syndrome state of the root @xmath189 of the @xmath190th tree with column @xmath190 of the corresponding parity - check matrix , that is , @xmath357 , @xmath248 .
then grow each tree in such a way , that every branch between any two nodes @xmath175 and @xmath225 is labeled by a column vector @xmath358 , @xmath359 , such that @xmath360 , where every branch label on the path @xmath252 does not occur more than once .
consider now a certain node @xmath175 with nonzero state @xmath361 , where @xmath362 , @xmath363 is a @xmath364 column vector .
if we assume that the @xmath365th position of the column - vector @xmath356 is nonzero , then there are at most @xmath366 columns which can cancel this nonzero position and have not been considered previously . therefore , every node @xmath175 has at most @xmath366 children nodes per nonzero position .
such a tree would grow until all possible linear combinations have been found .
therefore , we assume that the minimum distance is restricted by @xmath367 , that is , the maximum depth of the tree is @xmath368 .
consequently , a node @xmath175 at depth @xmath183 will not be extended , if the number of nonzero positions in its partial syndrome state column - vector @xmath356 exceeds @xmath369 , since at most @xmath3 ones can be canceled by each branch . by initially reordering the columns of the parity - check matrix @xmath47 in such a way that each of the @xmath174 nonoverlapping blocks of @xmath39 rows contains not more than a single one per column
, we can strengthen the stopping criterion as follows : a node @xmath175 at depth @xmath183 will not be extended , if the number of nonzero positions in each of its partial syndrome state column - vectors @xmath362 , @xmath363 exceeds @xmath370 , since at most @xmath128 one in each block can be canceled by each branch . in particular , such a reordering of the parity - check matrix @xmath47 corresponds directly to the parity - check matrix @xmath66 of the equivalent @xmath2-regular ldpc block code constructed from circulant matrices .
determine the minimum distance of a rate @xmath21 @xmath5-regular ldpc block code 1 .
assume a suitable restriction @xmath75 on the minimum distance @xmath367 .
2 . grow @xmath40 separate trees as follows : 1 .
initialize the root node of the @xmath190th tree by @xmath371 with depth @xmath192 .
2 . extend all nodes @xmath175 as long as the hamming weights of their partial syndrome states @xmath372 ( note , for codes with blocks of @xmath39 rows containing only a single one , this criterion can be strengthen to @xmath373 , @xmath363 ) .
3 . the minimum distance @xmath142 follows directly as + rcl d _ & = & _ \ { ( ) @xmath374 ( ) = @xmath35 } + if there is no node @xmath175 whose partial syndrome state @xmath375 , then the minimum distance is lower - bounded by @xmath376 .
c!c!c!c!l @xmath4 & @xmath328 & @xmath142 & @xmath39 & + + & & & & @xmath377 + & & & & @xmath378 + & @xmath379 & & & @xmath380 + & ( @xmath381@xcite ) & & & @xmath382 + & @xmath383 & & & @xmath384 + & ( @xmath385@xcite ) & & & @xmath386 + & & & & @xmath387 + & & & & @xmath388 + & & & & @xmath389 + & & & & @xmath390 + & & & & @xmath391 + & & & & @xmath392 + & & & & @xmath393 + & & & & @xmath394 + & & & & @xmath395 + & & & & @xmath396 + & & & & @xmath397 + & & & & @xmath398 + + & & & & @xmath399 + & & & & @xmath400 + & & & & @xmath348 + & & & & @xmath349 + & & & & @xmath350 + & & & & @xmath351 + & & & & @xmath401 + & & & & @xmath402 + & & & & @xmath403 + & & & & @xmath404 + & & & & @xmath405 + & & & & @xmath406 +
when presenting our search results for qc ( @xmath407)-regular ldpc block codes with different girth we will distinguish two cases .
we started by searching for qc ( @xmath407)-regular ldpc block codes using an all - one base matrix @xmath99 , applied the algorithms as described above and obtained qc ( @xmath408)-regular ldpc block codes with girth @xmath409 , and @xmath8 as given in tables [ tab : girth6 ] [ tab : girth12 ] .
these codes correspond to a parity - check matrix @xmath50 of a convolutional code with only monomial entries as given , for example , in .
however , according to theorem [ th : three - a ] the achievable girth @xmath125 of a qc @xmath2-regular ldpc code , constructed in such a way , is limited by @xmath410 .
thus , in order to find qc @xmath411-regular ldpc block codes with girth @xmath412 as presented in tables [ tab : girthgreater ] and [ tab : girth20 ] , we have to allow zero entries in our base matrix @xmath99 .
this is a straight - forward generalization of the restriction to only monomial entries in the parity - check matrix @xmath50 of the underlying convolutional code .
c!c!c!c!l @xmath4 & @xmath328 & @xmath142 & @xmath39 & + + & @xmath413 & & @xmath414 & @xmath415 + & ( @xmath416@xcite ) & & ( @xmath417@xcite ) & @xmath418 + & @xmath419 & & @xmath420 & @xmath421 + & ( @xmath422@xcite ) & & ( @xmath420@xcite ) & @xmath423 + & @xmath424 & & @xmath425 & @xmath426 + & ( @xmath427@xcite ) & & ( @xmath428@xcite ) & @xmath429 + & & & @xmath430 & @xmath431 + & & & ( @xmath432@xcite ) & @xmath433 + & & & @xmath434 & @xmath435 + & & & ( @xmath436@xcite ) & @xmath437 + & & & @xmath438 & @xmath439 + & & & ( @xmath440@xcite ) & @xmath441 + & & & & @xmath442 + & & & @xmath443 & @xmath444 + & & & ( @xmath445@xcite ) & @xmath446 + & & & & @xmath447 + & & & & @xmath448 + & & & @xmath449 & @xmath450 + & & & ( @xmath451@xcite ) & @xmath452 + & & & & @xmath453 + & & & & @xmath454 + & & & @xmath455 & @xmath456 + & & & ( @xmath457@xcite ) & @xmath458 + & & & & @xmath459 + + & & & & @xmath415 + & & & & @xmath418 + c!c!c!c!l @xmath4 & @xmath328 & @xmath142 & @xmath39 & + + & @xmath460 & & @xmath461 & @xmath462 + & ( @xmath463@xcite ) & & ( @xmath464@xcite ) & @xmath465 + & @xmath466 & & @xmath467 & @xmath468 + & ( @xmath469@xcite ) & & ( @xmath470@xcite ) & @xmath471 + & @xmath472 & & @xmath473 & @xmath474 + & ( @xmath475@xcite ) & & ( @xmath476@xcite ) & @xmath477 + & & & @xmath478 & @xmath479 + & & & ( @xmath476@xcite ) & @xmath480 + & & & @xmath481 & @xmath482 + & & & ( @xmath483@xcite ) & @xmath484 + & & & @xmath485 & @xmath486 + & & & ( @xmath487@xcite ) & @xmath488 + & & & & @xmath489 + & & & @xmath490 & @xmath491 + & & & ( @xmath492@xcite ) & @xmath493 + & & & & @xmath494 + & & & & @xmath495 + & & & & @xmath496 + & & & & @xmath497 + & & & & @xmath498 + & & & & @xmath499 + & & & & @xmath500 + & & & & @xmath501 + & & & & @xmath502 + & & & & @xmath503 + & & & & @xmath504 + & & & & @xmath505 + & & & & @xmath506 + in tables [ tab : girth6 ] [ tab : girth12 ] , parity - check matrices of short known qc @xmath411-regular ldpc block codes with girth @xmath409 , and @xmath8 together with those of large minimum distance are presented .
when searching for such codes , we applied the following restrictions to reduce the number of possible voltage assignments : * as the girth of a voltage graph is defined as the shortest cycle with voltage zero , and the sign of the voltage depends on the direction of the edge , we can add the same arbitrary offset to the voltage of all edges being connected to the same node .
thus , without loss of generality , we set the voltage of all edges connected to one specific symbol node as well as all edges connected to one specific constraint node to voltage zero .
( for consistency with codes constructed from steiner triple system , that will be introduced later , we choose the first symbol node and the last constraint node .
this corresponds directly to a degree matrix with zeros in its first column and last row . )
+ for example , the degree matrix of the @xmath507 qc ldpc block code with girth @xmath508 from table [ tab : girth8 ] follows directly as + rcl w & = & ( + cccc 0 & 1 & 4 & 6 + 0 & 5 & 2 & 3 + 0 & 0 & 0 & 0 + ) * furthermore , to reduce the number of only permuted degree matrices , we assume that * * the first row is sorted in ascending order . * * when sorting the first and the second row in ascending order , the second row is lexicographically less than the first row . * * the maximum degree is less than the tailbiting length @xmath39 for which there exists a qc @xmath411-regular ldpc block code with given girth @xmath125 . *
qc @xmath507-regular ldpc block codes were found by exhaustive search over the previously defined set of restricted edge voltages . *
qc @xmath509-regular ldpc block codes with @xmath510 were obtained by adding one additionally , randomly chosen column to the best degree matrices of codes with @xmath511 having the same girth @xmath125 .
the maximum degree in this additional column is limited by twice the maximum degree of the previous code . using these restrictions , the obtained qc @xmath411-regular
ldpc block codes with girth @xmath409 , and @xmath8 are presented in tables [ tab : girth6 ] [ tab : girth12 ] .
the first column @xmath4 denotes the number of nonzero elements per row , which corresponds to the number of columns in @xmath50 and @xmath63 , due to the all - ones base matrix @xmath99 . as all entries in the first column and the last row of the degree matrix @xmath63
are zero , they are omitted in the submatrix @xmath512 which is given in the corresponding last column .
consider now the parity - check matrix @xmath50 of the rate @xmath322 convolutional code @xmath22 , with only monomial entries corresponding to the degree matrix @xmath63 . by tailbiting the semi - infinite parity - check matrix @xmath34 to length @xmath39 ( given in the forth column ) , we obtain the parity - check matrix @xmath47 of an @xmath328 block code @xmath43 with minimum distance @xmath142 , where @xmath328 and @xmath142 follow from the second and third column , respectively .
note that due to linear dependent rows in @xmath47 the rank of @xmath43 might be less than @xmath513 .
the codes presented in tables [ tab : girth6 ] and [ tab : girth8 ] coincide with the qc ldpc block codes found by the `` hill - climbing '' algorithm @xcite , but we determined their minimum distance with our algorithm described in section [ sec : 7 ] .
tables [ tab : girth10 ] and [ tab : girth12 ] contain new qc @xmath411-regular ldpc block codes , which , to the best of our knowledge , are shorter than the previously known codes obtained from an all - ones base matrix @xcite . in particular , these codes are significantly shorter than those presented in @xcite , which are obtained from base matrices with zeros . however
, due to the zeros in their base matrix , the minimum distance of the ldpc block codes in @xcite can exceed @xmath20 .
for example , we determined the minimum distance of the @xmath514 qc @xmath130-regular ldpc block code with girth @xmath515 in @xcite to be @xmath516 , while the corresponding code in table [ tab : girth12 ] , that is , the @xmath460 qc @xmath130-regular ldpc block code , has only minimum distance @xmath517 , but shorter block length . using the beast @xcite
, we calculated the free distance of the corresponding parent convolutional code for the code in @xcite to be @xmath518 .
therefore , using our approach and a larger tailbiting length it would be possible to construct corresponding qc @xmath130-regular ldpc block codes with minimum distance up to @xmath519 . in order to find qc @xmath411-regular ldpc block codes with girth @xmath520
, we have to allow zero entries in our base matrix @xmath99 ; that is , relax the restriction from only monomial entries in @xmath50 to include also zero entries .
according to theorem [ th : two ] , a code with girth @xmath125 exists if the corresponding base graph has girth @xmath102 satisfying . additionally , as we are searching for codes with short block length , we consider the shortest possible base matrices @xmath99 .
c!c!c!c!c!l @xmath4 & @xmath125 & @xmath328 & @xmath39 & base graph & + & & @xmath521 & @xmath522 & @xmath523 & @xmath524 + & & ( @xmath525@xcite ) & ( @xmath526@xcite ) & @xmath527 & @xmath528 + & & @xmath529 & & @xmath530 & @xmath531 + & & ( @xmath532@xcite ) & & @xmath533 & @xmath534 + & & & & & @xmath535 + & & @xmath536 & @xmath537 & @xmath538 & @xmath539 + & & ( @xmath540@xcite ) & ( @xmath541@xcite ) & @xmath542 & @xmath543 + & & & & & @xmath544 + @xmath283 & @xmath10 & @xmath545 & @xmath546 & @xmath547 & _ available at @xcite _ + & & & & @xmath523 & @xmath548 + & & & & @xmath527 & @xmath549 + & & & & & @xmath550 + & & @xmath551 & & @xmath530 & @xmath552 + & & ( @xmath553@xcite ) & & @xmath533 & @xmath554 + & & & & & @xmath555 + & & & & & @xmath556 + & & @xmath557 & @xmath558 & @xmath538 & @xmath559 + & & ( @xmath560@xcite ) & ( @xmath561@xcite ) & @xmath542 & @xmath562 + & & & & & @xmath563 + & & & @xmath564 & @xmath523 & @xmath565 + & & & ( @xmath566@xcite ) & @xmath527 & @xmath567 + & & & & & @xmath568 + & & @xmath569 & & @xmath530 & @xmath570 + & & ( @xmath571@xcite ) & & @xmath533 & @xmath572 + & & & & & @xmath573 + when searching for qc @xmath411-regular ldpc block codes with girth @xmath574 , and @xmath11 , we started with a ( shortened ) base graph constructed by using steiner triple systems of order @xmath89 , that is , @xmath575 @xcite . for all @xmath89 , where @xmath576 is equal to @xmath128 or @xmath162
, there exists a steiner triple system of order @xmath89 .
then we construct a @xmath5-regular , @xmath31 base matrix @xmath99 with entries @xmath577 , @xmath52 and @xmath53 , in such a way that the positions of the nonzero entries in each column correspond to a steiner triple system of order @xmath578 .
denote such a @xmath31 base matrix @xmath579 .
using the obtained @xmath5-regular @xmath31 base matrix @xmath99 , we search for a set of edge labels , such that the corresponding voltage graph has at least girth @xmath125 . in general
, it is possible , without loss of generality , to label a certain subset of edges of the voltage graph simultaneously with zero voltage and thereby decreasing the number of possible labelings .
the following algorithm constructs a @xmath31 base matrix @xmath99 based on @xmath580 and reorders the matrix to maximize the number of zero entries in its lower left corner .
using such a base matrix , it is always possible to label the last nonzero entry in each column with degree zero .
moreover , in each of the remaining rows at the top of the base matrix , we can label at least one nonzero entry with degree zero .
( hereinafter we will always choose at least the first element in the remaining rows to be labeled with zero voltage ) .
construction of a @xmath5-regular @xmath31 base graph @xmath99 obtained from @xmath580 1 .
initialize a counter @xmath581 to zero and denote the current row and column by @xmath150 and @xmath75 , respectively , starting from the right - most entry in the last row , that is , @xmath582 and @xmath583 .
2 . set the @xmath584 elements in row @xmath150 and column @xmath585 to one , that is , @xmath586 for @xmath587 and @xmath588 .
3 . choose the remaining @xmath589 nonzero positions in each of those @xmath584 columns to fulfill the properties of a steiner triple system .
if possible , choose the positions @xmath577 to minimize @xmath190 .
in other words , try to avoid using the lowest rows @xmath590 , if possible , despite of the restrictions imposed by the steiner triple system .
4 . finally ,
decrease @xmath75 by @xmath584 , set @xmath150 to @xmath591 , denote the number of nonzero elements in the new row @xmath150 by @xmath584 and continue with step @xmath163 until all @xmath40 columns are used , that is , @xmath592 . by removing the last row and last @xmath4 columns of the @xmath5-regular @xmath31 base matrix @xmath99 constructed using @xmath580
, we obtain a shortened @xmath593 ( @xmath594)-regular base matrix @xmath158 , which we denote @xmath595 . by deleting columns and rows ,
it is also possible to obtain intermediate codes , which are , however , irregular .
[ ex : basests ] in the following we shall construct the @xmath596 ) base matrices @xmath99 of dimension @xmath597 @xmath598 , dimension @xmath599 @xmath600 and dimension @xmath601 @xmath602 . using algorithm sts , we obtain the following steiner triple systems of order @xmath296 ( @xmath523 ) , @xmath603 ( @xmath538 ) and @xmath604 ( @xmath547 ) .
rllll + \ { & \{2 , 3 , 5 } , & \{1 , 4 , 6 } , & \{1 , 3 , 7 } , & \{2 , 6 , 7 } , + & \{4 , 5 , 7 } , & \{1 , 2 , 8 } , & \{5 , 6 , 8 } , & \{3 , 4 , 8 } , + & \{1 , 5 , 9 } , & \{2 , 4 , 9 } , & \{3 , 6 , 9 } , & \{7 , 8 , 9 } } + + \ { & \{0 , 3 , 6 } , & \{0 , 2 , 7 } , & \{1 , 5 , 7 } , & \{3 , 4 , 7 } , + & \{3 , 5 , 8 } , & \{1 , 4 , 8 } , & \{2 , 6 , 8 } , & \{2 , 4 , 9 } , + & \{5 , 6 , 9 } , & \{0 , 1 , 9 } , & \{1 , 3 , 10 } , & \{0 , 4 , 10 } , + & \{6 , 7 , 10 } , & \{2 , 5 , 10 } , & \{8 , 9 , 10 } , & \{7 , 8 , 11 } , + & \{4 , 6 , 11 } , & \{1 , 2 , 11 } , & \{0 , 5 , 11 } , & \{3 , 9 , 11 } , rllll & \{10 , 11 , 12 } , & \{7 , 9 , 12 } , & \{0 , 8 , 12 } , & \{1 , 6 , 12 } , + & \{4 , 5 , 12 } , & \{2 , 3 , 12 } } + + \ { & \ { 4 , 5,10 } , & \ { 1 , 9,10 } , & \ { 7 , 8,11 } , & \ { 1 , 6,11 } , + & \ { 2 , 3,12 } , & \ { 0 , 9,12 } , & \ { 6 , 8,12 } , & \ { 8 , 9,13 } , + & \ { 6 , 7,13 } , & \ { 0 , 5,13 } , & \ { 2,10,13 } , & \ { 3 , 4,14 } , + & \ { 1,12,14 } , & \ { 0 , 2,14 } , & \ { 7 , 9,14 } , & \ { 5,11,14 } , + & \ { 5 , 6,15 } , & \ { 3,10,15 } , & \ { 4,12,15 } , & \ { 1 , 7,15 } , + & \ { 0 , 8,15 } , & \{11,13,16 } , & \ { 5 , 7,16 } , & \ { 6,10,16 } , + & \ { 2 , 8,16 } , & \ { 3 , 9,16 } , & \ { 0 , 4,16 } , & \ { 9,11,17 } , + & \{12,13,17 } , & \ { 1 , 3,17 } , & \ { 4 , 7,17 } , & \ { 0 , 6,17 } , + & \ { 2 , 5,17 } , & \ { 8,17,18 } , & \ { 3,11,18 } , & \ { 2 , 4,18 } , + & \{13,15,18 } , & \ { 0,10,18 } , & \ { 1,16,18 } , & \ { 6,14,18 } , + & \ { 9,18,19 } , & \ { 4 , 8,19 } , & \{14,15,19 } , & \{10,11,19 } , + & \ { 0 , 3,19 } , & \ { 2 , 7,19 } , & \{12,16,19 } , & \ { 1 , 5,19 } , + & \{17,19,20 } , & \ { 9,15,20 } , & \{10,12,20 } , & \ { 0,11,20 } , + & \ { 5 , 8,20 } , & \ { 1 , 4,20 } , & \{13,14,20 } , & \ { 3 , 7,20 } , + & \ { 2 , 6,20 } , & \ { 5,18,21 } , & \ { 4 , 6,21 } , & \ { 1,13,21 } , + & \{16,17,21 } , & \{10,14,21 } , & \ { 2 , 9,21 } , & \ { 3 , 8,21 } , + & \{11,15,21 } , & \
{ 7,12,21 } , & \{19,21,22 } , & \{18,20,22 } , + & \ { 0 , 7,22 } , & \{10,17,22 } , & \ { 3 , 5,22 } , & \ { 6 , 9,22 } , rllll & \ { 2,15,22 } , & \
{ 1 , 8,22 } , & \{11,12,22 } , & \ { 4,13,22 } , + & \{14,16,22 } , & \{20,21,23 } , & \ { 0 , 1,23 } , & \ { 6,19,23 } , + & \{15,16,23 } , & \ { 2,11,23 } , & \ { 7,18,23 } , & \ { 5,12,23 } , + & \{14,17,23 } , & \ { 4 , 9,23 } , & \ { 8,10,23 } , & \ { 3,13,23 } , + & \ { 0,21,24 } , & \{22,23,24 } , & \ { 1 , 2,24 } , & \{16,20,24 } , + & \ { 7,10,24 } , & \ { 8,14,24 } , & \{13,19,24 } , & \ { 3 , 6,24 } , + & \{12,18,24 } , & \{15,17,24 } , & \ { 5 , 9,24 } , & \ { 4,11,24 } } each number @xmath605 occurs @xmath4 times within the set of steiner triples .
however , the chosen steiner triples are not uniquely determined .
the corresponding base matrices of dimension @xmath597 @xmath523 , dimension @xmath599 @xmath538 , and dimension @xmath601 @xmath547 are sparse matrices with nonzero elements only in column @xmath190 and row @xmath233 , where the @xmath190th steiner triple contains the value @xmath233 . the @xmath597 @xmath130-regular base matrix constructed from @xmath523 denoted by @xmath606
is given , for example , by rcl [ eq : base_sts9 ] b _ & = & entries that correspond to edges in the base graph that can be , according to algorithm sts , labeled with zero voltage are marked in bold . by removing the last row and the last @xmath607 columns ,
the corresponding shortened @xmath608 @xmath609-regular base matrix @xmath610 follows directly as rcl [ eq : base_sts9_shortened ] b _ & = & this corresponds to removing the four steiner triples of @xmath523 containing the number of the last row . shortening the @xmath597 base matrix @xmath606 constructed from @xmath523 to obtain a shortened @xmath608 base matrix @xmath610 is unpractical as its code rate is @xmath611 . however , by shortening the @xmath612 base matrix @xmath613 in the same way we obtain a @xmath614 base matrix @xmath615 with the feasible code rate @xmath616 . in table
[ tab : girthgreater ] the obtained qc ( @xmath407)-regular ldpc block codes with girth @xmath574 , and @xmath11 constructed from steiner triple systems are presented .
while the number of nonzero elements in each column is fixed to @xmath16 , the number of nonzero elements in each row @xmath4 is specified in the first column .
the second column corresponds to the obtained girth @xmath125 , while in the third and forth columns we give the dimensions of the block code @xmath328 after tailbiting to length @xmath39 .
and the fifth column contains which steiner triple system ( @xmath575 ) is used .
finally , in the last column @xmath617 we give the degrees of the corresponding degree matrix @xmath63 in a compact way .
as we have constructed the base matrices in such a way that the last nonzero entry in each column and the first entry in all other rows of the base matrix is labeled with a zero voltage , these entries are omitted .
an entry of @xmath617 in column @xmath233 and row @xmath190 corresponds to the voltage degree of the @xmath618th nonzero entry in the @xmath190th row of the corresponding base matrix .
c!c!c!c!c @xmath4 & @xmath125 & @xmath619 & @xmath39 & base graph ( table [ tab : girth8 ] ) + @xmath164 & @xmath620 & @xmath621 & @xmath622 & @xmath623 , @xmath508 + @xmath270 & @xmath620 & @xmath624 & @xmath625 & @xmath626 , @xmath508 + @xmath97 & @xmath620 & @xmath627 & @xmath628 & @xmath629 , @xmath508 + @xmath164 & @xmath630 & @xmath631 & @xmath632 & @xmath623 , @xmath508 @xcite + @xmath270 & @xmath630 & @xmath633 & @xmath634 & @xmath626 , @xmath508 + @xmath164 & @xmath0 & @xmath635 & @xmath636 & @xmath623 , @xmath508 + when searching for qc @xmath411-regular ldpc block codes with girth @xmath637 and @xmath0 , we started with previously obtained qc @xmath411-regular ldpc block codes of smaller block size and smaller girth , and ( re-)applied our algorithms .
the obtained results for qc @xmath411-regular ldpc block codes with girth @xmath637 and @xmath0 are presented in table [ tab : girth20 ] .
they are all but one based on previously obtained @xmath411-regular ldpc block codes with girth @xmath508 ( _ cf . _
table [ tab : girth8 ] ) , as specified in their last column in table [ tab : girth20 ] .
as before , the first column @xmath4 denotes the number on nonzero elements in each column ; then we give the obtained girth @xmath125 and the dimensions of the block code @xmath328 after tailbiting to length @xmath39 .
the corresponding degree matrices are too large and are omitted in table [ tab : girth20 ] , but are available at @xcite .
these codes are ( probably ) unpractical due to their huge block length .
however , the table illustrates that by interpreting qc @xmath5-regular ldpc block codes as base matrices and re - applying our algorithms we can find qc @xmath5-regular ldpc block codes of `` any '' girth @xmath125 .
using the relation between the parity - check matrix of qc ldpc block codes and the biadjacency matrix of bipartite graphs , new searching techniques have been presented . starting from a base graph , a set of edge voltages
is used to construct the corresponding voltage graph with a given girth . by representing bipartite graphs in different ways , lower and
upper bounds on the girth as well as on the minimum distance of the corresponding tailbiting block code have been discussed .
new algorithms for searching iteratively for bipartite graphs with large girth and for determining the minimum distance of the corresponding qc ldpc block code have been presented .
depending on the given girth , the search algorithms are either based on all - ones matrices , steiner triple systems , or qc block codes . amongst others ,
new qc regular ldpc block codes with girth between @xmath18 and @xmath0 have been presented including their minimum distance if possible . in particular , the previously unknown minimum distance , for some known codes with girth @xmath97 and @xmath96 , has been determined .
this research was supported in part by the swedish research council under grant 621 - 2007 - 6281 .
d. j. mackay and m. c. davey , `` evaluation of gallager codes for short block length and high rate applications , '' in _ codes , systems and graphical models_.1em plus 0.5em minus 0.4emspringer - verlag , 1999 , pp . 113130 .
chung , j. g. david forney , t. j. richardson , and r. urbanke , `` on the design of low - density parity - check codes within 0.0045 db of the shannon limit , '' _ ieee commun .
_ , vol . 5 , no . 2 ,
5860 , feb .
m. lentmaier , a. sridharan , d. j. costello , jr . , and k. sh .
zigangirov , `` iterative decoding threshold analysis for ldpc convolutional codes , '' _ ieee trans .
inf . theory _ ,
56 , no .
52745289 , oct .
s. kudekar , t. richardson , and r. urbanke , `` threshold saturation via spatial coupling : why convolutional ldpc ensembles perform so well over the bec , '' in _ proc .
ieee international symposium on information theory ( isit10 ) _ , austin , texas , jun .
13 18 , 2010 , pp . 684688 .
g. schmidt , v. v. zyablov , and m. bossert , `` on expander codes based on hypergraphs , '' in _ proc .
ieee international symposium on information theory ( isit03 ) _ , yokohama , japan , jun .
29 jul . 4 , 2003 , p. 88 .
i. e. bocharova , b. d. kudryashov , r. v. satyukov , and s. stiglmayr , `` short quasi - cyclic ldpc codes from convolutional codes , '' in _ proc .
ieee international symposium on information theory ( isit09 ) _ , seoul , south - korea , jun .
28 jul . 3 , 2009 , pp .
551555 .
l. dolecek , p. lee , z. zhang , v. anantharam , b. nikolic , and m. wainwright , `` predicting error floors of structured ldpc codes : deterministic bounds and estimates , '' _ ieee j. sel .
areas commun . _ ,
27 , no . 6 , pp .
239246 , aug .
y. kou , s. lin , and m. p. c. fossorier , `` low - density parity - check codes based on finite geometries : a rediscovery and new results , '' _ ieee trans .
inf . theory _ ,
47 , no . 7 , pp . 27112736 , nov .
2001 .
, `` construction of low - density parity - check codes from kirkman triple systems , '' in _ proc .
ieee global telecommunications conference ( globecom01 ) _ , vol . 2 , san antonio , usa , nov . 2529 , 2001 , pp . 970974 .
r. m. tanner , `` on graph constructions for ldpc codes by quasi - cyclic extension , '' in _ information , coding and mathematics _ , m. blaum , p. g. farrel , and h. c. a. van tilbord , eds.1em plus 0.5em minus 0.4em norwell , ma : kluwer , 2002 , pp .
209219 .
r. m. tanner , d. sridhara , a. sridharan , t. e. fuja , and j. daniel j. costello , `` ldpc block and convolutional codes based on circulant matrices , '' _ ieee trans .
inf . theory _ ,
50 , no . 12 , pp .
29662984 , dec . 2004 .
m. esmaeili and m. gholami , `` structured quasi - cyclic ldpc codes with girth 18 and column - weight @xmath12 , '' _ international journal of electronics and communications ( aeu ) _ , vol .
64 , no . 3 , pp .
202217 , 2010 .
r. smarandache and p. o. vontobel , `` on regular quasi - cyclic ldpc codes from binomials , '' in _ proc .
ieee international symposium on information theory ( isit04 ) _ , chicago , usa , jun . 27 jul . 2 , 2004 ,
p. 274 .
y. wang , j. s. yedidia , and s. c. draper , `` construction of high - girth qc - ldpc codes , '' in _ proc .
5th international symposium on turbo codes and related topics _
, lausanne , switzerland , 0105 , 2008 , pp .
180185 .
i. e. bocharova , f. hug , r. johannesson , b. d. kudryashov , and r. v. satyukov , `` new low - density parity - check codes with large girth based on hypergraphs , '' in _ proc .
ieee international symposium on information theory ( isit10 ) _ , austin , texas , jun .
13 18 , 2010 , pp . 819823 .
j. thorpe , k. andrews , and s. dolinar , `` methodologies for designing ldpc codes using protographs and circulants , '' in _ proc .
ieee international symposium on information theory ( isit04 ) _ , chicago , usa , p. 238 . | the relation between parity - check matrices of quasi - cyclic ( qc ) low - density parity - check ( ldpc ) codes and biadjacency matrices of bipartite graphs supports searching for powerful ldpc block codes . using the principle of tailbiting , compact representations of bipartite graphs based on convolutional codes can be found .
bounds on the girth and the minimum distance of ldpc block codes constructed in such a way are discussed .
algorithms for searching iteratively for ldpc block codes with large girth and for determining their minimum distance are presented .
constructions based on all - ones matrices , steiner triple systems , and qc block codes are introduced .
finally , new qc regular ldpc block codes with girth up to @xmath0 are given .
ldpc code , convolutional code , tanner graph , biadjacency matrix , tailbiting , girth , minimum distance |
the simulations analyzed are a small subset of the simulations realized in the framework of the galmer project @xcite .
the 2d velocity maps and the rotation curves of the remnants of retrograde elliptical - spiral mergers are analyzed at least 400 myr after the coalescence of the two systems ( see figs.[maps ] and [ curves ] ) .
the rotation curves clearly show that the counter - rotating region is completely associated to stars initially belonging to the spiral galaxy and that counter - rotating cores are more extended in the case of e0-sa merger remnants rather than for e0-sb ones @xcite .
+ initially , the redistribution of the orbital angular momentum affects mostly the outer parts of the galaxies .
as the interaction proceeds toward the final merging phase , tidal torques begin to affect inner regions too .
if the tidal forces in the inner regions are not strong enough to reverse the initial spin of the spiral galaxy , the spiral continues to rotate with a spin parallel to the initial one , while , in turn , the outer parts can acquire an amount of orbital angular momentum sufficient to reverse their initial rotation .
this causes the emergence of a counter - rotating region , among stars initially in the disk galaxy @xcite . .
top panels correspond to e0-sa mergers , bottom panels to e0-sb ones . in each panel
three curves are shown : the line - of - sight velocity profile of the old stellar component ( black curve ) , the line - of - sight profile of stars initially belonging to the elliptical ( red curve ) , and that of stars initially belonging to the spiral ( blue curve ) .
[ curves],width=130 ] .
top panels correspond to e0-sa mergers , bottom panels to e0-sb ones . in each panel
three curves are shown : the line - of - sight velocity profile of the old stellar component ( black curve ) , the line - of - sight profile of stars initially belonging to the elliptical ( red curve ) , and that of stars initially belonging to the spiral ( blue curve ) .
[ curves],width=130 ] .
top panels correspond to e0-sa mergers , bottom panels to e0-sb ones . in each panel
three curves are shown : the line - of - sight velocity profile of the old stellar component ( black curve ) , the line - of - sight profile of stars initially belonging to the elliptical ( red curve ) , and that of stars initially belonging to the spiral ( blue curve ) .
[ curves],width=130 ] .
top panels correspond to e0-sa mergers , bottom panels to e0-sb ones . in each panel
three curves are shown : the line - of - sight velocity profile of the old stellar component ( black curve ) , the line - of - sight profile of stars initially belonging to the elliptical ( red curve ) , and that of stars initially belonging to the spiral ( blue curve ) .
[ curves],width=130 ] di matteo , p. , combes , f. , melchior , a. l. , & semelin b. 2007a , a&a , 468 , 61 di matteo , p. , combes , f. , melchior , a. l. , & semelin b. , proceedings of `` sf2a-2007 : semaine de lastrophysique francaise '' , ( j. bouvier , a. chalabaev , and c. charbonnel eds ) , astro - ph/0709.2553 di matteo , p. , combes , f. , melchior , a. l. , & semelin b. 2007c , a&a , in press di matteo , p. , combes , f. , melchior , a. l. , & semelin b. in preparation | we investigate , by means of numerical simulations , the kinematics of elliptical - spiral merger remnants .
counterrotation can appear both in coplanar and in non - coplanar retrograde mergers , and it is mostly associated to the presence of a disk component , which preserves part of its initial spin . in turn ,
the external regions of the two interacting galaxies acquire part of the orbital angular momentum , due to the action of tidal forces . |
@xmath0 surfaces with non - symplectic involution are basic objects in the study of @xmath0 surfaces .
they connect @xmath0 surfaces with rational surfaces , enriques surfaces , and low genus curves . in this article
we address the rationality problem for the moduli spaces of @xmath0 surfaces with non - symplectic involution . to be more precise , let @xmath1 be a complex @xmath0 surface with an involution @xmath2 .
when @xmath2 acts nontrivially on @xmath3 , @xmath2 is called _ non - symplectic _ , and the pair @xmath4 is called a _ 2-elementary @xmath0 surface_. by nikulin @xcite , the deformation type of @xmath4 is determined by its _ main invariant _
@xmath5 , a triplet of integers associated to the lattice of @xmath2-invariant cycles .
he classified all main invariants of 2-elementary @xmath0 surface , which turned out to be seventy - five in number ( see figure [ nikulin table ] ) .
for each main invariant @xmath5 , let @xmath6 be an even lattice of signature @xmath7 whose discriminant form is 2-elementary of length @xmath8 and parity @xmath9 .
yoshikawa @xcite , @xcite showed that the moduli space @xmath10 of 2-elementary @xmath0 surfaces of type @xmath5 is the complement of a heegner divisor in the arithmetic quotient defined by the orthogonal group @xmath11 of @xmath6 .
in particular , @xmath10 is irreducible of dimension @xmath12 .
our main result is the following .
[ main ] the moduli space @xmath10 of 2-elementary @xmath0 surfaces of type @xmath5 is rational , possibly except when @xmath13 and @xmath14 .
the remaining spaces @xmath15 and @xmath16 are respectively birational to the moduli spaces of plane sextics and of bidegree @xmath17 curves on @xmath18
. the rationality questions for them are rather classical , and remain open .
a few @xmath10 have been known to be rational : @xmath19 is the moduli of enriques surfaces and is rational by kond @xcite ; he also proved that @xmath20 is rational by identifying it with the moduli of certain trigonal curves @xcite ; @xmath21 is birational to the moduli of genus six curves ( see @xcite , @xcite ) , and hence rational by shepherd - barron @xcite ; dolgachev and kond @xcite recently proved the rationality of @xmath22 , which is the moduli of coble surfaces .
now theorem [ main ] shows that these results are not of sporadic nature .
in other terms , the arithmetic quotients defined by @xmath11 for the 2-elementary primitive sublattices @xmath6 of the @xmath0 lattice are mostly rational .
in @xcite we proved that all @xmath10 are unirational .
the approach there was to use some isogenies between @xmath10 to reduce the problem to fewer main invariants .
in fact , in @xcite we studied certain covers of @xmath10 rather than @xmath10 themselves .
but the rationality problem is much more subtle , requiring individual treatment and delicate analysis .
thus in this article we study the spaces @xmath10 one by one .
although our approach is case - by - case , the schemes of proofs are common for most cases , which we shall try to explain here .
we first find a _
birational _ period map @xmath23 from the quotient space @xmath24 of an explicit parameter space @xmath25 by an algebraic group @xmath26
. then we prove that @xmath24 is rational as a problem in invariant theory ( cf .
@xcite , @xcite ) .
the space @xmath25 parametrizes ( possibly singular ) @xmath27-curves @xmath28 on some smooth rational surfaces @xmath29 , and the map @xmath30 is defined by taking the desingularizations of the double covers of @xmath29 branched along @xmath28 .
thus our approach for 2-elementary @xmath0 surfaces is essentially from quotient surfaces and branch curves .
the fact that @xmath30 is birational means that our construction is canonical " for the generic member of @xmath10 .
the verifications of this property mostly fall into the following two patterns : @xmath31 when the curves @xmath28 are smooth , so that the desingularization process does not take place , it is just a consequence of the equivalence between ( variety , involution ) and ( quotient , branch ) .
this simple approach is taken when @xmath32 ( @xmath29 are del pezzo ) and when @xmath33 .
@xmath34 for many other @xmath5 , the surfaces @xmath29 are @xmath35 or hirzebruch surfaces @xmath36 , the curves @xmath28 have simple singularities of prescribed type , and the group @xmath26 is @xmath37 . for this type of period map
we can calculate its degree in a systematic manner .
a recipe of such calculation is presented in
[ ssec : recipe ] . by finding a period map of degree @xmath38 of this type
, our rationality problem is reduced to invariant theory over @xmath35 or @xmath36 .
it should be noted , however , that the existence of birational period map as above , especially of type @xmath34 , is a priori not clear .
indeed , when @xmath39 , @xmath40 , @xmath41 and when @xmath42 , we give up the above two approaches and study a _
non - birational _ period map @xmath43 of type @xmath34 .
using the recipe in [ ssec : recipe ] , we show that it is a quotient map by a weyl group , and then analyze that weyl group action to find a canonical construction .
the proof for these cases would be more advanced than the above cases @xmath31 ,
@xmath34 . on the other hand , when @xmath44 , some of our birational period maps seem related to nakayama s minimal model process for log del pezzo surfaces of index @xmath45 @xcite , which may explain their birationality . recall that for most @xmath5 the fixed locus @xmath46 of @xmath47 is the union of a genus @xmath48 curve @xmath49 and some other @xmath50-curves ( @xcite ) . as an application of our birational period maps
, we will determine the generic structure of the fixed curve map @xmath51 when @xmath52 .
the point is that @xmath53 is the composition of the inverse period map @xmath54 with the map @xmath55 that associates to a @xmath27-curve its component of maximal genus . we will find that the latter map identifies @xmath24 with a natural fibration over a sublocus of @xmath56 defined in terms of special line bundles and points .
thus , via @xmath53 , @xmath10 is in a nice relation to @xmath56 .
this generalizes the descriptions of @xmath20 and @xmath21 referred above .
finally , we should explain that this article has been developed in two - steps .
in an earlier version of this article , we could not prove the rationality of @xmath57 , @xmath58 , @xmath59 , @xmath60 , ( and @xmath22 ) by the above methods .
after that version had circulated , dolgachev and kond @xcite settled the case of @xmath22 using a lattice - theoretic trick .
then the referee and kond independently suggested that a similar technique is also applicable for @xmath58 , which is produced in the present article .
afterwards , we developed the method of dolgachev - kond further and proved the rationality of @xmath57 , @xmath59 , and @xmath60 . in view of this development , we shall treat those four cases separately in the appendix .
the rest of the article is organized as follows .
[ sec : invariant theory ] and [ sec : hirze ] are preliminaries on invariant theory . in [ sec : 2-ele k3 ] , after reviewing basic theory of 2-elementary @xmath0 surfaces , we explain how to calculate the degrees of certain period maps .
the proof of theorem [ main ] begins with [ sec : k=0 ] , where we treat the case @xmath32 using del pezzo surfaces .
the case @xmath61 is divided according to the genus @xmath62 of main fixed curves .
this division policy comes from our observation on the relation of @xmath10 to @xmath56 . in [ sec : g>6 , k>0 ] we treat the case @xmath63 where trigonal curves of fixed maroni invariant are central .
section @xmath64 is for the case @xmath65 . in
[ sec : g=1 ( 1 ) ] we treat the case @xmath66 , @xmath67 , @xmath41 , where del pezzo surfaces appear again in more nontrivial way .
we study in [ sec : g=1 ( 2 ) ] the rest cases with @xmath66 , @xmath68 , and in [ sec : g=0 ] the case @xmath69 , @xmath70 .
finally , we treat the cases @xmath71 , @xmath72 and @xmath73 in the appendix .
let @xmath25 be an irreducible variety acted on by an algebraic group @xmath26 . throughout this article we shall denote by @xmath24 a _ rational quotient _ ,
namely a variety whose function field is isomorphic to the invariant subfield @xmath74 of the function field @xmath75 of @xmath25 .
as @xmath74 is finitely generated over @xmath76 , a rational quotient always exists ( cf .
@xcite ) and is unique up to birational equivalence .
the inclusion @xmath77 induces a quotient map @xmath78 .
every @xmath26-invariant rational map @xmath79 is factorized as @xmath80 by a unique rational map @xmath81 .
we are interested in the problem when @xmath24 is rational .
we recall some results and techniques following @xcite and @xcite .
[ miyata ] if @xmath26 is a connected solvable group and @xmath25 is a linear representation of @xmath26 , the quotient @xmath24 is rational .
[ katsylo ] if @xmath25 is a linear representation of @xmath82 , the quotient @xmath24 is rational .
theorem [ katsylo ] is for the most part a consequence of katsylo s theorem in @xcite .
the exception in @xcite was settled by @xcite .
see also @xcite theorem 2.12 .
the @xmath26-action on @xmath25 is _ almost transitive _ if @xmath25 contains an open orbit .
the following is a special case of the slice method ( @xcite , @xcite ) .
[ slice ] let @xmath83 be a @xmath26-equivariant dominant morphism with almost transitive @xmath26-action on @xmath84 . if @xmath85 is the stabilizer group of a general point @xmath86 , then @xmath24 is birational to @xmath87 . indeed , the fiber @xmath88 is a slice of @xmath25 in the sense of @xcite with the stabilizer @xmath89 .
the @xmath26-action on @xmath25 is _ almost free _ if a general point of @xmath25 has trivial stabilizer .
[ no - name 1 ] if @xmath90 is a @xmath26-linearized vector bundle with almost free @xmath26-action on @xmath25 , then the induced map @xmath91 is birationally equivalent to a vector bundle over @xmath24 .
when @xmath26 is reductive , the no - name lemma is a consequence of the descent theory for principal @xmath26-bundle . a proof for general @xmath26 is given in @xcite lemma 4.4 .
we will use the no - name method also in the following form .
[ no - name 2 ] let @xmath90 be a @xmath26-linearized vector bundle and let @xmath92 .
suppose that @xmath93 @xmath94 acts almost freely on @xmath25 , @xmath95 @xmath96 acts on @xmath97 by a scalar multiplication @xmath98 , and @xmath99 there exists a @xmath26-linearized line bundle @xmath100 on which @xmath96 acts by @xmath101 .
then @xmath102 is birational to @xmath103
. apply proposition [ no - name 1 ] to the @xmath104-linearized vector bundle @xmath105 .
we have a canonical identification @xmath106 .
notice that the quotient map @xmath107 is linear on the fibers ( cf .
we prepare miscellaneous results concerning automorphism action on hirzebruch surfaces . for @xmath108
let @xmath109 be the vector bundle @xmath110 over @xmath111 . the projectivization @xmath112 is a hirzebruch surface .
our convention is that a point of @xmath36 represents a line in a fiber of @xmath109 .
then the section @xmath113 is a @xmath114-curve which we denote by @xmath115 .
let @xmath116 be the natural projection .
if @xmath53 is a @xmath117-fiber , we shall denote the line bundle @xmath118 by @xmath119 .
the picard group @xmath120 consists of the bundles @xmath119 , @xmath121 .
for example , the section @xmath122 belongs to @xmath123 ; the fiber @xmath53 belongs to @xmath124 ; the canonical bundle @xmath125 is isomorphic to @xmath126 .
we have @xmath127 and @xmath128 . _
except for [ ssec : trigonal ] , we assume @xmath129 in this section .
_ under this assumption , the action of @xmath130 preserves @xmath117 and @xmath115 .
consequently , we have the exact sequence @xmath131 where @xmath132 . via the given splitting @xmath133
, we may identify @xmath134 thus we have @xmath135 .
when @xmath136 is even , @xmath109 admits a @xmath137-linearization via that of @xmath138 , so that the sequence @xmath139 splits .
let @xmath140 where @xmath141 acts on the component @xmath142 of @xmath143 in the natural way . via the identification and the @xmath141-linearization of @xmath138 ,
the bundle @xmath109 is @xmath144-linearized .
this induces a surjective homomorphism @xmath145 , whose kernel is @xmath146 generated by @xmath147 when @xmath136 is odd , and by @xmath148 when @xmath136 is even .
[ cover linearization hirze ] every line bundle on @xmath36 is @xmath144-linearized .
the @xmath141-linearization of @xmath149 induces a @xmath144-linearization of @xmath150 .
moreover , @xmath151 is the dual of the tautological bundle over @xmath36 , which is the blow - up of @xmath109 along the zero section .
hence @xmath152 is @xmath144-linearized .
[ linearization hirze ] when @xmath136 is odd , every line bundle on @xmath36 is @xmath130-linearized .
when @xmath136 is even , the bundle @xmath153 admits an @xmath130-linearization if @xmath154 is even . in the @xmath144-linearization of @xmath109 , the element @xmath155 ( resp .
@xmath156 ) acts trivially when @xmath136 is even ( resp .
hence @xmath151 is always @xmath130-linearized .
also @xmath157 is @xmath130-linearized via the @xmath137-linearization of @xmath158 .
this proves the assertion for even @xmath136 .
the homomorphism @xmath159 defined by @xmath160 induces a @xmath38-dimensional representation @xmath84 of @xmath144 on which @xmath156 acts by @xmath161 .
hence the bundle @xmath162 is @xmath130-linearized for odd @xmath136 .
we study the @xmath130-action on the linear systems @xmath124 , @xmath123 , @xmath163 , and @xmath164 . recall that we are assuming @xmath129 . [
stabilizer of fiber ] the group @xmath130 acts on the linear system @xmath124 transitively , and the stabilizer of a point of @xmath124 is connected and solvable .
we have a canonical identification @xmath165 given by @xmath166 .
the sequence @xmath139 restricted to the stabilizer @xmath167 of a @xmath168 gives the exact sequence @xmath169 where @xmath170 is the stabilizer of @xmath171 in @xmath172 . since both @xmath143 and @xmath173 are connected and solvable , so is @xmath167 .
every smooth divisor @xmath174 in @xmath123 is a section of @xmath117 disjoint from @xmath115 .
hence it corresponds to another splitting @xmath175 for which the image of @xmath176 is @xmath174 .
[ stabilizer of section ] the group @xmath130 acts transitively on the open set @xmath177 of smooth divisors .
the sequence restricted to the stabilizer @xmath178 of an @xmath179 gives the exact sequence @xmath180 in fact , the subgroup @xmath181 acts transitively on @xmath25 because any two splittings @xmath175 are @xmath182-equivalent .
when @xmath174 corresponds to the original splitting of @xmath109 , we have @xmath183 .
the homomorphism @xmath184 is surjective thanks to the @xmath141-linearization of @xmath185 .
[ l_1,1 ] the group @xmath130 acts on @xmath163 almost transitively with the stabilizer of a general point being connected and solvable .
we first show that the subgroup @xmath186 acts on @xmath163 almost freely . indeed , if an element @xmath187 preserves ( and hence fixes ) a smooth @xmath188 , then for a general @xmath189 the @xmath190 points @xmath191 are fixed by @xmath62 . since @xmath192 , then @xmath62 fixes @xmath174 and
so must be trivial .
the group @xmath143 preserves the sub linear system @xmath193 of curves passing through each @xmath194 . since @xmath195 ,
the @xmath143-action on @xmath196 is also almost transitive .
thus @xmath130 acts on @xmath163 almost transitively .
let @xmath197 be the stabilizer of a general @xmath188 .
since @xmath198 , the natural homomorphism @xmath199 is injective .
its image is contained in the stabilizer @xmath170 of the point @xmath200 .
then the embedding @xmath201 is open because @xmath202 .
therefore @xmath26 is connected and solvable .
every smooth divisor in @xmath164 is a hyperelliptic curve of genus @xmath203 whose @xmath204 is given by the restriction of @xmath117 .
[ he moduli ] if @xmath205 is the open set of smooth divisors , a geometric quotient @xmath206 exists and is isomorphic to the moduli space @xmath207 of hyperelliptic curves of genus @xmath203 .
recall that @xmath207 is a normal irreducible variety .
we have a natural @xmath130-invariant morphism @xmath208 . in order to show that @xmath209 is surjective , let @xmath210 be a hyperelliptic curve of genus @xmath203 and let @xmath211 be its @xmath204 .
the curve @xmath210 is naturally embedded in @xmath212 over @xmath111 by the evaluation map .
let @xmath213 be the decomposition with respect to the hyperelliptic involution @xmath2 , where @xmath2 acts on @xmath214 by @xmath215 .
it is clear that @xmath216 , and a cohomology calculation shows that @xmath217 .
thus @xmath212 is isomorphic to @xmath36 .
then @xmath210 belongs to @xmath218 for some @xmath219 , and the genus formula shows that @xmath220 .
therefore @xmath209 is surjective .
conversely , for a @xmath221 the natural embedding @xmath222 extends to an isomorphism @xmath223 ( e.g. , by the evaluation at @xmath210 via the canonical identifications @xmath224 and @xmath225 ) .
this implies that the @xmath209-fibers are @xmath130-orbits .
now our assertion follows from @xcite proposition 0.2 . by the proof
, the hyperelliptic involution of a smooth @xmath226 uniquely extends to an involution @xmath227 of @xmath36 .
concretely , for each @xmath117-fiber @xmath53 consider the involution of @xmath53 which fixes the point @xmath228 and exchanges the two points @xmath229 ( or fixes @xmath229 when @xmath53 is tangent to @xmath210 ) .
this defines @xmath227 .
[ stab he ] when @xmath230 , the stabilizer in @xmath130 of a general @xmath231 is @xmath232 . by proposition
[ he moduli ] @xmath210 has no automorphism other than its hyperelliptic involution .
any automorphism of @xmath36 acting trivially on @xmath210 must be trivial because it fixes three points of general @xmath117-fibers .
_ we allow @xmath108 in this subsection .
_ it is known that trigonal curves are naturally related to hirzebruch surfaces .
we recall here some basic facts ( cf .
@xcite , @xcite ) . a canonically embedded trigonal curve @xmath233 of genus @xmath234 is contained in a unique rational normal scroll , that is , the image of a hirzebruch surface @xmath36 by a linear system @xmath235 .
the scroll is swept out by the lines spanned by the fibers of the trigonal map @xmath236 ( which is unique ) , and may also be cut out by the quadrics containing @xmath210 .
the number @xmath136 is the _ scroll invariant _ of @xmath210 , and the number @xmath237 is the _ maroni invariant _ of @xmath210 . if we regard @xmath210 as a curve on @xmath36 , the system @xmath235 gives the canonical system of @xmath210 , and @xmath124 gives the @xmath238 of @xmath210 .
then @xmath210 belongs to @xmath239 for @xmath240 .
the genus formula derives the relation @xmath241 .
conversely , for a smooth curve @xmath210 on @xmath36 with @xmath242 , the linear system @xmath235 with @xmath243 is identified with @xmath244 by restriction , and the curve @xmath245 in @xmath246 is a canonical model of @xmath210 which is contained in the scroll @xmath247 .
for @xmath234 we denote by @xmath248 the locus of trigonal curves of scroll invariant @xmath136 in the moduli space of genus @xmath62 curves . in some literature
, @xmath249 is called a _ maroni locus_. general trigonal curves have scroll invariant @xmath250 or @xmath38 depending on whether @xmath62 is even or odd .
the above facts infer the following . [ moduli of trigonal ] for @xmath234 and @xmath108 the space @xmath249 is naturally birational to a rational quotient @xmath251 , where @xmath252 and @xmath26 is the identity component of @xmath130 . the hirzebruch surface @xmath36 has a natural coordinate system .
let @xmath253 $ ] be the homogeneous coordinate of @xmath111 .
let @xmath254 and @xmath255 be open sets of @xmath111 .
we fix the original splitting @xmath256 and denote @xmath257 . an open covering @xmath258 of @xmath36
is defined by @xmath259 let @xmath260 be the section given by the constant function @xmath38 .
let @xmath261 be the sections given by @xmath262 and @xmath263 respectively .
note that @xmath264 where we regard @xmath265 .
we shall use @xmath266 as a local frame of @xmath109 over @xmath267 .
then we have an isomorphism @xmath268 ( resp .
@xmath269 ) given by @xmath270 , \ ; { { \mathbb{c}}}(a\mathbf{1}+bs_0 ) ) \mapsto ( y^{-1}x , b^{-1}a ) \qquad ( \textrm{resp . } \ : \ : \mapsto ( y^{-1}x , a^{-1}b)),\ ] ] and an isomorphism @xmath271 ( resp .
@xmath272 ) given by @xmath270 , \ ; { { \mathbb{c}}}(a\mathbf{1}+cs_1 ) ) \mapsto ( x^{-1}y , c^{-1}a ) \qquad ( \textrm{resp . }
: \mapsto ( x^{-1}y , a^{-1}c)).\ ] ] thus the open sets @xmath273 have coordinates @xmath274 glued by @xmath275 @xmath276 the @xmath114-curve @xmath115 is defined by the equation @xmath277 .
let us describe some of the @xmath130-action on @xmath36 in terms of these coordinates .
the elements @xmath278 leave @xmath279 invariant . if @xmath280 is written as @xmath281 , then @xmath282 acts by @xmath283 the group @xmath284 acts on @xmath36 via the @xmath284-linearization of @xmath109 .
then the elements @xmath285 and @xmath286 of @xmath284 act respectively by @xmath287 @xmath288 in , the second @xmath289 is the coordinate in @xmath290 .
this action is also regarded as the rational transformation @xmath291 .
we describe the linear systems @xmath292 in terms of these coordinates .
we may assume @xmath293 , which is equivalent to the condition that @xmath292 has no fixed component .
if @xmath294 , then @xmath295 is a curve on @xmath296 and thus defined by the equation @xmath297 for a polynomial @xmath298 .
[ def eqn in hirze ] for @xmath293 the linear system @xmath292 is identified by restriction to @xmath296 with the projectivization of the vector space @xmath299 of polynomials of @xmath300 .
let @xmath297 be an equation of @xmath301 for a general @xmath294 .
expanding @xmath302 , we have @xmath303 and @xmath304 . by substitution
, the curve @xmath305 on @xmath306 is defined by @xmath307 for some @xmath308 .
putting @xmath309 , we know that @xmath310 .
since @xmath311 should be a polynomial of @xmath312 , we must have @xmath313 .
then the equality @xmath314 concludes the proof .
a defining polynomial @xmath315 in @xmath296 for a curve @xmath316 is transformed into @xmath317 in @xmath306 , into @xmath318 in @xmath279 , and into @xmath319 in @xmath290 .
we review basic theory of 2-elementary @xmath0 surfaces following @xcite and @xcite .
let @xmath4 be a 2-elementary @xmath0 surface , namely @xmath1 is a complex @xmath0 surface and @xmath2 is a non - symplectic involution on @xmath1 .
the presence of @xmath2 implies that @xmath1 is algebraic .
we denote by @xmath320 the lattice of cohomology classes @xmath321 with @xmath322 .
then @xmath323 is the orthogonal complement of @xmath324 and contained in the nron - severi lattice @xmath325 .
if @xmath326 is the rank of @xmath323 , then @xmath323 and @xmath324 have signature @xmath327 and @xmath7 respectively .
let @xmath328 be the dual lattice of @xmath329 .
the discriminant form of @xmath329 is the finite quadratic form @xmath330 where @xmath331 and @xmath332 is induced by the quadratic form on @xmath328 .
we have a canonical isometry @xmath333 .
the abelian group @xmath334 is 2-elementary , namely @xmath335 for some @xmath336 .
the parity @xmath9 of @xmath337 is defined by @xmath338 if @xmath339 , and @xmath41 otherwise .
the triplet @xmath5 is called the _ main invariant _ of the lattice @xmath323 , and also of the 2-elementary @xmath0 surface @xmath4 . by @xcite , the isometry class of @xmath329 is uniquely determined by @xmath5 . [ compute main inv ]
let @xmath340 be the fixed locus of @xmath2 .
@xmath341 if @xmath342 , then @xmath343 .
@xmath344 if @xmath345 , then @xmath46 is a union of two elliptic curves .
@xmath346 in other cases , @xmath46 is decomposed as @xmath347 such that @xmath49 is a genus @xmath62 curve and @xmath348 are rational curves with @xmath349 one has @xmath338 if and only if the class of @xmath46 is divisible by @xmath45 in @xmath323 .
[ classify 2-ele k3 ] the deformation type of a 2-elementary @xmath0 surface @xmath4 is determined by its main invariant @xmath5 .
all possible main invariants of 2-elementary @xmath0 surfaces are shown on the following figure [ nikulin table ] ( which is identical to the table in @xcite page 31 ) .
, width=359 ] a moduli space of 2-elementary @xmath0 surfaces of type @xmath5 is constructed as follows .
let @xmath6 be an even lattice of signature @xmath7 whose discriminant form is 2-elementary of length @xmath8 and parity @xmath9 .
the orthogonal group @xmath350 of @xmath6 acts on the domain @xmath351 the quotient space @xmath352 turns out to be an irreducible , normal , quasi - projective variety of dimension @xmath12 .
the complex analytic divisor @xmath353 , where @xmath9 are @xmath50-vectors in @xmath6 , is the inverse image of an algebraic divisor @xmath354 .
we set @xmath355 for a 2-elementary @xmath0 surface @xmath4 of type @xmath5 we have an isometry @xmath356 .
then @xmath357 is contained in @xmath358 .
the period of @xmath4 is defined by @xmath359 \in { { \mathcal{m}_{r , a,\delta}}},\ ] ] which is independent of the choice of @xmath360 .
[ thm : moduli ] the variety @xmath10 is a moduli space of 2-elementary @xmath0 surfaces of type @xmath5 in the sense that , for a family @xmath361 of such 2-elementary @xmath0 surfaces the period map @xmath362 , @xmath363 , is a morphism of varieties , and via the period mapping the points of @xmath10 bijectively correspond to the isomorphism classes of such 2-elementary @xmath0 surfaces .
let @xmath56 be the moduli of genus @xmath62 curves .
when @xmath364 , setting @xmath48 , we have the _ fixed curve map _ @xmath365 where @xmath49 is the genus @xmath62 component of @xmath46 . in this article
we will determine the generic structure of @xmath53 in terms of @xmath56 for @xmath52 .
for example , one will find that * @xmath53 is generically injective for @xmath366 and for @xmath367 , @xmath368 . *
the members of @xmath369 have clifford index @xmath370 .
if @xmath371 in addition , they have clifford index @xmath372 .
* when @xmath373 , we have @xmath338 if and only if the generic member of @xmath369 possesses a theta characteristic of projective dimension @xmath374 .
we shall explain a generalized double cover construction of 2-elementary @xmath0 surfaces .
recall from @xcite that a _ dpn pair _ is a pair @xmath375 of a smooth rational surface @xmath29 and a bi - anticanonical curve @xmath376 with only a - d - e singularities .
when @xmath28 is smooth , @xmath375 is called a _ right dpn pair_.
2-elementary @xmath0 surfaces @xmath4 with @xmath377 are in canonical correspondence with right dpn pairs : for such an @xmath4 the quotient @xmath378 is a smooth rational surface , and the branch curve @xmath28 of the quotient map @xmath379 is a @xmath27-curve isomorphic to @xmath46 .
conversely , for a right dpn pair @xmath375 the double cover @xmath380 branched along @xmath28 is a @xmath0 surface , with the covering transformation being a non - symplectic involution . from @xmath28 one
knows the invariant @xmath381 of @xmath1 via proposition [ compute main inv ] .
also one has @xmath382 . the lattice @xmath323 is generated by the sublattice @xmath383 and the classes of components of @xmath46 ( cf .
@xcite ) . by @xcite , if @xmath384 is the irreducible decomposition of @xmath28 , then @xmath4 has parity @xmath338 if and only if @xmath385 for some @xmath386 .
let @xmath375 be a dpn pair .
right resolution _ of @xmath375 is a triplet @xmath387 such that @xmath388 is a right dpn pair and @xmath389 is a birational morphism with @xmath390 .
when @xmath29 is obvious from the context , we also call it a right resolution of @xmath28 .
a right resolution exists and is unique up to isomorphism .
it may be constructed explicitly as follows ( @xcite ) .
let @xmath391 be the blow - ups defined inductively by @xmath392 being the blow - up at the singular points of @xmath393 , and @xmath394 where @xmath395 is the strict transform of @xmath393 and @xmath396 are the @xmath397-curves over the triple points @xmath171 of @xmath393 .
each @xmath398 is also a dpn pair .
this process will terminate and we finally obtain a right dpn pair @xmath399 . let @xmath171 be a singular point of @xmath28 . according to the type of singularity ,
the dual graph of the curves on @xmath400 contracted to @xmath171 is as follows .
( 100,25)[graph a_n ] ( 10,-5)@xmath401 ( 60,0 ) ( 80,0 ) ( 100,0 ) ( 140,0 ) ( 64,0)(1,0)12 ( 84,0)(1,0)12 ( 100,-5)(40,10)@xmath402 ( 64,-11)1 ( 84,-11)2 ( 104,-11)3 ( 144,-11)@xmath403 $ ] ( 100,45)[graph d_2n ] ( 10,-5)@xmath404 ( 64,16 ) ( 64,-16 ) ( 80,0 ) ( 80,0 ) ( 100,0 ) ( 140,0 ) ( 140,0 ) ( 160,0 ) ( 77,3)(-1,1)10 ( 77,-3)(-1,-1)10 ( 84,0)(1,0)12 ( 100,-5)(40,10)@xmath405 ( 144,0)(1,0)12 ( 70,15)1 ( 70,-25)2 ( 84,-11)3 ( 104,-11)4 ( 164,-11)@xmath406 ( 100,60)[graph d_2n+1 ] ( 10,-5)@xmath407 ( 64,16 ) ( 80,0 ) ( 100,0 ) ( 100,0 ) ( 120,0 ) ( 160,0 ) ( 160,0 ) ( 180,0 ) ( 77,3)(-1,1)10 ( 84,0)(1,0)12 ( 104,0)(1,0)12 ( 120,-5)(40,10)@xmath405 ( 164,0)(1,0)12 ( 70,15)1 ( 84,-11)2 ( 104,-11)3 ( 124,-11)4 ( 184,-11)@xmath406 ( 100,50)[graph e_6 ] ( 10,-5)@xmath408 ( 60,0 ) ( 80,0 ) ( 100,0 ) ( 100,20 ) ( 100,20 ) ( 64,0)(1,0)12 ( 84,0)(1,0)12 ( 100,4)(0,1)12 ( 64,-11)1 ( 84,-11)2 ( 104,-11)3 ( 104,11)4 ( 100,45)[graph e_7 ] ( 10,-5)@xmath409 ( 60,0 ) ( 80,0 ) ( 80,0 ) ( 100,0 ) ( 120,0 ) ( 120,0 ) ( 120,20 ) ( 140,0 ) ( 160,0 ) ( 160,0 ) ( 64,0)(1,0)12 ( 84,0)(1,0)12 ( 104,0)(1,0)12 ( 120,4)(0,1)12 ( 124,0)(1,0)12 ( 144,0)(1,0)12 ( 64,-11)1 ( 84,-11)2 ( 104,-11)3 ( 124,-11)4 ( 124,11)5 ( 144,-11)6 ( 164,-11)7 ( 100,50)[graph e_8 ] ( 10,-5)@xmath410 ( 60,0 ) ( 60,0 ) ( 80,0 ) ( 100,0 ) ( 100,0 ) ( 120,0 ) ( 140,0 ) ( 140,0 ) ( 140,20 ) ( 160,0 ) ( 180,0 ) ( 180,0 ) ( 64,0)(1,0)12 ( 84,0)(1,0)12 ( 104,0)(1,0)12 ( 124,0)(1,0)12 ( 140,4)(0,1)12 ( 144,0)(1,0)12 ( 164,0)(1,0)12 ( 64,-11)1 ( 84,-11)2 ( 104,-11)3 ( 124,-11)4 ( 144,-11)5 ( 144,11)6 ( 164,-11)7 ( 184,-11)8 here black vertices represent @xmath50-curves , white vertices represent @xmath397-curves , and double circles represent @xmath411-curves .
the @xmath411-curves are components of @xmath412 , while the @xmath50-curves are disjoint from @xmath412 .
the @xmath397-curves intersect with @xmath412 transversely at two points unless @xmath171 is @xmath413-type ; when @xmath171 is an @xmath413-point , the @xmath397-curve is tangent to @xmath412 at one point .
the labeling for the vertices will be used later . note
that identification of the dual graph of the curves with the above abstract graph is not unique when @xmath171 is @xmath404-type . in that case ,
such an identification is obtained after one distinguishes the three branches ( resp .
two tangential branches ) of @xmath28 at @xmath171 when @xmath414 ( resp .
@xmath415 ) .
let @xmath375 be a dpn pair with a right resolution @xmath387 .
taking the double cover of @xmath400 branched over @xmath412 , we associate a 2-elementary @xmath0 surface @xmath4 to @xmath375 .
the composition @xmath379 of the quotient map @xmath416 and the blow - down @xmath117 is called the _
right covering map _ for @xmath375 .
note that @xmath4 is also the minimal desingularization of the double cover of @xmath29 branched over @xmath28 . by the above dual graphs , the invariant @xmath381 of @xmath4 is calculated in terms of @xmath375 as follows .
let @xmath417 ( resp .
@xmath418 , @xmath419 ) be the number of singularities of @xmath28 of type @xmath401 ( resp .
@xmath420 , @xmath421 ) .
then @xmath422 is given by @xmath423 if @xmath424 is the number of components of @xmath28 , the number @xmath425 of components of @xmath426 is given by @xmath427 the genus @xmath62 of the main component of @xmath46 is the maximal geometric genus of components of @xmath28 .
for the parity @xmath9 we will later use the following criteria .
[ delta=1 ] let @xmath4 and @xmath375 be as above .
then @xmath4 has parity @xmath41 if we have distinct irreducible components @xmath393 of @xmath28 with either of the following conditions .
@xmath428 @xmath429 contains a node of @xmath28 for every @xmath430 .
@xmath431 @xmath432 contains a node and a @xmath433-point of @xmath28 .
@xmath434 @xmath432 contains a node and a @xmath404-point of @xmath28 , in the latter of which @xmath435 and @xmath436 share a tangent direction .
@xmath437 @xmath435 has a node which is also a node of @xmath28 .
@xmath428 let @xmath438 be a node of @xmath28 contained in @xmath439 with @xmath440 .
let @xmath441 be the @xmath50-curve over @xmath438 .
it suffices to show that the @xmath442-cycle @xmath443 belongs to @xmath444 . if @xmath445 is the component of @xmath46 over @xmath393 , then @xmath446 for every @xmath447 .
we have @xmath448 for other components
@xmath53 of @xmath46 .
since @xmath449 is the pullback of a divisor on @xmath450 , @xmath451 has integral intersection pairing with the pullbacks of divisors on @xmath450 .
this proves the assertion .
@xmath431 let @xmath452 be the blow - up of @xmath29 at the @xmath433-point , @xmath453 the exceptional curve , and @xmath395 ( resp .
@xmath454 ) the strict transform of @xmath393 ( resp .
@xmath28 ) in @xmath452 .
then one may apply the case @xmath428 to the dpn pair @xmath455 and the components @xmath456 of @xmath457 .
@xmath434 blow - up @xmath29 at the @xmath404-singularity and use the induction on @xmath136 : the assertion is reduced to the case @xmath431 .
if @xmath458 is the @xmath50-curve over that node , we have @xmath459 . in certain cases ,
the right covering map @xmath380 may be recovered from a line bundle on @xmath1 .
[ covering map & lb ] let @xmath4 and @xmath375 be as above and suppose that @xmath29 is either @xmath35 or @xmath18 or @xmath36 with @xmath129 .
let @xmath460 be @xmath461 , @xmath462 , @xmath463 for respective case .
then the map @xmath464 is isomorphic . the bundle @xmath465 is nef of degree @xmath466 so that @xmath467 .
hence the assertion follows from the coincidence @xmath468 . by this lemma
, the morphism @xmath469 is identified with the composition @xmath470 .
the morphism @xmath471 is an embedding when @xmath472 or @xmath18 .
if @xmath473 with @xmath129 , @xmath471 contracts the @xmath114-curve @xmath115 .
therefore , for @xmath474 , one may recover the morphism @xmath380 from the bundle @xmath465 by desingularizing the surface @xmath475 .
for @xmath476 , @xmath477 is the blow - down of the @xmath397-curve @xmath115 , and one may recover @xmath478 from @xmath465 if one could identify the point @xmath479 in @xmath35 . in the rest of this subsection
we prepare some auxiliary results under the following assumption ( which is necessary if one wants to obtain general members of @xmath10 ) .
[ genericity assumption ] the singularities of @xmath28 are only of type @xmath480 .
for such a singularity @xmath171 of @xmath28 , the irreducible curves on @xmath400 contracted to @xmath171 are only @xmath411-curves and @xmath397-curves transverse to @xmath412 .
the reduced preimages of those curves in @xmath1 are @xmath2-invariant @xmath50-curves , and their dual graph is the dynkin graph of same type as the singularity of @xmath171 .
we denote by @xmath481 the generated root lattice .
let @xmath482 be the irreducible decomposition of @xmath28 , and let @xmath445 be the component of @xmath46 with @xmath483 .
[ l_+ and a - d - e ] the lattice @xmath323 is generated by the sublattice @xmath484 , where @xmath171 are the singularities of @xmath28 , and the classes of @xmath445 , @xmath485 .
let @xmath486 be the quotient map by @xmath2 . by the construction of @xmath487 ,
the lattice @xmath488 is contained in @xmath484 . also the components of @xmath46 other than @xmath489 are contracted by @xmath478 to the triple points of @xmath28 , so that their classes are contained in @xmath490 .
we shall construct an ample divisor class on @xmath1 using the objects in lemma [ l_+ and a - d - e ] .
let @xmath171 be a triple point of @xmath28 . choose an identification of the dual graph of the exceptional curves over @xmath171 with an abstract graph presented in figure [ dual graph ] . via the labeling given for the latter , we denote by @xmath491 the @xmath50-curves generating @xmath492 .
then we define a divisor @xmath493 on @xmath1 by @xmath494 when @xmath171 is @xmath404-type ; @xmath495 when @xmath171 is @xmath409-type ; and @xmath496 when @xmath171 is @xmath410-type .
this divisor is independent of the choice of an identification of the graphs .
[ ample class ] for an arbitrary ample class @xmath497 the divisor class @xmath498 where @xmath171 are the triple points of @xmath28 , is an @xmath2-invariant ample class on @xmath1 .
the class @xmath499 is the pullback of a divisor class @xmath500 on @xmath400 .
note the bounds @xmath501 and @xmath502 and apply the nakai criterion to @xmath500 .
by lemmas [ l_+ and a - d - e ] and [ ample class ] we have a basis and a polarization of the lattice @xmath323 defined explicitly in terms of @xmath375 .
let @xmath29 be one of the following rational surfaces : @xmath503 suppose that we are given an irreducible , @xmath37-invariant locus @xmath504 such that @xmath341 every member @xmath505 satisfies condition [ genericity assumption ] , @xmath344 the singularities of @xmath506 of each type form an etale cover of @xmath25 , and @xmath346 the number of components of @xmath506 is constant
. then the 2-elementary @xmath0 surfaces associated to the dpn pairs @xmath507 have constant main invariant @xmath5 , and we obtain a period map @xmath508 . since @xmath171 is @xmath37-invariant , it descends to a rational map @xmath509 here @xmath510 stands for a rational quotient ( see [ sec : invariant theory ] ) . in this section
we try to explain how to calculate the degree of @xmath30 .
such calculations have been done in some cases ( e.g. , @xcite , @xcite , @xcite ) , and the method to be explained is more or less a systemization of them .
roughly speaking , @xmath511 expresses for a general @xmath47 how many contractions @xmath512 exist by which the branch of @xmath513 is mapped to a member of @xmath25 .
below we exhibit a recipe of calculation supplemented by few typical examples .
the recipe will be applied in the rest of the article to about fifty main invariants @xmath5 . since the materials are diverse , it seems hard to formulate those calculations into some general proposition .
instead , we shall give sufficient instruction for each @xmath5 and then leave the detail to the reader by referring to the recipe below , which we believe is not difficult to master .
we use a certain cover of @xmath10 .
let @xmath6 be the 2-elementary lattice of signature @xmath7 used in the definition of @xmath10 , and let @xmath514 be the group of isometries of @xmath6 which act trivially on the discriminant group @xmath515 .
let @xmath516 be the arithmetic quotient @xmath517 , which is irreducible for @xmath514 has an element exchanging the two components of @xmath358 .
the natural projection @xmath518 is a galois covering .
the galois group is the orthogonal group @xmath519 of the discriminant form , for the homomorphism @xmath520 is surjective ( @xcite ) and @xmath521 acts trivially on @xmath515 . in particular , we have @xmath522 since @xmath515 is 2-elementary , one may calculate @xmath523 by using @xcite corollary 2.4 and lemma 2.5 .
we shall use standard notation for orthogonal / symplectic groups in characteristic @xmath45 : @xmath524 , @xmath525 , and @xmath526 ( see @xcite , @xcite ) .
the cover @xmath516 is birationally a moduli of 2-elementary @xmath0 surfaces with lattice - marking .
we fix a primitive embedding @xmath527 where @xmath528 , an even hyperbolic 2-elementary lattice @xmath529 of main invariant @xmath5 , and an isometry @xmath530 .
suppose that we have a 2-elementary @xmath0 surface @xmath47 and a lattice isometry @xmath531 . by nikulin
s theory @xcite , @xmath532 can be extended to an isometry @xmath533
. then @xmath534 belongs to @xmath358 , and we define the period of @xmath535 by @xmath536\in{{\widetilde{\mathcal{m}}_{r , a,\delta}}}$ ] .
since the restriction @xmath537 is fixed , this definition does not depend on the choice of @xmath360 .
two such objects @xmath535 and @xmath538 have the same period in @xmath516 if and only if there exists a hodge isometry @xmath539 with @xmath540 .
the open set of @xmath516 over @xmath10 parametrizes such periods of lattice - marked 2-elementary @xmath0 surfaces . in order to calculate @xmath541
, we construct a generically injective lift of @xmath30 , @xmath542 where @xmath543 is a certain cover of @xmath25 and @xmath26 is the identity component of @xmath37 .
then we compare the degree of the projection @xmath544 with @xmath523 .
here is a more precise procedure . 1 .
we define a cover @xmath545 in which @xmath543 parametrizes pairs @xmath546 where @xmath505 and @xmath547 is a reasonable " labeling of the singularities , the branches at the @xmath404-singularities , and the components of @xmath506 .
2 . let @xmath548 be the 2-elementary @xmath0 surface associated to @xmath507 .
by lemma [ l_+ and a - d - e ] , the labeling @xmath547 and a natural basis of @xmath549 induce a lattice marking @xmath532 of @xmath323 .
actually , lemma [ l_+ and a - d - e ] implies an appropriate definition of the reference lattice @xmath529 , and then @xmath532 should be obtained naturally . considering the period of @xmath535
as defined above , we obtain a lift @xmath550 of @xmath171 . by borel
s extension theorem ( @xcite ) , @xmath551 is a morphism of varieties .
one observes that the group @xmath26 acts on @xmath543 and that @xmath551 is @xmath26-invariant .
hence @xmath551 induces a rational map @xmath552 , which is a lift of @xmath30 .
we prove that @xmath553 is generically injective .
for that it suffices to show that the @xmath551-fibers are @xmath26-orbits .
if @xmath554 , we have a hodge isometry @xmath555 with @xmath556 for the associated @xmath535 and @xmath538 . by lemma [ ample class ] , @xmath360 turns out to preserve the ample cones .
then we obtain an isomorphism @xmath557 with @xmath558 by the torelli theorem .
let @xmath380 , @xmath559 be the right covering maps and let @xmath560 be the bundle as in lemma [ covering map & lb ] .
since @xmath561 , by lemma [ covering map & lb ] we obtain an automorphism @xmath209 of @xmath29 with @xmath562 .
this will imply that @xmath563 .
since @xmath209 acts trivially on @xmath549 , we have @xmath564 . 5 .
now assume that @xmath565 . since @xmath516 is irreducible , then @xmath553 is birational
. therefore @xmath511 is equal to @xmath523 divided by the degree of @xmath544 .
the latter may be calculated geometrically . in the above recipe the construction of @xmath543 and @xmath553
is left rather ambiguous .
it could be formulated generally using monodromies for the universal curve over @xmath25 .
but in order to give an effective account , we find it better to describe it by typical examples .
( when @xmath566 , the idea can also be found in @xcite . )
[ ex:1 ] we consider curves on @xmath567 .
let @xmath568 be the open set of pairs @xmath569 such that @xmath210 and @xmath174 are smooth and transverse to each other .
the space @xmath25 parametrizes the nodal @xmath570-curves @xmath571 , and we obtain a period map @xmath572 . in this case , the cover @xmath543 should be @xmath573 which parametrizes the curves @xmath571 endowed with a labeling of its six nodes @xmath574 .
the projection @xmath545 is an @xmath575-covering .
we shall prepare the reference lattice @xmath529 .
let @xmath576 be a natural orthogonal basis of the lattice @xmath577 and let @xmath578 be a natural basis of the lattice @xmath579 with @xmath580 , @xmath581 .
we define the vectors @xmath582 by @xmath583 and @xmath584 .
then the overlattice @xmath585 is even and 2-elementary of main invariant @xmath586 . for a @xmath587 ,
we let @xmath588 and @xmath589 be the right covering map .
the fixed curve @xmath590 is decomposed as @xmath591 such that @xmath592 and @xmath593 .
then we define an embedding @xmath531 of lattices by @xmath594 $ ] , @xmath595 $ ] , @xmath596 $ ] , and @xmath597 $ ] . by lemma [ l_+ and a - d - e ] we have @xmath598 .
this gives a lattice - marked 2-elementary @xmath0 surface @xmath535 , and we obtain a period map @xmath599 . the morphism @xmath551 is _ not _
@xmath600-invariant for @xmath600 may exchange @xmath601 and @xmath602 .
rather @xmath551 is invariant under the identity component @xmath603 of @xmath600 .
we prove that the @xmath551-fibers are @xmath26-orbits .
if two points @xmath604 and @xmath605 of @xmath543 have the same @xmath551-period , we have a hodge isometry @xmath555 with @xmath556 for the associated @xmath535 and @xmath538 .
in particular , @xmath360 maps @xmath606 , [ ( f')^{-1}(p_i ) ] , [ f_j']$ ] respectively to @xmath607 , [ f^{-1}(p_i ) ] , [ f_j]$ ] . by lemma [ ample class ] @xmath360 preserves the ample cones .
hence by the torelli theorem we obtain an isomorphism @xmath557 with @xmath558 .
then we have an automorphism @xmath209 of @xmath608 with @xmath609 by lemma [ covering map & lb ] . considering the branch loci of @xmath478 and @xmath610 , we have @xmath611 . also we have @xmath612 because @xmath613 . since @xmath209 leaves the two rulings of @xmath608 invariant
, we have @xmath564 .
this concludes that the @xmath551-fibers are @xmath26-orbits . in view of the equality @xmath614
, @xmath551 induces a birational lift @xmath615 of @xmath30 . since @xmath616
, we have @xmath617 by @xcite . on the other hand ,
the projection @xmath618 has degree @xmath619\cdot|\frak{s}_6|=2\cdot6!$ ] because @xmath600 acts on @xmath25 almost freely .
therefore @xmath30 is birational .
[ ex:2 ] we consider curves on @xmath620 . keeping the notation of [ sec : hirze ] , we let @xmath621 be the locus of pairs @xmath622 such that @xmath210 is smooth , transverse to @xmath53 and @xmath115 respectively , and passes through the point @xmath623 .
the locus @xmath25 parametrizes the @xmath624-curves @xmath625 , whose singularities are the @xmath433-point @xmath171 and the three nodes @xmath626 .
hence we obtain a period map @xmath627 . in this case , @xmath543 should be the double cover @xmath628 in @xmath543 only the two nodes @xmath629 are labelled , and the rest two singularities are left unmarked .
however , the node @xmath630 is distinguished from @xmath629 by the irreducible decomposition of @xmath625 , and the @xmath433-point @xmath171 is evidently distinguished from the nodes .
thus @xmath543 actually parametrizes the curves @xmath625 endowed with a complete and reasonable labeling of the singularities .
also the components of @xmath625 are distinguished by their classes in @xmath631 , and this distinguishes the three branches of @xmath625 at @xmath171 .
let us prepare the lattice @xmath529 .
let @xmath632 and @xmath633 be natural basis of the lattices @xmath634 and @xmath635 respectively .
we denote the root basis of the @xmath433-lattice by @xmath636 where @xmath637 and @xmath638 .
we put @xmath639 .
then we define the vectors @xmath640 by @xmath641 , @xmath642 , and @xmath643 .
the overlattice @xmath644 is even and 2-elementary of main invariant @xmath645 . for @xmath646 ,
let @xmath647 and @xmath648 be the right covering map .
we denote @xmath623 as above .
on @xmath1 we define line bundles and @xmath50-curves by @xmath649 , @xmath650 , @xmath651 for @xmath652 , and @xmath653 . by [ ssec : right resol ] , the four @xmath50-curves on @xmath1 contracted by @xmath478 to @xmath171 form a dynkin graph of type @xmath433 .
we denote by @xmath654 those over the infinitely near points of @xmath171 given by @xmath655 respectively .
the remaining one , denoted by @xmath656 , is a component of @xmath46 .
we have an embedding @xmath657 by @xmath658 , e\to[e ] , e_i\mapsto[e_i]$ ] , and @xmath659 $ ] .
the fixed curve @xmath46 is decomposed as @xmath660 such that @xmath592 , @xmath661 , and @xmath662 .
then @xmath663 extends to an isometry @xmath531 by sending @xmath664 $ ] .
thus we associate a lattice - marked 2-elementary @xmath0 surface @xmath535 , which defines a period map @xmath665 .
since @xmath666 acts trivially on @xmath631 , the morphism @xmath551 is @xmath666-invariant .
we show that the @xmath551-fibers are @xmath666-orbits . indeed ,
if @xmath667 , we will obtain an isomorphism @xmath557 with @xmath668 for the associated @xmath535 and @xmath538 , as in the previous example .
let @xmath648 and @xmath669 be the respective right covering maps .
we fix a contraction @xmath670 of @xmath115 .
since @xmath671 , by lemma [ covering map & lb ] we obtain an automorphism @xmath209 of @xmath35 such that @xmath672 .
the point is that @xmath209 fixes @xmath673 , which is the unique @xmath674-singularity of the branch curves of both @xmath675 and @xmath676 .
therefore @xmath209 lifts to an automorphism @xmath677 of @xmath620 with @xmath678 .
the rest of the proof is similar to the previous example .
since @xmath679 , @xmath551 descends to a birational lift @xmath680 of @xmath30 .
the projection @xmath681 is a double covering . on the other hand , since @xmath682 , we have @xmath683
. therefore @xmath30 is birational .
[ ex:3 ] we consider curves on @xmath684 .
we keep the notation in [ sec : hirze ] .
let @xmath685 be the open locus of pairs @xmath686 such that @xmath210 and @xmath451 are smooth and transverse to each other .
we consider the @xmath687-curves @xmath688 , which have the eight nodes @xmath689 and no other singularity .
the associated 2-elementary @xmath0 surfaces @xmath4 have parity @xmath338 because the strict transforms of @xmath690 in @xmath450 belong to @xmath691 .
thus we obtain a period map @xmath692 . in this case
, we should label the six nodes @xmath693 taking into account the irreducible decomposition of @xmath451 .
let @xmath694 be the locus of those @xmath695 such that @xmath696 and that @xmath697 belong to a same component of @xmath451 .
accordingly , we may distinguish the two components of @xmath451 as @xmath698 such that @xmath699 is the component through @xmath700 .
the rest two nodes of @xmath688 , @xmath701 , are automatically labelled as @xmath702 . in this way
the nodes and components of @xmath688 are labelled compatibly .
the natural projection @xmath545 is an @xmath703-covering .
we prepare the reference lattice @xmath529 as follows .
let @xmath578 and @xmath704 be natural basis of the lattices @xmath579 and @xmath705 respectively .
we define vectors @xmath706 by @xmath707 , @xmath708 , and @xmath709 .
then the overlattice @xmath710 is even and 2-elementary of main invariant @xmath711 . for @xmath712 ,
let @xmath713 and @xmath714 be the natural projection .
the fixed curve @xmath46 is decomposed as @xmath715 such that @xmath592 , @xmath716 , and @xmath717 .
the labelling @xmath718 induces a lattice - marking @xmath531 by @xmath719 $ ] , @xmath720 $ ] for @xmath721 , @xmath722 $ ] for @xmath723 , and @xmath724 $ ] .
thus we obtain a lift @xmath725 of @xmath30 .
one may recover the morphism @xmath478 from the class @xmath726 by lemma [ covering map & lb ] ( desingularize the quadratic cone ) , the points @xmath700 from @xmath727 , and the curve @xmath688 from @xmath478 as the branch locus .
as before , these imply that @xmath553 is generically injective .
since @xmath728 has the same dimension @xmath729 as @xmath730 , @xmath553 is actually birational . since @xmath731
, we have @xmath732 by @xcite . on the other hand ,
the projection @xmath733 has degree @xmath734 .
this concludes that @xmath30 has degree @xmath38 .
[ ex:4 ] let @xmath735 be the locus of triplets @xmath736 such that @xmath210 is smooth , @xmath737 is tangent to @xmath210 with multiplicity @xmath738 , and @xmath739 is transverse to @xmath210 and passes through @xmath740 .
the sextics @xmath741 have a @xmath742-singularity at @xmath743 , three nodes at @xmath744 , and no other singularity .
the associated 2-elementary @xmath0 surfaces have invariant @xmath745 , and thus we obtain a period map @xmath746 . in this case
we label the three nodes @xmath744 by considering the locus @xmath747 of those @xmath748 such that @xmath749 .
the rest singularity @xmath743 of @xmath741 is apparently distinguished from those nodes .
moreover , the two lines @xmath750 are distinguished by their intersection with @xmath210 , and the three branches of @xmath741 at @xmath743 are distinguished by the irreducible decomposition of @xmath741 .
thus , by @xmath543 the relevant datum for @xmath741 are completely labelled .
we prepare the reference lattice @xmath529 .
let @xmath751 and @xmath633 be natural basis of the lattices @xmath752 and @xmath635 respectively .
let @xmath753 be the root basis of the @xmath742-lattice whose numbering corresponds to the one given in the graph in figure [ dual graph ] .
we define @xmath754 , @xmath721 , inductively by @xmath755 .
then @xmath756 form an orthogonal basis of a sublattice of @xmath742 isometric to @xmath757 .
now we define vectors @xmath758 by @xmath759 , @xmath760 , and @xmath761 .
the overlattice @xmath762 is even and 2-elementary of main invariant @xmath763 . for @xmath764 ,
let @xmath765 and @xmath766 be the natural projection .
let @xmath767 be the sublattice of @xmath323 generated by the @xmath50-curves contracted by @xmath478 to the @xmath742-point @xmath743 .
assigning @xmath768 and @xmath769 respectively to the branches of @xmath210 and @xmath737 at @xmath743 , we have a canonical isometry @xmath770 which maps the root basis to the classes of @xmath50-curves .
then the labelling @xmath771 determines an embedding @xmath772 by @xmath773 $ ] and @xmath774 $ ] .
the fixed curve @xmath46 is decomposed as @xmath775 such that @xmath776 , @xmath777 for @xmath778 , and @xmath779 $ ] for @xmath780 .
the assignment @xmath781 $ ] extends the embedding @xmath663 to an isometry @xmath531 . considering the period of @xmath535
, we obtain a lift @xmath782 of @xmath30 . in this construction
, one may recover the morphism @xmath478 from the class @xmath783 by lemma [ covering map & lb ] , the points @xmath438 from @xmath784 , and the sextic @xmath741 as the branch curve of @xmath478 . checking that
@xmath785 , we deduce that @xmath553 is birational as before .
since @xmath786 , we have @xmath787 .
hence the two projections @xmath788 and @xmath789 are both @xmath790-coverings , so that @xmath30 is birational .
in this section we prove that the spaces @xmath792 with @xmath793 are rational .
the quotient @xmath378 of a general @xmath794 is a del pezzo surface of degree @xmath795 .
let @xmath796 be the moduli of del pezzo surfaces of degree @xmath797 ( we exclude @xmath18 ) . by the correspondence between 2-elementary @xmath0 surfaces and right
dpn pairs , we have a fibration @xmath798 whose fiber over @xmath29 is birational to @xmath799 . when @xmath366 , @xmath29 has no moduli so that @xmath800 .
the rationality of @xmath792 is reduced to the @xmath37-representation @xmath801 . by contrast , when @xmath802 , @xmath803 has positive dimension and @xmath37 is a small finite group .
we then analyze the fibration @xmath798 or the fixed curve map @xmath804 to deduce the rationality .
let @xmath29 be the blow - up of @xmath35 at one point @xmath806 . by associating to a smooth curve
@xmath807 the double cover of @xmath29 branched over @xmath28 , we have a birational map @xmath808 indeed , @xmath30 is generically injective by the correspondence between 2-elementary @xmath0 surfaces and right dpn pairs , and is dominant because @xmath799 has dimension @xmath809 .
one may identify @xmath810 with the linear system of plane sextics singular at @xmath171 , and @xmath37 with the stabilizer of @xmath171 in @xmath811 .
[ ( 2,2,1 ) ] the quotient @xmath799 is rational .
therefore @xmath805 is rational .
let @xmath812 be the @xmath397-curve .
let @xmath813 be the @xmath37-equivariant map defined by @xmath814 .
the group @xmath37 acts on @xmath815 almost transitively . for two disinct points @xmath816
, the fiber @xmath817 is an open set of the linear system @xmath818 of @xmath27-curves through @xmath819 and @xmath820 . if @xmath821 is the stabilizer of @xmath822 , by the slice method [ slice ] we have @xmath823 .
let @xmath824 be the strict transform of the line passing through @xmath171 with tangent @xmath438 .
set @xmath825 and consider the @xmath26-equivariant map @xmath826 the fiber @xmath827 over a general @xmath828 is an open set of a linear subspace of @xmath818 .
since @xmath27 is @xmath37-linearized , @xmath26 acts on @xmath84 so that @xmath209 is @xmath26-birational to the projectivization of a @xmath26-linearized vector bundle over an open set of @xmath829 .
the group @xmath26 acts on @xmath829 almost freely . indeed ,
for a general @xmath828 the four points @xmath830 on @xmath831 are not projectively equivalent to the four points @xmath832 on @xmath833 , and any nontrivial @xmath834 with @xmath835 does not fix @xmath819 .
clearly an automorphism of @xmath29 acting trivially on @xmath836 must be trivial .
thus we may apply the no - name method to see that @xmath837 .
since @xmath838 , @xmath839 is rational .
note that the natural moduli map @xmath840 is generically injective , for the normalization of a one - nodal plane sextic has only one @xmath841 , the restriction of @xmath842 ( see @xcite appendix a.20 ) .
the composition @xmath843 is nothing but the fixed curve map @xmath844 .
therefore [ fixed curve ( 2,2,1 ) ] the fixed curve map @xmath844 is generically injective with a generic image being the locus of non - hyperelliptic , non - trigonal , non - bielliptic curves possessing @xmath841 .
let @xmath29 be the blow - up of @xmath35 at two distinct points @xmath846 .
as in [ sec:(2,2,1 ) ] , we have a natural birational map @xmath847 by the double cover construction .
one may identify @xmath810 with the linear system of plane sextics singular at @xmath846 , and @xmath37 with the stabilizer of @xmath822 in @xmath811 . in this form ,
casnati and del centina @xcite proved that @xmath799 is rational .
their proof is based on a direct calculation of an invariant field . here
we shall present another simple proof . [ ( 3,3,1 ) ] the quotient @xmath799 is rational .
therefore @xmath845 is rational .
let @xmath848 be the @xmath397-curve over @xmath438 and let @xmath849 .
we consider the map @xmath850 , @xmath851 , which is @xmath37-equivariant .
the identifications @xmath852 show that @xmath37 acts on @xmath829 almost transitively .
if @xmath853 is a general point of @xmath829 , let @xmath854 be the line passing through @xmath438 with tangent @xmath855 .
then the stabilizer @xmath821 of @xmath856 is identified with the group of @xmath857 which preserve @xmath858 and @xmath822 . in particular ,
the fiber @xmath860 is an open set of the linear system @xmath818 of sextics through @xmath861 . by the slice method we have @xmath823 .
the net @xmath862 is a @xmath26-invariant linear subspace of @xmath818 .
since @xmath26 is finite , we can decompose the @xmath26-representation @xmath84 as @xmath863 .
the projection @xmath864 from @xmath865 is a @xmath26-linearized vector bundle . since @xmath26 acts on @xmath866
almost freely , by the no - name method we have @xmath867 .
the quotient @xmath868 is clearly rational . as in corollary
[ fixed curve ( 2,2,1 ) ] , we have the following . [ fixed curve ( 3,3,1 ) ] the fixed curve map @xmath869 is generically injective with a generic image being the locus of non - hyperelliptic , non - trigonal , non - bielliptic curves possessing @xmath841 . let @xmath871 be three linearly independent points in @xmath35 .
the blow - up @xmath29 of @xmath35 at @xmath872 is a sextic del pezzo surface .
as before , we have a birational equivalence @xmath873 .
one may identify @xmath810 with the linear system of plane sextics singular at @xmath872 , and @xmath37 with the group @xmath874 where @xmath26 is the stabilizer of @xmath872 in @xmath811 and @xmath875 is generated by the standard cremona transformation based at @xmath872 . in this form , @xmath799 is proved to be rational by casnati and del centina @xcite . [
( 4,4,1 ) ]
the space @xmath870 is rational .
[ another proof ( 4,4,1 ) ] the proof in @xcite is by a direct calculation of an invariant field .
actually , it is also possible to give a geometric proof as in the previous sections , but not so short . as noted in @xcite , the natural map @xmath876 is generically injective .
this is a consequence of the classical fact that a generic three - nodal plane sextic has exactly two @xmath877 , @xmath842 and its transformation by the cremona map .
thus [ fixed curve ( 4,4,1 ) ] the fixed curve map for @xmath870 is generically injective with a generic image being the locus of non - hyperelliptic , non - trigonal , non - bielliptic curves possessing @xmath877 .
let @xmath29 be a quintic del pezzo surface .
as in the previous sections , we have a birational map @xmath878 by the double cover construction ( see also @xcite ) . shepherd
- barron @xcite proved that @xmath799 is rational .
also it is classically known that the natural map @xmath879 is birational ( cf .
therefore [ rational ( 5,5,1 ) ] the space @xmath21 is rational .
the fixed curve map @xmath880 is birational .
let @xmath882 and let @xmath883 be the grassmannian of pencils of quadrics in @xmath884 .
let @xmath885 be the universal quotient bundle .
the fiber @xmath886 over a pencil @xmath887 is the linear space @xmath888 .
a general pencil @xmath321 defines a smooth @xmath889 complete intersection @xmath890 in @xmath884 , which is an anticanonical model of a quartic del pezzo surface ( cf .
@xcite ) . one has the identification @xmath891 by a general property of complete intersections .
hence a general point of the bundle @xmath892 corresponds to a pair @xmath893 of a quartic del pezzo surface @xmath890 and a @xmath894-curve @xmath28 .
this defines a period map @xmath895 , which descends to a rational map @xmath896 .
since @xmath897 is an anticanonical model , we see that @xmath30 is generically injective .
the equality @xmath898 shows that @xmath30 is birational .
the @xmath894-curve on @xmath890 defined by a general point of @xmath899 is a @xmath900 complete intersection in @xmath884 , which is a canonical genus five curve .
we study @xmath901 from this viewpoint .
let @xmath902 be the grassmannian of nets of quadrics in @xmath884 , and let @xmath903 be the universal sub bundle .
the fiber @xmath904 over a net @xmath905 is the linear subspace @xmath25 of @xmath84 .
the bundle @xmath892 parametrizes pairs @xmath906 of a @xmath45-plane @xmath907 and a line @xmath908 , while the bundle @xmath909 parametrizes pairs @xmath910 of a @xmath911-plane @xmath912 and a 2-plane @xmath913 . these two objects canonically correspond by @xmath914 and @xmath915 .
thus we have a canonical @xmath916-isomorphism @xmath917 .
let @xmath905 be a general point of @xmath902 , and @xmath918 be the @xmath919 curve defined by @xmath920 .
one may identify @xmath921 with @xmath922 , and @xmath923 with the kernel of the natural linear map @xmath924 , which is surjective by m. noether .
let @xmath925 be the universal curve ( over the open locus of curves with no automorphism ) , @xmath926 be the relative canonical bundle for @xmath117 , and @xmath927 be the kernel of the bundle map @xmath928 . since @xmath929 is naturally birational to @xmath930 , the above identification gives rise to the birational equivalence @xmath931 the moduli space @xmath930 is rational by katsylo @xcite .
therefore [ rational ( 6,6,1 ) ] the quotient @xmath932 is rational .
hence @xmath881 is rational .
the following assertion is also obtained . [
fixed curve ( 6,6,1 ) ] the fixed curve map @xmath933 is dominant , with the fiber over a general @xmath934 being birationally identified with @xmath935 where @xmath25 is the kernel of the natural map @xmath924 .
the first bundle structure @xmath936 corresponds to the quotient surface map @xmath937 , @xmath938 , while the second one @xmath939 corresponds to the fixed curve map .
the latter is easier to handle with thanks to the absence of automorphism of general @xmath919 curves .
if @xmath941 is a cubic surface , the restriction map @xmath942 is isomorphic .
hence a general point of @xmath943 corresponds to a pair @xmath375 of a smooth cubic surface @xmath29 and a smooth @xmath27-curve @xmath28 , which gives a 2-elementary @xmath0 surface of type @xmath944 .
this induces a period map @xmath945 . since cubic surfaces
are anticanonically embedded , @xmath30 is generically injective . by the equality @xmath946 , we see that @xmath30 is birational .
[ rational ( 7,7,1 ) ] the quotient @xmath947 is rational .
therefore @xmath940 is rational . for a general @xmath948
the intersection @xmath949 is a canonical @xmath950 curve .
this induces a rational map @xmath951 , which is identified with the fixed curve map @xmath952 .
the @xmath478-fiber over a general @xmath953 is the linear system of cubics containing @xmath28 , for the quadric containing @xmath28 is unique .
therefore , if @xmath954 is the universal curve ( over an open locus ) and @xmath955 is the kernel of the bundle map @xmath956 where @xmath926 is the relative canonical bundle , then we have the birational equivalence @xmath957 the rationality of @xmath958 is proved by shepherd - barron @xcite .
[ fixed curve ( 7,7,1 ) ] the fixed curve map @xmath952 is dominant , with the fiber over a general @xmath28 being birationally identified with the projectivization of the kernel of the natural map @xmath959 .
one may also deduce the rationality of @xmath940 by applying the no - name lemma to the projection @xmath960 and resorting to the rationality of @xmath961 ( cf .
@xcite ) . as before , one may use the fibration @xmath963 , @xmath938 , to reduce the rationality of @xmath962 to that of @xmath964 due to katsylo @xcite .
however , in order to describe the fixed curve map , we here adopt a more roundabout approach . recall that if @xmath29 is a quadric del pezzo surface , its anticanonical map @xmath965 is a double covering branched over a smooth quartic @xmath966 . the correspondence @xmath967 induces a birational map @xmath968 .
the covering transformation @xmath663 of @xmath969 is the _
geiser involution _ of @xmath29 .
its fixed curve @xmath970 belongs to @xmath810 .
let @xmath971 be the line given by @xmath970 and let @xmath972 .
we have the @xmath663-decomposition @xmath973 where @xmath974 .
[ basics of |-2k| ( 8,8,1 ) ] every smooth curve @xmath376 has genus @xmath911 , and the map @xmath975 is a canonical map of @xmath28 . in particular , @xmath28 is hyperelliptic if and only if @xmath976 .
the first sentence follows from the adjunction formula @xmath977 and the vanishings @xmath978 . by the @xmath2-decomposition of @xmath801 we have @xmath979 if and only if @xmath980 .
hence @xmath981 is generically injective ( actually an embedding ) unless @xmath976 , which proves the last assertion .
let @xmath982 be the cone over the locus of double conics @xmath983 , @xmath984 , with vertex @xmath985 .
it is the closure of the locus of smooth quartics tangent to @xmath966 at eight points lying on a conic .
[ cone structure ] the morphism @xmath986 induces an isomorphism between @xmath987 and @xmath988 which maps the pencils @xmath989 , @xmath984 , to the pencils @xmath990 .
if @xmath991 , then @xmath992 so that @xmath993 is tangent to @xmath966 at @xmath994 .
hence @xmath995 , and the equality @xmath996 proves the assertion .
now let @xmath997 be the locus of pairs @xmath998 such that @xmath966 and @xmath999 are smooth and tangent to each other at eight points lying on a conic . for a @xmath1000
we take the double cover @xmath965 branched over @xmath966 . by lemma [ cone structure ] we have @xmath1001 for a smooth @xmath376 where @xmath663 is the geiser involution of @xmath29 . taking the 2-elementary @xmath0 surface corresponding to @xmath1002 , we obtain a well - defined morphism @xmath1003 , which descends to a rational map @xmath1004 .
[ birat ( 8,8,1 ) ] the map @xmath30 is birational . by the birational equivalence @xmath1005 and lemma [ cone structure ] ,
the first projection @xmath1006 , @xmath1007 , induces a fibration @xmath1008 whose fiber over a general @xmath1009 is an open set of @xmath987 . since @xmath1010 for a general @xmath29
, this shows that @xmath30 is generically injective .
the equality @xmath1011 concludes the proof .
[ rational ( 8,8,1 ) ] the quotient @xmath1012 is rational .
hence @xmath962 is rational .
we have a @xmath811-equivariant dominant morphism @xmath1013 where @xmath608 is the unique conic with @xmath983 contained in the pencil @xmath1014 .
the @xmath1015-fiber over a general @xmath1016 is identified with an open set of the pencil @xmath1017 in @xmath1018 .
then we may use the no - name lemma [ no - name 2 ] to see that @xmath1019 the quotient @xmath1020 is rational by the slice method for the projection @xmath1021 and katsylo s theorem [ katsylo ] . by @xmath30
we identify @xmath962 birationally with @xmath1012 .
the first projection @xmath1006 , @xmath1007 , induces the quotient surface map @xmath1022 on the other hand , the second projection @xmath1006 , @xmath1023 , induces the fixed curve map @xmath1024 .
let @xmath1025 be the rational involution of @xmath962 induced by the involution @xmath1026 of @xmath25 .
now we know that [ fixed curve ( 8,8,1 ) ] the two fibrations @xmath1027 are exchanged by the involution @xmath1025 of @xmath962 . in particular ,
the generic fiber @xmath1028 of @xmath53 is birationally identified with the cone @xmath1029 in @xmath1030 .
the fixed locus of @xmath1025 contains the locus @xmath1031 of pairs @xmath1032 , @xmath985 .
generically , @xmath1033 may be characterized either as the locus of @xmath428 the vertices of @xmath53-fibers @xmath1029 , @xmath1034 , of @xmath431 the vertices of @xmath608-fibers @xmath987 , @xmath1009 , and of @xmath434 quadruple covers of @xmath35 branched over smooth quartics . via the last description
, @xmath1033 admits the structure of a ball quotient by a result of kond @xcite .
let @xmath29 be a del pezzo surface of degree @xmath38 .
the bi - anticanonical map @xmath1036 is a degree @xmath45 morphism onto a quadratic cone @xmath608 , which maps the base point of @xmath1037 to the vertex @xmath1038 of @xmath608 , and which is branched over a smooth curve @xmath1039 and @xmath1038 . in view of this
, we may define a map @xmath1040 as follows . for a general @xmath1041
we take the double cover @xmath1042 branched over @xmath210 and @xmath1038 , and let @xmath376 be the pullback of @xmath174 .
( equivalently , we take the double cover of the desingularization @xmath684 of @xmath608 branched over @xmath1043 , and then contract the @xmath397-curve over @xmath115 . )
then we associate the 2-elementary @xmath0 surface corresponding to the right dpn pair @xmath375 .
this construction , being @xmath600-invariant , defines a period map @xmath1044 . since the double cover @xmath1042 is a bi - anticanonical map of @xmath29 , @xmath30 is generically injective .
then @xmath30 is birational because @xmath1045 has dimension @xmath1046 .
[ rational ( 9,9,1 ) ] the quotient @xmath1045 is rational .
therefore @xmath1035 is rational .
we may apply the no - name lemma [ no - name 2 ] to the projection @xmath1047 to see that @xmath1048 .
the quotient @xmath1049 is birational to @xmath1050 , which is rational by dolgachev @xcite .
in this section we prove that the spaces @xmath10 with @xmath63 and @xmath371 are rational . in [ sec : period maps g>6 , k>0 ] we construct birational period maps using trigonal curves of fixed maroni invariant .
then we prove the rationality for each space .
we first construct birational period maps for @xmath1051 with @xmath1052 using curves on the hirzebruch surface @xmath1053 .
we keep the notation of [ sec : hirze ] .
let @xmath1054 be the open set of smooth curves which are transverse to the @xmath1055-curve @xmath115 . for each @xmath1056
the curve @xmath1043 belongs to @xmath1057 and has the @xmath1058 nodes @xmath1059 as the singularities .
considering the 2-elementary @xmath0 surfaces associated to the dpn pairs @xmath1060 , we obtain a period map @xmath1061 .
[ birational ( r , r-2,delta ) , r<6 ] the map @xmath1062 is birational .
of course one may follow the recipe in [ ssec : recipe ] , but here a more direct proof is possible . indeed , by proposition [ moduli of trigonal ] ,
@xmath1063 is naturally birational to the moduli @xmath1064 of trigonal curves of genus @xmath1065 and scroll invariant @xmath1066 .
then @xmath1062 is generically injective because the fixed curve map for @xmath1051 gives the left inverse . by comparison of dimensions
, @xmath1062 is birational .
[ fixed curve ( r , r-2 ) , r<6 ] the fixed curve map for @xmath1051 with @xmath1052 is generically injective , with a generic image the maroni locus @xmath1064 of maroni invariant @xmath45 . next we construct a birational period map for @xmath1067 using curves on @xmath1068 .
let @xmath1069 be the open set of pairs @xmath622 such that @xmath210 is smooth and transverse to @xmath53 . for each @xmath1070
the curve @xmath625 belongs to @xmath1071 and has the four nodes @xmath1072 as the singularities . by taking the right resolution of @xmath625
, we obtain a period map @xmath1073 .
[ birat ( 6,2,0 ) ] the map @xmath30 is birational .
we proceed as in example [ ex:2 ] .
consider the following @xmath790-cover of @xmath25 : @xmath1074 the variety @xmath543 parametrizes the curves @xmath625 equipped with labelings of the three nodes @xmath1075 .
the rest one node @xmath228 is distinguished from those three by the irreducible decomposition of @xmath625 .
therefore we will obtain a generically injective lift @xmath1076 of @xmath30 .
since @xmath1077 has dimension @xmath1078 , @xmath553 is birational .
the projection @xmath1079 is an @xmath790-covering , while the projection @xmath1080 is an @xmath1081-covering for the lattice @xmath1082 .
it is easy to see that @xmath1083 .
hence @xmath30 is birational .
the quotient @xmath1084 is birationally identified with the maroni locus @xmath1085 , and the fibers of the projection @xmath1086 are the trigonal pencils .
therefore [ fixed curve ( 6,2 ) ] the fixed curve map for @xmath1067 gives a dominant map @xmath1087 whose general fibers are identified with the trigonal pencils . the locus @xmath1085 is the complement of @xmath1088 in the moduli of trigonal curves .
corollary [ fixed curve ( 6,2 ) ] may also be obtained from corollary [ fixed curve ( r , r-2 ) , r<6 ] for @xmath1089 by extending the fixed curve map for @xmath1090 to a component of the discriminant divisor .
the maroni loci @xmath249 in this section may be characterized in terms of special divisors : a trigonal curve @xmath210 of genus @xmath1091 ( resp .
@xmath1092 ) has maroni invariant @xmath45 ( resp .
@xmath38 ) if and only if @xmath1093 ( resp .
@xmath1094 ) is irreducible .
this follows from maroni s description of @xmath1095 ( see @xcite proposition 1 ) . by a _ jacobian @xmath0 surface _ we mean a @xmath0 surface endowed with an elliptic fibration and with its zero section .
[ rational ( 2,2,0 ) ] the space @xmath1096 is birational to the moduli space of jacobian @xmath0 surfaces .
therefore @xmath1096 is rational . in proposition
[ birational ( r , r-2,delta ) , r<6 ] we saw that a general member @xmath4 of @xmath1096 is canonically a double cover @xmath1097 branched over a smooth curve @xmath1098 .
the natural projection @xmath1099 gives rise to an elliptic fibration @xmath1100 , and the @xmath50-curve @xmath1101 is its section .
the involution @xmath2 is the inverse map of the fibration with respect to @xmath97 . in this way a general member of @xmath1096 is equipped with the structure of a jacobian @xmath0 surface .
conversely , for a jacobian @xmath0 surface @xmath1102 such that every singular fiber is irreducible , the inversion map @xmath2 of @xmath1103 with respect to @xmath97 is a non - symplectic involution of @xmath1 , and the quotient of @xmath1100 by @xmath2 is the natural projection @xmath1104 of a hirzebruch surface @xmath36 . as the image of @xmath97 in @xmath36 is a @xmath411-curve , we have @xmath1105 .
thus our first assertion is verified .
it is known that the moduli of jacobian @xmath0 surfaces is birational to the quotient @xmath1106 via the weierstrass forms of elliptic fibrations ( for example , see @xcite ) . by katsylo s theorem [ katsylo ]
this quotient is rational . by proposition [ birational ( r , r-2,delta ) , r<6 ] we have a birational equivalence @xmath1108 for the bundle @xmath1109 on @xmath1068 .
[ rational ( 3,1,1 ) ] the quotient @xmath1110 is rational .
hence @xmath1107 is rational .
we apply the slice method to the map @xmath1111 , @xmath1112 .
let @xmath1113 be the stabilizer of a point @xmath194 , and let @xmath1114 be the linear system of curves through @xmath171 .
then we have @xmath1115 .
the group @xmath26 is connected and solvable by proposition [ stabilizer of fiber ] , and @xmath26 acts linearly on @xmath84 by proposition [ linearization hirze ] .
hence @xmath1116 is rational by miyata s theorem [ miyata ] . by proposition [ birational ( r , r-2,delta ) , r<6 ] we have a birational equivalence @xmath1118 for the bundle @xmath1119 on @xmath684 .
[ rational ( 4,2,1 ) ] the quotient @xmath1120 is rational .
hence @xmath1117 is rational .
first we define an @xmath1121-equivariant map @xmath1122 to the symmetric product of @xmath684 as follows . for a general @xmath1123 ,
let @xmath1124 and @xmath445 be the @xmath117-fiber through @xmath438 .
we have a unique involution @xmath1125 of @xmath445 which fixes @xmath438 and exchanges the two points @xmath1126 .
then we let @xmath1127 be the fixed point of @xmath1125 other than @xmath438 , and set @xmath1128 .
the @xmath1129-fiber over a general @xmath1130 is the _ linear _ space @xmath1131 of curves @xmath210 with @xmath1124 and @xmath1132 , where @xmath445 is the @xmath117-fiber through @xmath1133 ,
@xmath438 is @xmath1134 , and @xmath1135 is the involution of @xmath445 fixing @xmath1133 and @xmath438 .
if @xmath1136 is the stabilizer of @xmath1137 , by the slice method we have @xmath1138 next we consider the @xmath1139-equivariant map @xmath1140 the @xmath1141-fiber over a general @xmath1142 is the sub linear system @xmath1143 of curves passing through @xmath1144 or singular at @xmath438 .
one checks that @xmath1139 acts almost transitively on @xmath1145 . if @xmath1146 is the stabilizer of @xmath1142 , by the slice method we have @xmath1147 we analyze the @xmath1148-representation @xmath1149 by using the coordinate system @xmath258 of @xmath684 introduced in [ ssec : coordinate ] .
we may assume that @xmath1150 ( resp .
@xmath1151 ) is the origin @xmath1152 in @xmath279 ( resp .
@xmath1153 in @xmath290 ) .
then @xmath438 is the origin @xmath1154 in @xmath1155 , @xmath778 .
we may also assume that @xmath1156 is expressed as @xmath1157 by the coordinate @xmath274 of @xmath1155 around @xmath438 .
let @xmath1158 , and @xmath532 be the automorphisms of @xmath684 described in the equations @xmath1159 , @xmath1160 , and @xmath1161 respectively .
we set @xmath1162 . for @xmath1163 as above
, the stabilizer @xmath1148 is given by @xmath1164 here @xmath532 acts on @xmath1165 by @xmath1166 , and @xmath156 acts on @xmath76 by @xmath161 .
first observe that @xmath1167 , and @xmath1168 are contained in @xmath1148 .
this is clear for @xmath532 and @xmath156 . by , @xmath1168 acts on @xmath296 and @xmath306 by @xmath1169 , @xmath1170 , @xmath778 .
hence it acts on both @xmath1171 trivially .
conversely , we set @xmath1172 .
the quotient @xmath1173 is @xmath146 generated by @xmath532 .
the @xmath1174-action on @xmath115 is contained in @xmath1175 .
hence by the sequence @xmath139 every element of @xmath1174 is written as @xmath1176 for some @xmath278 and @xmath1177 .
since @xmath1178 and @xmath1179 fix @xmath1133 , we have @xmath1180 so that @xmath1181 for some @xmath1182 .
this implies that @xmath1183 .
then @xmath1184 acts on @xmath296 by @xmath1185 , and on @xmath306 by @xmath1186 .
thus we must have @xmath1187 and @xmath1188 .
next we describe the linear system @xmath1189 by the coordinates . by proposition
[ def eqn in hirze ] the space @xmath1190 is isomorphic to the vector space @xmath1191 by restriction to @xmath296 .
we shall express the polynomials @xmath1192 as @xmath1193 .
the linear subspace @xmath1194 is defined by the equations @xmath1195 let @xmath1196 be defined by @xmath1197 .
the condition that @xmath1198 ( resp .
@xmath1199 ) is expressed by @xmath1200 ( resp .
@xmath1201 ) .
the condition on the tangent @xmath1202 ( resp .
@xmath1203 ) is written as @xmath1204 ( resp .
@xmath1205 ) .
the restriction of @xmath210 to the fiber @xmath1206 ( resp .
@xmath1207 ) is given by @xmath1208 ( resp .
@xmath1209 ) .
since these polynomials should be anti - symmetric , we have @xmath1210 and @xmath1211 . using the coordinate @xmath279
, we identify @xmath1149 with the subspace of @xmath1212 defined by , and calculate the @xmath1148-action on it .
the coefficients @xmath1213 of the polynomials are basis of the dual space @xmath1214 .
the subspace @xmath1215 is @xmath1148-invariant .
indeed , we have @xmath1216 therefore the annihilator of @xmath1217 , @xmath1218 is a @xmath1148-invariant subspace of @xmath1149 .
we want to apply the slice method to the projection @xmath1219 from @xmath1220 , which is @xmath1148-equivariant .
we have the coordinate @xmath1221 for @xmath1222 .
then @xmath1223 is an inhomogeneous coordinate of @xmath1224 . by the above calculation , @xmath1148 acts on @xmath1224 by @xmath1225 this shows that @xmath1148 acts on the affine line @xmath1226 transitively with the stabilizer of the point @xmath1227 being @xmath1228 .
the fiber of the projection @xmath1229 over @xmath1038 is the hyperplane @xmath1230 minus @xmath1231 , which is naturally @xmath1232-isomorphic to the @xmath1232-representation @xmath1233 .
hence by the slice method we have @xmath1234 let @xmath1235 be the subspace @xmath1236 one checks that @xmath1237 is @xmath1232-invariant , and that with respect to the given basis @xmath1232 acts on @xmath1237 by @xmath1238 and @xmath1239 .
in particular , @xmath1232 acts on @xmath1237 effectively and so almost freely . since @xmath1232 is a finite group , we have a @xmath1232-decomposition @xmath1240 . applying the no - name lemma [ no - name 1 ] to the projection @xmath1241 from @xmath1242 , we have @xmath1243 since @xmath1244 ,
the quotient @xmath1245 is rational .
this completes the proof of proposition [ rational ( 4,2,1 ) ] . by corollary [ fixed curve ( r , r-2 ) , r<6 ]
, @xmath1090 is birational to @xmath1088 . in @xcite
we proved that the moduli of trigonal curves of genus @xmath1246 is rational .
therefore [ rational ( 5,3,1 ) ] the moduli space @xmath1090 is rational .
recall from proposition [ birat ( 6,2,0 ) ] that we have a birational equivalence @xmath1247 .
[ rational ( 6,2,0 ) ] the quotient @xmath1248 is rational .
therefore @xmath1067 is rational . applying the slice method to the projection @xmath1249
, we have @xmath1250 where @xmath1251 is the stabilizer of a point @xmath194 . as in the proof of proposition [ rational ( 3,1,1 ) ] , we see that @xmath1252 is rational .
in this section we prove that the spaces @xmath10 with @xmath1253 and @xmath371 are rational .
we will find curves of clifford index @xmath38 , i.e. , trigonal curves and plane quintics as the main fixed curves .
let @xmath1255 be the open set of pairs @xmath1256 such that @xmath210 is smooth and transverse to @xmath500 .
the 2-elementary @xmath0 surface associated to the sextic @xmath1257 has parity @xmath338 . indeed ,
if @xmath1258 is the corresponding right dpn pair , we have @xmath1259 .
thus we obtain a period map @xmath1260 .
[ birational ( 6,4,0 ) ] the period map @xmath30 is birational .
we consider the following @xmath1261-cover of @xmath25 .
@xmath1262 by @xmath543 the sextics @xmath1257 are endowed with complete labelings of the nodes . since @xmath1263
, this induces a birational lift @xmath1264 of @xmath30 by the recipe in [ ssec : recipe ] .
the projection @xmath1265 has degree @xmath1266 because @xmath811 acts almost freely on @xmath25 . on the other hand
, @xmath1267 is an @xmath1081-cover of @xmath1254 for the lattice @xmath1268 .
by @xcite we have @xmath1269 .
this proves the proposition .
[ rational ( 6,4,0 ) ] the quotient @xmath1270 is rational .
therefore @xmath1254 is rational .
the vector bundle @xmath1271 over @xmath1272 is @xmath1273-linearized in which the element @xmath1274 acts by the scalar multiplication by @xmath1275 . on the other hand ,
the tautological bundle @xmath1276 over @xmath1272 is also @xmath1273-linearized where @xmath1277 acts by the multiplication by @xmath1278 .
hence @xmath1279 is @xmath811-linearized .
note that @xmath892 is canonically isomorphic to @xmath1280 .
since @xmath811 acts almost freely on @xmath1272 , we may apply the no - name lemma to @xmath1279 to see that @xmath1281 the quotient @xmath1282 is rational by shepherd - barron @xcite .
[ curve moduli ( 6,4,0 ) ] the fixed curve map for @xmath1254 is a dominant map onto the locus of plane quintics , i.e. , non - hyperelliptic curves having @xmath1283 , and its general fibers are identified with the @xmath1283 .
note that a smooth plane quintic has only one @xmath1284 .
let @xmath567 and let @xmath1286 be the open set of pairs @xmath622 such that @xmath210 is smooth and transverse to @xmath53 .
the group @xmath1287 acts on @xmath25 . for a @xmath1070 the 2-elementary @xmath0 surface @xmath4 associated to the bidegree @xmath17 curve @xmath1288 has invariant @xmath1289 . in order to calculate @xmath9
, we blow - up a point @xmath171 in @xmath1075 and then contract the two ruling fibers through @xmath171 to obtain an irreducible sextic with a node or a cusp and with a @xmath433-singularity . by lemma [ l_+ and a - d - e ]
this shows that @xmath1290 , and so @xmath4 has @xmath41 .
thus we obtain a period map @xmath1291 .
[ birat ( 6,4,1 ) ] the period map @xmath30 is birational .
let @xmath1292 be the locus of @xmath1293 such that @xmath1294 .
as in example [ ex:1 ] , we see that @xmath30 lifts to a birational map @xmath1295 .
note that @xmath1296 .
the projection @xmath1297 is an @xmath1298-covering , while @xmath1299 is an @xmath1081-cover of @xmath1285 for the lattice @xmath1300 .
it is easy to calculate that @xmath1301 .
[ rational ( 6,4,1 ) ] the quotient @xmath1302 is rational .
therefore @xmath1285 is rational .
we denote @xmath1303 .
the vector bundle @xmath1304 over @xmath818 admits a natural linearization by @xmath1305 , in which @xmath1306 acts by the multiplication by @xmath161 . on the other hand , the hyperplane bundle @xmath1307 over @xmath818
is also @xmath1305-linearized where @xmath1166 acts by @xmath161 .
hence the bundle @xmath1308 is @xmath26-linearized .
we have a canonical isomorphism @xmath1309 .
the @xmath26-action on @xmath818 is almost free , with the quotient @xmath1116 naturally birational to the moduli @xmath1310 of trigonal curves of genus @xmath1311 ( see proposition [ moduli of trigonal ] ) .
hence by the no - name lemma for @xmath1308 we have @xmath1312 the space @xmath1310 is rational by shepherd - barron @xcite .
[ curve moduli ( 6,4,1 ) ] the fixed curve map for @xmath1285 gives a dominant map @xmath1313 , whose fiber over a general @xmath210 is identified with the pencil @xmath1314 where @xmath1315 is the trigonal bundle
. one might also deduce corollaries [ curve moduli ( 6,4,0 ) ] and [ curve moduli ( 6,4,1 ) ] from the birational map @xmath1316 in [ sec:(5,5,1 ) ] , by extending it to the discriminant divisor ( cf .
@xcite ) .
we construct 2-elementary @xmath0 surfaces using curves on @xmath684 .
we keep the notation of [ sec : hirze ] . for @xmath1319
let @xmath1320 be the locus of pairs @xmath622 such that @xmath341 @xmath210 is smooth , @xmath344 @xmath53 intersects with @xmath210 at @xmath1059 with multiplicity @xmath1321 , and @xmath346 @xmath1322 . for @xmath1323 , these conditions mean that @xmath53 does not pass @xmath1059 and is transverse to @xmath210 .
in particular , @xmath306 is open in @xmath1324 . for @xmath1325 ,
@xmath1326 is regarded as a sublocus of @xmath1327 because @xmath53 is uniquely determined by the point @xmath1059 .
[ lemma ( 5+k,5-k ) ] the variety @xmath1326 has the expected dimension @xmath1328 .
let @xmath1329 be a defining equation of a smooth @xmath1330 as in proposition [ def eqn in hirze ] .
we normalize @xmath1331
. then we have @xmath1332 if and only if the cubic polynomial @xmath1333 of @xmath1334 is factorized as @xmath1335 such that @xmath1336 and @xmath62 has no multiple root .
this proves the assertion . for a pair @xmath1332
, the curve @xmath1337 belongs to @xmath1338 and is singular at @xmath1339 . for @xmath1323
those points are distinct nodes of @xmath28 . for @xmath1325
the point @xmath1340 is a @xmath1341-singularity of @xmath28 , and the rest @xmath1342 points @xmath1343 are nodes .
hence the 2-elementary @xmath0 surface associated to @xmath28 has invariant @xmath1344 , and we obtain a period map @xmath1345 where @xmath41 for @xmath1346 and @xmath338 for @xmath1347 . [ birational ( 5+k,5-k ) ] the map @xmath1348 is birational .
first we treat the case @xmath1323 using the recipe in [ ssec : recipe ] .
let @xmath1349 be the locus of @xmath1350 such that @xmath1351 .
the space @xmath1352 parametrizes the @xmath687-curves @xmath1337 endowed with labelings of the three nodes @xmath1075 .
the rest two nodes , @xmath228 and @xmath1059 , are distinguished by the irreducible decomposition of @xmath28 , and the components of @xmath28 are identified by their classes in @xmath1353 . thus we will obtain a birational lift @xmath1354 of @xmath1355 .
the projection @xmath1356 is an @xmath790-covering , while @xmath1357 is an @xmath1081-cover of @xmath1358 for the lattice @xmath1359 .
since @xmath1083 , the map @xmath1355 is birational . for @xmath1325
a similar argument is possible , but we may also proceed as in the proof of proposition [ birational ( r , r-2,delta ) , r<6 ] : since @xmath1326 may be regarded as a sublocus of @xmath1327 , by proposition [ moduli of trigonal ] the quotient @xmath1360 is naturally birational to a sublocus of the maroni divisor @xmath1361 . considering the fixed curve map for @xmath1317 , we see that @xmath1348 is generically injective . by lemma [ lemma ( 5+k,5-k ) ]
, @xmath1348 is birational .
[ rational ( 5+k,5-k ) ] the quotient @xmath1360 is rational
. therefore @xmath1317 with @xmath1319 is rational .
this is analogous to the proofs of propositions [ rational ( 3,1,1 ) ] and [ rational ( 6,2,0 ) ] : we apply the slice method to the projection @xmath1362 , whose general fiber is an open set of a linear subspace @xmath1363 by the proof of lemma [ lemma ( 5+k,5-k ) ] .
note that although @xmath1109 may not be @xmath1121-linearized , one may instead use the group @xmath1364 in [ sec : hirze ] : it acts on @xmath1109 by lemma [ cover linearization hirze ] , and the stabilizer @xmath1365 of a point of @xmath124 is still connected and solvable by the proof of proposition [ stabilizer of fiber ] .
therefore we may use miyata s theorem for the @xmath26-representation @xmath1366 to see that @xmath1367 is rational .
we shall study the fixed curve maps .
for @xmath1325 we let @xmath1368 be the closure of the image of natural morphism @xmath1369 . in particular , @xmath1370 coincides to @xmath1361 . by proposition [ birational ( 5+k,5-k ) ]
we have [ fixed curve ( 7,3 ) ( 8,2 ) ] the fixed curve map for @xmath1358 is a dominant map @xmath1371 whose general fibers are birationally identified with the @xmath238 .
the fixed curve map for @xmath1317 with @xmath1325 is generically injective with a generic image @xmath1372 .
in fact , it is more natural to identify the fibers of @xmath1371 with the non - free pencils @xmath1314 where @xmath1315 is the trigonal bundle , whose free part is @xmath1373 ( cf .
corollary [ curve moduli ( 6,4,1 ) ] ) . the loci @xmath1372 may be described in terms of special divisors as follows .
[ fixed curve locus g=6 ] let @xmath210 be a trigonal curve of genus @xmath1311 with the trigonal bundle @xmath1315
. @xmath428 the curve @xmath210 has scroll invariant @xmath45 if and only if @xmath1374 is irreducible .
@xmath431 when @xmath210 has scroll invariant @xmath45 , we have @xmath1375 if and only if @xmath1376 is contained in @xmath1377
. @xmath434 the locus @xmath1378 is the intersection of @xmath1379 with the theta - null divisor .
@xmath428 this follows from @xcite proposition 1 ( or @xcite v. a-9 ) .
specifically , @xmath1374 consists of the curve @xmath1380 and the point @xmath1381 , and @xmath1381 is not contained in @xmath1380 if and only if @xmath210 has scroll invariant @xmath250 .
when @xmath210 is an @xmath1109-curve on @xmath684 , we have @xmath1382 for the point @xmath1383 .
@xmath431 the locus @xmath1384 consists of those @xmath210 whose trigonal map ramifies at @xmath1038 . by @xcite proposition 1
we have @xmath1385 where @xmath1386 and @xmath1387 is the residual of @xmath1388 .
let @xmath1389 and @xmath1390 be the residual of @xmath1391 .
then we see that @xmath1392 if @xmath1393 , we have @xmath1394 .
since @xmath1395 , this proves the assertion .
@xmath434 a curve @xmath210 in @xmath1361 has an effective even theta characteristic if and only if the residual involution on @xmath1094 has a fixed point . by the structure of @xmath1094 described above
, this is exactly when @xmath1396 is one point , i.e. , @xmath1397 . when @xmath1398 , this is equivalent to the condition @xmath1399 .
in this section we prove that the spaces @xmath10 with @xmath919 and @xmath371 are rational .
the case @xmath1400 was settled by kond @xcite using trigonal curves with vanishing theta - null .
it turns out that the main fixed curves for other @xmath1401 are also trigonal .
we construct 2-elementary @xmath0 surfaces using curves on @xmath620 .
let @xmath1403 be the open set of pairs @xmath569 such that @xmath210 and @xmath174 are smooth and transverse to each other .
the 2-elementary @xmath0 surface associated to the @xmath624-curve @xmath571 has invariant @xmath1404 .
thus we obtain a period map @xmath1405 .
[ birat ( 7,5,1 ) ] the period map @xmath30 is birational . as before , we consider an @xmath1261-cover @xmath543 of @xmath25 whose fiber over a @xmath1406 corresponds to labelings of the five nodes @xmath574 .
noticing that the blow - down @xmath1407 contracts the @xmath397-curve to the unique node of @xmath1408 , one may proceed as in example [ ex:2 ] to obtain a birational lift @xmath1409 of @xmath30 .
the projection @xmath1410 is an @xmath1081-covering for the lattice @xmath1411 .
by @xcite we have @xmath1269 , so that @xmath30 is birational .
[ rational ( 7,5,1 ) ] the quotient @xmath1412 is rational .
hence @xmath1402 is rational . by proposition [ moduli of trigonal ] , @xmath1413 is canonically birational to the moduli @xmath1414 of trigonal curves of genus @xmath1415 .
since @xmath463 is @xmath666-linearized , we may apply the no - name lemma [ no - name 2 ] to the projection @xmath1416 .
then we have @xmath1417 .
the space @xmath1414 is rational by @xcite . for every smooth @xmath1123 ,
the restriction of @xmath123 to @xmath210 is identified with the linear system @xmath1418 where @xmath1315 is the trigonal bundle .
therefore [ fixed curve ( 7,5,1 ) ] the fixed curve map for @xmath1402 gives a dominant map @xmath1419 whose general fibers are birationally identified with the residuals of the @xmath238 .
we consider curves on @xmath620 .
for @xmath1319 let @xmath1421 be the locus of pairs @xmath622 such that @xmath341 @xmath210 is smooth and transverse to the @xmath397-curve @xmath115 , @xmath344 @xmath53 intersects with @xmath210 at one of @xmath1059 with multiplicity @xmath1321 , and @xmath346 @xmath1422 .
when @xmath1323 , the conditions @xmath344 and @xmath346 simply mean that @xmath53 is transverse to @xmath1043 .
in particular , @xmath306 is open in @xmath1423 . the projection @xmath1424 is dominant for @xmath1425 , and generically injective for @xmath1426 . as in the proof of lemma [ lemma ( 5+k,5-k ) ] ,
one checks that @xmath1326 has the expected dimension @xmath1427 . for a @xmath1332
, the curve @xmath1337 belongs to @xmath1428 .
when @xmath1323 , @xmath28 has only nodes ( the intersections of the components ) as the sigularities . when @xmath1325 , @xmath28 has the @xmath1341-point @xmath228 , the @xmath1342 nodes @xmath1429 , the one node @xmath1430 , and no other singularity .
thus the 2-elementary @xmath0 surface associated to @xmath28 has invariant @xmath1431 . when @xmath1432 , we have parity @xmath41 by lemma [ delta=1 ] @xmath434 .
hence we obtain a period map @xmath1433 .
[ birational ( 6+k,6-k ) ] the map @xmath1348 is birational . when @xmath1426 , the singularities of the curves @xmath1337 are a priori distinguished by their type and by the irreducible decomposition of @xmath28 .
also the three branches at the @xmath1341-point are distinguished by the decomposition of @xmath28 .
therefore @xmath1348 lifts to a birational map @xmath1434 by the recipe in [ ssec : recipe ] .
one checks that @xmath1435 in these cases .
the case @xmath1436 is treated in example [ ex:2 ] . for @xmath1323 , we label the three nodes @xmath1075 and the two nodes @xmath1059 independently .
this is realized by an @xmath1437-cover @xmath1352 of @xmath306 .
the rest node @xmath228 of @xmath28 is distinguished from those five .
then we will obtain a birational lift @xmath1438 of @xmath1355 .
by @xcite we have @xmath1439 for the lattice @xmath1440 .
therefore @xmath1355 is birational .
[ rational g=5 , k>1 ] the quotient @xmath1441 is rational .
therefore @xmath1420 with @xmath1318 is rational . the same proof as for proposition [ rational ( 5+k,5-k ) ] works . for completeness
, we briefly explain the birational period map for @xmath20 constructed by kond @xcite .
let @xmath1442 be the locus of smooth curves @xmath210 which are tangent to @xmath115 .
the quotient @xmath1412 is identified with the theta - null divisor @xmath1443 in the trigonal locus . for a @xmath221 let @xmath53 be the @xmath117-fiber through @xmath1059 . the 2-elementary @xmath0 surface associated to
the @xmath624-curve @xmath625 belongs to @xmath20 .
then the period map @xmath1444 is birational .
kond s proof is essentially along the line of [ ssec : recipe ] .
it can also be deduced using the fixed curve map .
we describe the fixed curve maps as rational maps @xmath1445 to the trigonal locus . by proposition [ birational ( 6+k,6-k ) ] we have [ fixed curve ( 6+k,6-k ) ver1 ] the map @xmath1446 is dominant for @xmath1425 , and is generically injective for @xmath1426 .
the general fibers of @xmath833 are birationally identified with the @xmath238 , and the general fibers of @xmath1447 are identified with the two points @xmath1448 , @xmath1449 .
the @xmath1447-fiber over a general @xmath1123 is identified with the two points @xmath1450 .
note that @xmath1451 for the trigonal bundle @xmath1315 .
by @xcite , @xmath1374 consists of two residual components , @xmath1452 and @xmath1387 .
then @xmath1448 is the intersection @xmath1453 . a generic image of @xmath656 ( resp .
@xmath1454 ) is the locus where the trigonal map ramifies ( resp . totally ramifies ) at the base point of one of @xmath1448 .
we also note that the theta - null divisor @xmath1443 is exactly the locus where @xmath1448 is one point .
thus the double covering @xmath1455 is the quotient by the residuation and is ramified at @xmath20 over @xmath1443
in this section we study the case @xmath950 , @xmath371 .
our constructions are related to canonical models of genus @xmath738 curves , i.e. , curves on quadratic surfaces cut out by cubics . except for
[ ssec:(10,4,0 ) ] we shall use the following notation : @xmath608 is the surface @xmath18 , @xmath119 is the bundle @xmath1456 , and @xmath1457 is the identity component of @xmath600 .
let @xmath1459 be the open set of pairs @xmath569 such that @xmath210 and @xmath174 are smooth and transverse to each other .
considering the 2-elementary @xmath0 surfaces associated to the @xmath570-curves @xmath571 , we obtain a period map @xmath572 .
in example [ ex:1 ] we proved that @xmath30 is birational .
[ rational ( 8,6 ) ] the quotient @xmath1460 is rational .
hence @xmath1458 is rational .
recall that the quotient @xmath1461 is naturally birational to the moduli @xmath958 of genus @xmath738 curves .
this is a consequence of the fact that a general canonical genus @xmath738 curve is a complete intersection of a cubic and a unique smooth quadric . for a smooth @xmath1462
the linear system @xmath163 is identified with @xmath244 by restriction .
then let @xmath954 be the universal genus @xmath738 curve ( over an open locus ) , and @xmath955 be the bundle @xmath1463 over @xmath958 .
the above remark implies that @xmath1464 is birational to @xmath892 .
since @xmath958 is rational ( @xcite ) , so is @xmath892 . the fixed curve map @xmath1465 is dominant with the general fibers birationally identified with the canonical systems . for a point @xmath1467
we denote by @xmath493 the union of the two ruling fibers meeting at @xmath171 .
let @xmath1468 be the open set of pairs @xmath1469 such that @xmath210 is smooth and transverse to @xmath493 .
taking the right resolution of the @xmath570-curves @xmath1470 , we obtain a period map @xmath1471 .
[ birat ( 9,5 ) ] the map @xmath30 is birational .
let @xmath543 be the space of those @xmath1472 such that @xmath1473 and that @xmath871 lie on the @xmath1474-component of @xmath493 .
the space @xmath543 parametrizes the curves @xmath1470 endowed with labelings of the six nodes @xmath1475 which take into account the decomposition of @xmath493 .
the rest node of @xmath1470 , the point @xmath171 , is clearly distinguished from those six .
note that @xmath600 does not act on @xmath543 , for the definition of @xmath543 involves the distinction of the two rulings .
rather @xmath543 is acted on by @xmath1476 . as in example
[ ex:1 ] , @xmath30 lifts to a birational map @xmath1477 .
the projection @xmath1478 has degree @xmath1479 , while @xmath1480 is an @xmath1081-cover of @xmath1466 for the lattice @xmath1481 .
by @xcite we have @xmath732 . [ rational ( 9,5 ) ]
the quotient @xmath1482 is rational .
therefore @xmath1466 is rational . applying the slice method to the projection @xmath1483
, we have @xmath1484 where @xmath1485 is the stabilizer of a point @xmath1467 .
let @xmath1486 be respectively the @xmath1474- and the @xmath1487-fiber through @xmath171 .
we denote @xmath1488 .
we want to apply the no - name lemma to the @xmath26-equivariant map @xmath1489 , @xmath1490 .
it is birationally the projectivization of a subbundle @xmath955 of the vector bundle @xmath1491 over @xmath829 .
let @xmath144 be the preimage of @xmath1485 in @xmath1492 .
since @xmath1492 acts on @xmath1493 , the bundle @xmath955 is @xmath144-linearized . on the other hand , consider the natural line bundle @xmath1494 over @xmath829 which is also @xmath144-linearized .
the kernel of @xmath1495 is generated by the elements @xmath1496 of @xmath1497 , which act on both @xmath955 and @xmath1276 by the multiplication by @xmath161 .
thus the bundle @xmath1279 is @xmath26-linearized .
the group @xmath26 acts on @xmath829 almost freely : this follows by noticing that for a general point @xmath1498 of @xmath829 there is no isomorphism @xmath1499 mapping @xmath1500 to @xmath1501 , and that @xmath137 acts on @xmath1502 almost freely .
therefore we may apply the no - name lemma to @xmath1279 to deduce that @xmath1503 the quotient @xmath839 is rational because @xmath838 .
[ fixed curve ( 9,5 ) ] the fixed curve map @xmath1504 is dominant with the general fibers birationally identified with the products of the two trigonal pencils . as in [ ssec:(9,5,1 ) ] , for a point @xmath1467 we denote by @xmath493 the reducible bidegree @xmath1506 curve singular at @xmath171 . let @xmath1468 be the locus of pairs @xmath1469 such that @xmath210 is smooth , passes @xmath171 , and is transverse to each component of @xmath493 .
the space @xmath25 is an open set of the universal bidegree @xmath1507 curve .
the bidegree @xmath17 curve @xmath1470 has the @xmath433-point @xmath171 , the four nodes @xmath1508 , and no other singularity .
the associated 2-elementary @xmath0 surface has invariant @xmath1509 .
it has parity @xmath41 by lemma [ delta=1 ] @xmath431 .
thus we obtain a period map @xmath1510 .
[ birat ( 10,4,1 ) ] the map @xmath30 is birational .
as in [ ssec:(9,5,1 ) ] , we consider the locus @xmath1292 of those @xmath1511 such that @xmath1512 and that @xmath846 lie on the @xmath1474-component of @xmath493 .
since the three branches of @xmath1470 at @xmath171 are distinguished by the irreducible decomposition of @xmath1470 , then @xmath30 lifts to a birational map @xmath1513 .
the space @xmath1514 is an @xmath1515-cover of @xmath1460 , while @xmath1516 is an @xmath1081-cover of @xmath1505 for the lattice @xmath1517 .
by @xcite we have @xmath1518 .
since @xmath1461 is naturally biational to @xmath958 , the quotient @xmath1460 is birational to the universal genus @xmath738 curve , which is rational by catanese @xcite .
therefore [ rational ( 10,4,1 ) ] the moduli space @xmath1505 is rational .
[ fixed curve ( 10,4,1 ) ] the fixed curve map @xmath1519 identifies @xmath1505 birationally with the universal genus @xmath738 curve .
we consider curves on @xmath684 , keeping the notation of [ sec : hirze ] .
let @xmath1521 be the open set of pairs @xmath686 such that @xmath210 and @xmath451 are smooth and transverse to each other .
in example [ ex:3 ] , we showed that the 2-elementary @xmath0 surfaces associated to the @xmath687-curves @xmath688 have main invariant @xmath711 , and that the induced period map @xmath1522 is birational .
the quotient @xmath1523 is naturally birational to the theta - null divisor @xmath1524 in @xmath958 .
indeed , recall that the morphism @xmath1525 associated to the bundle @xmath463 is the minimal desingularization of the quadratic cone @xmath1526 , and that the restriction of @xmath1527 to each smooth @xmath1528 is a canonical map of @xmath210 .
then our claim follows from the fact that a non - hyperelliptic genus @xmath738 curve has an effective even theta characteristic if and only if its canonical model lies on a singular quadric . in that case
the half canonical pencil is given by the pencil of lines on @xmath608 , or equivalently , the pencil @xmath124 .
[ rational ( 10,4,0 ) ] the quotient @xmath1529 is rational .
hence @xmath1520 is rational .
since both @xmath1530 and @xmath1531 are @xmath1121-linearized ( proposition [ linearization hirze ] ) and since a general @xmath1532 has no automorphism , we may apply the no - name lemma [ no - name 2 ] to the projection @xmath1533 to see that @xmath1534 the space @xmath1524 is rational by dolgachev @xcite .
[ fixed curve ( 10,4,0 ) ] the fixed curve map for @xmath1520 is a dominant map onto @xmath1524 whose general fibers are birationally identified with the symmetric products of the half - canonical pencils .
let @xmath567 .
for a point @xmath1467 we denote by @xmath493 the reducible bidegree @xmath1506 curve singular at @xmath171 . for @xmath1426 let @xmath1537 be the locus of pairs @xmath1469 such that @xmath341 @xmath210 is smooth with @xmath1538 , @xmath344 the @xmath1474-component of @xmath493 is tangent to @xmath210 at @xmath171 with multiplicity @xmath1321 , and @xmath346 the @xmath1487-component of @xmath493 is transverse to @xmath210 .
the space @xmath1326 is acted on by @xmath1476 ( but not by @xmath600 ) .
the bidegree @xmath17 curve @xmath1470 has the @xmath1341-singularity @xmath171 , the @xmath1539 nodes @xmath1508 , and no other singularity .
taking the right resolution of @xmath1470 , we obtain a period map @xmath1540 .
[ birational ( 11,3)&(12,2 ) ] the map @xmath1348 is birational .
we only have to distinguish the two intersection points other than @xmath171 of @xmath210 and the @xmath1487-component of @xmath493 .
this defines a double cover @xmath1541 .
the rest singularities of @xmath1470 and the branches of @xmath1470 at @xmath171 are a priori labeled as before .
checking @xmath1542 , we see that @xmath1348 lifts to a birational map @xmath1543 .
the variety @xmath1544 is an @xmath519-cover of @xmath1545 for the lattice @xmath1546 .
then @xmath1547 for both @xmath1426 , so that @xmath1348 has degree @xmath38 .
[ rational ( 11,3)&(12,2 ) ] the quotient @xmath1548 is rational .
therefore @xmath1535 and @xmath1536 are rational .
this is a consequence of the slice method for the projection @xmath1549 and miyata s theorem .
the stabilizer of a point @xmath1467 in @xmath1476 is isomorphic to @xmath1550 , which is connected and solvable .
[ fixed curve ( 11,3 ) ] the fixed curve map @xmath1551 is finite and dominant , with the fiber over a general @xmath1552 being the ramification points of the two trigonal maps @xmath1553 . for @xmath1347 the image of the natural map @xmath1554
consists of curves such that one of its trigonal maps has a total ramification point .
such a point is nothing but a weierstrass point whose first non - gap is @xmath911 .
therefore [ fixed curve ( 12,2 ) ] the fixed curve map for @xmath1536 is generically injective with a generic image being the locus of curves having a weierstrass point whose first non - gap is @xmath911 .
in this section we study the case @xmath1555 , @xmath371 . when @xmath1556 , we use plane quartics to construct general members of @xmath10 . for @xmath1557 we find hyperelliptic curves as the main fixed curves .
let @xmath1559 be the open locus of pairs @xmath1560 such that @xmath210 and @xmath608 are smooth and transverse to each other .
the 2-elementary @xmath0 surfaces associated to the sextics @xmath1561 have invariant @xmath1562 , and we obtain a period map @xmath1563 . [ birat ( 9,7 ) ] the map @xmath30 is birational .
we consider an @xmath1564-covering @xmath545 whose fiber over a @xmath1565 corresponds to labelings of the eight nodes @xmath1566 of @xmath1561 . as before , we see that @xmath30 lifts to a birational map @xmath1567 .
the projection @xmath1568 is an @xmath1081-covering for the lattice @xmath1569 .
by @xcite we calculate @xmath1570 .
recall that we have a natural birational equivalence @xmath1571 , for a canonical model of a non - hyperelliptic genus @xmath911 curve is a plane quartic .
[ rational ( 9,7 ) ] the quotient @xmath1020 is rational .
therefore @xmath1558 is rational .
let @xmath1572 be the universal curve ( over an open set ) , @xmath926 be the relative canonical bundle for @xmath117 , and @xmath955 be the bundle @xmath1573 . as the restriction map @xmath1574 for a smooth quartic @xmath210 is isomorphic ,
the quotient @xmath1020 is birational to @xmath892 . since @xmath1575 is rational by katsylo @xcite , so is @xmath892 .
[ fixed curve ( 9,7 ) ] the fixed curve map @xmath1576 is dominant with general fibers being birationally identified with the bi - canonical systems . let @xmath1559 be the locus of pairs @xmath1560 such that @xmath608 is a union of two distinct lines , and @xmath210 is smooth and transverse to @xmath608 .
the 2-elementary @xmath0 surfaces associated to the nodal sextics @xmath1561 have invariant @xmath1578 , and parity @xmath41 by lemma [ delta=1 ] ( 1 ) .
hence we obtain a period map @xmath1579 .
[ birat ( 10,6,1 ) ] the period map @xmath30 is birational .
we define an @xmath1580-cover of @xmath25 as the locus @xmath1581 of those @xmath1582 such that @xmath1583 and that @xmath1584 belong to a same component of @xmath608 . by considering @xmath543 , the eight nodes @xmath1566 and the two components of @xmath608 are labelled compatibly ( cf . example [ ex:3 ] ) .
the rest node of @xmath1561 , @xmath1585 , is distinguished from those eight by the decomposition of @xmath1561
. then @xmath30 will lift to a birational map @xmath1586 .
the projection @xmath789 is an @xmath1580-covering , while the galois group of @xmath1587 is @xmath1081 for the lattice @xmath1588 . since @xmath1589 by @xcite , @xmath30 is then birational .
[ rational ( 10,6,1 ) ] the quotient @xmath1012 is rational .
hence @xmath1577 is rational .
we keep the notation in the proof of proposition [ rational ( 9,7 ) ] .
let @xmath1590 be the bundle @xmath1591 over @xmath1575 . by restriction ,
the space of singular plane conics is identified with the symmetric product of the canonical system of every smooth quartic .
this implies that @xmath1012 is birational to the symmetric product of @xmath1592 over @xmath1575 .
since @xmath1593 , we have @xmath1594 .
then @xmath1595 is birational to the quotient of @xmath1596 by the permutation , and so is rational . since @xmath1575 is rational ( @xcite ) ,
our assertion is proved .
[ fixed curve ( 10,6,1 ) ] the fixed curve map @xmath1597 is dominant with general fibers being birationally identified with the symmetric products of the canonical systems .
we construct 2-elementary @xmath0 surfaces using curves on the hirzebruch surface @xmath1599 .
we keep the notation of [ sec : hirze ] .
let @xmath1600 be the open set of pairs @xmath569 such that @xmath210 and @xmath174 are smooth and transverse to each other . the 2-elementary @xmath0 surface @xmath4 associated to the nodal @xmath1601-curve @xmath1602 has invariant @xmath1578 .
since the strict transform of @xmath1603 in @xmath1604 belongs to @xmath1605 , @xmath4 has parity @xmath338 .
thus we obtain a period map @xmath1606 .
[ birat ( 10,6,0 ) ] the map @xmath30 is birational .
we consider an @xmath1564-covering @xmath545 whose fiber over a @xmath1406 corresponds to labelings of the eight nodes @xmath574 of @xmath1602 . in the familiar way we will obtain a birational lift @xmath1607 of @xmath30 .
the variety @xmath1608 is an @xmath1081-cover of @xmath1598 for the lattice @xmath1609 .
by @xcite we have @xmath1570 .
[ rational ( 10,6,0 ) ] the quotient @xmath1610 is rational .
therefore @xmath1598 is rational .
this is a consequence of the slice method for the projection @xmath1611 , proposition [ stabilizer of section ] , and katsylo s theorem [ katsylo ] .
recall that the quotient @xmath1612 is birational to the moduli @xmath1613 of genus @xmath911 hyperelliptic curves ( proposition [ he moduli ] ) , and that the stabilizer in @xmath1614 of a general @xmath226 is generated by its hyperelliptic involution ( corollary [ stab he ] ) . since @xmath1615
, we have @xmath1616 .
then the restriction map @xmath1617 is isomorphic because @xmath1618 .
these show that [ fixed curve ( 10,6,0 ) ] the fixed curve map for @xmath1598 gives a dominant morphism @xmath1619 whose general fibers are birationally identified with the quotients of the bi - canonical systems by the hyperelliptic involutions .
we note that the hyperelliptic locus @xmath1613 is the theta - null divisor of @xmath1575 .
let @xmath1559 be the locus of pairs @xmath1560 such that @xmath210 is smooth , @xmath608 is the union of distinct lines with @xmath1621 lying on @xmath210 , and each component of @xmath608 is transverse to @xmath210 .
the point @xmath171 is a @xmath433-singularity of the sextic @xmath1561 , and the rest singularities of @xmath1561 are the six nodes @xmath1622 .
the 2-elementary @xmath0 surface associated to @xmath1561 has invariant @xmath1623 .
thus we obtain a period map @xmath1624 .
[ birat ( 11,5,1 ) ] the map @xmath30 is birational .
let @xmath1625 be the locus of those @xmath1626 such that @xmath1627 and that @xmath871 lie on the same component of @xmath608 . for a point @xmath1628 of @xmath543 , the six nodes of @xmath1561 and the two components of @xmath608 are labelled in a compatible way .
in particular , the three tangents at the @xmath433-point of @xmath1561 are also distinguished .
thus @xmath30 lifts to a birational map @xmath1629 .
the space @xmath1630 is an @xmath703-cover of @xmath1012 , while @xmath1631 is an @xmath1081-cover of @xmath1620 for the lattice @xmath1632 .
by @xcite we calculate @xmath732 .
[ rational ( 11,5 ) ] the quotient @xmath1012 is rational .
hence @xmath1620 is rational .
let @xmath1572 be the universal genus @xmath911 curve ( over an open locus ) , and let @xmath955 be the subbundle of @xmath1633 whose fiber over a @xmath1469 is @xmath1634 .
the natural map @xmath1635 , @xmath1636 , shows that @xmath1012 is birational to the symmetric product of @xmath892 over @xmath1637 .
in particular , @xmath1012 is birational to @xmath1638 .
the space @xmath1637 is known to be rational : see @xcite where this fact is attributed to shepherd - barron .
it is a consequence of the slice method and miyata s theorem ( associate to a pointed quartic @xmath1469 the pointed tangent line @xmath1639 ) .
[ fixed curve ( 11,5 ) ] the fixed curve map for @xmath1620 lifts to a dominant map @xmath1640 whose fiber over a general @xmath1469 is birationally identified with the symmetric product of the pencil @xmath1641 .
one can also prove propositions [ rational ( 10,6,1 ) ] and [ rational ( 11,5 ) ] by considering the configuration @xmath1642 of points and using the no - name lemma .
this avoids resorting to the rationality of @xmath1575 .
for @xmath1645 let @xmath1646 be the locus of triplets @xmath736 such that @xmath341 @xmath210 is smooth and passes @xmath1647 , @xmath344 @xmath739 is transverse to @xmath210 , and @xmath346 @xmath737 is tangent to @xmath210 at @xmath171 with multiplicity @xmath1321 and transverse to @xmath210 elsewhere . for @xmath1648 the point @xmath171
is an inflectional point of @xmath210 of order @xmath1321 .
the sextic @xmath1649 has the @xmath1341-point @xmath171 , the @xmath1650 nodes @xmath1651 , and no other singularity .
taking the right resolution of @xmath28 , we obtain a period map @xmath1652 . here
@xmath41 for @xmath1426 , and @xmath338 for @xmath1653 . [ birat ( 8+k,8-k ) ] the map @xmath1348 is birational .
see example [ ex:4 ] for the case @xmath1653 . for @xmath1432 , we label the three nodes @xmath1654 and the two nodes @xmath1655 independently .
this is realized by an @xmath1437-covering @xmath1656 as before .
note that @xmath737 and @xmath739 are distinguished by their intersection with @xmath210 .
then we will obtain a birational lift @xmath1657 of @xmath1658 .
the invariant lattice @xmath529 is isometric to @xmath1659 .
by @xcite we have @xmath1660 .
therefore @xmath1658 is birational . for @xmath1347
, we label only the three nodes @xmath1654 , which defines an @xmath790-covering @xmath1661 .
the rest one node @xmath1655 is obviously distinguished from those three .
therefore we will obtain a birational lift @xmath1662 of @xmath1663 .
the invariant lattice @xmath529 is isometric to @xmath1664 .
then we have @xmath1665 , which proves the proposition . [ rational ( 8+k,8-k ) ]
the quotient @xmath1666 is rational .
therefore @xmath1667 , @xmath1668 , and @xmath1669 are rational .
consider the projection @xmath1670 .
the group @xmath811 acts almost transitively on @xmath1671 with the stabilizer of a general @xmath1672 being isomorphic to @xmath1550 .
the @xmath117-fiber over @xmath1672 is an open set of the linear system of quartics which are tangent to @xmath737 at @xmath1673 with multiplicity @xmath1674 .
thus our assertion follows from the slice method for @xmath117 and miyata s theorem .
the fixed curve maps are described as follows .
recall that the ordinary inflectional points of a smooth quartic @xmath1675 are just the normal weierstrass points of @xmath210 , and the inflectional points of order @xmath738 are just the weierstrass points of weight @xmath45 .
then let @xmath1637 be the universal genus @xmath911 curve , @xmath1676 the locus of normal weierstrass points , and @xmath1677 the locus of weierstrass points of weight @xmath45 . the fixed curve map for @xmath1667 ( resp .
@xmath1643 with @xmath1648 ) lifts to a dominant map @xmath1678 ( resp .
@xmath1679 ) whose general fibers are birationally identified with the pencils @xmath1641 .
in this section we treat the case @xmath1680 , @xmath371 . in view of the unirationality result @xcite
, we may assume @xmath1681 .
the case @xmath1682 is reduced to the rationality of @xmath1520 via the structure as an arithmetic quotient . the case @xmath1683 is settled by analyzing the quotient rational surfaces . the cases @xmath1684 and @xmath1685 are studied using genus @xmath45 curves on @xmath1068 , and the case @xmath1686 with @xmath41 is studied using cuspidal plane quartics . here
we may take the same approach as the one for @xmath19 by kond @xcite . recall that @xmath1687 is an open set of the modular variety @xmath1688 where @xmath6 is the 2-elementary lattice @xmath1689 .
since we have canonical isomorphisms @xmath1690 , the variety @xmath1688 is isomorphic to @xmath1691 .
the lattice @xmath1692 is isometric to @xmath1693 , so that @xmath1691 is birational to @xmath1520 . in [ ssec:(10,4,0 ) ] we proved that @xmath1520 is rational , and hence @xmath1687 is rational .
we construct the corresponding right dpn pairs starting from the hirzebruch surface @xmath684 . for a smooth curve
@xmath1695 transverse to @xmath115 , let @xmath1696 be the double cover branched along @xmath210 . since we have a conic fibration @xmath1697 , the surface @xmath29 is rational .
the curve @xmath1698 is a @xmath411-curve on @xmath29 .
[ lemma 1 ( 10,8,1 ) ] we have @xmath1699 . by the ramification formula we have @xmath1700 .
since @xmath1701 and @xmath1702 , it suffices to show that @xmath1703 .
we take a smooth curve @xmath189 transverse to @xmath210 .
the inverse image @xmath1704 is a smooth genus @xmath45 curve disjoint from @xmath1698 . since @xmath1705
, we have @xmath1706 by the adjunction formula .
then the exact sequence @xmath1707 shows that @xmath1708 .
thus the resolution of the bi - anticanonical map @xmath1709 of @xmath29 is given by the composition of @xmath1696 and the morphism @xmath1710 associated to @xmath463 .
the image @xmath1711 is a quadratic cone with vertex @xmath1712 . in this way
the quotient morphism @xmath478 is recovered from the bi - anticanonical map of @xmath29 .
it follows that [ lemma 2 ( 10,8,1 ) ] for two smooth curves @xmath1713 transverse to @xmath115 , the associated double covers @xmath1714 are isomorphic if and only if @xmath210 and @xmath1715 are @xmath1121-equivalent .
now we let @xmath1716 be the open locus of pairs @xmath569 such that @xmath210 and @xmath174 are smooth and that @xmath210 is transverse to @xmath1717 . to a @xmath1406 we associate the right dpn pair @xmath375 where @xmath1696 is the double cover branched along @xmath210 and @xmath1718 .
since @xmath375 has invariant @xmath1719 , we obtain a period map @xmath1720 . by lemmas
[ lemma 1 ( 10,8,1 ) ] and [ lemma 2 ( 10,8,1 ) ] the induced map @xmath1721 is generically injective . in view of the equality @xmath1722
, we have a birational equivalence @xmath1723 .
[ rational ( 10,8,1 ) ] the quotient @xmath1724 is rational .
hence @xmath1694 is rational .
this follows from the slice method for the projection @xmath1725 , proposition [ stabilizer of section ] , and katsylo s theorem [ katsylo ] .
[ second proof ( 10,8,1 ) ] the above @xmath375 are ( generically ) right resolution of the plane sextics @xmath1561 where @xmath210 is a one - nodal quartic and @xmath608 is a smooth conic transverse to @xmath210 .
although the period map for this sextic model has degree @xmath1726 , we can analyze its fibers to derive the rationality of @xmath1694 .
this alternative approach may also be of some interest .
we here give an outline for future reference .
let @xmath1727 be the locus of @xmath1728 such that @xmath210 is one - nodal and @xmath1729 for a smooth conic @xmath608 . using the labeling @xmath1730
, we obtain a birational map @xmath1731 as before .
the projection @xmath1732 is an @xmath1733-covering , which exceeds the the obvious @xmath1564-symmetry of @xmath1734 . in order to find the rest @xmath1735-symmetry ,
let @xmath174 be the group of even cardinality subsets of @xmath1736 with the symmetric difference operation .
for @xmath1737 and @xmath1738 , consider the quadratic transformation @xmath1739 based at @xmath1740 , and @xmath1741 .
we set @xmath1742 , and @xmath1743 for @xmath1744 .
then we have @xmath1745 , and this defines an action of @xmath174 on @xmath1734
. the element @xmath1746 acts on @xmath1734 trivially ( it gives the covering transformation of the above @xmath1747 ) .
now the period map @xmath1748 is @xmath1749-invariant , so that @xmath1694 is birational to @xmath1750 by a degree comparison .
we can prove that @xmath1750 is rational .
we construct 2-elementary @xmath0 surfaces using curves on @xmath1068 .
let @xmath1752 be the open set of pairs @xmath686 such that @xmath210 and @xmath451 are smooth and transverse to each other .
then the curves @xmath688 belong to @xmath1071 . taking the right resolution of @xmath688
, we obtain a period map @xmath1753 .
the map @xmath30 is birational .
we consider an @xmath1564-covering @xmath545 whose fiber over a @xmath1754 corresponds to labelings of the eight nodes @xmath693 .
the rest node of @xmath688 is the one point @xmath1755 , which is obviously distinguished from those eight .
thus we obtain a birational lift @xmath1756 of @xmath30 .
the projection @xmath1757 is an @xmath1081-covering for the lattice @xmath1758 .
by @xcite we have @xmath1570 , which implies that @xmath30 is birational .
[ rational ( 11,7 ) ] the quotient @xmath1759 is rational .
therefore @xmath1751 is rational .
we apply the slice method to the projection @xmath1760 , and then use proposition [ l_1,1 ] and miyata s theorem . for every smooth @xmath226
the linear system @xmath163 is identified with @xmath1761 by restriction .
indeed , we have @xmath1762 by the adjunction formula , and @xmath1763 because @xmath1764 .
then the vanishings @xmath1765 prove our claim .
in view of proposition [ he moduli ] and corollary [ stab he ] , we have [ fixed curve ( 11,7 ) ] the fixed curve map @xmath1766 is dominant with a general fiber being birationally identified with the quotient of @xmath1761 by the hyperelliptic involution .
we construct 2-elementary @xmath0 surfaces using curves on @xmath1068 .
for @xmath1770 let @xmath1771 be the locus of triplets @xmath1772 such that @xmath341 @xmath210 and @xmath174 are smooth and transverse to each other , @xmath344 @xmath53 intersects with @xmath210 at @xmath1773 with multiplicity @xmath1774 in case @xmath1426 , and is transverse to @xmath571 in case @xmath1436 .
it is easy to calculate @xmath1775 . for a @xmath1776
the curve @xmath1777 belongs to @xmath1071 .
when @xmath1436 , @xmath28 has only nodes as the singularities . when @xmath1426 , @xmath28 has the @xmath1778-singularity @xmath171 , the nodes @xmath1779 , @xmath1780 , @xmath228 , and no other sigularity . the 2-elementary @xmath0 surface @xmath4 associated to @xmath28 has invariant @xmath1781 .
when @xmath1347 , @xmath4 has parity @xmath338 .
indeed , let @xmath388 be the corresponding right dpn pair .
the curve @xmath412 has two components over @xmath171 , say @xmath1782 and @xmath1783 , whose numbering corresponds to the one for the vertices of the @xmath674-graph in figure [ dual graph ] .
then the sum of @xmath1784 and the strict transform of @xmath1785 belongs to @xmath1605 , which proves @xmath338 .
thus we obtain a period map @xmath1786 where @xmath41 if @xmath1787 and @xmath338 if @xmath1347 .
[ birat g=2 ii ] the map @xmath1348 is birational . for @xmath1436
we label the six nodes @xmath574 and the two nodes @xmath1075 independently , which is realized by an @xmath1788-covering @xmath1789 . the rest two nodes , @xmath1790 and @xmath228 ,
are distinguished by the irreducible decomposition of @xmath28 .
this defines a birational lift @xmath1791 of @xmath1792 .
the variety @xmath1793 is an @xmath1081-cover of @xmath1767 for the lattice @xmath1794 .
by @xcite we have @xmath617 .
therefore @xmath1792 has degree @xmath38 . for @xmath1426
we consider labelings the five nodes @xmath1795 , which defines an @xmath1261-cover @xmath1796 of @xmath1326 .
the rest singularities of @xmath28 and the branches at the @xmath1778-point are a priori distinguished .
thus we obtain a birational lift @xmath1797 of @xmath1348 .
the variety @xmath1798 is an @xmath519-cover of @xmath1799 for the lattice @xmath1800 .
we have @xmath1801 for both @xmath1426 .
hence @xmath1348 is birational .
[ rational g=2 ii ] the quotient @xmath1802 is rational .
therefore @xmath1767 , @xmath1768 , and @xmath1769 are rational .
we apply the slice method to the projection @xmath1803 . by propositions [ stabilizer of section ] and [ stabilizer of fiber ] the group @xmath1804 acts on @xmath1805 almost transitively with the stabilizer @xmath26 of a general
@xmath1806 being connected and solvable .
the fiber @xmath1807 is an open set of a linear subspace @xmath1808 of @xmath164 .
then @xmath1809 is rational by miyata s theorem . as in the paragraph just before corollary [ fixed curve ( 11,7 ) ] , we see that for every smooth @xmath226 the linear system @xmath123 is identified with @xmath1810 by restriction . when @xmath1436 , we obtain a pointed genus @xmath45 curve @xmath1469 by considering either of the two points @xmath229 . when @xmath1644 , the point @xmath1811 determines @xmath53 , and @xmath174 is a general member of @xmath1812 . for @xmath1347 ,
@xmath171 is a weierstrass point of @xmath210 .
these infer the following . [
fixed curve g=2 , 2<k<6 ] let @xmath1813 be the moduli of pointed genus @xmath45 curves @xmath1469 , and let @xmath1814 be the divisor of weierstrass points .
the fixed curve maps for @xmath1767 , @xmath1768 , and @xmath1769 lift to rational maps @xmath1815 .
then @xmath428 @xmath1816 is dominant with a general fiber birationally identified with the quotient of @xmath1810 by the hyperelliptic involution .
@xmath431 @xmath1817 is dominant with a general fiber birationally identified with @xmath1812 .
@xmath434 @xmath1818 is a dominant map onto @xmath1819 whose general fiber is birationally identified with the quotient of @xmath1812 by the hyperelliptic involution .
let @xmath1821 be the locus of plane quartics with an ordinary cusp and with no other singularity . for @xmath1822
we denote by @xmath1823 the locus of pairs @xmath1256 such that if @xmath1824 is the tangent line of @xmath210 at the cusp , then @xmath500 intersects with @xmath210 at @xmath1825 with multiplicity @xmath1826 , and is transverse to @xmath210 elsewhere .
the space @xmath1326 is of dimension @xmath1827 .
this is obvious for @xmath1828 . for @xmath1829 , if we take the homogeneous coordinate @xmath1830 $ ] of @xmath35 and normalize @xmath1831 $ ] , @xmath1832 , and @xmath1833 , then quartics @xmath210 having cusp at @xmath171 with @xmath1834 and @xmath1835 are defined by the equations @xmath1836 in which the coefficients @xmath1837 may be taken general .
this shows that @xmath1838 is of the expected dimension . for a @xmath1839 the sextic
@xmath1840 has an @xmath409-singularity at @xmath1841 , @xmath1650 nodes at @xmath1842 , a @xmath1843-singularity at @xmath1844 ( resp .
two nodes at @xmath1844 and @xmath1845 ) in case @xmath1846 ( resp .
@xmath1347 ) , and no other singularity .
hence the 2-elementary @xmath0 surface @xmath4 associated to @xmath28 has invariant @xmath1781 .
when @xmath1347 , @xmath4 has parity @xmath41 by lemma [ delta=1 ] @xmath428 .
thus we obtain a period map @xmath1847 .
[ birat g=2 iii ] the map @xmath1348 is birational .
we consider an @xmath1848-covering @xmath1541 whose fiber over a @xmath1839 corresponds to labelings the @xmath1650 nodes @xmath1842 . as described above , the rest singular points of @xmath28 are a priori distinguished by the type of singularity and by the irreducible decomposition of @xmath28 .
also note that the two lines @xmath500 and @xmath1849 are distinguished by their intersection with @xmath210 .
therefore we will obtain a birational lift @xmath1850 of @xmath1348 .
the anti - invariant lattice @xmath6 is isometric to @xmath1851 .
it is easy to check that @xmath1852 . since @xmath811 acts on @xmath25
almost freely , this shows that @xmath1348 has degree @xmath38 .
[ rational g=2 iii ] the quotient @xmath1666 is rational .
therefore @xmath1820 with @xmath1822 is rational .
let @xmath1853 be the locus of triplets @xmath1854 such that @xmath1855 .
we have the @xmath811-equivariant map @xmath1856 , @xmath1857 , where @xmath1849 is the tangent line of @xmath210 at the cusp .
the group @xmath811 acts on @xmath84 almost transitively with the stabilizer of a general point isomorphic to @xmath1858 . since a general @xmath1859-fiber is an open set of a sub - linear system of @xmath1018 , the assertion follows from the slice method for @xmath1859 and miyata s theorem .
let @xmath1813 be the moduli of pointed genus @xmath45 curves @xmath1469 .
recall ( cf .
@xcite ) that we have a birational map @xmath1860 by associating to a pointed curve @xmath1469 the image @xmath1861 by the linear system @xmath1862 .
when @xmath171 is not a weierstrass point , @xmath1408 is a quartic with a cusp @xmath1863 , and the projection from @xmath1863 gives the hyperelliptic map of @xmath210 .
thus we see the following .
the fixed curve map for @xmath1820 with @xmath1822 lifts to a rational map @xmath1864 .
then @xmath428 @xmath1818 is dominant with general fibers birationally identified with @xmath1862 .
@xmath431 @xmath1865 is dominant with general fibers birationally identified with @xmath1866 .
@xmath434 @xmath1867 is birational .
@xmath437 @xmath1868 is generically injective with a generic image being the divisor of those @xmath1469 with @xmath1869 effective minus the divisor of weierstrass points .
in this section we study the case @xmath66 , @xmath1870 , @xmath41 .
it seems ( at least to the author ) difficult to use the previous methods for this case .
we obtain general members of @xmath10 from plane sextics of the form @xmath1871 where @xmath1872 is a nodal cubic and @xmath1873 is a smooth cubic intersecting with @xmath1872 at the node with suitable multiplicity .
but the period map for this sextic model has degree @xmath1726 . in order to pass to a `` canonical '' construction
, we find a weyl group symmetry in this sextic model , which reflects the fact that the galois group of @xmath1874 is the weyl group or its central quotient .
this leads to regard the cubics @xmath1875 as anticanonical curves on del pezzo surfaces of degree @xmath308 .
we begin with constructing a universal family of marked del pezzo surfaces with an equivariant action by the weyl group . for @xmath1876
let @xmath1877 be the open set of ordered @xmath797 points in general position in the sense of @xcite .
there exists a geometric quotient @xmath1878 of @xmath1879 by @xmath811 ( see @xcite ) .
since any @xmath1880 has trivial stabilizer , the quotient map @xmath1881 is a principal @xmath811-bundle by luna s etale slice theorem .
hence by @xcite proposition 7.1 the projection @xmath1882 descends to a smooth morphism @xmath1883 where @xmath1884 is a geometric quotient of @xmath1885 by @xmath811 . here
@xmath811 acts on @xmath35 in the natural way .
for @xmath1886 we have the section @xmath1887 of @xmath1883 induced by the @xmath663-th projection @xmath1888 , @xmath1889 .
blowing - up the @xmath797 sections @xmath1887 , we obtain a family @xmath1890 of del pezzo surfaces of degree @xmath1891 .
the exceptional divisor over @xmath1887 , denoted by @xmath1892 , is a family of @xmath397-curves .
let @xmath1893 be the lattice @xmath1894 with a natural orthogonal basis @xmath1895 . for each @xmath1896 ,
the picard lattice of the @xmath478-fiber @xmath1897 over @xmath1898 is isometric to @xmath1893 by associating @xmath751 to the pullback of @xmath461 and @xmath1899 to the class of the @xmath397-curve @xmath1900 .
this gives a trivialization @xmath1901 of the local system .
thus we have a universal family @xmath1902 of marked del pezzo surfaces .
recall that the weyl group @xmath1903 is the group of isometries of @xmath1893 which fix the vector @xmath1904 .
let @xmath1905 . for each @xmath1896 the marked del pezzo surface @xmath1906 is transformed by @xmath1907 to @xmath1908 , which is isomorphic to @xmath1909 for a @xmath1910 .
indeed , the classes @xmath1911 are represented by disjoint @xmath397-curves @xmath1912 on @xmath1897 which define a blow - down @xmath1913
. then @xmath856 is the blown - down points of @xmath1914 .
we set @xmath1915 . by construction
, we have an isomorphism @xmath1916 with @xmath1917 .
the last equality characterizes @xmath1918 uniquely because the cohomological representation @xmath1919 is injective .
this ensures that @xmath1920 and that @xmath1921 for @xmath1922 . in this way we obtain an equivariant action of @xmath1903 on the family @xmath1890 .
the @xmath1903-action on @xmath1923 is the cremona representation . in the following
we will refer to @xcite for the basic properties of the cremona representation for each @xmath797 .
on @xmath1924 we have two natural @xmath1903-linearized vector bundles : the relative tangent bundle @xmath1925 and the relative anticanonical bundle @xmath1926 .
the direct image @xmath1927 is a @xmath1903-linearized bundle over @xmath1923 .
the fiber of the @xmath1928-bundle @xmath1929 over a point @xmath1930 $ ] of @xmath1923 is identified with the linear system of plane cubics through the @xmath797 points @xmath1931 of @xmath35 .
let @xmath1933 be the family of marked del pezzo surfaces of degree @xmath38 constructed above .
the cremona representation of the weyl group @xmath1934 has kernel @xmath1935 whose generator @xmath1936 acts on the @xmath478-fibers by the bertini involutions . the quotient @xmath1937 is isomorphic to @xmath1938 and acts on @xmath1939 almost freely .
we let @xmath1940 be the locus of singular anticanonical curves .
the locus @xmath1941 is invariant under the @xmath1934-action on @xmath1929 , and the projection @xmath1942 is of degree @xmath1943 .
we denote by @xmath955 the pullback of the bundle @xmath1927 by @xmath1942 .
an open set of the variety @xmath892 parametrizes triplets @xmath1944 of a marked del pezzo surface @xmath1945 , a nodal @xmath1946-curve @xmath1872 , and a smooth @xmath1946-curve @xmath1873 .
notice that @xmath1872 is irreducible and transverse to @xmath1873 because @xmath1947 . taking a right resolution of the dpn pair @xmath1948
, we obtain a 2-elementary @xmath0 surface of invariant @xmath1949 .
this defines a period map @xmath1950 .
[ birat ( 11,9 ) ] the map @xmath30 is @xmath1934-invariant and descends to a birational map @xmath1951 .
two @xmath1934-equivalent points of @xmath892 give rise to isomorphic dpn pairs so that @xmath30 is @xmath1934-invariant .
we shall show that @xmath30 lifts to a birational map @xmath1952 .
for a point @xmath1953 of @xmath892 , we have the marking @xmath1954 of the picard lattice of @xmath1955 induced by @xmath1898 .
the curve @xmath1871 has two nodes , namely the intersection point @xmath1956 and the node of @xmath1872 , which are clearly distinguished .
this induces a marking of the invariant lattice of @xmath1957 , which defines a lift @xmath1958 of @xmath30 . in order to show that @xmath553 is birational , the key point is that the blow - down @xmath1959 defined by @xmath1898 translates the triplet @xmath1953 into the plane sextic @xmath1960 endowed with a labeling of the nine nodes @xmath1961 .
indeed , @xmath1961 consists of the eight ordered points @xmath1898 and the rest one point @xmath1962 .
the first eight nodes recover the marked del pezzo surface , and the ninth is determined as the base point of the associated cubic pencil .
note also that the unlabeled node @xmath1963 of @xmath1960 is clearly distinguished from those nine . in this way
@xmath892 is birationally identified with the @xmath811-quotient of the space of such nodal sextics with labelings .
now one may follow the recipe in [ ssec : recipe ] to see that @xmath553 is birational .
we compare the two projections @xmath1964 and @xmath1965 .
the bertini involution @xmath1966 acts trivially on @xmath892 because it acts trivially on the anticanonical pencils @xmath1967 .
hence @xmath1964 is an @xmath1938-covering . on the other hand
, @xmath1968 is an @xmath1081-cover of @xmath1932 for the lattice @xmath1969 .
since @xmath1970 , this finishes the proof .
the @xmath1934-action on @xmath892 gets rid of the markings of del pezzo surfaces .
this implies that the quotient @xmath1971 is birationally a moduli of triplets @xmath1972 where @xmath29 is an ( unmarked ) del pezzo surface of degree @xmath38 , and @xmath1872 ( resp .
@xmath1873 ) is a singular ( resp .
smooth ) @xmath1946-curve on @xmath29 .
we consider the blow - up @xmath1973 at the base point @xmath1956 of @xmath1967 .
the quotient of @xmath1974 by the bertini involution is the hirzebruch surface @xmath684 , and the quotient morphism @xmath1975 is branched over the @xmath50-curve @xmath115 and over a smooth @xmath1530-curve @xmath966 .
note that @xmath1976 is the exceptional curve of @xmath1973 .
the pencil @xmath124 on @xmath684 is pulled - back by @xmath969 to the pencil @xmath1977 .
then @xmath1978 is an @xmath1979-fiber tangent to @xmath966 , and @xmath1980 is an @xmath1979-fiber transverse to @xmath966 . if we let @xmath1981 be the locus of triplets @xmath1982 such that @xmath966 is smooth and @xmath1983 ( resp .
@xmath833 ) is tangent ( resp .
transverse ) to @xmath966 , we thus obtain a rational map @xmath1984 since this construction may be reversed , the map is birational .
[ rational ( 11,9 ) ] the quotient @xmath1724 is rational .
thus @xmath1932 is rational .
let @xmath84 be the fiber product @xmath1985 , where @xmath1986 is the natural projection .
we consider the @xmath1121-equivariant map @xmath1987 where @xmath171 is the point of tangency of @xmath966 and @xmath1983 , and @xmath1988 . by
[ sec : hirze ] , the group @xmath1121 acts on @xmath1989 almost transitively .
if we normalize @xmath1990)$ ] and @xmath1991)$ ] , the stabilizer @xmath26 of a general point @xmath1992 with @xmath1993 is described by the exact sequence @xmath1994 , where @xmath1995)=0\}$ ] . in particular , @xmath26 is connected and solvable . by the slice method
we have @xmath1996 .
the fiber @xmath1997 is an open set of the linear system @xmath1998 of curves @xmath966 with @xmath1999 . by miyata
s theorem the quotient @xmath2000 is rational .
let us begin with few remarks .
let @xmath29 be a quadric del pezzo surface and let @xmath965 be its anticanonical map , which is branched along a smooth quartic @xmath966 .
the pullback @xmath2002 of a line @xmath2003 is singular if and only if @xmath500 is tangent to @xmath966 .
therefore the reduced curve @xmath2004 , the fixed locus of the geiser involution , is the locus of singular points of anticanonical curves on @xmath29 . for every @xmath2005 we have a unique singular @xmath1946-curve @xmath2006 with @xmath2007 , which is the pullback of the tangent line @xmath2008 of @xmath966 at @xmath1863 .
the singular curve @xmath2006 is irreducible and nodal if and only if @xmath2008 is an ordinary tangent line . in this case
, the geiser involution exchanges the two tangents of @xmath2006 at @xmath171 . on the other hand , @xmath2006 has only one tangent at @xmath171 if and only if @xmath1863 is an inflectional point , i.e. , a weierstrass point of @xmath966 .
now let @xmath2009 be the family of marked quadric del pezzo surfaces constructed in
[ ssec : equiv weyl ] .
the cremona representation of the weyl group @xmath2010 has kernel @xmath1935 whose generator @xmath1936 acts on the @xmath478-fibers by the geiser involutions . the quotient @xmath2011 is isomorphic to @xmath2012 and acts on @xmath2013 almost freely .
let @xmath2014 be the fixed locus of @xmath1936 .
we define a double cover @xmath2015 of @xmath2016 as the locus in @xmath2017 of triplets @xmath2018 such that @xmath2019 is a tangent at @xmath171 of the anticanonical curve on @xmath2020 singular at @xmath171 .
the locus @xmath2015 is invariant under the @xmath2010-action on @xmath2017 .
in particular , the geiser involution @xmath1936 acts on @xmath2015 by the covering transformation of @xmath2021 .
the branch divisor of @xmath2022 is the family of the weierstrass points .
we pull back the bundle @xmath1927 on @xmath2013 by the projection @xmath2023 , and consider its subbundle @xmath955 whose fiber over a @xmath2024 is the vector space of anticanonical forms on @xmath2020 vanishing at @xmath171 . an open set of the variety @xmath892 parametrizes quadruples @xmath2025 such that @xmath2026 is a marked quadric del pezzo surface , @xmath1872 is an irreducible nodal @xmath1946-curve , @xmath2027 is one of the tangents of @xmath1872 at the node , and @xmath1873 is a smooth @xmath1946-curve passing through the node of @xmath1872 .
then @xmath1871 is a @xmath27-curve with the @xmath433-singularity @xmath1956 and with no other singularity .
its two tangents given by @xmath1872 at the @xmath433-point are distinguished by @xmath2027 .
the 2-elementary @xmath0 surface associated to the dpn pair @xmath1948 has invariant @xmath2028 .
thus we obtain a period map @xmath2029 .
[ birat ( 12,8 ) ] the map @xmath30 is @xmath2010-invariant and descends to a birational map @xmath2030 .
the @xmath2010-invariance of @xmath30 is straightforward .
the marking @xmath1954 of @xmath2031 and the labeling @xmath2027 for the @xmath433-point @xmath1956 induce a marking of the invariant lattice of @xmath2032 .
this defines a lift @xmath2033 of @xmath30 . as in [ ssec:(11,9 ) ] , the blow - down @xmath2034 determined by @xmath1898 translates the quadruple @xmath2035 into the plane sextic @xmath1960 endowed with a labeling of its seven nodes , @xmath1898 , and of the two tangents at its @xmath433-point @xmath1962 given by @xmath2036 . applying the recipe in [ ssec : recipe ] for such sextics with labelings , we see that @xmath553 is birational .
the projection @xmath2037 is an @xmath519-covering for the lattice @xmath2038 .
by @xcite we have @xmath2039 .
since @xmath2010 acts on @xmath892 almost freely , this concludes that the induced map @xmath2030 is birational .
we shall prove that @xmath2040 is rational .
the weyl group @xmath2010 acts on the markings of del pezzo surfaces , and the geiser involution @xmath2041 is the covering transformation of @xmath2042 .
these facts infer that @xmath2040 is birationally a moduli space of triplets @xmath2043 where @xmath29 is an ( unmarked ) quadric del pezzo surface , @xmath2044 is a fixed point of the geiser involution , and @xmath210 is a @xmath1946-curve passing through @xmath171 .
the anticanonical map @xmath2045 translates @xmath2043 into the triplet @xmath2046 such that @xmath966 is a smooth plane quartic , @xmath2047 , and @xmath500 is a line through @xmath743 . if @xmath2048 denotes the space of such triplets @xmath2046 , we obtain a birational equivalence @xmath2049 in proposition [ rational ( 8+k,8-k ) ] ( for @xmath1432 ) we proved that this quotient @xmath1012 is rational .
therefore [ rational ( 12,8 ) ] the moduli space @xmath2001 is rational .
[ 2nd map ] by the final step of the proof , we have a natural birational map @xmath2050 via @xmath1012 .
on the other hand , the referee suggested that we have another birational map @xmath2050 as follows ( cf .
appendix ) . for the anti - invariant lattice @xmath2038 for @xmath2001 , consider the odd lattice @xmath2051 .
its even part @xmath1849 is isometric to @xmath2052 by example [ i_n&d_n ] .
one checks that the element in @xmath2053 giving the extension @xmath2054 is invariant under @xmath2055 .
therefore we have an isomorphism @xmath2056 of modular varieties by proposition [ even part ] , and @xmath2057 is birational to @xmath1667
. it would be interesting whether these two maps @xmath2050 coincide .
anyway , the second construction and the rationality of @xmath1667 ( [ ssec:(8+k,8-k ) ] ) provide another proof of the rationality of @xmath2001 .
we begin with the remark that for every point @xmath171 of a cubic del pezzo surface @xmath29 , there uniquely exists a @xmath1946-curve @xmath2006 singular at @xmath171 . indeed , if we embed @xmath29 in @xmath2059 naturally , @xmath2006 is the intersection of @xmath29 with the tangent plane of @xmath29 at @xmath171 .
when @xmath171 is generic , @xmath2006 is irreducible and nodal . for later use
, we also explain an alternative construction .
the blow - up @xmath1974 of @xmath29 at a general @xmath2044 is a quadric del pezzo surface .
let @xmath2060 be the @xmath397-curve over @xmath171 , and let @xmath2061 be the anticanonical map of @xmath1974 which is branched over a smooth quartic @xmath966 .
then @xmath2062 is a bitangent of @xmath966 , and @xmath2063 where @xmath2 is the geiser involution of @xmath1974 .
since @xmath2064 is a @xmath2065-curve , the image of @xmath2066 in @xmath29 gives the desired curve @xmath2006 .
the fact that @xmath969 and @xmath2 are given by the projection @xmath2067 from @xmath171 connects these two constructions .
let @xmath2068 be the family of marked cubic surfaces constructed in
[ ssec : equiv weyl ] .
the weyl group @xmath2069 acts on @xmath2070 almost freely because a generic cubic surface has no automorphism .
we shall denote a point of the variety @xmath2071 as @xmath2018 where @xmath2072 , @xmath2073 , and @xmath2074 .
then we let @xmath2075 be the locus of @xmath2018 such that @xmath2027 is one of the tangents of the anticanonical curve @xmath2006 singular at @xmath171 .
the locus @xmath2076 is @xmath2069-invariant and is a double cover of @xmath2077 .
the branch divisor of @xmath2078 is the family of the hessian quartics restricted to the marked cubic surfaces .
we pull - back the vector bundle @xmath1927 on @xmath2070 by the projection @xmath2079 , and consider its subbundle @xmath955 whose fiber over a @xmath2080 is the space of anticanonical forms on @xmath2081 which vanish at @xmath171 and whose first derivatives at @xmath171 vanish at @xmath2027 .
then @xmath955 is a @xmath2069-linearized vector bundle over @xmath2076 .
an open set of the variety @xmath892 parametrizes triplets @xmath2082 where @xmath2083 is a marked cubic surface , @xmath2044 is such that the singular @xmath1946-curve @xmath2006 is irreducible and nodal , and @xmath210 is a smooth @xmath1946-curve with @xmath2084 . by @xmath2027
is chosen which branch of @xmath2006 at @xmath171 is tangent to @xmath210 .
the curve @xmath2085 has the @xmath674-singularity @xmath171 as its only singularity .
thus , taking the right resolution of the dpn pairs @xmath2086 , we obtain a period map @xmath2087 .
[ birat ( 13,7 ) ] the map @xmath30 is @xmath2069-invariant and descends to a birational map @xmath2088 .
this is analogous to propositions [ birat ( 11,9 ) ] and [ birat ( 12,8 ) ] : for a triplet @xmath2089 in @xmath892 , the blow - down @xmath2090 determined by @xmath1898 translates @xmath2089 into the plane sextic @xmath2091 endowed with the labeling @xmath1898 of its six nodes .
the rest singularity of @xmath2091 is the @xmath674-point @xmath2092 , at which the three branches of @xmath2091 are a priori distinguished by the irreducible decomposition of @xmath2091 and by the intersection multiplicity at @xmath2092 . by the recipe in [ ssec : recipe ]
we see that @xmath30 lifts to a birational map @xmath2093 .
the variety @xmath2094 is an @xmath519-cover of @xmath2058 for the lattice @xmath2095 . since @xmath2096
, our assertion is proved .
we shall prove that @xmath2097 is rational .
first we apply the no - name method to the @xmath2069-linearized bundle @xmath955 over @xmath2076 to see that @xmath2098 the variety @xmath2076 is a moduli of triplets @xmath2099 where @xmath2083 is a marked cubic surface , @xmath2044 , and @xmath2100 is one of the tangents of the singular @xmath1946-curve @xmath2006 .
the @xmath2069-action takes off the markings of surfaces .
let @xmath1973 be the blow - up at @xmath171 , and let @xmath97 be the @xmath397-curve over @xmath171 .
when @xmath171 is generic , @xmath1974 is a quadric del pezzo surface .
as explained in the beginning of [ sssec:(13,7 ) ] , if we regard @xmath2027 as a point @xmath743 of @xmath97 , then @xmath743 is contained in @xmath2101 where @xmath2 is the geiser involution of @xmath1974 . by applying the anticanonical map @xmath2102 of @xmath1974 , the new triplet @xmath2103 is then translated into a triplet @xmath2104 of a smooth quartic @xmath966 , a bitangent @xmath500 of @xmath966 , and a point @xmath920 in @xmath2105 .
therefore , if @xmath2106 denotes the space of pairs @xmath2107 of a smooth quartic @xmath966 and a tangent point @xmath920 of a bitangent of @xmath966 , we have a rational map @xmath2108 conversely , for a point @xmath2107 of @xmath25 , we take the double cover @xmath2061 branched along @xmath966 . if @xmath500 is the tangent line of @xmath966 at @xmath920 , we have @xmath2109 for a @xmath397-curve @xmath97 and the geiser involution @xmath2 of @xmath1974 .
since the triplet @xmath2110 is isomorphic to @xmath2111 by @xmath2 , we may contract either @xmath97 or @xmath2066 to obtain a well - defined inverse map of .
thus @xmath2112 is birational to @xmath1012 .
[ rational ( 13,7 ) ] the quotient @xmath1012 is rational .
hence @xmath2058 is rational . for a point @xmath2107 of @xmath25 ,
let @xmath608 be the another tangent point of the bitangent at @xmath920 .
this defines a @xmath811-equivariant map @xmath2113 , @xmath2114 .
the @xmath811-action on @xmath2115 is almost transitive with the stabilizer @xmath26 of a general @xmath2116 isomorphic to @xmath1550 . by the slice method
we have @xmath2117 .
the fiber @xmath2118 is an open set of the linear system @xmath818 of quartics @xmath966 with @xmath2119 . by miyata
s theorem @xmath1116 is rational .
we first note the following .
[ nodal -k curve ( 14,1 ) ] let @xmath29 be a quartic del pezzo surface and let @xmath2121 be a general point of @xmath2122 , where @xmath2044 and @xmath2100 .
then there uniquely exists an irreducible nodal @xmath1946-curve @xmath2123 whose node is @xmath171 with one of the tangents being @xmath2027 .
let @xmath2124 be the blow - up at @xmath171 with the exceptional curve @xmath2125 , and @xmath2126 be the blow - up at @xmath2127 with the exceptional curve @xmath97 .
then @xmath1974 is the blow - up of @xmath35 at seven points in almost general position in the sense of @xcite , and the strict transform @xmath451 of @xmath2125 is the unique @xmath50-curve on @xmath1974 .
hence the anticanonical map @xmath2061 is the composition of the contraction @xmath2128 of @xmath451 and a double covering @xmath2129 branched along a quartic @xmath966 with exactly one node .
the curve @xmath2130 is a line passing through the node and tangent to @xmath966 elsewhere .
we have @xmath2131 where @xmath2 is the involution of @xmath1974 induced by the covering transformation of @xmath2129 .
then the image of @xmath2066 in @xmath29 is the desired curve @xmath2123 .
the uniqueness of @xmath2123 follows from intersection calculation .
pulling back the pencil of lines through @xmath2132 , we obtain the pencil of @xmath2065-curves through @xmath2101 . then its image in @xmath1037
is the pencil @xmath2133 of @xmath1946-curves whose general member @xmath210 is smooth and passes @xmath171 and @xmath2027 with @xmath2134 .
( there is another pencil of @xmath1946-curves passing through @xmath2121 with @xmath2134 : the one of those singular at @xmath171 . )
now let @xmath2135 be the family of marked quartic del pezzo surfaces constructed in
[ ssec : equiv weyl ] .
the kernel of the cremona representation of the weyl group @xmath2136 is isomorphic to @xmath2137 , which is the automorphism group of a general @xmath478-fiber .
let @xmath2138 be the open set of triplets @xmath2018 such that there exists an anticanonical curve @xmath2123 on @xmath2139 as in lemma [ nodal -k curve ( 14,1 ) ] .
let @xmath1590 be the pullback of the bundle @xmath1927 by the natural projection @xmath2140 .
we consider the subbundle @xmath955 of @xmath1590 such that the fiber of @xmath892 over a @xmath2018 is the pencil @xmath2133 described above . by the uniqueness of @xmath2123 ,
@xmath955 is invariant under the @xmath2136-action on @xmath1590 .
an open set of @xmath892 parametrizes triplets @xmath2141 such that @xmath2142 is a marked quartic del pezzo surface , @xmath2123 is an irreducible nodal @xmath1946-curve , and @xmath210 is a smooth @xmath1946-curve with @xmath2134 .
the @xmath27-curve @xmath2143 has the @xmath2144-point @xmath171 as its only singularity .
the 2-elementary @xmath0 surface @xmath4 associated to the dpn pair @xmath2145 has invariant @xmath2146 .
we show that @xmath4 has parity @xmath41 .
if @xmath2147 is the blow - down given by @xmath1898 , the sextic @xmath2148 has the five nodes @xmath1898 , the @xmath2144-point @xmath2092 , and no other singularity . then we may apply lemma [ delta=1 ] @xmath434 .
thus we obtain a period map @xmath2149 .
[ birat ( 14,6,1 ) ] the period map @xmath30 is @xmath2136-invariant and descends to a birational map @xmath2150 .
this is proved in the same way as propositions [ birat ( 11,9 ) ] , [ birat ( 12,8 ) ] , and [ birat ( 13,7 ) ] .
the projection @xmath2151 has degree @xmath523 for the lattice @xmath2152 , which by @xcite is equal to @xmath2153 .
we shall prove that @xmath2154 is rational . by the no - name method for the @xmath2136-linearized bundle @xmath955 over @xmath2155
, we have @xmath2156 the variety @xmath2157 is the moduli of triplets @xmath2158 where @xmath29 is an ( unmarked ) quartic del pezzo surface , @xmath2044 , and @xmath2100 . as in the proof of lemma [ nodal -k curve ( 14,1 ) ] , for a general @xmath2158 we blow - up @xmath29 at @xmath171 and @xmath2027 and then apply the anticanonical map to obtain a one - nodal quartic @xmath966 and a line @xmath500 passing through the node of @xmath966 and tangent to @xmath966 elsewhere .
if @xmath2159 is the space of such pairs @xmath2160 , we thus obtain a rational map @xmath2161 conversely , given a @xmath2162 , we take the double cover @xmath2129 branched along @xmath966 and then the minimal desingularization @xmath2128 .
the covering transformation of @xmath2129 induces an involution @xmath2 of @xmath1974 .
the pullback of @xmath500 to @xmath1974 is written as @xmath2163 where @xmath451 is the @xmath50-curve over the double point of @xmath2164 , and @xmath97 is a @xmath397-curve with @xmath2165 .
notice that @xmath2 exchanges @xmath97 and @xmath2066 , and leaves @xmath451 invariant
. then we contract @xmath97 ( or equivalently , @xmath2066 ) and @xmath451 successively to obtain a point @xmath2158 of @xmath2157 .
this gives a well - defined inverse map of , and thus @xmath2157 is birational to @xmath1012 .
[ rational ( 14,6,1 ) ] the quotient @xmath1012 is rational .
hence @xmath2120 is rational .
we have the @xmath811-equivariant morphism @xmath2166 as in the proof of proposition [ rational ( 13,7 ) ] , we have @xmath2167 for a general point @xmath2168 .
then @xmath2118 is identified with an open set of a linear system of quartics , so that our assertion follows from miyata s theorem .
del pezzo surfaces in this section are not the quotient surfaces of 2-elementary @xmath0 surfaces , but rather their `` canonical '' blow - down compare with [ sec : k=0 ] where del pezzo surfaces appear as the quotient surfaces .
these two series are related by mirror symmetry for lattice - polarized @xmath0 surfaces .
the weyl group symmetry translated to the moduli of labelled sextics @xmath1871 is generated by the renumbering of labelings and by quadratic transformations based at ordinary intersection points of @xmath1872 and @xmath1873 ( cf .
remark [ second proof ( 10,8,1 ) ] ) .
in this section we treat the case @xmath66 , @xmath1644 , @xmath2169 . by the unirationality result
@xcite we may assume @xmath2170 .
for @xmath1686 we use plane cubics with a chosen inflectional point to construct birational period maps . together with [ sec : g=1 ( 1 ) ] , we now cover the range @xmath66 , @xmath371 . among the rest two , @xmath19 is the moduli of enriques surfaces and is rational by kond @xcite , while @xmath57 will be treated in appendix . the space @xmath2171 is proven to be rational in the same way as for @xmath1687 .
the anti - invariant lattice @xmath6 for @xmath2171 is isometric to @xmath2172 .
then we have @xmath2173 , so that @xmath2171 is birational to @xmath1669 . in [ ssec:(8+k,8-k ) ] we proved that @xmath1669 is rational .
let @xmath2175 be the locus of triplets @xmath2176 such that @xmath341 @xmath210 is smooth , @xmath344 @xmath171 is an inflectional point of @xmath210 , and @xmath346 @xmath608 is smooth , passes @xmath171 , and is transverse to @xmath210 .
if @xmath2003 is the tangent line of @xmath210 at @xmath171 , the sextic @xmath2177 has a @xmath2144-singularity at @xmath171 , five nodes at @xmath1622 , one node at @xmath2178 , and no other singularity . the 2-elementary @xmath0 surface associated to the sextic @xmath28 has invariant @xmath2179 .
thus we obtain a period map @xmath2180 .
[ birat ( 15,5 ) ] the period map @xmath30 is birational .
we consider an @xmath1261-cover @xmath543 of @xmath25 whose fiber over a @xmath2181 corresponds to labelings of the five nodes @xmath1622 . as in the previous sections
, @xmath30 lifts to a birational map @xmath2182 .
the variety @xmath2183 is an @xmath519-cover of @xmath2174 for the lattice @xmath2184 . since @xmath811 acts on @xmath25
almost freely and since @xmath1801 by @xcite , @xmath30 has degree @xmath38 .
[ rational ( 15,5,1 ) ] the quotient @xmath1012 is rational .
hence @xmath2174 is rational .
let @xmath2185 be the locus of triplets @xmath2186 such that @xmath2187 .
we have a @xmath811-equivariant map @xmath2188 defined by @xmath2189 , where @xmath500 is the tangent line of @xmath210 at @xmath171 .
a general @xmath117-fiber is an open set of a linear system @xmath2190 of cubics .
the group @xmath811 acts on @xmath84 almost transitively with the stabilizer @xmath26 of a general point @xmath2186 isomorphic to @xmath2191 . indeed , since @xmath608 is anti - canonically embedded , @xmath26 is identified with the group of automorphisms of @xmath2192 fixing the two points @xmath2193 . by the slice method
we have @xmath2194 , and @xmath2000 is clearly rational . for @xmath2197 ,
let @xmath2198 be the locus of quadruplets @xmath2199 such that @xmath341 @xmath210 is smooth , @xmath344 @xmath2200 are linearly independent , @xmath346 @xmath1647 is an inflectional point of @xmath210 with tangent line @xmath737 , @xmath2201 @xmath2202 are respectively transverse to @xmath210 , and @xmath2203 @xmath210 passes ( resp
. does not pass ) @xmath2204 for @xmath2205 ( resp .
@xmath1653 ) . when @xmath1653 , the singularities of the sextic @xmath2206 are @xmath2207 where @xmath171 is a @xmath2144-singularity and the rest points are nodes . when @xmath2205 , denoting @xmath2208 , we have @xmath2209 where @xmath171 is a @xmath2144-singularity , @xmath743 is a @xmath433-singularity , and the rest points are nodes .
the 2-elementary @xmath0 surface associated to @xmath28 has invariant @xmath2210 , and we obtain a period map @xmath2211 .
[ birat ( 16,4)&(17,3 ) ] the map @xmath1348 is birational .
for @xmath1653 we label the three nodes @xmath2212 and the two nodes @xmath2213 independently , which is realized by an @xmath1437-cover @xmath2214 of @xmath2215 .
for @xmath2205 we distinguish the two nodes @xmath2216 , which defines a double cover @xmath2217 of @xmath2218 .
note that in both cases , the three lines @xmath2219 are distinguished by their intersection properties , and hence the rest nodes are a priori labelled .
as before , we see that @xmath1348 lifts to a birational map @xmath2220 .
then @xmath2221 is an @xmath519-cover of @xmath2222 for the lattice @xmath2223 .
by @xcite we have @xmath2224 for @xmath2197 respectively , which concludes the proof .
[ rational ( 16,4)&(17,3 ) ] the quotient @xmath1666 is rational .
therefore @xmath2195 and @xmath2196 are rational .
this follows from the slice method for the projection @xmath2225 .
general @xmath117-fibers are open subsets of linear systems of cubics , and @xmath811 acts on @xmath2226 almost transitively with a general stabilizer isomorphic to @xmath2227 . degenerating the above sextic models
, one sees that 1 .
@xmath2228 is birational to the kummer modular surface for @xmath2229 , 2 .
@xmath2230 is birational to the pullback of the kummer modular surface for @xmath2231 by the fricke involution , 3 .
@xmath2232 is birational to the modular curve for @xmath2233 , via their fixed curve maps .
note that @xmath2228 and @xmath2230 are heegner divisors of @xmath2196 , though they look like boundary divisors of toroidal compactification .
in this section we study the case @xmath69 with @xmath1644 .
dolgachev and kond @xcite proved the rationality of @xmath22 , which is the moduli of coble surfaces and for which @xmath2234 .
the remaining case @xmath2235 will be studied in appendix . here
we treat the case @xmath1432 by a similar approach as in
[ sec : g=1 ( 1 ) ] , and the cases @xmath1648 using cuspidal plane cubics .
the range @xmath2236 is covered by the unirationality result @xcite .
let @xmath2238 be the family of marked cubic surfaces considered in [ ssec:(13,7 ) ] .
let @xmath2239 be the locus of those @xmath2240 such that the @xmath1946-curve @xmath2006 on @xmath2241 singular at @xmath171 is irreducible and cuspidal .
( @xmath1819 is the branch divisor of the double cover @xmath2078 in [ ssec:(13,7 ) ] ) .
then we consider the locus @xmath2242 in @xmath2243 of those @xmath2244 such that the @xmath1946-curve @xmath2245 singular at @xmath743 is irreducible , nodal , and intersects with @xmath2006 at @xmath171 with multiplicity @xmath911 .
the curves @xmath2006 and @xmath2245 do not intersect outside @xmath171 . then the @xmath27-curves @xmath2246 have an @xmath409-singularity at @xmath171 , a node at @xmath743 , and no other singularity .
taking the right resolution of the dpn pairs @xmath2247 , we obtain a period map @xmath2248 .
this is similar to the arguments in [ sec : g=1 ( 1 ) ] , so we shall be brief .
the natural blow - down @xmath2250 determined by @xmath1898 maps @xmath2251 to the irreducible cubic pair @xmath2252 such that ( i ) @xmath2253 is cuspidal , ( ii ) @xmath2254 is nodal and intersects with @xmath2253 at its cusp with multiplicity @xmath911 , and ( iii ) @xmath2253 and @xmath2254 are transverse outside the cusp of @xmath2253 . the six nodes of @xmath2255 are the blown - up points of @xmath117 , and hence naturally ordered by @xmath1898 .
one recovers @xmath2244 from the labelled sextic @xmath2255 by blowing - up the ordered nodes .
thus , if @xmath2256 is the space of such labelled sextics , @xmath2242 is birationally identified with @xmath1630 .
then we can apply our familiar recipe to @xmath543 to see that @xmath30 lifts to a birational map @xmath2257 .
the map @xmath30 is clearly @xmath2069-invariant .
we also note that @xmath2069 acts on @xmath2242 almost freely . on the other hand
, @xmath2258 has the galois group @xmath519 for the lattice @xmath2259 .
by @xcite we have @xmath2260 .
therefore the induced map @xmath2249 has degree @xmath38 .
let @xmath2261 be the locus of triplets @xmath2262 such that ( i ) @xmath29 is a smooth cubic surface containing @xmath171 and @xmath743 , ( ii ) the tangent plane section @xmath2263 is irreducible and cuspidal , and ( iii ) the tangent plane section @xmath2264 is irreducible , nodal , and intersects with @xmath2263 at @xmath171 with multiplicity @xmath911 .
getting rid of markings and considering anti - canonical models , we see that @xmath2265 is birationally identified with @xmath2266 . for @xmath2267 , the point @xmath171 is contained in both tangent planes @xmath2268 , @xmath2269
moreover , the common tangent line of @xmath2263 and @xmath2264 at @xmath171 is given by @xmath2270 .
thus , if we consider the space @xmath2271 of quadruplets @xmath2272 such that @xmath2273 and @xmath2274 , we have the @xmath2275-equivariant map @xmath2276 by our second remark , the @xmath1015-fiber over a general @xmath2277 is the locus @xmath2278 of smooth cubic surfaces @xmath29 such that ( i ) the plane cubic @xmath2279 is nodal at @xmath820 and passes through @xmath819 with tangent line @xmath2280 , and ( ii ) the plane cubic @xmath2281 is cuspidal at @xmath819 with tangent line @xmath2280 . then @xmath25 is an open set of a linear system .
it is straightforward to see that @xmath2282 acts on @xmath829 almost transitively , with the stabilizer @xmath2283 of @xmath2272 being connected and solvable . by the slice method we have @xmath2284 and
the quotient @xmath24 is rational by miyata s theorem .
let @xmath2286 be the variety of cuspidal plane cubics .
let @xmath2287 be the open set of pairs @xmath1560 such that @xmath608 is smooth and transverse to @xmath1257 , where @xmath500 is the tangent line of @xmath210 at the cusp .
then the sextic @xmath2177 has an @xmath409-singularity at the cusp of @xmath210 , eight nodes at @xmath1566 and @xmath2288 , and no other singularity .
the associated 2-elementary @xmath0 surface has invariant @xmath2289 , and we obtain a period map @xmath2290 .
we label the six nodes @xmath1566 and the two nodes @xmath2288 independently .
this defines an @xmath1788-cover @xmath543 of @xmath25 . by the familiar method
we see that @xmath30 lifts to a birational map @xmath2291 .
the variety @xmath2292 is an @xmath519-cover of @xmath2285 for the lattice @xmath2293 .
by @xcite we calculate @xmath2294 , which implies that @xmath2295 .
general members of @xmath2285 can also be obtained from six general lines on @xmath35 .
matsumoto - sasaki - yoshida @xcite studied this sextic model , and showed that the period map is the quotient map by the association involution .
therefore @xmath2285 is birational to the moduli of double - sixers ( cf .
@xcite ) , which is rational by coble @xcite ( see @xcite for a proof by dolgachev ) .
this is an alternative approach for the rationality of @xmath2285 .
conversely , this section may offer another proof of the result of coble .
let @xmath84 be the space of cuspidal plane cubics and @xmath2287 be the locus of pairs @xmath1560 such that @xmath608 is the union of distinct lines and transverse to @xmath1257 , where @xmath500 is the tangent line of @xmath210 at the cusp .
the sextic @xmath2302 has an @xmath409-singularity at the cusp of @xmath210 , nine nodes at @xmath1566 , @xmath2288 , and @xmath1585 , and no other singularity .
hence the associated 2-elementary @xmath0 surface has invariant @xmath2303 , and we obtain a period map @xmath2304 .
let @xmath1625 be the locus of @xmath2305 such that @xmath2306 and that @xmath871 belong to the same component of @xmath608 .
by @xmath543 , the six nodes @xmath1566 and the two components of @xmath608 are labeled compatibly . as before , @xmath30 lifts to a birational map @xmath2307 .
since @xmath811 acts on @xmath25 almost freely , @xmath1630 is an @xmath703-cover of @xmath1012 . on the other hand , @xmath2308 is an @xmath519-cover of @xmath2301 for the lattice @xmath2309 .
by @xcite we calculate @xmath2310 .
consider the @xmath811-equivariant map @xmath2311 , @xmath2312 .
a general fiber @xmath2313 is an open set of the net of conics singular at @xmath171 .
it is straightforward to see that @xmath811 acts on @xmath2314 almost freely .
then we may apply the no - name lemma [ no - name 2 ] to see that @xmath2315 the quotient @xmath2316 is of dimension @xmath38 and so is rational .
let @xmath500 be an odd lattice .
the _ even part _ @xmath2317 of @xmath500 is the kernel of the natural homomorphism @xmath2318 which is the maximal even sublattice of @xmath500 .
the index @xmath45 extension @xmath2319 is given by an element @xmath2320 of order @xmath45 in the discriminant group @xmath2321 of @xmath2317 : that is , we have @xmath2322 inside @xmath2323 . since @xmath500 is odd , @xmath2320 must have norm @xmath38 modulo @xmath2324 with respect to the discriminant form .
[ i_n&d_n ] let @xmath2325 denote the odd unimodular lattice @xmath2326 . when @xmath2327 , the even part of @xmath2328 is isometric to @xmath420 .
indeed , if we take natural orthogonal basis @xmath2329 of @xmath2328 , the even part is expressed as @xmath2330 .
it is generated by the following vectors : @xmath2331 one checks that these form a root system of @xmath420-type for @xmath2327 .
when @xmath2332 , @xmath2325 is isometric to @xmath2333 .
therefore the even part of @xmath2325 is isometric to @xmath2334 for @xmath2335 .
@xmath428 when @xmath2338 is odd , @xmath2339 is isomorphic to @xmath2340 generated by @xmath2341 . in particular , @xmath2342 as a finite quadratic form . in this case
@xmath2320 is the unique element of order @xmath45 in @xmath2339 .
@xmath431 when @xmath2343 is even , @xmath2339 is isomorphic to @xmath2344 generated by @xmath2345 and @xmath2346 . as a quadratic form
, we have @xmath2347 when @xmath237 is odd , and @xmath2348 when @xmath237 is even . in case @xmath237
is odd or divisible by @xmath738 , @xmath2320 is the unique element with norm @xmath2349 . in case @xmath2350 mod @xmath738 ,
@xmath2320 is one of the three non - zero elements of @xmath2339 , which all have norm @xmath2349 .
if a lattice @xmath500 has signature @xmath2351 , we let @xmath2352 be the domain defined as in . given a finite - index subgroup @xmath966 of @xmath2353 , we shall denote , but may be slightly different from some other literatures where the notation `` @xmath2354 '' is used instead .
] by @xmath2355 the modular variety @xmath2356 where @xmath2357 is a connected component of @xmath2352 and @xmath2358 is the stabilizer of @xmath2357 .
we sometimes abbreviate @xmath2355 as @xmath2359 .
the key proposition in this appendix is the following .
[ even part ] let @xmath2317 be the even part of an odd lattice @xmath500 of signature @xmath2351 , and @xmath2360 be the element giving the extension @xmath2319 .
let @xmath2361 be the stabilizer of @xmath2320 . then the modular varieties @xmath2362 and @xmath2363 are naturally isomorphic .
since any @xmath2364 preserves the canonical sublattice @xmath2319 , we have the inclusion @xmath2365 inside @xmath2366 .
if we regard @xmath2364 as an isometry of @xmath2317 , it stabilizes the overlattice @xmath2367 and hence preserves the subgroup @xmath2368 .
conversely , any @xmath2369 preserving @xmath2368 must be contained in @xmath2353 .
this establishes the equality @xmath2370 inside @xmath2371 .
we shall apply this proposition for @xmath500 a scaling of the anti - invariant lattice or its dual lattice for @xmath58 , @xmath59 , @xmath60 , and @xmath57 .
consequently , those @xmath10 get realized as a finite cover or a heegner divisor of another @xmath2373 , or even become birational to @xmath2373 .
then we prove the rationality of @xmath10 by making use of the geometric description of @xmath2373 obtained in the earlier sections .
the case of heegner divisor is similar to @xcite , while the case of finite cover is novelty of this appendix .
the following argument was independently suggested by the referee and kond .
the anti - invariant lattice for @xmath58 is isometric to the scaling @xmath2374 of the unimodular lattice @xmath2375 .
hence @xmath58 is naturally birational to @xmath2376 . by example
[ i_n&d_n ] , the even part of @xmath2375 is isometric to @xmath2377 , and the element in @xmath2378 giving the extension @xmath2379 is invariant under @xmath2380 .
therefore we have @xmath2381 by proposition [ even part ] .
then @xmath2377 is the anti - invariant lattice for @xmath1536 , so that @xmath2382 is birational to @xmath1536 . in proposition
[ rational ( 11,3)&(12,2 ) ] we proved that @xmath1536 is rational .
hence @xmath58 is rational too .
the anti - invariant lattice for @xmath59 is isometric to @xmath2383 . scaling it by @xmath2384 , we see that @xmath59 is birational to @xmath2385 . by example
[ i_n&d_n ] , the even part of @xmath2386 is isometric to @xmath2387 and the extension @xmath2388 is given by the unique element of order @xmath45 in @xmath2389 .
therefore we have @xmath2390 by proposition [ even part ] .
in order to study @xmath2057 , we shall exploit the fact that the @xmath2391-lattice is naturally the orthogonal complement of a @xmath411-vector @xmath321 in the @xmath674-lattice ( take @xmath2392 where @xmath2393 are the root basis as in ) .
hence @xmath1849 can be embedded in the lattice @xmath2394 as the orthogonal complement of a @xmath411-vector .
then we want to realize @xmath2057 as a heegner divisor of @xmath2395 . for
that we need the following lattice - theoretic properties .
[ structure as heegner div ] let @xmath2396 be a @xmath411-vector with @xmath2397
. then its orthogonal complement @xmath2398 is isometric to @xmath1849 , and every isometry of @xmath2399 extends to that of @xmath500 .
moreover , such @xmath411-vectors are all equivalent under @xmath2353 .
we first note that @xmath2400 is primitive .
let @xmath2401 ( resp .
@xmath2402 ) be the subgroup corresponding to the image of the orthogonal projection @xmath2403 ( resp .
@xmath2404 ) .
then we have canonical isomorphisms @xmath2405 .
in other words , we have an anti - isometry @xmath2406 whose graph @xmath2407 gives the overlattice @xmath2408 ( cf .
the discriminant form of @xmath500 is identified with the quadratic form on @xmath2409 .
we have @xmath2415 by the assumption @xmath2397 . hence @xmath2416 .
if @xmath2417 , then we would have @xmath2418 , which contradicts the 2-elementariness of @xmath2419 .
hence @xmath2410 .
since the discriminant form of @xmath2399 is non - degenerate , we may find an element @xmath2420 with @xmath2421 . then @xmath2422 is perpendicular to @xmath2423 . since @xmath2409 is 2-elementary , we have @xmath2424 , which means that @xmath2425
. therefore @xmath2426 so that @xmath2427 for some @xmath2428 .
in particular , the quadratic form on @xmath2429 is non - degenerate .
hence we have the orthogonal splitting @xmath2430 for a subgroup @xmath2431 .
since @xmath2432 , we actually have @xmath2433 .
now @xmath2434 is represented by the following elements : @xmath2435 in view of the structure of @xmath2436 ( example [ i_n&d_n ] ) , the elements @xmath2437 should have norm @xmath2384 .
therefore @xmath308 is determined as @xmath1436 . by this claim the discriminant form of @xmath2399
is uniquely determined , so that we may apply nikulin s theory @xcite to see that the isometry class of @xmath2399 is uniquely determined .
looking the discriminant form of @xmath1849 at example [ i_n&d_n ] , we see that @xmath2438 .
the claim also says that the extension @xmath2439 is uniquely determined inside @xmath2440 .
it follows that for any isometry @xmath2441 of @xmath2399 , the isometry @xmath2442 of @xmath2443 preserves the overlattice @xmath500 .
finally , the uniqueness of @xmath2399 and of the extension @xmath2439 imply our uniqueness assertion for @xmath321 .
let @xmath2444 be the heegner divisor defined by @xmath411-vectors as in lemma [ structure as heegner div ] , that is , the image of the analytic divisor @xmath2445 where @xmath321 range over @xmath411-vectors in @xmath500 with @xmath2397 .
noticing that @xmath2446 contains an element exchanging the two components of @xmath2447 , we deduce from lemma [ structure as heegner div ] that @xmath2448 is irreducible and birational to @xmath2057 : more precisely , we have a natural projection @xmath2449 which gives the normalization of @xmath2448 .
now @xmath500 is the anti - invariant lattice for @xmath1536 , so that @xmath2395 is birational to @xmath1536 .
we shall study the heegner divisor @xmath2450 through the geometric description of @xmath1536 given in [ ssec:(11,3)&(12,2 ) ] .
this requires to refine the period map there , especially to extend it .
let @xmath567 and @xmath2451 be the locus of smooth curves @xmath210 such that there exists exactly one @xmath1474-fiber @xmath1983 tangent to @xmath210 with multiplicity @xmath911 ( compare @xmath25 with the locus @xmath2452 in [ ssec:(11,3)&(12,2 ) ] ) .
let @xmath833 be the @xmath1487-fiber through @xmath2453 .
taking the right resolution of @xmath2454 , we obtain a period map @xmath2455 as a morphism of varieties . in proposition
[ birational ( 11,3)&(12,2 ) ] , we showed that @xmath171 descends to a birational map from a rational quotient of @xmath25 by @xmath1457 .
we refine it as follows .
we first check the existence of a geometric quotient @xmath2457 .
let @xmath2458 be the affine open locus in @xmath2459 of smooth curves , of which @xmath25 is a locally closed subset .
we have a geometric quotient of @xmath2458 by @xmath600 as an open set of @xmath958 .
by @xcite converse 1.12 , this assures the existence of an @xmath1476-linearized line bundle @xmath1276 over @xmath2458 with respect to which every point of @xmath2458 is stable .
now we can restrict @xmath1276 to @xmath25 and apply @xcite theorem 1.10 .
next we show that the induced morphism @xmath30 from the geometric quotient has finite fibers . indeed ,
if @xmath2460 for @xmath221 , then @xmath210 can be recovered from the natural projection @xmath589 , which in turn from the line bundle @xmath2461 by lemma [ covering map & lb ] .
thus we have a surjective map onto @xmath2462 from a subset of @xmath323 , so that @xmath2462 is countable
. then we can apply zariski s main theorem to conclude that @xmath30 is an open immersion .
since @xmath2448 is irreducible , it suffices to see that @xmath2465 . for @xmath2298 , the tangent point @xmath743 of @xmath210 and @xmath833
is a tacnode of @xmath2454 .
if @xmath2460 and @xmath589 is the natural projection , then @xmath2466 is an @xmath2467-configuration of @xmath50-curves .
we write @xmath2468 such that @xmath2469 and @xmath2470 . by figure
[ dual graph ] , @xmath2 acts on @xmath2466 by @xmath2471 and @xmath2472 .
then we have the @xmath2-anti - invariant cycle @xmath2473 as an algebraic cycle on @xmath1 .
this means that the period of @xmath4 is in the heegner divisor defined by @xmath451 . since @xmath2474
, it remains to check that @xmath2475 .
this follows from @xmath2476 where the first equality holds because @xmath2477 .
let @xmath2479 be the locus of pairs @xmath2480 such that @xmath171 and @xmath743 lie on a same @xmath1487-fiber .
we have the @xmath1476-equivariant map @xmath2481 where @xmath1983 and @xmath833 are the ruling fibers defined from @xmath210 as above .
the group @xmath1497 acts on @xmath829 almost transitively , with the stabilizer of a general point @xmath2480 isomorphic to @xmath1858 .
the fiber @xmath2482 is an open set of a sub - linear system of @xmath2459 .
therefore @xmath2464 is rational by the slice method for @xmath1015 and miyata s theorem . as before , @xmath60 is birational to @xmath2483 . by example
[ i_n&d_n ] , the even part @xmath500 of @xmath2484 is isometric to @xmath2485 , with the extension @xmath2486 given by one of the three non - zero elements of @xmath2487 , say @xmath2320 .
let @xmath2488 be the stabilizer of @xmath2320 .
then we have @xmath2489 by proposition [ even part ] .
note that @xmath500 is the anti - invariant lattice for @xmath1669 . since @xmath2490 , the modular variety @xmath2355 is realized as an intermediate cover @xmath2491 the galois group of @xmath788 is @xmath2492 , which is isomorphic to @xmath790 by the permutation action on the three non - zero elements of @xmath2419 .
if @xmath2493 is the stabilizer of any one , then we have @xmath2494 in example [ ex:4 ] , we gave a geometric description of @xmath788 in terms of plane quartics .
below let us reuse the notation there .
we considered a space @xmath25 parametrizing smooth plane quartics @xmath210 with a tangent line @xmath737 of multiplicity @xmath738 and with a line @xmath739 through @xmath2495 .
we also considered an @xmath790-covering @xmath545 whose fibers consist of the numberings of the three nodes @xmath2496 .
then we obtained a birational lift @xmath2497 of the period map @xmath2498 .
[ equivariance ( 14,2,0 ) ] the map @xmath553 is @xmath790-equivariant . here
@xmath790 acts on @xmath543 by renumbering of the three nodes , and on @xmath2499 by permutation of the non - zero elements of @xmath2419 .
this follows simply because any automorphism of @xmath790 is inner , but here we shall provide geometric reasoning . for a point @xmath2500 of @xmath543 ,
let @xmath4 be the associated 2-elementary @xmath0 surface with the lattice - marking @xmath531 . if @xmath766 is the natural projection , then @xmath784 is the class of the @xmath50-curve @xmath2501 . by lemma [ l_+ and a - d - e ] ,
the vectors @xmath2502 have integral pairing with @xmath529 and hence are contained in @xmath2503 .
they represent the three non - zero elements of @xmath334 . if @xmath2504 is the marking associated to the labelling @xmath2505 for @xmath2506 , then @xmath2507 maps @xmath2508 to @xmath2509 .
now choose isometries @xmath2510 such that @xmath2511 and @xmath2512 extend to @xmath2513 respectively .
if we denote by @xmath2514 the image of @xmath2508 by the natural anti - isometry @xmath2515 , then @xmath2516 maps @xmath2517 to @xmath2518 by nikulin @xcite . in view of the definition of @xmath553 , this proves our assertion .
let @xmath2527 be the locus of triplets @xmath2528 such that @xmath2529 .
we consider the @xmath811-equivariant map @xmath2530 since the line @xmath739 is recovered as @xmath2531 , the @xmath1015-fiber over a general @xmath2528 is identified with an open set of a sub - linear system of @xmath1018 .
the group @xmath1273 acts on @xmath84 almost transitively with connected and solvable stabilizer .
hence @xmath2526 is rational by the slice method for @xmath1015 and miyata s theorem .
the lines of argument for @xmath58 , @xmath59 , and @xmath60 would apply more generally to @xmath2532 until @xmath2533 , implying that those modular varieties are rational for @xmath2534 and unirational for @xmath2535 : * when @xmath136 is odd , @xmath2532 is realized as a @xmath411-heegner divisor of @xmath2536 , where @xmath2537 . for @xmath2538
we then study a locus of @xmath27-curves that acquire tacnode as extra point of tangency ( and finally use miyata s theorem ) , while @xmath2539 gets birational to @xmath2540 ; * when @xmath2541 , @xmath2542 is realized as a triple cover of @xmath1067 ; * for other even @xmath136 , @xmath2532 becomes birational to @xmath2543 , @xmath2544 .
the anti - invariant lattice @xmath6 for @xmath57 is isometric to @xmath2547 .
we consider the odd lattice @xmath2548 , for which we have @xmath2549 the even part of @xmath1692 is @xmath2550 with the extension @xmath2551 given by an element @xmath2552 of norm @xmath38 . by proposition
[ even part ] , we have @xmath2553 for the stabilizer @xmath2488 of @xmath2320
. since @xmath500 is the anti - invariant lattice for @xmath1520 , we can realize @xmath2355 as an intermediate cover @xmath2554 as in .
let us describe the galois group @xmath2492 of @xmath2555 .
let @xmath2556 and @xmath2557 be natural basis of the two summands @xmath2558 of @xmath500 with @xmath2559 and @xmath2560 .
we define norm @xmath38 elements of @xmath2419 by @xmath2561 then we have the orthogonal decomposition @xmath2562 with @xmath2563 .
it follows that the six elements @xmath2564 are all of the norm @xmath38 elements in @xmath2419 , and the decomposition is canonical .
therefore in example [ ex:3 ] , we studied the covering @xmath2555 using curves on @xmath684 .
below we shall reuse the notation there .
we considered an open locus @xmath685 , and an @xmath703-covering @xmath545 whose fiber over @xmath1754 consists of labellings of the six points @xmath693 that are compatible with the decomposition of @xmath451 .
then we obtained a birational lift @xmath725 of the period map @xmath2569 .
[ equivariance ( 10,4,0 ) ] the map @xmath553 is @xmath703-equivariant . here
@xmath703 acts on @xmath543 by relabelling of the six points , and on @xmath730 by permutation of the norm @xmath38 elements of @xmath2419 .
this is similar to the proof of lemma [ equivariance ( 14,2,0 ) ] . for a point @xmath2570 of @xmath543 ,
let @xmath4 be the associated 2-elementary @xmath0 surface with the lattice - marking @xmath531 .
if @xmath714 is the natural projection , the class of the @xmath50-curve @xmath2571 is given by @xmath727 .
one checks that the six vectors @xmath2572 are contained in @xmath2503 .
they represent the six norm @xmath38 elements of @xmath334 with @xmath2573 .
choosing an isometry @xmath2574 suitably in advance , we may assume that the image of @xmath2575 by the natural anti - isometry @xmath2515 is @xmath2576 .
if we transform the labelling @xmath718 by an element @xmath155 of @xmath703 , then the new marking @xmath2504 satisfies @xmath2577 . as in the proof of lemma [ equivariance ( 14,2,0 ) ] , this shows the @xmath703-equivariance of @xmath553 . for @xmath2584 ,
let @xmath53 be the component of @xmath451 _ not _ through @xmath171 .
we have the @xmath1121-equivariant map @xmath2585 the @xmath1015-fiber over a general @xmath2586 is identified with an open set of a hyperplane @xmath818 of @xmath2587 .
since @xmath1121 acts on @xmath2588 almost transitively , by the slice method we have @xmath2589 for the stabilizer @xmath2590 of @xmath2586 . by and , @xmath174 is an extension of @xmath2191 by @xmath2591 and hence connected and solvable .
since @xmath1530 is @xmath1121-linearized , @xmath174 acts on @xmath84 linearly .
thus we may apply miyata s theorem to see that @xmath2592 is rational .
* acknowledgements .
* i would like to express my gratitude to professor ken - ichi yoshikawa for introducing me to this rich subject .
i am also indebted to professors i. dolgachev and s. kond , whose paper @xcite inspired me to resume the study after a period of abandonment .
i received invaluable suggestions from the referee and prof .
kond , especially the proof for @xmath58 ( p. , from both ) and the second construction of @xmath2050 ( p. , from the referee ) .
i should like to thank them for their generous permission to include those in this article .
finally , i am grateful to the editors for their help and patience .
casnati , g. ; del centina , a. _ on certain spaces associated to tetragonal curves of genus @xmath1246 and @xmath2593 .
_ commutative algebra and algebraic geometry ( ferrara ) , 3545 , lect .
notes in pure and appl .
math * 206 * , dekker , 1999 .
matsumoto , k. ; sasaki , t. ; yoshida , m. _ the monodromy of the period map of a @xmath738-parameter family of @xmath0 surfaces and the hypergeometric function of type @xmath2594 .
_ internat . j. math .
* 3 * ( 1992 ) , no .
1 . morrison , d. r. ; sait , m .- h . _ cremona transformations and degrees of period maps for @xmath0 surfaces with ordinary double points . _ algebraic geometry , sendai , 1985 , 477513 , adv .
pure math . * 10 * , 1987 .
shepherd - barron , n. i. _ the rationality of certain spaces associated to trigonal curves . _ algebraic geometry , bowdoin , 1985 , 165171 , proc .
pure math .
, 46 , part 1 , amer .
soc . , providence , ri , 1987 . | @xmath0 surfaces with non - symplectic involution are classified by open sets of seventy - five arithmetic quotients of type iv .
we prove that those moduli spaces are rational except two classical cases . and grant - in - aid for scientific research ( s ) , no 22224001 . ] |
in a previous paper ( font and ibez 1997 , hereafter paper i ) we studied the morphology and dynamics of relativistic bondi - hoyle accretion in axisymmetric flows past a schwarzschild black hole .
the main conclusion of that work was to extend the validity of the bondi - hoyle accretion picture ( hoyle and lyttleton , 1939 ; bondi and hoyle , 1944 ) to the relativistic regime , finding that the matter is always accreted onto the hole in a stationary way .
furthermore , if the flow was , initially , supersonic , the main feature of the accretion pattern was the presence of a shock cone in the solution .
at the same time , we checked the validity of our numerical code revisiting an existing previous calculation ( petrich et al . , 1989 ) using more accurate numerical techniques specifically designed to capture discontinuities . in the present investigation
we have extended those studies to account for non - axisymmetric configurations .
we have only considered uniform flows at infinity focussing on studying whether or not the flow pattern reaches , ultimately , a stationary state .
the main purpose of this work is , then , to extend previous non - axisymmetric newtonian computations to the realm of general relativity .
as far as we know , this has not been studied so far . in particular
, we plan to address if non - steady patterns , so prominent in some of the 2d non - axisymmetric simulations in the classical regime , also arise here . during the last years
, a large number of wind accretion simulations past a gravitating body have been performed in newtonian hydrodynamics ( see , e.g. , the list of references in paper i ) .
one of the most interesting features , only revealed by numerical simulations , was the appearance of unstable accretion patterns , contrary to the theoretical ( and simplified ) bondi - hoyle ( 1944 ) accretion picture .
this highly non - steady behaviour on the wake of the accretor was only found in 2d non - axisymmetric simulations , especially when the resolution employed was fine enough .
these unstable patterns are characterized by the shock wave moving from side to side ( the so - called flip - flop instability ) and also by the appearance of transient phases of disc formation where the angular momentum accreted by the central object significantly increases .
this kind of behaviour has been found not only assuming local density and velocity gradients ( fryxell and taam , 1988 ; taam and fryxell , 1989 ) but also when considering accretion of uniform flows at infinity ( matsuda et al . , 1991 ) .
however , in detailed 3d computations performed more recently ( ruffert and arnett , 1994 ; ruffert , 1994a , b , 1995 , 1996 ) the accretion cone remains quite stable and no sign of flip - flop instability appears unless the flow is assumed to have explicit density gradients at infinity ( ruffert , 1997 ) . in all previous studies
it was shown that the key parameter which controls the appearance of the instability was the _ size of the accretor_. in particular , this size should be a very small fraction of the accretion radius in order to find any evidence of flip - flop .
for instance , in sawada et al .
( 1989 ) and matsuda et al .
( 1991 ) computations , the flip - flop instability appeared , in 2d , for a central size object of @xmath3 , where @xmath4 is the accretion radius , the natural length scale of this problem , defined as @xmath5 here , @xmath6 is the mass of the accreting object , @xmath7 is the gravitational constant and @xmath8 is the asymptotic velocity of the fluid .
they also found that the flow was stable if the size was larger than @xmath9 .
more recently , benensohn , lamb and taam ( 1997 ) have performed detailed 2d non - axisymmetric computations with a smaller accretor of size @xmath10 finding unstable behaviour as well .
in addition , there are other parameters that , combined with the size of the accretor , can play a role in the development of the instability .
in particular , one of these is the _ mach number _ of the flow .
the larger the mach number at infinity is , the more turbulent the wake becomes .
finally , another extremely important consideration is the _ position of the shock _ , namely if it is attached ( tail shock ) or detached ( bow shock ) with respect to the central object .
this is basically controlled by the values of the asymptotic mach number and the adiabatic index of the gas . for low mach number and high adiabatic index flows ( say @xmath11 ) ,
the shock is mainly detached at the front while the opposite trend produces a tail shock .
all previous studies where the instability appeared needed the formation of a detached shock in front of the accretor .
as the evolution procceeded , this detached shock transformed into a dome - shaped shock and , eventually , this was followed by a violent flip - flop instability ( sawada et al . , 1989 ;
matsuda et al . , 1991 ; livio et al . , 1991
; see also ruffert , 1997 ) .
in particular , livio et al .
( 1991 ) demonstrated , using a simplified analytic treatment , the existence of a wake instability to small deflections when the shock opening angle is large .
this can happen even for an initially homogeneous flow if the accretion process has low efficiency .
however , the flow can be stabilized if the accreting object is larger than some critical dimension ( livio et al . , 1991 ) .
according to these previous classical results , there are several factors that can lead to the non - appearance of this instability in the relativistic regime when the accretor is a black hole .
first , and most important , in this regime there is a physical minimum value for the size of the accreting object , given by the gravitational ( schwarzschild ) radius .
this , of course , places some constraints in the possibility of extending previous newtonian studies to very small values of the accretor , assuming that we also need an asymptotically supersonic flow ( in order to have any shock wave ) and avoiding , at the same time , flow velocities larger than the speed of light .
the second factor has to do with the position of the shock . as we will show below
( see also paper i ) , all models that we have evolved are characterized by the presence of a _ tail _ shock in the rear part of the hole .
hence , according to the newtonian studies , this would not help the development of the flip - flop instability .
the paper is organized as follows : in next section ( @xmath12 ) we briefly present the system of equations of general relativistic hydrodynamics written as a hyperbolic system of conservation laws .
the numerical code and the initial setup are also described here .
the results of the simulations are presented and analyzed in section @xmath13 .
finally , section @xmath14 summarizes the main conclusions of this work .
in order to study non - axisymmetric patterns in two dimensions we have restricted ourselves to an infinitesimally thin disk in the equatorial plane of the black hole .
therefore , we are using ( @xmath15 ) coordinates , instead of ( @xmath16 ) used previously in paper i. although this configuration is somehow artificial , it suffices to try to understand the stability of the flow .
this kind of setup has also been used in newtonian simulations , assuming flow past an infinite cylinder ( fryxell and taam , 1988 ; taam and fryxell , 1989 ; benensohn et al . , 1997 ) .
with the same definitions introduced in paper i and using geometrized units ( @xmath17 with @xmath18 the speed of light ) , the equations of ( adiabatic ) general relativistic hydrodynamics can be written , in the equatorial plane ( @xmath19 ) of the schwarzshild metric , as @xmath20 in this equation , the vector of _ primitive variables _ is defined as @xmath21 where @xmath22 and @xmath23 are , respectively , the rest - mass density and the specific internal energy , related to the pressure @xmath24 via an equation of state which we chose that of an ideal gas law @xmath25 with @xmath26 being the constant adiabatic index .
in addition , @xmath27 and @xmath28 denote the radial and azimuthal covariant components of the velocity . on the other hand ,
the vector of unknowns ( evolved quantities ) in equation ( 2 ) is @xmath29 the explicit relations between both sets of variables are @xmath30 @xmath31 @xmath32 the quantity @xmath33 stands for the lorentz factor , which satisfies @xmath34 with @xmath35 , where @xmath36 is the 3-velocity of the fluid , defined , for the case of a zero shift vector , according to @xmath37 and @xmath38 are the spatial components of the spacetime metric where the fluid evolves .
the quantity @xmath39 appearing in equations ( 7 ) and ( 8) is the specific enthalpy , defined as @xmath40 .
the corresponding fluxes in the radial and azimuthal directions are , respectively @xmath41 @xmath42 and the corresponding vector of sources @xmath43 is @xmath44 , -{{m}\over{\alpha r^2}}(\tau + d ) + \alpha \left[{{1}\over{r}}s_{\phi}v^{\phi } - { { 2}\over{r}}s_r v^r \right ] , \right .
\nonumber \\ & s_{\phi}v^r \left[{{m}\over{\alpha r^2 } } - { { 2\alpha}\over{r}}\right ] , \left . -{{2\alpha}\over{r } } s^r - d v^r \left[{{m}\over{\alpha r^2 } } - { { 2\alpha}\over{r}}\right ] \right ) \end{aligned}\ ] ] where @xmath45 .
the models that we have evolved numerically are summarized in table 1 .
some of them were already evolved in axisymmetry in paper i. now , we have added some higher mach number models and we have not considered subsonic and low mach number flows . according to newtonian studies , this is the appropriate initial condition to look for wake instabilities .
we must also mention that , contrary to the axisymmetric results , it was now extremely difficult to evolve low mach number models without corrupting the asymptotic initial values in the upstream region of the flow . in this part of the domain , all fluid variables should increase monotonically . in practice
, we found that the local velocity and mach number in the upstream direction fell below the initial values .
this was still the case even when we placed the outer boundary further out ( although the evolution could be accurately followed for much longer times ) .
this sort of behaviour has not appeared in the high mach number models , as we will show in detail below .
this inherent difficulty in approximate the asymptotic conditions to a finite distance has also been pointed out in previous newtonian simulations ( ruffert and arnett , 1994 ; benensohn et al . , 1997 ) .
the flow pattern is completely characterized by some initial conditions at infinity upstream the hole , namely the velocity @xmath8 , sound speed @xmath46 and adiabatic index , @xmath26 .
the first two parameters fix the asymptotic mach number , @xmath47 .
the covariant components of the initial velocity are given in terms of its asymptotic value : @xmath48 @xmath49 where @xmath50 and @xmath51 .
as mentioned in the introduction , it is convenient to stress the differences that our initial setup has with respect to those of classical simulations where the tangential wake instability arose . for this purpose ,
we plot in fig . 1 the ratio between the size of the accretor , @xmath52 , the schwarzschild radius , and the accretion radius , as a function of the asymptotic flow velocity . for the accretion radius we are using the same definition of paper
i @xmath53 the curved lines represent that quotient for different values of the sound speed at infinity .
the solid line corresponds to zero sound speed and if we move upward on the plot , we go to increasingly larger values .
the last three curves correspond to the thermodinamically maximum permitted values , @xmath54 , for @xmath55 , @xmath56 and @xmath11 , respectively . on the other hand , the straight lines represent the fluid mach number ( divided by 5 ) as a function of the velocity , parametrized for the same set of sound speeds .
therefore , the lines are drawn accordingly with the same line style , with the exception of the solid straight line which corresponds to @xmath57 ( instead of zero sound speed ) . from that figure
, one can see that , in order to have tiny accretors , one has to move to the lower left corner .
however , in that region , the asymptotic initial velocity has to be quite small ( newtonian flow ) .
furthermore , in order to have an initial supersonic , or even better , hypersonic flow , the initial sound speed has to be considerably small .
this kind of initial data is quite difficult to evolve for any numerical code , as the internal energy density is very close to zero . on the other hand ,
if we move to larger values of the velocity , for the sample of sound speeds considered here , the size of the accretor has to be huge . in the limiting case of maximum sound speed and maximum velocity ,
the minimum size of the accretor is larger than @xmath58 for @xmath55 and @xmath56 and even larger than @xmath59 for @xmath60 . according to this , and having in mind the classical results
, it would not be too surprising that the relativistic patterns would appear much more stable .
finally , concerning the numerical grid , the spatial numerical domain , @xmath61 , has been covered by a canonical grid of @xmath62 numerical zones .
the radial and angular discretizations lie , respectively , in the interval @xmath63 and @xmath64 , where @xmath65 and @xmath66 depend on the particular model .
the specific values are given in table 1 .
note that , with the coordinates we are using , @xmath65 will always be slightly larger than @xmath52 in order to avoid divergences at the horizon in some hydrodynamical quantities such as , e.g. , the lorentz factor . for the angular direction we have used an equally spaced grid while in the radial direction we have employed the schwarzschild _ tortoise coordinate _ defined by @xmath67 .
we have solved the general relativistic hydrodynamic equations with the same numerical code described in paper i. we briefly summarize its main features here .
the code belongs to the so called tvd ( total variation diminishing ) schemes ( harten , 1983 ) and makes use of a linearized riemann solver ( roe , 1981 ) as an accurate procedure to capture discontinuities .
it also uses a monotonic upstream reconstruction procedure ( van leer , 1979 ) to extrapolate the variables at the cell interfaces to solve for the riemann problem and a third order runge - kutta scheme , which preserves the tvd property , to update the solution in time .
further details of the code can be found in paper i and font et al .
( 1994 ) .
the boundary conditions employed in the simulations are , basically , the same as we used in paper i. for the innermost radial boundary we have considered now two different sets of conditions , namely _ outflow _ , where all variables are linearly extrapolated to the boundary zones , and _ absorbing conditions _ , where the velocities are reset to zero and the density and internal energy are fixed to very small values . in practice , we have not found any difference between using one condition or the other .
both types of conditions give the maximum accretion rates of mass and momentum ( benensohn et al . , 1997 ) .
for the outermost radial boundary we have , again , imposed different conditions depending on the upstream or downstream position of the flow with respect to the hole . in the upstream part , the initial asymptotic values for all variables have been used at every time step . for the downstream part
, we have performed a linear extrapolation .
this boundary condition turns out to be the most critical one as , if the mach number of the flow is not high enough , unphysical information can be reflected back to the computational domain .
this could be the reason why we did not succeed in evolving low mach number models . in future work
we plan to use characteristic conditions in this boundary to see if they work better .
finally , we do not need to impose boundary conditions on the angular direction as no boundaries exist .
we have tried different angular resolutions to analyze the convergence of the numerical scheme .
we chose model mb2 of table 1 for this purpose , as it showed signs of unstable patterns in the wake .
this resolution study was motivated by the recent work by benensohn et al .
( 1997 ) where the unstable patterns only appeared if the number of angular zones to describe the entire circle was fine enough ( roughly of the order of 200 zones ) .
we performed 4 runs with 40 , 80 , 160 and 320 angular zones .
in all cases the initial conditions were identical and the flow was completely uniform and totally symmetric with respect to the @xmath68 line . with 40 zones the flow started stable and symmetric but after some time , due to purely numerical reasons , mainly accumulated roundoff error , an unstable pattern developed which even broke the wake .
this was also visible with 80 angular zones as we will show below .
however , the numerical instability appeared later and was much less violent ( meaning that the solution was converging ) . with 160 and 320 zones we have not found any difference ( even looking at the specific numbers in the data files ) between the two regions above and below @xmath68 . for practical purposes
we have always used a grid of 80 angular zones as most of the models showed good late time behaviour with this resolution .
the evolution of the models listed in table 1 is discussed here . in figs . 2 and 3 we show the final state of the first six models .
we plot in fig .
2 isocontours of the logarithmic of the density normalized to the asymptotic value while fig .
3 shows isocontours of the total velocity of the flow .
we note in both figures that all models present two clearly defined different regions upstream and downstream the accreting hole . in the upstream part of the flow , the isocontours are almost perfectly rounded , indicating slight deviation from spherical accretion .
this is clearly demonstrated in fig .
4 where we plot a radial cut along the line passing through the points @xmath68 and @xmath69 for several quantities and only for model mb3 ( the rest of models show a completely similar behaviour ) .
we can not directly compare the curves for the density and total velocity with the analytic values expected for spherical accretion ( see , e.g. , hawley , smarr and wilson 1984 ) : @xmath70 @xmath71 where @xmath72 is a constant .
the reason is that we are assuming a fixed value for the velocity at the edge of the grid . for zero velocity at infinity
this value is the free - fall velocity at the given radius ( assuming that this radius is smaller than the transonic radius ) .
however , the velocity pattern is clearly spherical as , just by simply adding a radial term to the exact velocity , of , e.g. , the form @xmath73 , where @xmath74 is a constant , it is possibly to get the numerical spherical solution .
it is worth to mention , however , that the maximum values for the velocity at the innermost point for all models completely agree with the expected free - fall values as can be read off table 2 . from fig .
4 one can also notice that , at the rear part of the hole , the density has always its maximum values , noticeably larger than the values at the front part .
this means that the majority of the material is accreted onto the hole precisely at its rear part .
the velocity plot reaches a zero value at the _ stagnation point _ behind the hole .
all material inside the radius delimited by this point is ultimately captured by the hole . on the other hand , downstream the accretor , all models show the presence of a well - defined tail shock that , starting very close to the axis , propagates all the way up to the outer boundary of the domain .
this shock wave starts at the axis as a consequence of the gravitational bending that the matter feels , which originates the collision and piling - up of the upper and lower parts of the flow .
then , after developing the bondi - hoyle accretion _ column _
, the shock starts to continuosly open to wider angles ( due to the large pressure gradient between the inner and outer parts of the wake ) until it finally acommodates itself in a fixed position .
the particular models that we have considered are always characterized by a tail - shock .
this shock never propagates to the upstream part of the flow .
as we can see in both figures the wake appears to be completely symmetric with respect to the @xmath68 line for all models .
as mentioned earlier , only slight asymmetries develop for @xmath75 models ( mb2 and mb3 ) .
this is purely a numerical instability .
results with larger resolutions ( 160 or 320 angular zones ) , not shown here , reveal a fully converged and symmetric pattern .
however , with inadequate resolutions , this numerical instability can trigger asymmetric patterns which can , eventually , broke the wake .
these asymmetries are more clearly noticeable in the velocity contours of fig .
the final model in table 1 , model ua2 , is plotted in figs . 5 and 6 .
5 shows isocontours of density and velocity while fig .
6 shows radial cuts along the line connecting @xmath68 and @xmath76 .
this model differs from the rest in having a larger value for the asymptotic sound speed .
this value is very close to its maximum given by @xmath54 , @xmath26 being the adiabatic index of the fluid .
hence , this is an ultrarelativistic model , both from the thermodynamic and kinematic points of view ( note that the asymptotic initial velocity is also very close to one ) .
the isocontour plots reveal a completely similar flow pattern to those of figs . 2 and 3 .
note that the steady - state solution is achieved here at much earlier times .
clearly there is a correlation between flow velocity and stability : the larger the speed , the more stable the pattern becomes ( and the larger the size of the accretor is ) .
let us also mention that model ua2 is a large accretor , with @xmath77 , which can help in obtaining extremely stable patterns .
the stability of the large speed flows has also been confirmed with accretion rate plots as we will show below .
in table 2 we present the main quantitative results concerning the evolution of these models . comparing the morphology of the flow with the axisymmetric configurations of paper i we notice that the shock opening angles are now smaller and , somehow closer to the expected analytic values at large distances from the object @xmath78 we can also compute this analytic estimate using the _
relativistic definition _ for the mach number ( knigl , 1980 ) @xmath79 where @xmath33 is the lorentz factor of the flow and @xmath80 is the lorentz - like factor for the sound speed , i.e. , @xmath81 , with the sound speed defined as @xmath82 . with this definition , the expected values for
the _ mildly _ relativistic models ( ma2-mc3 ) are slightly modified : from @xmath83 to @xmath84 for models labeled ` 2 ' and from @xmath85 to @xmath86 for models labeled ` 3 ' .
the corresponding value for ua2 , which has a larger asymptotic initial velocity , is now much different changing from @xmath87 to @xmath88 . in either case , the agreement with the exact solution is not too good .
there should also be mentioned that all the computed angles are quite similar despite the different mach numbers .
the dependence of the shock opening angle with the asymptotic mach number is plotted in fig .
the position of the stagnation point presents some similitudes and some differences with respect to the axisymmetric models .
the first thing we notice is that the position has very little dependence on the incident mach number , as can be seen comparing models labeled 2 with models labeled 3. on the other hand there is a clear dependence with the adiabatic index of the flow .
the position decreases as @xmath26 increases .
this behaviour was also found in the axisymmetric models .
comparing with those models of paper i we find now considerable smaller values .
all models that can be directly compared ( ma2 , mb2 and mc2 ) are now roughly a factor 2 smaller .
this is however not the case for model ua2 .
we have computed the accretion rates of mass , radial momentum ( drag rate ) and angular momentum .
the results are summarized in table 2 and in figs .
these quantities are excellent indicators to see if the solution ever approaches to a steady - state . in particular
, monitoring the angular momentum rate during the whole simulation is interesting to see if any transient accretion disc forms .
following the same procedure of paper i we now compute the mass accretion rate in the equatorial plane according to @xmath89 to compute the radial and angular momentum rates we have followed petrich et al .
( 1989 ) defining the variation in time of the @xmath90 momentum as @xmath91 where the flux integral is evaluated along the surface @xmath92 of a given volume @xmath93 and the integral including the source terms is a full volume integral . in practice ,
if we have the outer boundary sufficiently far out , one can neglect the second integral as the source terms go to zero in flat space .
however , as this is not the case in our finite numerical domain , we have always added the contribution of the sources in our results . in the above expresion
, @xmath94 are the spatial components of the stress - energy tensor and @xmath95 is the determinant of the four metric .
after some algebra , the final expresions we use are @xmath96 d\phi + \nonumber \\ \int_{r_{min}}^{r_c } \int_{0}^{2 \pi } [ -{m\over\alpha}(\tau+d ) + \alpha r(s_{\phi}v^{\phi } - 2s_r v^r ) ] dr d\phi \end{aligned}\ ] ] for the radial momentum rate , and @xmath97 for the angular momentum rate . in these expresions , @xmath98 is the critical radius at which we have computed the rates . in practice , the mass accretion rate has been obtained at the accretion radius for all models except for model ua2 where it was computed at @xmath99 ( the accretion radius is inside the horizon in this case ) .
the momentum rates have been calculated at three different parts of the grid just to check the influence of the source terms integral .
these three radii are @xmath65 , @xmath4 and @xmath66 .
we only plot the results corresponding to the accretion radius .
all mass accretion rates have been normalized to the canonical value proposed by petrich et al .
we can , thus , compare our results for the mass accretion rates with those previously found in axisymmetric computations .
in addition , the radial momentum rates for models ma2-mc3 have been scaled to one and the values corresponding to model ua2 are displayed with respect to the same scale factor .
the evolution in time of the different accretion rates appears in figs .
8 shows the mass accretion rates for models ma2 to mc3 and fig .
9 shows the same quantity for model ua2 .
this mass rate has been scaled by the same canonical value used in paper i. the scaled radial momentum rates for models ma2 to mc3 are plotted in fig .
10 while fig .
11 shows this quantity for model ua2 .
finally , the angular momentum accretion rates for the different models appear in figs . 12 and 13 . from these figures
we can again conclude that all models approach to a remarkably well defined final steady - state .
all the conclusions derived in the axisymmetric simulations are still valid here . we find again that , given a fixed value of @xmath47 , the mass accretion rate increases as @xmath26 goes from 1.1 to @xmath11 .
the dependence with @xmath47 for a fixed value of @xmath26 is , however , not too clear , and the numerical values show only small discrepancies .
this is particularly true for models mb2 and mb3 which give the same final value . as in the axisymmetric simulations of paper i , the model with an almost maximum sound speed , model ua2 , displays much larger values . comparing the values of table 2 for the mass accretion rates with the axisymmetric results
, we found that , for models ma2 , mb2 and mc2 now we get roughly 10 times larger values .
for model ua2 the difference is even bigger , a factor 16 .
the radial momentum accretion rate plots clearly show the same behaviour than the mass accretion plots . after some initial time the solution relaxes to a constant value .
this happens after @xmath100 for all models with low sound speed ( ma2-mc3 ) and much earlier , @xmath101 , for model ua2 .
as mentioned previously , we have computed the radial momentum rate at three different locations .
although we do not plot it here , we have verified that , as expected , for larger values of the radial coordinate , the integral containing the source terms in equation ( 18 ) becomes less relevant and the solution converges to a constant value , both in radius and time .
it is also interesting to look at the mass and radial momentum accretion rates plots to compare the stability of the different models .
first thing to notice is that both types of figures display the same behaviour .
focussing on models ma2-mc3 , we can see that those with @xmath55 are the most smooth and stable ones .
as @xmath26 increases to @xmath56 there are some small and irregular oscillations .
finally , for @xmath60 , the oscillations around the central equilibrium value are clearly of constant period and their amplitude is fairly small .
we have also found that models mb2 and mb3 with higher resolution ( 160 angular zones ) show the same periodic feature than models mc2 and mc3 . on the other hand , our ultrarelativistic model , ua2 , shows , apart from an oscillatory phase around @xmath102 , a very smooth constant pattern .
this is due to the large velocities and large size of this accretor , and is also in agreement with the rest of @xmath55 models ( ma2 and ma3 ) .
finally , we have computed the angular momentum accretion rates . by looking at figs .
12 and 13 one can clearly notice that the flow never succeeds in getting high angular momentum values , except for some large amplitude oscillations at late times for model mb3 and at an intermediate time for model ua2 .
this means that no transient accretion discs actually form , which would have not been the case if any flip - flop instability would have developed .
all models show , basically , the same behaviour : an initial constant , almost zero value , followed by a period of oscillation around that central value . in some cases , as model mc3 , this oscillatory phase shows signs of growing linearly , although at the time the simulation was stopped , the amplitude of the oscillations was still small .
in particular , models ma2 , ma3 , mc2 and mc3 display very small maximum amplitudes .
models with @xmath75 , mb2 and mb3 , show , however , larger oscillations and , in particular , model mb3 shows a clear deviation from zero at very late times . as mentioned previously , the reason for this is again purely numeric and directly related to the unresolved wake patterns obtained with our canonical grid size .
the corresponding plot for model ua2 appears in fig .
although at some point of the evolution it gets some large oscillations , it eventually settles down to small values .
we have performed a detailed numerical study of relativistic bondi - hoyle accretion onto a moving schwarzshild black hole .
we have extended our previous work to include non - axisymmetric configurations .
this has been motivated by the possibility of finding unstable accretion patterns in the flow .
we have only considered flows which are initially uniform .
we have found that , contrary to newtonian simulations , the non - axisymmetric solution is always stable to tangetial oscillations . however , this statement should be relaxed , as the corresponding initial setup in our relativistic wind simulations can not reproduce , for physical reasons , the one used in classical simulations . in particular , we have considered values of the accretion radius which , when used in the classical regime , would have not given rise to any instability as well .
this may indicate that , in the relativistic regime , a uniform flow accreting onto a black hole could always remain symmetric and stationary
. it might be worthwhile to go to much larger values of the asymptotic mach number to see if any unstable pattern actually develops .
these are , however , quite challenging simulations that would push the code to its limits , as the asymptotic sound speed should be considerably small .
we have checked the stability and stationarity of our relativistic accretion patterns by analyzing the broad morphology of the flow with isocontour plots as well as by integrating , over all time of the simulation , global quantities as the accretion rates of mass and momentum . in all cases considered ,
we have found that the mass and radial momentum accretion rates approach asymptotically to very accurate constant values in time .
in addition , the angular momentum accretion rate always remains in very low values , indicating the total absence of rotating flows around the hole which could have triggered any tangetial instability .
it maybe worthwhile to mention that , though we have restricted in the present work to homogeneous flows , the code is capable of evolving more complicated non - axisymmetric configurations where the flow has some amount of angular momentum at infinity .
we are currently working in a parametric study of this kind of flows ( pons et al . , 1998 ) .
in particular , we plan to gradually increase the angular momentum of the incident flow and study how it is transfered onto the hole via alternate spiral shocks . at the same time
, we also plan to add rotation to the hole a feature that the actual code used in this work already incorporates extending thus the relativistic bondi - hoyle accretion to moving kerr black holes .
ultimately , we plan to perform these simulations in three spatial dimensions .
this work has been supported by the spanish dgicyt ( grant pb94 - 0973 ) . j.a.f has benefited from a european contract ( nr .
erbfmbict971902 ) .
all computations have been performed at the aei in potsdam .
+ + model & @xmath46 & @xmath26 & @xmath47 & @xmath103 & @xmath8 & @xmath104 & @xmath105 & @xmath106 & @xmath107 ma2 & @xmath108 & @xmath109 & @xmath110 & @xmath111 & @xmath112 & @xmath113 & @xmath114 & @xmath115 & @xmath116 @xmath46 is the asymptotic sound speed , @xmath26 is the adiabatic exponent , @xmath47 is the asymptotic mach number ( newtonian definition ) , @xmath103 is the asymptotic relativistic mach number ( see definition in the text ) , @xmath8 is the asymptotic flow velocity , @xmath4 is the accretion radius , @xmath65 and @xmath66 are the minimum and maximum radial values of the computational domain and @xmath129 is the final time , in units of @xmath6 , at which the simulation is stopped .
the specific names of the models are chosen to be in agreement with those of paper i. + + model & @xmath130 & @xmath131 & @xmath132 & @xmath133 & @xmath134 & @xmath135 & @xmath136 & @xmath137 & @xmath138 & @xmath139 & @xmath140 & @xmath141 ma2 & @xmath83 & @xmath84 & @xmath113 & @xmath142 & @xmath143 & @xmath144 & @xmath145 & @xmath146 & @xmath147 & @xmath148 & @xmath149 & @xmath150 the asterisk in models mb2 and mb3 indicates that the results have been obtained with @xmath203 angular zones .
@xmath204 is the analytical value for the shock opening angle at large distances according to @xmath205 ( newtonian definition for the mach number ) , @xmath131 is the numerical result , @xmath132 is the mean value of the stagnation point position , @xmath133 is the maximum in the density in the upstream direction along the line @xmath76 , @xmath134 is the corresponding maximum in the density downstream @xmath68 , @xmath135 is the maximum upwind velocity , @xmath136 is the maximum downwind velocity , @xmath137 is the maximum value of the upwind mach number and @xmath138 is its corresponding maximum value in the downwind direction . in addition
@xmath139 is the normalized mass accretion rate ( to the canonical value used in paper i ) and @xmath140 and @xmath141 are the scaled rates of radial and angular momentum , respectively . for the latter quantity
we only indicate the maximum values . | we present the results of an exhaustive numerical study of fully relativistic non - axisymmetric bondi - hoyle accretion onto a moving schwarzschild black hole .
we have solved the equations of general relativistic hydrodynamics with a high - resolution shock - capturing numerical scheme based on a linearized riemann solver .
the numerical code was previously used to study axisymmetric flow configurations past a schwarzschild hole .
we have analyzed and discussed the flow morphology for a sample of asymptotically high mach number models .
the results of this work reveal that initially asymptotic uniform flows always accrete onto the hole in a stationary way which closely resembles the previous axisymmetric patterns .
this is in contrast with some newtonian numerical studies where violent flip - flop instabilities were found .
as discussed in the text , the reason can be found in the initial conditions used in the relativistic regime , as they can not exactly duplicate the previous newtonian setups where the instability appeared .
the dependence of the final solution with the inner boundary condition as well as with the grid resolution has also been studied . finally , we have computed the accretion rates of mass and linear and angular momentum .
-.5 in 0.0 in 0.0 in 6.4 in 8.75 in 0.5 in = 24pt * non - axisymmetric relativistic bondi - hoyle accretion onto a schwarzschild black hole * jos a. font@xmath0 and j.m@xmath1 ibez@xmath2 @xmath0 max - planck - institut fr gravitationsphysik , albert - einstein - institut + schlaatzweg 1 , 14473 potsdam , germany + e - mail : [email protected] @xmath2 departamento de astronoma y astrofsica + universidad de valencia , 46100 burjassot ( valencia ) , spain + e - mail : [email protected] * key words * accretion hydrodynamics methods : numerical relativity shock waves |
one of the fundamental ideas of modern cosmology is that there was an epoch early in the history of the universe when potential , or vacuum , energy dominated other forms of energy densities such as matter or radiation . during such a vacuum - dominated era
the scale factor grew exponentially ( or nearly exponentially ) in some small time . during this phase , dubbed inflation @xcite , a small , smooth spatial region of size of order the hubble radius grew so large that it easily could encompass the comoving volume of the entire presently observable universe .
if the universe underwent such a period of rapid expansion , one can understand why the observed universe is homogeneous and isotropic to such high accuracy .
one of the predictions of the simplest models of inflation is a spatially flat universe , _
@xmath4 , with great precision .
inflation has also become the dominant paradigm for understanding the initial conditions for structure formation and for cosmic microwave background ( cmb ) anisotropy . in the inflationary picture ,
primordial density and gravity - wave ( tensor ) fluctuations are created from quantum fluctuations and `` redshifted '' out of the horizon during an early period of superluminal expansion of the universe , where they are `` frozen '' as perturbations in the background metric @xcite .
metric perturbations at the surface of last scattering are observable as temperature anisotropies in the cmb .
the first and most impressive confirmation of the inflationary paradigm came when the cmb anisotropies were detected by the cosmic background explorer ( cobe ) satellite in 1992 @xcite .
subsequently , it became clear that the measurements of the spectrum of the cmb anisotropy can provide very detailed information about fundamental cosmological parameters @xcite and other crucial parameters for particle physics .
in the past few years , a number of balloon - borne and terrestrial experiments have mapped out the cmb angular anisotropies @xcite , revealing a remarkable agreement between the data and the inflationary predictions of a flat universe with a nearly scale - invariant spectrum of adiabatic primordial density perturbations , ( see _ e.g. , _
@xcite ) . despite the simplicity of the inflationary paradigm ,
the number of inflation models that have been proposed in the literature is enormous @xcite .
this is true even if we limit ourselves to models with only one scalar field ( the _ inflaton _ ) . with the previous data on cmb anisotropies from balloon and terrestrial experiments it has been possible for the first time to place interesting constraints on the space of possible inflation models @xcite
however , the quality of the data were not good enough to rule out entire classes of models .
a boost along these lines has been very recently provided by the data of the wilkinson microwave anisotropy probe ( wmap ) mission , which has marked the beginning of the precision era of the cmb measurements in space @xcite .
the wmap collaboration has produced a full - sky map of the angular variations in the microwave flux , in particular the cosmic microwave background , with unprecedented accuracy .
wmap data support the inflationary mechanism for the generation of curvature superhorizon fluctuations and provide a strong bound on the possible admixture of isocurvature modes @xcite .
furthermore , consistent with the simplest single - field models of inflation @xcite , no evidence of nongaussianity is found @xcite .
the goal of this paper is to use the wmap data to discriminate among the various single - field inflationary models . to obtain some indication of the robustness of our analysis
, we also consider a data set consisting of wmap augmented with several other cmb experiments . for single - field inflation models , the relevant parameter space for distinguishing among models is defined by the scalar spectral index @xmath5 , the ratio of tensor to scalar fluctuations @xmath6 , and the running of the scalar spectral index @xmath7 .
we employ _ monte carlo reconstruction _ , a stochastic method for `` inverting '' observational constraints to determine an ensemble of inflationary potentials compatible with observation @xcite .
in addition to encompassing a broader set of models than usually considered ( large - field , small - field , hybrid and linear models ) , monte carlo reconstruction makes it possible easily to incorporate constraints on the running of the spectral index as well as to include effects to higher order in slow roll . since studies on the implications of wmap data for inflation @xcite have already appeared , we briefly mention the different elements between our analysis and others .
( we will elaborate on these differences later . )
the wmap collaboration analysis @xcite included wmap data , additional cmb data ( cbi @xcite and acbar @xcite ) , large - scale structure data ( 2dfgrs @xcite ) , as well as lyman-@xmath8 power spectrum data @xcite .
they used a markov chain monte carlo technique to explore the likelihood surface .
et al . _
@xcite , considered wmap data only , but with a top - hat prior on the hubble constant @xmath9 ( @xmath10 100 @xmath9 km sec@xmath11 mpc@xmath11 ) from the hst key project @xcite . also ,
_ did not consider a running of the scalar spectral index .
we only consider cmb data .
we first analyze just the wmap results .
we then analyze the wmap data set in conjunction with other cmb data sets ( boomerang-98 @xcite , maxima-1 @xcite , dasi @xcite , cbi @xcite , acbar @xcite , vsae @xcite , and archeops @xcite ) .
we employ a grid of models in the likelihood analysis , which differs from the method used by the wmap team .
the paper is organized as follows . in sec .
ii we discuss single - field inflation and the relevant observables in more detail . in sec .
iii we discuss the inflationary model space , and in sec .
iv we describe the monte carlo reconstruction technique .
section v describes the methods used for the cmb analysis . in sec .
vi we present the constraints from the cmb anisotropy data sets . in sec .
vii we present our conclusions .
in this section we briefly review scalar field models of inflationary cosmology , and explain how we relate model parameters to observable quantities .
inflation , in its most general sense , can be defined to be a period of accelerating cosmological expansion during which the universe evolves toward homogeneity and flatness .
this acceleration is typically a result of the universe being dominated by vacuum energy , with an equation of state @xmath12 . within this broad framework ,
many specific models for inflation have been proposed .
we limit ourselves here to models with `` normal '' gravity ( _ i.e. , _ general relativity ) and a single order parameter for the vacuum , described by a slowly rolling scalar field @xmath13 , the inflaton .
these assumptions are not overly restrictive ; the most widely studied inflation models fall within this category , including linde s `` chaotic '' inflation scenario @xcite , inflation from pseudo nambu - goldstone bosons ( `` natural '' inflation @xcite ) , dilaton - like models involving exponential potentials ( power - law inflation ) , hybrid inflation @xcite , and so forth .
other models , such as starobinsky s @xmath14 model @xcite and versions of extended inflation , can , through a suitable transformation , be viewed in terms of equivalent single - field models .
of course in single - field models of inflation , the inflaton `` field '' need not be a fundamental field at all .
also , some `` single - field '' models require auxiliary fields .
hybrid inflation models @xcite , for example , require a second field to end inflation .
what is significant is that the inflationary epoch be described by a single dynamical order parameter , the inflaton field .
a scalar field in a cosmological background evolves with an equation of motion @xmath15 the evolution of the scale factor is given by the scalar field dominated frw equation , @xmath16,\cr \left(\ddot a \over a\right ) & & = { 8 \pi \over 3 m_{\rm pl}^2 } \left[v\left(\phi\right ) - \dot\phi^2\right ] .
\label{eqbackground}\end{aligned}\ ] ] here @xmath17 is the planck mass , and we have assumed a flat friedmann - robertson - walker metric , @xmath18 where @xmath19 is the scale factor of the universe .
_ inflation _ is defined to be a period of accelerated expansion , @xmath20 . a powerful way of describing the dynamics of a scalar field - dominated cosmology is to express the hubble parameter as a function of the field @xmath13 , @xmath21 , which is consistent provided @xmath13 is monotonic in time .
the equations of motion become @xcite : @xmath22 ^ 2 - { 12 \pi \over m_{\rm pl}^2 } h^2(\phi ) = - { 32 \pi^2 \over m_{\rm pl}^4 } v(\phi).\label{eqbasichjequations}\end{aligned}\ ] ] these are completely equivalent to the second - order equation of motion in eq.([eqequationofmotion ] ) .
the second of the above equations is referred to as the _ hamilton - jacobi _ equation , and can be written in the useful form @xmath23 = \left({8 \pi \over 3 m_{\rm pl}^2}\right ) v(\phi),\label{eqhubblehamiltonjacobi}\ ] ] where @xmath24 is defined to be @xmath25 the physical meaning of @xmath24 can be seen by expressing eq.([eqbackground ] ) as @xmath26,\ ] ] so that the condition for inflation @xmath27 is given by @xmath28 .
the scale factor is given by @xmath29,\ ] ] where the number of e - folds @xmath30 is @xmath31 to create the observed flatness and homogeneity of the universe , we require many e - folds of inflation , typically @xmath32 . this figure varies somewhat with the details of the model .
we can relate a comoving scale @xmath33 in the universe today to the number of e - folds @xmath30 before the end of inflation by @xcite @xmath34 here @xmath35 is the potential when the mode leaves the horizon , @xmath36 is the potential at the end of inflation , and @xmath37 is the energy density after reheating .
scales of order the current horizon size exited the horizon at @xmath38 . since this number depends , for example , on the details of reheating , we will allow @xmath30 to vary within the range @xmath39 for any given model in order to consider the most general case .
( dodelson and hui have recently argued that the value of @xmath30 corresponding to the current horizon size can be no larger than 60 @xcite , so in this sense we are being more general than is necessary . ) we will frequently work within the context of the _ slow roll _
approximation @xcite , which is the assumption that the evolution of the field is dominated by drag from the cosmological expansion , so that @xmath40 and @xmath41 the equation of state of the scalar field is dominated by the potential , so that @xmath42 , and the expansion rate is approximately @xmath43 the slow roll approximation is consistent if both the slope and curvature of the potential are small , @xmath44 . in this case
the parameter @xmath24 can be expressed in terms of the potential as @xmath45 we will also define a second `` slow roll parameter '' @xmath46 by : @xmath47.\end{aligned}\ ] ] slow roll is then a consistent approximation for @xmath48 .
inflation models not only explain the large - scale homogeneity of the universe , but also provide a mechanism for explaining the observed level of _ inhomogeneity _ as well . during inflation , quantum fluctuations on small scales
are quickly redshifted to scales much larger than the horizon size , where they are `` frozen '' as perturbations in the background metric @xcite .
the metric perturbations created during inflation are of two types : scalar , or _ curvature _
perturbations , which couple to the stress - energy of matter in the universe and form the `` seeds '' for structure formation , and tensor , or gravitational wave perturbations , which do not couple to matter . both scalar and tensor perturbations
contribute to cmb anisotropy .
scalar fluctuations can also be interpreted as fluctuations in the density of the matter in the universe .
scalar fluctuations can be quantitatively characterized by perturbations @xmath49 in the intrinsic curvature scalar . as long as the equation of state @xmath24 is slowly varying
, the curvature perturbation can be shown to be @xcite @xmath50_{k = a h}.\ ] ] the fluctuation power spectrum is in general a function of wavenumber @xmath33 , and is evaluated when a given mode crosses outside the horizon during inflation , @xmath51 . outside the horizon , modes do not evolve , so the amplitude of the mode when it crosses back _ inside _ the horizon during a later radiation- or matter - dominated epoch is just its value when it left the horizon during inflation . the _ spectral index _
@xmath5 for @xmath49 is defined by @xmath52 so that a scale - invariant spectrum , in which modes have constant amplitude at horizon crossing , is characterized by @xmath53 .
some inflation models predict running of the spectral index with scale @xcite or even sharp features in the power spectrum @xcite .
we will consider the running of the spectral index in more detail on sec .
[ secmontecarlorecon ] . instead of specifying the fluctuation amplitude directly as a function of @xmath33 ,
it is often convenient to specify it as a function of the number of e - folds @xmath30 before the end of inflation at which a mode crossed outside the horizon .
scales of interest for current measurements of cmb anisotropy crossed outside the horizon at @xmath54 , so that @xmath49 is conventionally evaluated at @xmath55 .
the power spectrum of tensor fluctuation modes is given by @xcite @xmath56_{n}.\ ] ] the ratio of tensor to scalar modes is then @xmath57 so that tensor modes are negligible for @xmath58 .
tensor and scalar modes both contribute to cmb temperature anisotropy .
if the contribution of tensor modes to the cmb anisotropy can be neglected , normalization to the cobe four - year data gives @xcite @xmath59 . in the next section
, we will describe the predictions of various models in this parameter space .
to summarize the results of the previous section , inflation generates scalar ( density ) and tensor ( gravity wave ) fluctuations which are generally well approximated by power laws : @xmath60 in the limit of slow roll , the spectral indices @xmath5 and
@xmath61 vary slowly or not at all with scale . we can write the spectral indices @xmath5 and @xmath61 to lowest order in terms of the slow roll parameters @xmath24 and @xmath46 as @xcite : @xmath62 the tensor spectral index is _ not _ an independent parameter , but is proportional to the tensor / scalar ratio , given to lowest order in slow roll by @xmath63 this is known as the _ consistency relation _ for inflation .
( this relation holds only for single - field inflation , and weakens to an inequality for inflation involving multiple degrees of freedom @xcite . ) a given inflation model can therefore be described to lowest order in slow roll by three independent parameters , @xmath49 , @xmath64 , and @xmath5 .
if we wish to include higher - order effects , we have a fourth parameter describing the running of the scalar spectral index , @xmath65 .
the tensor / scalar ratio is frequently expressed as a ratio of their contributions to the cmb quadrupole , @xmath66 the relation between @xmath6 and ratio of amplitudes in the primordial power spectra @xmath67 depends on the background cosmology , in particular the densities of matter ( @xmath68 ) and cosmological constant ( @xmath69 ) .
for the currently favored values of @xmath70 and @xmath71 , the relation is approximately @xmath72 to lowest order in slow roll .
conventions for the normalization of this parameter vary widely in the literature .
in particular , peiris et al .
@xcite use @xmath73 . calculating the cmb fluctuations from a particular inflationary model
reduces to the following basic steps : ( 1 ) from the potential , calculate @xmath24 and @xmath46 .
( 2 ) from @xmath24 , calculate @xmath30 as a function of the field @xmath13 . ( 3 )
invert @xmath74 to find @xmath75 .
( 4 ) calculate @xmath49 , @xmath5 , and @xmath76 as functions of @xmath13 , and evaluate them at @xmath77 . for the remainder of the paper ,
all parameters are assumed to be evaluated at @xmath78 . even restricting ourselves to a simple single - field inflation scenario ,
the number of models available to choose from is large @xcite .
it is convenient to define a general classification scheme , or `` zoology '' for models of inflation .
we divide models into three general types : _ large - field _
, _ small - field _ , and _ hybrid _ , with a fourth classification , _ linear _ models , serving as a boundary between large- and small - field .
a generic single - field potential can be characterized by two independent mass scales : a `` height '' @xmath79 , corresponding to the vacuum energy density during inflation , and a `` width '' @xmath80 , corresponding to the change in the field value @xmath81 during inflation : @xmath82 different models have different forms for the function @xmath83 .
the height @xmath84 is fixed by normalization , so the only free parameter is the width @xmath80 .
with the normalization fixed , the relevant parameter space for distinguishing between inflation models to lowest order in slow roll is then the @xmath85 plane .
( to next order in slow - roll parameters , one must introduce the running of @xmath5 . )
different classes of models are distinguished by the value of the second derivative of the potential , or , equivalently , by the relationship between the values of the slow - roll parameters @xmath24 and @xmath46 .
model @xcite and the `` dual inflation '' model @xcite are characterized by @xmath86 , but are `` small - field '' in the sense that @xmath87 , with @xmath88 and negligible tensor modes . ]
each class of models has a different relationship between @xmath6 and @xmath5 . for a more detailed discussion of these relations ,
the reader is referred to refs.@xcite .
first order in @xmath24 and @xmath46 is sufficiently accurate for the purposes of this section , and for the remainder of this section we will only work to first order .
the generalization to higher order in slow roll will be discussed in sec .
[ secmontecarlorecon ] .
large - field models have inflaton potentials typical of `` chaotic '' inflation scenarios @xcite , in which the scalar field is displaced from the minimum of the potential by an amount usually of order the planck mass .
such models are characterized by @xmath90 , and @xmath91 .
the generic large - field potentials we consider are polynomial potentials @xmath92 , and exponential potentials , @xmath93 .
for the case of an exponential potential , @xmath94 , the tensor / scalar ratio @xmath6 is simply related to the spectral index as @xmath95 this result is often incorrectly generalized to all slow - roll models , but is in fact characteristic _ only _ of power - law inflation . for inflation with a polynomial potential , @xmath96 , we again have @xmath97 , @xmath98 so that tensor modes are large for significantly tilted spectra .
we will be particularly interested in models with @xmath99 as a test case for our ability to rule out models . for @xmath99 ,
the observables are given in terms of the number of e - folds @xmath30 by @xmath100 small - field models are the type of potentials that arise naturally from spontaneous symmetry breaking ( such as the original models of `` new '' inflation @xcite ) and from pseudo nambu - goldstone modes ( natural inflation @xcite ) .
the field starts from near an unstable equilibrium ( taken to be at the origin ) and rolls down the potential to a stable minimum .
small - field models are characterized by @xmath102 and @xmath101 . typically @xmath24 ( and hence the tensor amplitude )
is close to zero in small - field models .
the generic small - field potentials we consider are of the form @xmath103 $ ] , which can be viewed as a lowest - order taylor expansion of an arbitrary potential about the origin .
the cases @xmath104 and @xmath105 have very different behavior . for @xmath104 , @xmath106,\ ] ] where @xmath30 is the number of e - folds of inflation . for @xmath105 ,
the scalar spectral index is @xmath107 _ independent _ of @xmath6 . assuming @xmath108 results in an upper bound on @xmath6 of @xmath109 the hybrid scenario @xcite frequently appears in models which incorporate inflation into supersymmetry . in a typical hybrid inflation model , the scalar field responsible for inflation evolves toward a minimum with nonzero vacuum energy .
the end of inflation arises as a result of instability in a second field .
such models are characterized by @xmath90 and @xmath110 .
we consider generic potentials for hybrid inflation of the form @xmath111.$ ] the field value at the end of inflation is determined by some other physics , so there is a second free parameter characterizing the models . because of this extra freedom ,
hybrid models fill a broad region in the @xmath85 plane ( see fig . [ figregions ] )
. there is , however , no overlap in the @xmath85 plane between hybrid inflation and other models . the distinguishing feature of many hybrid models is a _
scalar spectral index , @xmath112 .
this corresponds to the case @xmath113 .
hybrid models can also in principle have a red spectrum , @xmath88 .
linear models , @xmath115 , live on the boundary between large - field and small - field models , with @xmath116 and @xmath117 .
the spectral index and tensor / scalar ratio are related as : @xmath118 this enumeration of models is certainly not exhaustive .
there are a number of single - field models that do not fit well into this scheme , for example logarithmic potentials @xmath119 typical of supersymmetry @xcite .
another example is potentials with negative powers of the scalar field @xmath120 used in intermediate inflation @xcite and dynamical supersymmetric inflation @xcite . both of these cases require an auxiliary field to end inflation and are more properly categorized as hybrid models , but fall into the small - field region of the @xmath85 plane
however , the three classes categorized by the relationship between the slow - roll parameters as @xmath121 ( large - field ) , @xmath122 ( small - field , linear ) , and @xmath110 ( hybrid ) , cover the entire @xmath85 plane and are in that sense complete . for large - field and @xmath123 for small - field . ] figure 1 @xcite shows the @xmath85 plane divided into regions representing the large field , small - field and hybrid cases .
figure 2 shows a `` zoo plot '' of the particular potentials considered here plotted on the @xmath85 plane .
in this section we describe _ monte carlo reconstruction _ , a stochastic method for `` inverting '' observational constraints to determine an ensemble of inflationary potentials compatible with observation .
the method is described in more detail in refs .
in addition to encompassing a broader set of models than we considered in sec .
[ seczoology ] , monte carlo reconstruction allows us easily to incorporate constraints on the running of the spectral index @xmath124 as well as to include effects to higher order in slow roll .
we have defined the slow roll parameters @xmath24 and @xmath46 in terms of the hubble parameter @xmath125 as @xmath126 these parameters are simply related to observables @xmath127 , and @xmath128 to first order in slow roll .
( we discuss higher order expressions for the observables below . )
taking higher derivatives of @xmath129 with respect to the field , we can define an infinite hierarchy of slow roll parameters @xcite : @xmath130,\cr { } ^\ell\lambda_{\rm h } & \equiv & \left({m_{\rm pl}^2 \over 4 \pi}\right)^\ell { \left(h'\right)^{\ell-1 } \over h^\ell } { d^{(\ell+1 ) } h \over d\phi^{(\ell + 1)}}.\end{aligned}\ ] ] here we have chosen the parameter @xmath131 to make comparison with observation convenient .
it is convenient to use @xmath30 as the measure of time during inflation .
as above , we take @xmath132 and @xmath133 to be the time and field value at end of inflation .
therefore , @xmath30 is defined as the number of e - folds before the end of inflation , and increases as one goes _
backward _ in time ( @xmath134 ) : @xmath135 where we have chosen the sign convention that @xmath136 has the same sign as @xmath137 : @xmath138 then @xmath24 itself can be expressed in terms of @xmath129 and @xmath30 simply as , @xmath139 similarly , the evolution of the higher order parameters during inflation is determined by a set of `` flow '' equations @xcite , @xmath140 \left({}^\ell\lambda_{\rm h}\right ) + { } ^{\ell+1}\lambda_{\rm h}.\label{eqfullflowequations}\end{aligned}\ ] ] the derivative of a slow roll parameter at a given order is higher order in slow roll .
a boundary condition can be specified at any point in the inflationary evolution by selecting a set of parameters @xmath141 for a given value of @xmath30 .
this is sufficient to specify a `` path '' in the inflationary parameter space that specifies the background evolution of the spacetime .
taken to infinite order , this set of equations completely specifies the cosmological evolution , up to the normalization of the hubble parameter @xmath129 .
furthermore , such a specification is exact , with no assumption of slow roll necessary . in practice
, we must truncate the expansion at finite order by assuming that the @xmath142 are all zero above some fixed value of @xmath143 .
we choose initial values for the parameters at random from the following ranges : @xmath144\cr \epsilon & = & \left[0,0.8\right]\cr \sigma & = & \left[-0.5,0.5\right]\cr { } ^2\lambda_{\rm h } & = & \left[-0.05,0.05\right]\cr { } ^3\lambda_{\rm h } & = & \left[-0.025,0.025\right],\cr & \cdots&\cr { } ^{m+1}\lambda_{\rm h } & = & 0.\label{eqinitialconditions}\end{aligned}\ ] ] here the expansion is truncated to order @xmath145 by setting @xmath146 . in this case
, we still generate an exact solution of the background equations , albeit one chosen from a subset of the complete space of models .
this is equivalent to placing constraints on the form of the potential @xmath147 , but the constraints can be made arbitrarily weak by evaluating the expansion to higher order .
for the purposes of this analysis , we choose @xmath148 .
the results are not sensitive to either the choice of order @xmath145 ( as long as it is large enough ) or to the specific ranges from which the initial parameters are chosen .
once we obtain a solution to the flow equations @xmath149 $ ] , we can calculate the predicted values of the tensor / scalar ratio @xmath6 , the spectral index @xmath5 , and the `` running '' of the spectral index @xmath150 .
to lowest order , the relationship between the slow roll parameters and the observables is especially simple : @xmath151 , @xmath152 , and @xmath153 . to second order in slow roll ,
the observables are given by @xcite , @xmath154,\label{eqrsecondorder}\ ] ] for the tensor / scalar ratio , and @xmath155 for the spectral index .
the constant @xmath156 , where @xmath157 is euler s constant . , given by @xmath158 .
the effect of this error is significant at second order in slow roll . ]
derivatives with respect to wavenumber @xmath33 can be expressed in terms of derivatives with respect to @xmath30 as @xcite @xmath159 the scale dependence of @xmath5 is then given by the simple expression @xmath160 which can be evaluated by using eq .
( [ eqnsecondorder ] ) and the flow equations .
for example , for the case of @xmath161 , the observables to lowest order are @xmath162 the final result following the evaluation of a particular path in the @xmath145-dimensional `` slow roll space '' is a point in `` observable parameter space , '' i.e. , @xmath163 , corresponding to the observational prediction for that particular model .
this process can be repeated for a large number of models , and used to study the attractor behavior of the inflationary dynamics .
in fact , the models cluster strongly in the observable parameter space @xcite .
figure [ figzooplot2 ] shows an ensemble of models generated stochastically on the @xmath164 plane , along with the predictions of the specific models considered in sec .
[ seczoology ] . figure [ nnk ] shows an ensemble of models generated stochastically on the @xmath165 plane . as one can see , and contrary to what commonly believed , there are single - field models of inflation which predict a significant running of the spectral index .
the same can be appreciated in fig .
[ figrnk ] , where we plot an ensemble of models generated stochastically on the @xmath166 plane .
the reconstruction method works as follows : 1 .
specify a `` window '' of parameter space : _
e.g. , _ central values for @xmath167 , @xmath6 , or @xmath168 and their associated error bars .
2 . select a random point in slow roll space , @xmath169 $ ] , truncated at order @xmath145 in the slow roll expansion .
3 . evolve forward in time ( @xmath170 ) until either ( a ) inflation ends ( @xmath171 ) , or ( b ) the evolution reaches a late - time fixed point ( @xmath172 ) .
4 . if the evolution reaches a late - time fixed point , calculate the observables @xmath6 , @xmath173 , and @xmath174 at this point .
if inflation ends , evaluate the flow equations backward @xmath30 e - folds from the end of inflation .
calculate the observable parameters at that point .
if the observable parameters lie within the specified window of parameter space , compute the potential and add this model to the ensemble of `` reconstructed '' potentials . 7 .
repeat steps 2 through 6 until the desired number of models have been found .
the condition for the end of inflation is that @xmath175 . integrating the flow equations forward in time will yield two possible outcomes .
one possibility is that the condition @xmath175 may be satisfied for some finite value of @xmath30 , which defines the end of inflation .
we identify this point as @xmath176 so that the primordial fluctuations are actually generated when @xmath177 .
alternatively , the solution can evolve toward an inflationary attractor with @xmath178 and @xmath112 , in which case inflation never stops . in reality
, inflation must stop at some point , presumably via some sort of instability , such as the `` hybrid '' inflation mechanism @xcite . here
we make the simplifying assumption that the observables for such models are the values at the late - time attractor .
given a path in the slow roll parameter space , the form of the potential is fixed , up to normalization @xcite .
the starting point is the hamilton - jacobi equation , @xmath179.\label{eqhjpotential}\ ] ] we have @xmath180 trivially from the flow equations . in order to calculate the potential
, we need to determine @xmath181 and @xmath182 . with
@xmath24 known , @xmath181 can be determined by inverting the definition of @xmath24 , eq.([eqepsilonfromn ] ) .
similarly , @xmath182 follows from the first hamilton - jacobi equation ( [ eqbasichjequations ] ) : @xmath183 using these equations and eq .
( [ eqhjpotential ] ) , the form of the potential can then be fully reconstructed from the numerical solution for @xmath180 .
the only necessary observational input is the normalization of the hubble parameter @xmath129 , which enters the above equations as an integration constant . here
we use the simple condition that the density fluctuation amplitude ( as determined by a first - order slow roll expression ) be of order @xmath184 , @xmath185 a more sophisticated treatment would perform a full normalization to the cobe cmb data @xcite .
the value of the field , @xmath13 , also contains an arbitrary , additive constant .
our analysis method is based on the computation of a likelihood distribution over a fixed grid of pre - computed theoretical models .
we restrict our analysis to a flat , adiabatic , @xmath84-cdm model template computed with cmbfast ( @xcite ) , sampling the parameters as follows : @xmath186 , in steps of @xmath187 ; @xmath188 , in steps of @xmath189 and @xmath190 , in steps of @xmath191 .
the value of the hubble constant is not an independent parameter , since : @xmath192 and we use the further prior : @xmath193 .
we allow for a reionization of the intergalactic medium by varying the compton optical depth parameter @xmath194 in the range @xmath195 in steps of @xmath191 .
our choice of the above parameters is motivated by big bang nucleosynthesis bounds on @xmath196 ( both from d @xcite and @xmath197 @xcite ) , from supernovae ( @xcite ) and galaxy clustering observations ( see _ e.g. , _
@xcite ) , and by the wmap temperature - polarization cross - correlation data , which indicate an optical depth @xmath198 ( @xcite ) .
our choice for an upper limit of @xmath199 is very conservative respect to the maximum values expected in numerical simulations ( see e.g. , @xcite ) even in the case of non - standard reionization processes . from the grid above ,
we only consider models with an age of the universe in excess of 11 gyr .
variations in the inflationary parameters @xmath5 , @xmath6 and @xmath200 are not computationally relevant , and for the range of values we considered they can be assumed as free parameters . the tensor spectral index @xmath201
is determined by the consistency relation .
for the wmap data we use the recent temperature and cross polarization results from ref .
@xcite and compute the likelihood @xmath202 for each theoretical model as explained in ref .
@xcite , using the publicly available code on the lambda web site ( http://lambda.gsfc.nasa.gov/ ) .
we further include the results from seven other experiments : boomerang-98 @xcite , maxima-1 @xcite , dasi @xcite , cbi @xcite , acbar @xcite , vsae @xcite , and archeops @xcite .
the expected theoretical gaussian signal inside the bin @xmath203 is computed by using the publicly available window functions and lognormal prefactors as in ref.@xcite .
the likelihood @xmath204 from this dataset and for a given theoretical model is defined by @xmath205 where @xmath206 is the gaussian curvature of the likelihood matrix at the peak and @xmath207 is the experimental signal in the bin .
we consider @xmath208 , @xmath209 , @xmath210 , @xmath211 , @xmath211 , @xmath211 and @xmath211 gaussian distributed calibration errors for the archeops , boomerang-98 , dasi , maxima-1 , vsae , acbar , and cbi , experiments respectively and include the beam uncertainties using the analytical marginalization method presented in @xcite .
we use as combined likelihood just the normalized product of the two likelihood distributions : @xmath212 . in order to constrain a set of parameters
@xmath213 we marginalize over the values of the remaining `` nuisance '' parameters @xmath214 .
this yields the marginalized likelihood distribution @xmath215 in the next section we will present constraints in several two - dimensional planes . to construct plots in two dimensions ,
we project ( not marginalize ) over the third , `` nuisance '' parameter .
for example , likelihoods in the @xmath164 plane are calculated for given choice of @xmath6 and @xmath5 by using the value of @xmath216 which maximizes the likelihood function .
the error contours are then plotted relative to likelihood falloffs of @xmath217 , @xmath218 and @xmath219 as appropriate for @xmath220 , @xmath221 , and @xmath3 contours of a _ three - dimensional _ gaussian . in effect
we are taking the shadow of the three - dimensional error contours rather than a slice , which makes clear the relationship between the error contours and the points generated by monte carlo .
all likelihoods used to constrain models are calculated relative the the full three - dimensional likelihood function .
we will plot the likelihood contours obtained from our analysis on three different planes : @xmath216 vs. @xmath5 , @xmath6 vs. @xmath5 and @xmath6 vs.@xmath216 . presenting our results on these planes
is useful for understanding the effects of theoretical assumptions and/or external priors .
we do this in fig .
[ fig_like ] for two cases : the wmap dataset alone ( left column ) , and wmap plus the additional cmb experiments boomerang-98 , maxima-1 , dasi , cbi , acbar , vsae , and archeops ( right column ) .
by analyzing these different datasets we can check the consistency of the previous experiments with wmap .
the dots superimposed on the likelihood contours show the models sampled by the monte carlo reconstruction . in the top row of fig .
[ fig_like ] , we show the 68% , 95% , and 99% likelihood contours on the @xmath5 vs. @xmath216 plane ( refer to the end of sec . [ seccmbanalysis ] for a discussion of the method used to plot the contours . )
the pivot scale @xmath222 is @xmath223 mpc@xmath11 .
as we can see , both datasets are consistent with a scale invariant @xmath0 power law spectrum with no further scale dependence ( @xmath224 ) .
a degeneracy is also evident : an increase in the spectral index @xmath5 is equivalent to a negative scale dependence ( @xmath225 ) .
we emphasize , however , that this beahavior depends strictly on the position of the pivot scale @xmath222 : choosing @xmath226 mpc@xmath11 would change the direction of the degeneracy .
models with @xmath227 need a negative running at about the @xmath3 level .
it is interesting also to note that models with lower spectral index , @xmath228 , are in better agreement with the data with a zero or positive running . for @xmath229 ,
the running is bounded by @xmath230 at @xmath220 . in the center row of fig .
[ fig_like ] we plot the 68% , 95% , and 99% likelihood contours on the @xmath6 vs. @xmath5 plane .
as we can see , the present data only weakly constrain the presence of tensor modes , although a gravity wave component is not preferred . models with @xmath231 must have a negligible tensor component , while models with @xmath232 can have @xmath6 larger than @xmath233 ( @xmath221 c.l . ) . however , as we can see from the bottom row of fig .
[ fig_like ] , there is no correlation between the tensor component and the running of the scalar index .
figures [ fig_potentials ] , [ fig_xsigma ] , and [ fig_xsigma_rescaled ] show a subset of 300 reconstructed potentials selected from the sampled set .
note in particular the wide range of inflationary energy scales compatible with the observational constraint .
this is to be expected , since there was no detection of tensor modes in the wmap data , which would be seen here as a detection of a nonzero tensor / scalar ratio @xmath6 .
in addition , the _ shape _ of the inflationary potential is also not well constrained by wmap .
figure [ fig_like_zoo ] shows the models sampled by the monte carlo categorized by their `` zoology '' , _
i.e. , _ whether they fit into the category of small - field , large - field , or hybrid .
we see that all three types of potential are compatible with the data , although hybrid class models are preferred by the best - fit region .
figures [ fig_v_zoo ] , [ fig_v_zoo_type ] , and [ fig_v_zoo_type_rescaled ] show the reconstructed potentials divided by type .
perhaps the only conclusion to be drawn here is that the wmap data places no significant constraint on the shape of the inflationary potential ; many `` reasonable '' potentials are consistent with the data .
however , significant portions of the observable parameter space are ruled out by wmap , and future observations can be expected to significantly tighten these constraints @xcite . for the particular example of @xmath161 , using eq .
[ eqphi4obs ] we find that this choice of potential ruled out to @xmath3 only for @xmath234 for the wmap data set .
this constraint is even weaker than that claimed by barger _
et al . _
@xcite , most likely because we allow for a running of the spectral index in our constraint .
when augmenting the wmap data set with the data from seven other cmb experiments , the most noticeable improvement in the constraints is a better upper limit to the tensor / scalar ratio @xmath6 , which results in a slightly improved upper limit on the height of the potential .
also , the width of the reconstructed potentials in planck units is somewhat less than in the case of the wmap - only constraint , showing that the additional data more strongly limit the form of the inflationary potential .
the combined data rules out a potential with @xmath161 for @xmath235 to @xmath3 , effectively killing such models as observationally viable candidates for the inflaton potential .
in this paper , we presented an analysis of the wmap data set with an emphasis on parameters relevant for distinguishing among the various possible models for inflation .
in contrast to previous analyses , we confined ourselves to cmb data only and performed a likelihood analysis over a grid of models including the full error covariance data from the wmap satellite alone , and in conjunction with measurements from boomerang-98 , maxima-1 , dasi , cbi , acbar , vsae , and archeops .
we found that the wmap data alone are consistent with a scale - invariant power spectrum , @xmath53 , with no running of the spectral index , @xmath236 .
however , a great number of models indicates a compatibility of the data with a blue spectral index and a substantial negative running .
this is consistent with the result published by the wmap team including data from large - scale structure measurements and the lyman-@xmath8 forest .
the wmap result is also consistent with previous cmb experiments .
the inclusion of previous datasets in the analysis has the effect of reducing the error bars and give a better determination of the inflationary parameters .
still , no clear evidence for the running is present in the combined analysis .
this result differs from the result obtained in peiris _
et al . _ in the case of combined ( wmap+cbi+acbar ) analysis , where a mild ( about @xmath237 ) evidence for running was reported .
the different and more conservative method of analysis adopted here , and the larger cmb dataset used in our paper can explain this difference .
in addition , we applied the monte carlo reconstruction technique to generate an ensemble of inflationary potentials consistent with observation .
of the three basic classes of inflation model , small - field , large - field , and hybrid , none are conclusively ruled out , although hybrid models are favored by the best fit region .
the reconstruction process indicates that a wide variety of smooth potentials for the inflaton are consistent with the data , indicating that the wmap result is too crude to significantly constrain either the height or the shape of the inflaton potential .
in particular , the lack of evidence for tensor fluctuations makes it impossible to constrain the energy scale at which inflation takes place .
nonetheless , wmap rules out a large portion of the available parameter space for inflation , itself a significant improvement over previous measurements . for the particular case of a potential of the form @xmath238 , wmap rules out
all such potentials for @xmath234 at the @xmath3 level , which means that @xmath239 potentials are not conclusively ruled out by wmap alone .
the combined data set , however , rules out @xmath239 models for @xmath235 , which kills such potentials as viable candidates for inflation .
after this paper first appeared , leach and liddle also released a reanalysis of the wmap data @xcite , which is in general agreement with our results here .
w.h.k . is supported by iscap and the columbia university academic quality fund .
iscap gratefully acknowledges the generous support of the ohrstrom foundation . e.w.k . is supported in part by nasa grant nag5 - 10842 .
we thank richard easther for helpful conversations and for the use of computer code .
a. melchiorri and j. silk , astro - ph/0203200 ; s. h. hansen , g. mangano , a. melchiorri , g. miele and o. pisanti , phys .
d * 65 * , 023511 ( 2002 ) , astro - ph/0105385 ; a. melchiorri , l. mersini , c. j. odman and m. trodden , astro - ph/0211522 .
w. h. kinney , a. melchiorri , a. riotto , phys .
d * 63 * , 023505 ( 2001 ) , astro - ph/0007375 ; s. hannestad , s. h. hansen , and f. l. villante , astropart . phys . * 17 * 375 ( 2002 ) ; s. hannestad , s. h. hansen and f. l. villante , astropart .
* 16 * , 137 ( 2001 ) ; d. j. schwarz , c. a. terrero - escalante , and a. a. garcia , phys .
b * 517 * , 243 ( 2001 ) ; x. wang , m. tegmark , b. jain and m. zaldarriaga , astro - ph/0212417 .
p.m. garnavich et al , ap.j .
letters * 493 * , l53 - 57 ( 1998 ) ; s. perlmutter et al , ap .
j. * 483 * , 565 ( 1997 ) ; s. perlmutter et al ( the supernova cosmology project ) , nature * 391 * 51 ( 1998 ) ; a.g .
riess et al , ap .
j. * 116 * , 1009 ( 1998 ) ; models generated by monte carlo plotted on the @xmath164 plane ( black dots ) .
the colored lines are the same models as in fig .
2 . for comparison with the models ,
points are plotted to first order in slow roll.,width=384 ] likelihood contours for wmap alone ( left column ) and wmap plus seven other experiments ( right column ) , plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling of inflationary models consistent with @xmath220 ( red , dots ) , @xmath221 ( blue , triangles ) , and @xmath3 ( black , stars ) contours .
all points in this and subsequent figures are plotted to second order in slow roll.,title="fig:",width=259 ] likelihood contours for wmap alone ( left column ) and wmap plus seven other experiments ( right column ) , plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling of inflationary models consistent with @xmath220 ( red , dots ) , @xmath221 ( blue , triangles ) , and @xmath3 ( black , stars ) contours .
all points in this and subsequent figures are plotted to second order in slow roll.,title="fig:",width=259 ] likelihood contours for wmap alone ( left column ) and wmap plus seven other experiments ( right column ) , plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling of inflationary models consistent with @xmath220 ( red , dots ) , @xmath221 ( blue , triangles ) , and @xmath3 ( black , stars ) contours .
all points in this and subsequent figures are plotted to second order in slow roll.,title="fig:",width=259 ] likelihood contours for wmap alone ( left column ) and wmap plus seven other experiments ( right column ) , plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling of inflationary models consistent with @xmath220 ( red , dots ) , @xmath221 ( blue , triangles ) , and @xmath3 ( black , stars ) contours .
all points in this and subsequent figures are plotted to second order in slow roll.,title="fig:",width=259 ] likelihood contours for wmap alone ( left column ) and wmap plus seven other experiments ( right column ) , plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling of inflationary models consistent with @xmath220 ( red , dots ) , @xmath221 ( blue , triangles ) , and @xmath3 ( black , stars ) contours . all points in this and
subsequent figures are plotted to second order in slow roll.,title="fig:",width=259 ] likelihood contours for wmap alone ( left column ) and wmap plus seven other experiments ( right column ) , plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling of inflationary models consistent with @xmath220 ( red , dots ) , @xmath221 ( blue , triangles ) , and @xmath3 ( black , stars ) contours .
all points in this and subsequent figures are plotted to second order in slow roll.,title="fig:",width=259 ] three hundred reconstructed potentials chosen from the sampling shown in fig .
[ fig_like ] for the wmap dataset ( left column ) and the wmap data set plus seven other experiments ( right column ) .
the potentials are color - coded according to their likelihoods : @xmath220 ( red ) @xmath221 ( blue ) , and @xmath3 ( black ) .
the top figure shows the potentials with height and width plotted in units of @xmath244 , and the bottom figure shows the same potentials rescaled to all have the same height and width.,title="fig:",width=288 ] three hundred reconstructed potentials chosen from the sampling shown in fig .
[ fig_like ] for the wmap dataset ( left column ) and the wmap data set plus seven other experiments ( right column ) .
the potentials are color - coded according to their likelihoods : @xmath220 ( red ) @xmath221 ( blue ) , and @xmath3 ( black ) .
the top figure shows the potentials with height and width plotted in units of @xmath244 , and the bottom figure shows the same potentials rescaled to all have the same height and width.,title="fig:",width=288 ] three hundred reconstructed potentials chosen from the sampling shown in fig .
[ fig_like ] for the wmap dataset ( left column ) and the wmap data set plus seven other experiments ( right column ) .
the potentials are color - coded according to their likelihoods : @xmath220 ( red ) @xmath221 ( blue ) , and @xmath3 ( black ) .
the top figure shows the potentials with height and width plotted in units of @xmath244 , and the bottom figure shows the same potentials rescaled to all have the same height and width.,title="fig:",width=288 ] three hundred reconstructed potentials chosen from the sampling shown in fig .
[ fig_like ] for the wmap dataset ( left column ) and the wmap data set plus seven other experiments ( right column ) . the potentials are color - coded according to their likelihoods : @xmath220 ( red ) @xmath221 ( blue ) , and @xmath3 ( black ) .
the top figure shows the potentials with height and width plotted in units of @xmath244 , and the bottom figure shows the same potentials rescaled to all have the same height and width.,title="fig:",width=288 ] reconstructed potentials from fig .
[ fig_potentials ] plotted separately by likelihood : @xmath220 ( red , top ) , @xmath221 ( blue , center ) , and @xmath3 ( black , bottom ) .
the wmap results are in the left column , and wmap plus seven other experiments are in the right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_potentials ] plotted separately by likelihood : @xmath220 ( red , top ) , @xmath221 ( blue , center ) , and @xmath3 ( black , bottom ) .
the wmap results are in the left column , and wmap plus seven other experiments are in the right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_potentials ] plotted separately by likelihood : @xmath220 ( red , top ) , @xmath221 ( blue , center ) , and @xmath3 ( black , bottom ) . the wmap results are in the left column , and wmap plus seven other experiments are in the right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_potentials ] plotted separately by likelihood : @xmath220 ( red , top ) , @xmath221 ( blue , center ) , and @xmath3 ( black , bottom ) .
the wmap results are in the left column , and wmap plus seven other experiments are in the right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_potentials ] plotted separately by likelihood : @xmath220 ( red , top ) , @xmath221 ( blue , center ) , and @xmath3 ( black , bottom ) . the wmap results are in the left column , and wmap plus seven other experiments are in the right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_potentials ] plotted separately by likelihood : @xmath220 ( red , top ) , @xmath221 ( blue , center ) , and @xmath3 ( black , bottom ) . the wmap results are in the left column , and wmap plus seven other experiments are in the right column.,title="fig:",width=259 ] same as fig .
[ fig_xsigma ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_xsigma ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_xsigma ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_xsigma ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_xsigma ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_xsigma ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] likelihood contours plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling , color coded by model type : small - field ( red , dots ) , large - field ( green , triangles ) , and hybrid ( blue , crosses ) .
the left column is wmap only while the right column is wmap plus seven other experiments , title="fig:",width=259 ] likelihood contours plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling , color coded by model type : small - field ( red , dots ) , large - field ( green , triangles ) , and hybrid ( blue , crosses ) .
the left column is wmap only while the right column is wmap plus seven other experiments , title="fig:",width=259 ] likelihood contours plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling , color coded by model type : small - field ( red , dots ) , large - field ( green , triangles ) , and hybrid ( blue , crosses ) .
the left column is wmap only while the right column is wmap plus seven other experiments , title="fig:",width=259 ] likelihood contours plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling , color coded by model type : small - field ( red , dots ) , large - field ( green , triangles ) , and hybrid ( blue , crosses ) .
the left column is wmap only while the right column is wmap plus seven other experiments , title="fig:",width=259 ] likelihood contours plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling , color coded by model type : small - field ( red , dots ) , large - field ( green , triangles ) , and hybrid ( blue , crosses ) .
the left column is wmap only while the right column is wmap plus seven other experiments , title="fig:",width=259 ] likelihood contours plotted in the ( @xmath241 ) plane ( top ) , the ( @xmath242 ) plane ( center ) , and the ( @xmath243 ) plane ( bottom ) .
the points represent the results of the monte carlo sampling , color coded by model type : small - field ( red , dots ) , large - field ( green , triangles ) , and hybrid ( blue , crosses ) . the left column is wmap only while the right column is wmap plus seven other experiments , title="fig:",width=259 ] three hundred reconstructed potentials chosen from the sampling shown in fig.[fig_like ] for wmap ( left column ) and wmap plus seven other experiments ( right column ) .
the potentials are color - coded according to model type : small - field ( red ) , large - field ( green ) , and hybrid ( blue ) .
the top figure shows the potentials with height and width plotted in units of @xmath244 , and the bottom figure shows the same potentials rescaled to all have the same height and width.,title="fig:",width=288 ] three hundred reconstructed potentials chosen from the sampling shown in fig.[fig_like ] for wmap ( left column ) and wmap plus seven other experiments ( right column ) .
the potentials are color - coded according to model type : small - field ( red ) , large - field ( green ) , and hybrid ( blue ) .
the top figure shows the potentials with height and width plotted in units of @xmath244 , and the bottom figure shows the same potentials rescaled to all have the same height and width.,title="fig:",width=288 ] three hundred reconstructed potentials chosen from the sampling shown in fig.[fig_like ] for wmap ( left column ) and wmap plus seven other experiments ( right column ) .
the potentials are color - coded according to model type : small - field ( red ) , large - field ( green ) , and hybrid ( blue ) .
the top figure shows the potentials with height and width plotted in units of @xmath244 , and the bottom figure shows the same potentials rescaled to all have the same height and width.,title="fig:",width=288 ] three hundred reconstructed potentials chosen from the sampling shown in fig.[fig_like ] for wmap ( left column ) and wmap plus seven other experiments ( right column ) .
the potentials are color - coded according to model type : small - field ( red ) , large - field ( green ) , and hybrid ( blue ) .
the top figure shows the potentials with height and width plotted in units of @xmath244 , and the bottom figure shows the same potentials rescaled to all have the same height and width.,title="fig:",width=288 ] reconstructed potentials from fig .
[ fig_v_zoo ] plotted separately by model type : small - field ( red , top ) , large - field ( green , center ) , and hybrid ( blue , bottom ) .
the wmap results are shown in the left column while wmap plus seven other experiments are shown onthe right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_v_zoo ] plotted separately by model type : small - field ( red , top ) , large - field ( green , center ) , and hybrid ( blue , bottom ) .
the wmap results are shown in the left column while wmap plus seven other experiments are shown onthe right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_v_zoo ] plotted separately by model type : small - field ( red , top ) , large - field ( green , center ) , and hybrid ( blue , bottom ) .
the wmap results are shown in the left column while wmap plus seven other experiments are shown onthe right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_v_zoo ] plotted separately by model type : small - field ( red , top ) , large - field ( green , center ) , and hybrid ( blue , bottom ) .
the wmap results are shown in the left column while wmap plus seven other experiments are shown onthe right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_v_zoo ] plotted separately by model type : small - field ( red , top ) , large - field ( green , center ) , and hybrid ( blue , bottom ) .
the wmap results are shown in the left column while wmap plus seven other experiments are shown onthe right column.,title="fig:",width=259 ] reconstructed potentials from fig .
[ fig_v_zoo ] plotted separately by model type : small - field ( red , top ) , large - field ( green , center ) , and hybrid ( blue , bottom ) .
the wmap results are shown in the left column while wmap plus seven other experiments are shown onthe right column.,title="fig:",width=259 ] same as fig .
[ fig_v_zoo_type ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_v_zoo_type ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_v_zoo_type ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_v_zoo_type ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_v_zoo_type ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] same as fig .
[ fig_v_zoo_type ] , but with the potentials rescaled to all have the same height and width.,title="fig:",width=259 ] | we extract parameters relevant for distinguishing among single - field inflation models from the wilkinson microwave anisotropy probe ( wmap ) data set , and from a combination of the wmap data and seven other cosmic microwave background ( cmb ) experiments .
we use only cmb data and perform a likelihood analysis over a grid of models including the full error covariance matrix . we find that a model with a scale - invariant scalar power spectrum ( @xmath0 ) , no tensor contribution , and no running of the spectral index , is within the 1-@xmath1 contours of both data sets .
we then apply the monte carlo reconstruction technique to both data sets to generate an ensemble of inflationary potentials consistent with observations .
none of the three basic classes of inflation models ( small - field , large - field , and hybrid ) are completely ruled out , although hybrid models are favored by the best - fit region .
the reconstruction process indicates that a wide variety of smooth potentials for the inflaton are consistent with the data , implying that the first - year wmap result is still too crude to constrain significantly either the height or the shape of the inflaton potential . in particular
, the lack of evidence for tensor fluctuations makes it impossible to constrain the energy scale of inflation . nonetheless , the data rule out a large portion of the available parameter space for inflation .
for instance , we find that potentials of the form @xmath2 are ruled out to @xmath3 by the combined data set , but not by the wmap data taken alone . |
ultraluminous and luminous infrared galaxies ( ulirgs / lirgs ) have strong infrared ( ir ) luminosities of @xmath5 and @xmath6 , respectively @xcite .
their intense ir luminosities are mainly due to starburst activities , and the contribution of active galactic nuclei ( agn ) to ir luminosity seems to be rather limited , i.e. , @xmath7 @xcite . the starburst activities in ( u)lirgs are likely to be triggered by mergings of galaxies , in particular gas - rich galaxies , since observations suggest that ( u)lirgs have complex and disturbed morphologies @xcite . in the @xmath8-band ( f814w filter )
images of 100 sampled ( u)lirgs at @xmath9 taken with _ hubble space telescope _
( _ hst _ ) , about @xmath10 of ulirgs have multiple ( @xmath0 ) nuclei @xcite .
the fraction increases to more than @xmath11 , if all probable galaxies are included @xcite .
the origin of the multiple nuclei has been thought to be multiple major merger events . in order to explain multiple mergers
, @xcite argued that the progenitors of ( u)lirgs , were compact groups of galaxies .
however , the evolution timescale of compact groups is very long ( @xmath12 ) @xcite .
therefore , the probability that two galaxies merge in a compact group and the merging galaxy has double cores is @xmath13 @xmath14 @xmath15 , where @xmath16 is the merging timescale of the galactic cores .
the probability that another galaxy merges with such merging galaxy in the compact group and the merging galaxy has triple cores is @xmath17 , where @xmath18 ( @xmath19 ) is the typical number of galaxies in a compact group @xcite .
then , the number density of ulirgs with multiple cores is @xmath20 @xmath21 , where @xmath22 ( @xmath23 ) is the number density of galaixes in a compact group @xcite .
since the number density of ulirgs is @xmath24 @xcite , the fraction of multiple merger to ulirgs is @xmath25 .
thus , it is unliky that multiple merger is a major cause of the observed multiple nuclei in ( u)lirgs .
an alternative explanation for the origin of multiple nuclei in ( u)lirgs seems necessary .
previous numerical simulations of merging galaxies @xcite have shown that starbursts occurred during the merging process . however , these simulations failed to reproduce formation of star clusters as observed in many interacting and merging galaxies @xcite , since their spatial and mass resolutions ( _ e.g. _ , @xmath26 and @xmath27 ) were not sufficiently high to distinguish star forming region and formation of star clusters whose sizes are less than @xmath26 .
in addition , previous simulations did not allow the interstellar medium ( ism ) to cool below @xmath28 .
this is another reason why formation of star clusters was not reproduced in the previous simulations . in order to reproduce formation of star clusters ,
high resolution simulations have been attempted .
@xcite have performed the simulation of a merging galaxy using tree+sph code gadget . in this simulation , mass and spatial resolutions were @xmath29 and @xmath30 , respectively .
they assumed an isothermal ism and used sink particles , that absorb their surrounding gas , to represent clusters .
they have shown that a number of massive clusters form in the merging process . in this simulation ,
the most massive cluster has mass of @xmath31 . @xcite
have performed the simulations of merging galaxies using sticky particles , that collide inelastically , instead of solving hydrodynamics . in these simulations ,
mass and spatial resolutions were @xmath32 and @xmath33 , respectively .
they have also shown formations of massive star clusters with masses of @xmath34 .
these simulations are , however , unrealistic in a sense that dynamical evolution of star clusters themselves can not be properly followed due to the limitation of the sticky and sink particle methods .
@xcite improved the spatial and mass resolutions ( @xmath35-@xmath36 and @xmath37 ) and the ism model that is allowed to cool to @xmath38 .
these simulations showed that the behavior of the multiphase ism in the merging galaxies is considerably altered and the formation of shock - induced star clusters is naturally reproduced @xcite .
there are several high resolution simulations of merging galaxies , which resolve the low temperature gas ( @xmath39-@xmath40 ) using adaptive mesh refinement ( amr ) methods @xcite .
these simulations also showed the difference in the behavior of the multiphase ism from that of ism used in the previous simulations . @xcite considered the merger of low - mass galaxies ( @xmath41 ) .
several spiky peaks of star formation rate were seen in thier simulation although the most prominet starburst was found at the beginning the simualtions and the significant fraction of the ism was blown away by energetic wind before the merging event .
thus , it would be hard to investigate detailed evolution of the ism in merging galaxies .
in @xcite , the polytropic equation of state was used instead of solving the energy equation , which is essentially different from @xcite and present paper .
the validity of this approximation is unclear for understanding the evolution of the ism in the merging galaxies .
high - resolution images of the local ( u)lirgs obtained by the integral field spectroscopy using _ william herschel telescope _ @xcite and _ vlt - vimos _ @xcite showed that ( u)lirgs generally have very complex structures , such as h@xmath42 bright knots , rings , and tidal tails .
as in the case of formation of star clusters , the reason why these structures have not been reproduced in numerical simulations might be just the inadequate treatment of ism and limited resolution .
thus , high resolution simulations resolving multiphase ism are essential to comprehend the complex structures in ( u)lirgs .
we have performed high - resolution simulations of merging galaxies with sufficiently high spatial resolution and cooling function of ism that covers a wide range of temperature ( @xmath43 ) . in this paper , we focus on the origin of the observed ulirgs with multiple nuclei . the detailed evolution of merging galaxies will be described in forthcoming papers .
the structure of this paper is as follows .
we first describe the numerical methods and models in [ sec : methods ] .
numerical results are shown in [ sec : results ] . in [
sec : discussion ] , we compare our results with observations .
conclusions and discussion are presented in
[ sec : conclusions ] .
the model parameters of the initial disk galaxy are the same as those used in @xcite .
the masses of the dark matter halo and old stellar and gas disks are @xmath44 , @xmath45 , and @xmath46 , respectively .
the gas fraction in the disk is @xmath47% of the total disk mass .
the gas fraction is set to be slightly higher than that of local spiral galaxies since the gas fraction decreases due to star formations during the isolated phase for @xmath48 before interactions ( see below ) .
we also assume the halo gas component of which total mass is @xmath49 .
it has the same distribution as the dark matter halo .
the initial gas temperatures are @xmath50 in the disk and @xmath51 in the halo .
the scale radii of the stellar and gas disks are @xmath52 and @xmath53 , respectively .
the dark - matter halo and the stellar disk are expressed by @xmath54-body particles , whereas the gas disk is expressed by smoothed particle hydrodynamics ( sph ) particles . unlike @xcite , before starting the simulations , we let an isolated disk galaxy evolve for @xmath48 in order to stabilize the gas component .
after this evolution , the amount of disk and halo gas in each galaxy is @xmath55 .
the sfr during this period is @xmath56 . during the first @xmath48 ,
the disk galaxy holds a quasi - steady state and is free from a global instability due to gravity .
the evolution of the isolated disk is different from that of a gas rich disk at high-@xmath57 .
simulations of the gas - rich disk , in which gas mass fraction of a disk is @xmath58 % , have shown that massive clumps form from gravitational instabillity of the gas rich disk @xcite .
in contrast to the simulations of merging of such clumpy disks at high - z by @xcite , we focus on the merging process of two local disk galaxies
. we let the pre - evolved two galaxies collide in a parabolic orbit with the pericentric distance of @xmath59 .
the simulations start with the initial separation of @xmath60 , where the separation distance is measured between the mass centers of the two galaxies .
we have performed nine runs .
three of them are high - resolution runs ( @xmath61 in tab .
[ param_tab ] ) and the rests are low - resolultion runs ( @xmath62 ) .
the initial masses of sph , star , and dark matter particles are the same .
they are @xmath63 and @xmath64 , for @xmath61 and @xmath62 runs , respectively .
subscripts p , t , and r denote prograde , tilted , and retrograde spin axes of two galaxies .
we performed six possible combinations of the spin orientations in low resolution and only pp and tt cases in high resolution .
the gravitational softening length , @xmath65 , is set to be @xmath36 for eight runs except for @xmath66 run in which the softening length is @xmath67 .
the orientation of the disk axis is specified by two angles , @xmath68 and @xmath69 . here , @xmath68 is the inclination and @xmath69 is the argument of pericenter .
they are defined as follows .
we use the coordinate system in which the orbital plane of two galaxies is in the @xmath70-@xmath71 plane .
the laplace - runge - lenz vector of the orbit is @xmath72 .
the spin axis of a galaxy is then given by @xmath73 .
these angles are defined relative to the orbit of each galaxy .
thus , in the case of `` tt '' runs , spin axes of two galaxies are parallel to each other .
lccccccccc run & @xmath74}$ ] & @xmath75}$ ] & @xmath76}$ ] & @xmath77}$ ] & @xmath78 $ ] & @xmath79 & @xmath80 & @xmath81 & @xmath82 $ ] + @xmath83 & @xmath84 & - & @xmath84 & - & @xmath85 & @xmath86 & @xmath87 & @xmath88 & 20 + @xmath89 & @xmath90 & @xmath91 & @xmath92 & @xmath91 & @xmath85 & @xmath86 & @xmath87 & @xmath88 & 20 + @xmath93 & @xmath84 & - & @xmath84 & - & @xmath94 & @xmath95 & @xmath96 & @xmath97 & 20 + @xmath98 & @xmath84 & - & @xmath99 & - & @xmath94 & @xmath95 & @xmath96 & @xmath97 & 20 + @xmath100 & @xmath99 & - & @xmath99 & - & @xmath94 & @xmath95 & @xmath96 & @xmath97 & 20 + @xmath101 & @xmath84 & - & @xmath92 & @xmath102 & @xmath94 & @xmath95 & @xmath96 & @xmath97 & 20 + @xmath103 & @xmath90 & @xmath102 & @xmath99 & - & @xmath94 & @xmath95 & @xmath96 & @xmath97 & 20 + @xmath104 & @xmath90 & @xmath91 & @xmath92 & @xmath91 & @xmath94 & @xmath95 & @xmath96 & @xmath97 & 20 + @xmath66 & @xmath90 & @xmath91 & @xmath92 & @xmath91 & @xmath85 & @xmath86 & @xmath87 & @xmath88 & 5 + @xmath105 the disk inclination for the galaxy @xmath106 . @xmath107
the argument of the pericenter for the galaxy @xmath106 .
@xmath108 the disk inclination for the galaxy @xmath109 .
@xmath110 the argument of the pericenter for the galaxy @xmath109
. @xmath111 the mass of sph , star , and dark matter particles . @xmath112 the number of sph particles .
@xmath113 the number of star particles .
@xmath114 the number of dark matter particles . @xmath115
the gravitational softening length .
numerical simulations were performed by an @xmath54-body / sph code asura @xcite that can optionally utilize grape @xcite in order to accelerate calculation of gravitational force . we adopted a time - step limiter @xcite for the time - integration of sph particles with individual time - steps , which keeps the differences in time - steps of neighboring particles to be small enough to handle the strong shocks correctly .
we also used the fast scheme @xcite that allows self - gravitating sph particles to use different time - steps for integrations of hydrodynamics and gravity .
this method accelerates simulations without losing the accuracy of the time - integration .
the leapfrog method is used for the time integrations of both gravity and hydrodynamics .
these improvements allow us to follow the formation of cold and dense gas clumps and their expansion by supernovae ( sne ) feedback without numerical problems . for the radiative cooling and heating
, we follow the treatment of @xcite , in which the gas is allowed to cool down to @xmath38 .
the treatment of star formation and sn feedback are also the same as those in @xcite .
when an sph particle satisfied ( 1 ) high number density ( @xmath116 ) , ( 2 ) a low temperature ( @xmath117 ) , and ( 3 ) a collapsing region ( @xmath118 ) , the sph particle spawns a star particle of which mass is one - third of the original sph particle mass .
we assume simple stellar polulation approximation for newly formed star particles .
salpeter s initial mass function with mass range of @xmath119 is adopted @xcite .
stars with mass @xmath120 in each star particle explode as type ii sne and release @xmath121 of thermal energy per one sn into the surrounding ism .
initially , the two galaxies move on the parabolic orbits and approach to each other . at around @xmath122 , they reach the pericenter ( first encounter ) . since orbital angular momenta of their main bodies are converted into the disk internal spin , they lose their orbital angular momenta @xcite and their trajectories deviate from the parabolic orbits .
the main bodies approach to each other again ( second encounter ) .
finally , after several encounters , their cores completely merge .
figures [ snapshot_hpp ] and [ snapshot_htt ] show the evolution of distribution of gas and stellar particles for runs @xmath83 and @xmath89 .
gas disks are strongly disturbed at the first encounter ( the panels at @xmath123 , @xmath124 , and @xmath125 ) and at the second encounter ( the panels at @xmath126 and @xmath127 ) in run @xmath83 and at the second encounter ( the panels at @xmath128 and @xmath129 ) in run @xmath89 . because of the strong perturbation
, gas is compressed by large - scale shocks .
the dense gas radiatively cools and becomes gravitationally unstable . as a result ,
massive star clusters form ( see also * ? ? ?
* ; * ? ? ?
figure [ image ] shows the synthesized @xmath8-band images from our merger simulations after the massive star clusters formed . here
, the @xmath8-band luminosity emitted from newly formed stars was calculated using population synthesis code pgase @xcite , and effects of dust absorption and re - emission were neglected .
the two top panels show the high resolution runs , and the six lower panels show the low resolution runs . in all images ,
multiple compact sources appear in the central regions of galaxies .
two of them with arrows are the progenitor galactic cores , and the others are newly - formed massive star clusters .
the absolute magnitudes of these clusters in @xmath8-band are @xmath130 , which are comparable to those of the galactic central cores .
the effect of dust extinction is discussed in 4.2 . since the cores form through the gravitational instability of strongly perturbed gas disks
, they do not contain dark matter and consist only of young stars .
on the other hand , the galactic luminous nuclei contain dark matter and a large amount of old stars that have formed before the merger event .
the properties of newly formed cores are , therefore , different from those of the galactic nuclei .
in addition , such cores are more massive than usual star clusters with @xmath131 , by an order of magnitude .
we , therefore , call the newly formed cores `` hypermassive star clusters ''
. since the hypermassive star clusters form in the galactic central region ( @xmath3 ) , they are different from tidal dwarves that are formed in tidal arms ( e.g. , * ? ? ?
* ; * ? ? ?
the reason why they are extremely massive can be explained as follows .
the total mass of the gas which becomes gravitationally unstable is very large ( @xmath132 ) because of the strong large - scale gravitational and hydrodynamical disturbances from another galaxy at the encounter phase .
the strong disturbances deform the gravitational potential into strongly non - axisymmentric shape .
therefore , the rotation of the disk becomes ineffective in stabilizing long - wavelength perturbations .
even though the gas disk fragments to a number of small clumps with @xmath133 , these small clumps are still gravitationally bound to each other and eventually merge to form larger clusters .
figure [ snapshot ] shows the snapshots during the formation of the most massive cluster from @xmath134 to @xmath135 in run @xmath89 .
multiple gas clumps formed from the disturbed gas disk merge with each other and become one massive cluster .
in other words , the merging of smaller clumps within the short timescale causes quick growth of the mass of star clusters . as a result , hypermassive star clusters form . in fig .
[ mass_evolution ] , we show the evolution of the masses of these hypermassive star clusters ( upper panels ) , and their distances from the center of the galaxy ( lower panels ) . here , the definition of the distance is the smaller of the two distances from the two galactic centers . the clusters first grow quickly , reaching about a half of the final mass in around @xmath136 . after that
, the growth of the mass slows down , and the clusters fall to the center of the galaxy due to the dynamical friction . as a result
, the merger remnant has a very compact and luminous core similar to those observed in a large number of elliptical galaxies ( e.g. , * ? ? ?
figure [ sfr ] shows the evolution of the sfr and bolometric luminosity coming from the stars in the merging event .
the effects of dust absorption and re - emission are not taken into account .
if we take them into account , the ir luminosity would become higher since ultraviolet and optical photons are absorbed by dust and re - emitted in the ir regime .
therefore , we expect that the bolometric luminosity of our runs is a good indicator of the ir luminosity . in all runs , the maximum sfr reaches @xmath137 , and the luminosity reaches @xmath138 , which is about an order of magnitude higher than that before the encounter .
the figure shows that the evolution of bolometric luminosity reflects the sfr .
the peak luminosity is comparable to those of lirgs @xcite .
if the progenitor galaxies are more massive and/or the initial gas fraction is much higher than those used in this simulation , the peak of sfr would become higher and the merger galaxy would become ulirgs with infrared emission @xmath139
. agns would also provide an additional ir luminosity , although our simulations do not take into account the effect of agns .
the influence of agns to the total ir luminosity is typically @xmath47 % of the total one @xcite .
the peak mass - to - light ratio of our run , @xmath140 , is less than @xmath141 and is comparable to that of ulirgs @xcite .
the period of the high luminosity phase is @xmath142 .
the hypermassive star clusters after their formation contain a large amount of dusty gas and young stars as shown in fig .
[ snapshot ] .
therefore , a strong infrared emission from the hypermassive star clusters is likely to be observable . in order to study the dependence on the numerical resolution of the formation of hypermassive star clusters
, we have performed simulations with different mass resolutions . in both high and low resolution runs ,
hypermassive star clusters with @xmath143 are formed as shown in fig .
[ image ] .
their formation process is the same as that of the high resolution runs .
in addition , the time evolution of sfr is also similar to that of high resolution runs . in both runs ,
the peak of sfr is @xmath144 , and the duration time of the active star formation is about @xmath145 .
here , we compare our numerical results with observations of ulirgs with nuclei .
@xcite observed nuclei with absolute magnitude from @xmath146 to @xmath147 in @xmath8-band .
the @xmath8-band luminosity of the identified putative nuclei is typically around a few percent of far - infrared luminosity of their host galaxy @xcite .
we estimated the absolute @xmath8-band magnitude of hypermassive star clusters formed in our simulations using pgase .
some hypermassive star clusters have luminosity comparable to the observed putative nuclei . in the upper panels of fig
. [ image ] , there is one hypermassive star cluster with @xmath148 in run @xmath83 , and there are three hypermassive star clusters with @xmath149 , @xmath150 , and @xmath150 in run @xmath89 , respectively .
these @xmath8-band absolute magnitudes are comparable to those of the galactic cores , which are @xmath151 and @xmath152 in run @xmath83 and @xmath153 and @xmath154 in run @xmath89 .
these results indicate that the galactic nuclei and hypermassive star clusters are not distinguishable from their luminosities .
the @xmath8-band ( bolometric ) luminosity of hypermassive star clusters is a few percent ( @xmath155% ) of the bolometric luminosity of the host galaxy with @xmath156 .
these characteristics are in good agreement with the observations .
we also compare the spatial distribution of the observed putative nuclei with that of simulated hypermassive star clusters .
if we select relatively compact ulirgs , the average separation between nuclei is about @xmath157 @xcite . in our simulations ,
the separation between hypermassive star clusters and galactic cores depends on the evolution phase . however , the typical separation is a few @xmath158 in the most luminous phase of the galaxies .
thus , the spatial distribution also agrees with the observations . in this subsection
, we investigate the effect of the dust extinction in @xmath8-band . here
, we use the high spatial resolution run , @xmath66 , in which the softening length is @xmath67 for mock observatations , since the detailed structure of the ism is important to estimate the extinction . in this run , similarly to run @xmath89 , hypermassive star clusters with @xmath159 form after the second encounter .
the masses of the first , second , third most massive star clusters are @xmath160 , @xmath161 , and @xmath162 , respectively .
the flux , taking dust extinction into account , from a star is estimated by @xmath163 . here ,
@xmath164 is the flux expected without dust extinction , and @xmath165 is optical depth for each sph particle and @xmath166 @xcite , where @xmath167 , @xmath168 , @xmath169 , and @xmath170 represent the clumpiness parameter , the dust cross - section per h atom at @xmath8-band for the milkyway dust model , column number density of atomic h of a sph particle , and the metallicity of a sph particle , respectively . the sph particles between the star and the observer are used for calculating the optical depth , @xmath171 .
the clumpiness parameter is introduced in order to represent the clumpiness of the dust distribution in the ism .
it is assumed to be @xmath172 , @xmath173 , and @xmath174 . the value of @xmath175 is used in ly@xmath42 emitters @xcite , whereas that of @xmath176 is the case without the dust clumpiness under the spatial resolution . here
, we do not estimate the `` clumpiness '' of sph particles along the line of sight directly , since the `` real '' clumpiness comes from much smaller scale in the ism .
we adopt @xmath177 @xcite .
the column density of an sph particle is given by @xmath178 , where @xmath179 and @xmath180 are the mass and 2-dimensional kernel function of the sph particle , respectively , where @xmath181 is the 2-dimensional distance between the sph and star particles in the projected plane , and @xmath182 is the kernel size of the sph particle . figure [ extinction_map ] illustrates the effect of dust extinctions .
the upper row shows the time evolution of the @xmath8-band map without extinction .
the middle row represents the expected observation taking dust extinction with @xmath183 into account for run @xmath66 . here
, the map is projected to the orbital plane of the galaxies .
after formation of hypermassive star clusters , strong dust extinctions take place in the clusters since there is a large amount of dusty gas .
the dust extinction reduces the flux from the formed hypermassive star clusters .
the extinction of the clusters in @xmath8-band is @xmath184 magnitude in the @xmath185 case as shown in the bottom panels of fig .
[ extinction_map ] . in the models of @xmath186 and @xmath187 ,
the extinctions are @xmath188 and @xmath189 , respectively .
the strong extinction continues till @xmath190 ( the left and the middle panels ) , which corresponds to the epoch about @xmath191 after the formation of the clusters . in this dusty phase ,
multiple core structures are buried by dust .
after @xmath192 ( the right panels ) , the clusters become optically thin because of consumption of dusty gas by star formation and escape of dusty gas from the cluster by sn feedback . as a result , the multiple cores appear .
the extinction becomes @xmath193 magnitude . on the other hand , the galactic central region
is still buried by dust due to the continuous gas accretion . in this phase ,
the extinction does not depend strongly on the clumpiness parameter since dusty gas is restricted to the central region of the cluster rather than distributed in the broad region of the cluster .
the timescale in which hypermassive star clusters become optically thin in @xmath8-band is less than the ulirg lifetime ( @xmath142 as shown in fig .
[ sfr ] ) .
therefore , it is possible for the clusters to be observed as ulirgs .
note , however , that the estimate of the dust extinction depends strongly on the distribution of the dust and that resolutions of current simulations are insufficient to resolve the `` real '' structure of the ism .
futher high resolution simulations are necessary to involve the clumpiness of the dust distribution in the ism .
the fraction of ulirgs with multiple nuclei strongly depends on the wavelength used for observation . in the analysis of _ hst
_ @xmath8-band data of @xmath194 ulirgs samples by @xcite and @xcite , the fraction of the ulirgs with multiple nuclei is @xmath195 .
in addition , @xcite have analyzed the nine samples of @xcite in @xmath8-band .
two ulirgs are found to have multiple nuclei , i.e. , the fraction is consistent with the results of @xcite and @xcite . on the other hand , @xcite and @xcite
have claimed that the fraction of ulirgs with multiple nuclei is less than @xmath196 , based on their analysis of the data observed in near ir bands . using the _ hst _ @xmath197-band data , @xcite reanalyzed 27 samples randomly selected from the 123 @xmath8-band samples , which include a part of borne s and cui s samples with multiple nuclei . in their analysis
, only one sample is classified as a ulirg with multiple nuclei .
the other sample includes some ulirgs that have been already classified as ulirgs with multiple nuclei by @xmath8-band image analysis @xcite .
the reasons for excluding them from ulirgs with multiple nuclei are as follows : ( 1 ) the nucleus seems to locate on the tidal arm and is thought to be the tidal arm or ( 2 ) multiple nuclei morphology do not appear in the @xmath197-band images .
furthermore , @xcite analyzed the @xmath198-band data of 118 ulirgs taken by the keck observatory .
their samples also include candidates of ulirgs with multiple nuclei in hst @xmath8-band data .
some of them are excluded since multiple cores are not apparent in @xmath198-band in spite of their appearance in @xmath8-band . as a result , only 5 ulirgs ( @xmath199 )
are classified as ulirgs with multiple nuclei . in order to understand the difference between the results obtained using images taken in different bands
, we made @xmath8- , @xmath197- , and @xmath200-band images of our simulation data using pgase .
we assumed that galaxies were at @xmath201 ( @xmath202 ) .
the point spread function was assumed to be gaussian , and its dispersion was calculated by @xmath203 , where @xmath204 is the spatial resolution .
the spatial resolutions of @xmath8- and @xmath197-band images were given by the diffraction limit since we assumed the observations by _
hst_. the spatial resolutions of @xmath8- and @xmath197-bands are @xmath205 and @xmath206 , respectively . for @xmath200-band images ,
the seeing determines the spatial resolution since we assumed ground - based observations by the university of hawaii @xmath207 telescope or the keck spectroscopy .
the spatial resolution , therefore , is set to be @xmath208 @xcite . for comparison
, we also made the emulated images with @xmath200-band by _ james web space telescope _ ( _ jwst _ ) or ground - based @xmath209 m telescopes with adaptive optics ( ao ) .
we assume that the angular resolution of these future instruments is @xmath210 .
figure [ multibands ] shows @xmath8- , @xmath197- , and @xmath200-band images , in which dust extinction are not taken into account , of runs @xmath83 and @xmath89 . in @xmath8-band images , multiple core - like structures are clearly visible , while such cores are blurred in other bands due to the limited resolution and the contribution of the luminosity of old stars . in @xmath197-band images ,
cores are connected one another and form arm - like structures . as a result ,
the cores seem to be in arms . in @xmath200-band images ,
multiple core structures are not resolved at all . as a result ,
ulirgs seem to have only single or double cores .
thus , multiple cores were identified as a single or double cores in @xmath197- and @xmath198-bands .
our analysis indicates that ulirgs with multiple cores can be misclassified as ones with single or double cores . the apparent discrepancy in the observational fractions of ulirgs with multiple nuclei
might have been caused by this difference in the spatial resolution . in the rightmost column in fig .
[ multibands ] , we show the emulated @xmath200-band images of our simulations with the angular resolution of @xmath211 .
we can clearly see the multiple bright sources in these @xmath200-band images that were not visible in grand - base observations , since the resolution of the image is remarkably improved .
we expect that , in the near future , observations of ( u)lirgs by _ jwst _ or ground - based @xmath209 m telescopes with ao will settle the problem of the band - to - band difference in the fraction of ulirgs with multiple nuclei .
figure [ multibands_dust ] shows the observed @xmath8- , @xmath197- , and @xmath200-band images taking dust extinction into account . here
, the cross - sections per h atom at @xmath197- and @xmath200-bands are assumed to be @xmath212 and @xmath213 , respectively @xcite .
we adopt @xmath183 as a fiducial value . in the @xmath8-band image ,
the hypermassive star clusters and the galactic cores are obscured because of dust extinction although the resolution is enough to observe them . on the other hand , in the @xmath197- and
ground - based @xmath200-band images , dust extinction becomes weaker although the clusters and the galactic cores are not resolved . in the @xmath200-band image by _ jwst _ or ground - based @xmath209 m telescopes with ao , the clusters and the galactic cores are clearly observable since not only dust extinction is small but also they are resolved .
when dust extinction is significant in @xmath8-band , we need the observations by _ jwst _ or ground - based @xmath209 m telescopes with ao .
we have performed high resolution @xmath54-body / sph simulations of merging galaxies .
we found that hypermassive star clusters with @xmath143 form from disturbed gas disks in the central region ( @xmath3 ) . in these clusters ,
active star formations take place , so that some bright core structures appear in the merging galaxy .
the features of formed hypermassive star clusters agree with the observations of ulirgs with multiple nuclei .
although dust extinction may obscure the clusters after their formation , the clusters become optically thin within the ulirg lifetime .
these results indicate that one major merger of two spiral galaxies can explain complex structures of multiple nuclei in ulirgs .
rare multiple major merger events on a short timescale are not necessary . in the high - redshift universe , gas - rich major mergers must occur frequently .
our result suggests that these merging galaxies have multiple hypermassive star clusters that would be observed as bright sources in either near or far infrared bands .
future observations using e.g. , _ jwst _ or ground - based @xmath209 m telescopes with ao are expected to find a number of such merging galaxies with multiple bright sources . in our simulations
, we assume that main feedback source is type ii sne , and radiative feedback from newly formed stars is not taken into account .
if the radiative feedback is effective , gas in hypermassive star clusters is heated and is ejected from the cluster @xcite .
this effect may reduce the cluster mass . in order to study the effect ,
we need radiation - hydrodynamic simulations as performed by @xcite . since the hypermassive star clusters are compact ( @xmath214 within @xmath2 ) , it is possible that star - star collisions and mergings occur in these clusters @xcite . if the stellar mass loss in the main sequence phase is not very large , such merging of stars might result in the runaway growth of supermassive stars and the formation of intermediate - mass black holes ( imbhs ) @xcite .
these imbhs are carried to the center of the galaxy together with the parent hypermassive star cluster by dynamical friction .
if there is already a supermassive black hole at the center of the progenitor galaxies , the imbh merges with the smbh through dynamical friction and eccentricity evolution @xcite .
if there was no supermassive black hole , multiple imbhs conveyed to the center by the parent cluster finally merge each other to form the seed of a smbh .
the authours thank the anonymous referee for many useful comments .
the authors also thank junichi baba , nozomu kawakatu , masakazu a. r. kobayashi , and tohru nagao for helpful discussions .
numerical computations were carried out on cray xt4 and grape system at center for computational astrophysics ( cfca ) of national astronomical observatory of japan .
this project is supported by nec corporation , molecular - based new computational science program of nins , grant - in - aid for scientific research ( 17340059 ) of jsps , and hpci strategic program field 5 `` the origin of matter and the universe '' .
is financially supported by a research fellowship from the japan society for the promotion of science for young scientists .
is financially supported by grant - in - aid for scientific research ( s ) by jsps ( 20224002 ) and by grant - in - aid for young scientists ( 21840015 ) .
, t. r. , daisaka , h. , kokubo , e. , makino , j. , oakmoto , t. , tomisaka , k. , wada , k. , & yoshida , n. 2010 , in astronomical society of the pacific conference series , vol .
423 , astronomical society of the pacific conference series , ed .
b. smith , j. higdon , s. higdon , & n. bastian , 185+ | ultraluminous infrared galaxies ( ulirgs ) with multiple ( @xmath0 ) nuclei are frequently observed .
it has been suggested that these nuclei are produced by multiple major mergers of galaxies
. the expected rate of such mergers is , however , too low to reproduce the observed number of ulirgs with multiple nuclei .
we have performed high - resolution simulations of the merging of two gas - rich disk galaxies .
we found that extremely massive and compact star clusters form from the strongly disturbed gas disks after the first or second encounter between the galaxies .
the mass of such clusters reaches @xmath1 , and their half - mass radii are @xmath2 .
since these clusters consist of young stars , they appear to be several bright cores in the galactic central region ( @xmath3 ) .
the peak luminosity of these clusters reaches @xmath4 of the total luminosity of the merging galaxy .
these massive and compact clusters are consistent with the characteristics of the observed multiple nuclei in ulirgs .
multiple mergers are not necessary to explain multiple nuclei in ulirgs . |
one - dimensional systems of @xmath0 particles mutually interacting through gravitational forces have been of interest in astrophysics for more than three decades . while used primarily as prototypes for the behaviour of gravity in higher dimensions , one - dimensional self - gravitating systems ( ogs s ) also approximate the behaviour of some physical systems in 3 spatial dimensions .
these include the dynamics of stars in a direction orthogonal to the plane of a highly flattened galaxy and the collisions of flat parallel domain walls moving in directions perpendicular to their surfaces .
furthermore , very long - lived core - halo structures in the ogs phase space are known to exist , reminiscent of structures observed in globular clusters , in which a dense massive core in near equilibrium is surrounded by a halo of stars with high kinetic energy that interact only weakly with the core yawn .
the statistical properties of the ogs are particularly intriguing . despite extensive study
, many unanswered questions remain .
for example it is not clear if the ogs can attain a true equilibrium state from arbitrary initial conditions .
its ergodic and equipartition properties are still not well understood .
this is primarily because the particle interactions of the ogs ( as with any self - gravitating system ) are attractive and cumulatively long range , in strong contrast to typical thermodynamic systems for which such interactions are repulsive and short range . for the ogs
the macroscopic dynamics does not decouple from the microscopic dynamics , and the usual thermodynamic analysis does not apply . however there are some established features of the ogs .
rybicki rybicki derived in closed form the single - particle distribution function in both the canonical and microcanonical ensembles . in the large @xmath0 limit
these distribution functions reduce to the isothermal solution of the vlasov equation .
all studies to date have neglected relativistic effects .
this limitation is understandable since no relativistic @xmath0 particle hamiltonian was available for analysis .
however this situation changed recently when a prescription for obtaining the hamiltonian for a relativistic one - dimensional self - gravitating system ( rogs ) was given by mann and ohta @xcite .
this hamiltonian can be rigorously derived from a generally covariant system coupling relativistic gravity in one spatial dimension ( i.e. a 1 + 1 dimensional theory of gravity @xcite ) to @xmath0 point particles . in the non - relativistic limit
, the hamiltonian reduces to that of the ogs .
although not available in closed form , the hamiltonian can be obtained as a series expansion in inverse powers of the speed of light @xmath4 to arbitary order .
we consider in this paper the one - particle distribution function for the rogs .
our work here is a natural extension of previous work on the @xmath0-body problem in relativistic gravity . in three spatial dimensions an exact solution to this problem
is known for pure newtonian gravity ( and a series solutions has been constructed for arbitrary @xmath0 ) . in the general theory of relativity dissipation of energy in the form of gravitational radiation
has obstructed progress toward obtaining exact solutions to the @xmath0-body problem even when @xmath5 .
however for the rogs an exact solution to the @xmath6-body problem was recently obtained @xcite , and generalizations including a cosmological constant and/or charge subsequently followed 2bdcossh,2bdcoslo,2bdchglo .
these solutions include both an explicit expression for the proper separtion of the two bodies as a function of time and an explicit expression for the hamiltonian for the @xmath6-body rogs as a function of the proper separation and the centre - of - inertia momentum of the bodies . encouraged by these results ,
we here begin a first attempt to understand the basic features of the @xmath0-body rogs .
we shall recapitulate the canonical formalism used in ref .
@xcite to derive the hamiltonian for the @xmath0-body rogs .
we then compute the partition function and canonical distribution functions . using an integral transform
we then calculate the microcanonical distribution functions .
all results are in closed form to leading order in @xmath7 .
we consider the limit of large @xmath0 and compare the rogs and the ogs .
we close with a few remarks .
lengthy intermediate calculations are confined to appendices .
the ogs hamiltonian is @xmath8where the summation is over all @xmath0 particles , located at positions @xmath9 along the spatial axis .
the potential term straightforwardly follows upon solving the newtonian equation @xmath10 in one spatial dimension , where @xmath11 is the mass density of the @xmath0 point particles .
our task in this section is to find a prescription for obtaining a relativistic generalization of ( [ ogs ] ) .
the hamiltonian for the rogs that we use is that of a @xmath12 dimensional theory ( a lineal gravity theory ) that models @xmath13 dimensional general relativity in that it sets the ricci scalar equal to the trace of the stress - energy of prescribed matter fields and sources .
hence matter governs the evolution of spacetime curvature which reciprocally governs the evolution of matter @xcite . we refer to this theory as @xmath14 theory . apart from being able to model a number of textbook scenarios in general relativity @xcite , it has the attractive feature of having a consistent newtonian limit @xcite .
this limit , essential for our purposes , is problematic in a generic @xmath15-dimensional theory of gravity theory @xcite .
since the einstein action is a topological invariant in @xmath15 dimensions , a scalar ( dilaton ) field must be included in the action @xcite .
its coupling to the curvature is chosen so that only the trace of the stress energy of matter ( @xmath0 point particles here ) is set equal to the ricci scalar .
this action will form the basis for the rogs we consider .
upon canonical reduction of the action @xcite , the rogs hamiltonian is given in terms of a spatial integral of the second derivative of the dilaton field , which is a function of the coordinates and momenta of the particles and is determined from the constraint equations .
the action integral for the gravitational field coupled to @xmath0 point particles is @xmath16{}\left .
+ \sum_{a}\int d\tau _ { a}\left\ { -m_{a}\left ( -g_{\mu \nu } ( x)\frac{dz_{a}^{\mu } } { d\tau _ { a}}\frac{dz_{a}^{\nu } } { d\tau _ { a}}% \right ) ^{1/2}\right\ } \delta ^{2}(x - z_{a}(\tau _ { a}))\right ] \ ; , \label{act1}\end{aligned}\ ] ] where @xmath17 is the dilaton field , @xmath18 and @xmath19 are the metric and its determinant , @xmath20 is the ricci scalar , and @xmath21 is the proper time of @xmath22-th particle whose mass is @xmath23 , with @xmath24 .
we use @xmath25 to denote the covariant derivative associated with @xmath26 .
the field equations derived from the action ( [ act1 ] ) are @xmath27 where @xmath28 is the stress - energy due to the point masses .
eq.([eq - psi ] ) guarantees the conservation of @xmath29 . inserting the trace of eq.([eq - psi ] ) into eq.([eq - r ] )
yields @xmath30 eqs .
( [ eq - psi ] ) , ( [ eq - z ] ) and ( [ rt ] ) form a closed sytem of equations for gravity and matter . in order to obtain the hamiltonian in canonical form
, we first decompose the scalar curvature in terms of the extrinsic curvature @xmath31 via @xmath32\ ; , \label{extk}\ ] ] where the metric is @xmath33 with @xmath34 , so that @xmath35 and @xmath36 . rewriting the action ( [ act1 ] ) in first - order form yields @xmath37 where @xmath38 and @xmath39 are the respective conjugate momenta to @xmath40 and @xmath17 . here
@xmath41 with the symbols @xmath42 and @xmath43 denoting @xmath44 and @xmath45 , respectively .
variation of the action ( [ act2 ] ) yields the set of equations @xmath46@xmath47{}+n_{1}\frac{p_{a}}{\gamma
^{2}}\frac{\partial \gamma } { % \partial z_{a}}=0\ ; , \label{e - p } \\ & & \dot{z_{a}}-n_{0}\frac{\frac{p_{a}}{\gamma } } { \sqrt{\frac{p_{a}^{2}}{% \gamma } + m_{a}^{2}}}+\frac{n_{1}}{\gamma } = 0\;. \label{e - z}\end{aligned}\]]all metric components ( @xmath48 , @xmath49 , @xmath40 ) in equations ( [ e - p ] ) and ( [ e - z ] ) are evaluated at the point @xmath50 , where @xmath51the quantities @xmath48 and @xmath49 are lagrange multipliers which yield the constraint equations ( [ e - r0 ] ) and ( [ e - r1 ] ) .
the above set of equations can be proved to be equivalent to the set of equations ( [ eq - r ] ) , ( [ eq - psi ] ) and ( [ eq - z ] ) @xcite .
an examination of the generator of space and time transformations or,2bd indicates that we find that we can consistently choose the coordinate conditions @xmath52{}\mbox{and}\makebox[2em]{}\pi = 0\ ; \label{cc}\]]upon which the action ( [ act2 ] ) reduces to @xmath53after elimination of the constraints , where @xmath54is the hamiltonian for the rogs .
the field @xmath17 is no longer arbitrary , but is instead a function of @xmath55 and @xmath56 that is determined by solving the constraints which are now @xmath57 @xmath58 once the coordinate conditions ( [ cc ] ) are imposed
. equation ( [ psi ] ) is an energy - balance equation which states that the energy of the particles plus the ( negative ) gravitational energy must vanish
. equation ( [ pi ] ) states that the total momentum of the gravitational field and the particles must vanish .
the consistency of this canonical reduction was proved in ref .
@xcite by showing that the canonical equations of motion derived from the reduced hamiltonian ( [ ham1 ] ) are identical with the equations ( [ e - p ] ) and ( [ e - z ] ) .
hence the hamiltonian ( [ ham1 ] ) is a function only of the coordinates and momenta of the @xmath0 particles in the system .
we turn next to its evaluation .
although the constraint equations are straightforward to solve in the regions between the particles , the matching conditions of these solutions at the juncture of the particles are quite non - trivial . for
the @xmath6-body rogs their enforcement yields an equation which determines the hamiltonian in terms of the remaining degrees of freedom of the system .
while this procedure holds in principle for the @xmath0-body rogs , we have not found a tractable means of obtaining an analogous determining equation for the hamiltonian .
however it is possible to straightforwardly and rigourously construct approximation schemes for computing the rogs hamiltonian for @xmath0 particles .
for example , the post - linear approximation is an expansion of the hamiltonian in powers of the gravitational coupling @xmath59 , obtained by writing @xmath60where@xmath61 is defined by @xmath62 .
insertion of these expansions into eqs .
( [ psi],[pi ] ) yields @xmath63 ^{2}\right
. \nonumber \\ & & \left . -\sum_{a}p_{a}\left
[ \sum_{b}p_{b}\left| r_{ab}\right| + \epsilon \;\sum_{b}\sqrt{p_{b}^{2}+m_{b}^{2}}r_{ab}\right ] \left [ \sum_{c}\sqrt{% p_{c}^{2}+m_{c}^{2}}\;\left| r_{ac}\right| + \epsilon \;\sum_{c}p_{c}r_{ac}% \right ] \right .
\nonumber \\ & & \left .
+ \sum_{a}\sum_{b}\left [ \sqrt{p_{a}^{2}+m_{a}^{2}}\sqrt{% p_{b}^{2}+m_{b}^{2}}\;\left| r_{ab}\right| -\epsilon \;p_{a}\sqrt{% p_{b}^{2}+m_{b}^{2}}r_{ab}\right ] \right .
\nonumber \\ & & \makebox[8em]{}\left
. \times \left [ \sum_{c}\sqrt{p_{c}^{2}+m_{c}^{2}}% \;\left| r_{bc}\right| + \epsilon \;\sum_{c}p_{c}r_{bc}\right ] \right .
\nonumber \\ & & \left . -\sum_{a}\sum_{b}\left
[ \sqrt{p_{a}^{2}+m_{a}^{2}}\;p_{b}\left| r_{ab}\right| -\epsilon \;p_{a}p_{b}r_{ab}\right ] \left [ \sum_{c}p_{c}\left| r_{bc}\right| + \epsilon \;\sum_{c}\sqrt{p_{c}^{2}+m_{c}^{2}}r_{bc}\right ] \right\ } \;\ ; \nonumber \\ & & \label{plinh}\end{aligned}\]]upon insertion into ( [ ham1 ] ) , where @xmath64 is the relative separation between particles @xmath22 and @xmath65 .
it can be shown that the solutions to eqs .
( [ psi],[pi ] ) must satisfy the boundary condition @xmath66 in the regions @xmath67 in order for the hamiltonian to be finite @xcite .
the @xmath59-expansion is appropriate for describing relativistic fast - motion of the particles and can be carried out to any desired order . however to compare the rogs from @xmath14 theory with the ogs , we turn to the post - newtonian expansion , which is an expansion of the hamiltonian in powers of @xmath68 . since both @xmath69 and @xmath70 are of the order of @xmath71 all terms up to the order of @xmath72 are included in the post - linear hamiltonian ( [ plinh ] ) .
the post - newtonian hamiltonian to this order is therefore @xcite @xmath73 + \cdots \label{rogs}\end{aligned}\]]where the explicit powers of @xmath2 have been restored .
the first term in ( [ rogs ] ) is the total rest energy of the particles , and the second two terms are the ogs hamiltonian ( [ ogs ] )
. the remaining terms are all relativistic corrections to the ogs to order @xmath71 .
the first of these corrections is a special - relativistic one , whereas the remaining corrections are due to relativistic gravity in one spatial dimension .
note that gravity not only modifies the potential to a quadratic form , but also includes couplings between particle momenta and their positions .
these features modifications of the distance behaviour of the potential and position - momentum couplings are fully analogous to those in general relativity in three spatial dimensions .
we next find the equations of motion for the position of the @xmath22th particle .
this follows straightforwardly from hamilton s principle .
we have @xmath74 \label{e2}\end{aligned}\ ] ] and @xmath75 \right ] \label{e3}\end{aligned}\ ] ] where @xmath76
. we can solve ( [ e2 ] ) for @xmath56 : @xmath77 + \cdots \label{e4}\ ] ] and then insert this into the equation ( [ e3 ] ) for @xmath78@xmath79 \nonumber \\ & = & -2\pi g\sum_{b=1}^{n}m_{a}m_{b}\varepsilon _ { ab}-\frac{1}{c^{2}}\left [ \pi g\sum_{b=1}^{n}\left ( m_{a}m_{b}\left [ \dot{z}_{b}^{2}+\dot{z}_{a}^{2}% \right ] \right ) \varepsilon _ { ab}-2\pi g\sum_{b=1}^{n}m_{a}m_{b}\dot{z}_{a}% \dot{z}_{b}\varepsilon _ { ab}\right .
\nonumber \\ & & \left .
+ 2\left ( \pi g\right ) ^{2}\sum_{b=1}^{n}\sum_{c=1}^{n}m_{a}m_{b}m_{c}\left [ \varepsilon _ { ab}\left| r_{ac}\right| + \varepsilon _ { ab}\left| r_{bc}\right| -r_{ab}% \right ] \right ] \label{e5}\end{aligned}\ ] ] which simplifies to @xmath80 ^{2}\right\ } \varepsilon _ { ab}\right ) \nonumber \\ & & -\left ( \frac{2\pi g}{c}\right ) ^{2}\sum_{b=1}^{n}\sum_{c=1}^{n}m_{a}m_{b}m_{c}\left ( \varepsilon _ { ac}-\varepsilon _ { bc}\right ) \left| r_{ab}\right| \nonumber \\ & & -2\left ( \frac{\pi g}{c}\right ) ^{2}\sum_{b=1}^{n}\sum_{c=1}^{n}m_{a}m_{b}m_{c}\left [ \varepsilon _ { ab}\left| r_{ac}\right| + \varepsilon _ { ab}\left| r_{bc}\right| -r_{ab}% \right ] \label{e6}\end{aligned}\ ] ] upon an iterative substitution of @xmath81 in powers of @xmath68 .
these equations of motion reduce to those of the ogs in the limit @xmath82 , and may be shown to be equivalent to the geodesic equations to this order @xcite .
we wish to investigate the instrinsic structure of the system described by the hamiltonian ( [ rogs ] ) .
however because of the translation invariance of the system , two phase - space degrees of freedom are redundant , and so must be factored out ; otherwise certain average properties such as density would be uniform throughout space . using equation([e3 ] )
, it is straightforward to show that @xmath83 and ( using also ( [ e2 ] ) ) that the hamiltonian is time - independent .
this means that we can perform the phase - space integration subject to the constraint @xmath84 since we can choose a frame of reference in which the centre of inertia is constant .
removing the redundant position degree - of - freedom is somewhat more delicate . although the system is invariant under the translation @xmath85 , eq . ( [ e4 ] )
implies @xmath86 \label{e9}\ ] ] which can not be written as a total time derivative .
physically , the centre of inertia is the relativistically well - defined concept , whereas the centre of mass is not .
however we can deal with this problem by inserting a factor of unity in all phase - space averages in the form @xmath87 where @xmath88 , with @xmath89 @xmath90 . if the @xmath91-dependence of any integral trivially factors out ( or can be removed by a shift of variable in the integrand ) , then we regard the remaining quantity as the physically relevant one to describe the system . for a canonical ensemble ,
all phase - space averages are carried out with a weighting function @xmath92 $ ] , where @xmath93 is the temperature multiplied by boltzmann s constant @xmath94 .
for the microcanonical ensemble , an additional constraint of fixed total energy @xmath95 must be included , consistent with the time - independence of the hamiltonian ( [ rogs ] ) . since the system is in momentum isolation , it is difficult to see how it can be in energy contact with a heat bath , and so the physical relevance of the canonical ensemble is somewhat unclear .
however an evaluation of quantities within the canonical ensemble is instructive in its own right and is a necessary preliminary to computing quantities in the more realistic microcanonical ensemble , and so we include it in the present discussion
. henceforth we set @xmath96 , so that @xmath97 .
we consider in this section the relativistic corrections to the canonical one - particle distribution function @xmath98 , which is defined to be the phase - space average of the quantity @xmath99weighted by @xmath100 with the constraint @xmath101 hence @xmath102where @xmath103is the partition function and where the 2nd line in ( [ can2 ] ) follows from the indistinguishability of the particles .
note that a shift of integration variable @xmath104renders the partition function in the form @xmath105where the @xmath91-dependence is seen to trivially factor out .
it can therefore be dropped ( along with the prime notation ) from further consideration in the evaluation of @xmath106 .
similarly , the single - particle distribution function becomes @xmath107which is of the form @xmath108 .
we therefore regard @xmath109 as the physically relevant quantity , where @xmath110and where the primes will hencefore be dropped . unlike the non - relativistic case ,
neither the partition function nor @xmath111 are separable .
we proceed by first evaluating the partition function .
we first write the hamiltonian ( [ rogs ] ) in the following form @xmath112 \label{can7}\end{aligned}\ ] ] so that @xmath113which is valid to the order in which we are working . writing @xmath114
\label{can9}\]]we have @xmath115 consider first the integral @xmath116which has integrals that are at most quartic in the momenta .
straightforward gaussian integration yields @xmath117hence we obtain @xmath118 \right ) \right .
\nonumber \\ & & \left .
-\frac{\pi g\beta } { c^{2}}\left ( \frac{m}{\beta } \right ) \frac{n-1}{% n}\sum_{a=1}^{n}\sum_{b=1}^{n}\left| r_{ab}\right| + \frac{2\pi g\beta } { c^{2}% } \left ( -\frac{m}{n\beta } \right ) \sum_{a > b}^{n}\left| r_{ab}\right| + \frac{% \beta } { 8m^{3}c^{2}}\frac{3\left ( n-1\right ) ^{2}}{n^{2}}\left ( \frac{m}{% \beta } \right ) ^{2}\left ( \sum_{a=1}^{n}1\right ) \right\ } \nonumber \\ & = & \left ( 1-\frac{\beta \left ( \pi g\right ) ^{2}}{c^{2}}\sum_{a=1}^{n}% \sum_{b=1}^{n}\sum_{c=1}^{n}m^{3}\left [ \left| r_{ab}\right| \left| r_{ac}\right| -r_{ab}r_{ac}\right ] -\frac{2\pi gm}{c^{2}}\sum_{a >
b}^{n}% \left| r_{ab}\right| + \frac{3\left ( n-1\right ) ^{2}}{8n\beta mc^{2}}\right ) \frac{1}{\sqrt{n}}\left ( \frac{2\pi m}{\beta } \right ) ^{\left ( n-1\right ) /2 } \nonumber \\ & & \label{can13}\end{aligned}\ ] ] we next consider the integration over the spatial variables . introducing @xmath119 we have @xmath120 where without loss of generality the particles are ordered in the sequence @xmath121 , and the overall result is then multiplied by @xmath122 this gives @xmath123 provided the function @xmath124 is symmetric under interchange of any pair of variables , which is the case here .
the inverse transformation reads @xmath125 and so we have @xmath126 & = & 2\left ( \sum_{b > a >
c}^{n}r_{ba}r_{ac}+\sum_{b > a > c}^{n}r_{ca}r_{ab}\right ) \nonumber \\ & = & 4\sum_{k=1}^{n-2}\sum_{l = k+1}^{n-1}\left ( n - l\right ) \left ( l - k\right ) ku_{l}u_{k } \label{can18}\end{aligned}\ ] ] consequently the partition function is @xmath127the three basic integrals in ( [ can19 ] ) are @xmath128 ^{2}% } \label{can20a } \\
\int d{\bf u}\sum_{l=1}^{n-1}k\left ( n - k\right ) u_{k}\exp \left ( -\lambda \sum_{n=1}^{n-1}n\left ( n - n\right ) u_{n}\right ) & = & \frac{n-1}{\lambda ^{n}% \left [ \left ( n-1\right ) !
\right ] ^{2 } } \label{can20 } \\
\int d{\bf u}\sum_{k=1}^{n-2}\sum_{l = k+1}^{n-1}\left ( n - l\right ) \left ( l - k\right ) ku_{l}u_{k}\exp \left ( -\lambda \sum_{n=1}^{n-1}n\left ( n - n\right ) u_{n}\right ) & = & \frac{\sum_{k=1}^{n-2}\sum_{l = k+1}^{n-1}\frac{% \left ( l - k\right ) } { l\left ( n - k\right ) } } { \lambda ^{n+1}\left [ \left ( n-1\right ) !
\right ] ^{2}}\mbox { \ \ \ \ \ \ \ } \label{can20c}\end{aligned}\]]yielding @xmath129 ^{2}}\right .
\nonumber \\ & & \mbox { \ \ \ } \left .
-\frac{4\beta } { c^{2}}\frac{\left ( \pi g\right ) ^{2}m^{3}}{\left ( 2\pi g\beta m^{2}\right ) ^{n+1}\left [ \left ( n-1\right ) ! % \right ] ^{2}}\sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}\frac{\left ( l - k\right ) } { % l\left ( n - k\right ) } -\frac{2\pi gm}{c^{2}}\frac{n-1}{\left ( 2\pi g\beta m^{2}\right ) ^{n}\left [ \left ( n-1\right ) !
\right ] ^{2}}\right ) \nonumber \\ & = & \frac{e^{-\beta mc^{2}}\left ( \frac{2\pi m}{\beta } \right ) ^{\left ( n-1\right ) /2}}{\sqrt{n}\left
( 2\pi g\beta m^{2}\right ) ^{n-1}\left [ \left ( n-1\right ) !
\right ] ^{2}}\left [ 1-\frac{1}{\beta mc^{2}}\left\ { \frac{\left ( 5n+3\right ) \left ( n-1\right ) + 8n\sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}\frac{% \left ( l - k\right ) } { l\left ( n - k\right ) } } { 8n}\right\ } \right ] \nonumber \\ & = & \frac{\exp \left [ -\beta mc^{2}-\frac{3\left ( n-1\right ) } { 2}\ln \left ( \beta mc^{2}\right ) -\left\ { \frac{\left ( 5n+3\right ) \left ( n-1\right ) + 8n\sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}\frac{\left ( l - k\right ) } { l\left ( n - k\right ) } } { 8n\beta mc^{2}}\right\ } \right ] } { \sqrt{n}\left ( \sqrt{2\pi } % g / c^{3}\right ) ^{\left ( n-1\right ) } \left [ \left ( n-1\right )
! \right ] ^{2}}% \mbox { } \label{can21}\end{aligned}\]]which is the partition function to lowest relativistic order .
the average energy is @xmath130to the relevant order in @xmath71 .
the relativistic correction grows quadratically with @xmath0 ( for fixed @xmath97 ) and is negative .
hence the average energy of a relativistic gravitating system is lower than its non - relativistic counterpart at the same temperature .
consider next the one - particle distribution function , which is @xmath131where @xmath132 in which @xmath133and @xmath134where eq .
( [ can14 ] ) has been used to express @xmath135 evaluation of @xmath136 is somewhat lengthy , and so we relegate its computation to the appendix .
we obtain @xmath137 \right ) \right .
\nonumber \\ & & + \frac{1}{2\beta mc^{2}}\left [ \frac{\beta ^{2}p^{4}n\left ( n^{2}-3n+3\right ) } { ( 2m)^{2}\left ( n-1\right ) ^{3}}+\frac{3\beta p^{2}\left ( n-2\right ) } { 2m\left ( n-1\right ) ^{5/2}}+\frac{3\left ( n-2\right ) ^{2}}{% 4\left ( n-1\right ) ^{3/2}}\right ] \label{can25bb } \\ & & + \left ( \frac{\mbox{2}\pi mg}{c^{2}}\right ) \left [ -\frac{1}{2}% \sum_{a=1}^{n}\sum_{b=1}^{n}\left| r_{ab}\right| -\left ( \frac{n^{2}\beta p^{2}}{2m\left ( n-1\right ) ^{2}}-\frac{n}{2(n-1)}\right ) \left ( \sum_{c=1}^{n}\left| r_{cn}\right| \right ) \right ] \nonumber\end{aligned}\]]or alternatively , in terms of the @xmath138 variables @xmath139 \label{can25bc } \\ & & -\left ( \frac{2\pi mg}{c^{2}}\right ) \left [ \sum_{l=1}^{n-1}l\left ( n - l\right ) u_{l}+\left ( \frac{n^{2}\beta p^{2}}{2m\left ( n-1\right ) ^{2}}-% \frac{n}{2(n-1)}\right ) \left ( \sum_{s=1}^{n-1}su_{s}+\sum_{s = n}^{n-1}(n - s)u_{s}\right ) \right ] \nonumber\end{aligned}\ ] ] now consider eq .
( [ can24 ] ) , which can be rewritten as @xmath140where @xmath141the integration now involves straightforward integrations over the @xmath138 variables , after which an evaluation of the @xmath142-integral using jordan s lemma must be performed .
this involves some rather tedious manipulations which we describe in appendix ? ? ?
the final result is @xmath143 \nonumber \\ & & \mbox { \ } \times \sum_{l=1}^{n-1}\left [ \left\ { a_{l}^{n}\left ( 1+\frac{1}{% 2\beta mc^{2}}\left ( \frac{\beta ^{2}p^{4}\left ( 1+\left ( n-1\right ) ^{3}\right ) } { \left ( 2m\right ) ^{2}\left ( n-1\right ) ^{3}}+\frac{3\left ( n-2\right ) \beta p^{2}}{2m\left ( n-1\right ) ^{2}}+\frac{3\left ( n-2\right ) ^{2}}{4\left ( n-1\right ) } \right ) \right ) \right .
\nonumber \\ & & \mbox{\ } -\frac{1}{\beta mc^{2}}\left [ b_{l}^{n}-a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \nonumber \\ & & + \frac{2}{\beta mc^{2}}\left ( \frac{n}{2\left ( n-1\right ) } -\frac{\beta p^{2}}{2m}\left [ \frac{n^{2}}{\left ( n-1\right ) ^{2}}\right ] \right ) \mbox{\ } \left [ c_{l}^{n}-\frac{1}{l}a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \label{can34 } \\ & & \mbox { \ \ \ \ } \left .
-\frac{1}{\beta mc^{2}}\left [ d_{l}^{n}+k_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \right\ } \exp \left ( -\frac{n\beta p^{2}}{2m\left ( n-1\right ) } -2\pi gn\beta m^{2}l\left| z\right| \right ) \right ] \mbox { \ \ } \nonumber\end{aligned}\]]integration over @xmath144 yields the canonical density distribution function @xmath145 \nonumber \\ & & \times \sum_{l=1}^{n-1}\left\ { a_{l}^{n}+\frac{1}{\beta mc^{2}}\left ( \frac{3}{8}\frac{\left ( n-1\right ) ^{2}}{n}a_{l}^{n}-b_{l}^{n}-d_{l}^{n}% \right ) \right .
\nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ } \left .
+ \frac{1}{\beta mc^{2}}\left [ a_{l}^{n}-k_{l}^{n}\right ] \left ( 1 - 2\pi gm^{2}n\beta l\left| z\right| \right ) \right\ } \exp \left ( -2\pi gm^{2}n\beta l\left| z\right| \right ) \mbox { \ \ \ \ \ } \label{can35}\end{aligned}\]]whereas integration over @xmath146 yields the canonical momentum distribution function @xmath147 \nonumber \\ & & \times \left ( 1+\frac{1}{\beta mc^{2}}\left ( \frac{n\beta ^{2}p^{4}\left ( n^{2}-3n+3\right ) } { 8m^{2}\left ( n-1\right ) ^{3}}-\frac{\beta p^{2}\left ( 4n^{2}-7n+6\right ) } { 4m\left ( n-1\right ) ^{2}}+\frac{5n(n-1)+3}{8n\left ( n-1\right ) } \right ) \right ) \mbox { \ \ \ \ } \label{can36}\end{aligned}\ ] ] from the results in the appendix , it can be shown that @xmath148 which can be used to show that @xmath149 we also have the relations @xmath150 the first of which was demonstrated by rybicki @xcite .
the results for the canonical ensemble obtained in the previous section are for a system of relativistic gravitating particles coupled to a heat bath which keeps the system at a constant temperature @xmath151 .
in such a situation the energy of the system ill - defined , and undergoes fluctuations of the order @xmath152 . an isolated system , on the other hand , would have its total energy conserved , and this is the more realistic astrophysical case .
this entails usage of the microcanonical ensemble , in which phase space integrations are carried out by constraining the total energy to be @xmath153 the weighting function @xmath154 in the phase space integral is therefore replaced with @xmath155 .
fortunately it is straightforward to compute the relevant microcanonical quantities from the canonical ones . using the same reasoning that led to ( [ can2a ] ) ,
the microcanonical single - particle distribution function is @xmath156 where @xmath157 note that @xmath158 and @xmath106 are related by the laplace transforms @xmath159where in contour @xmath160 in the latter integral extends from @xmath161 to @xmath162 to the right of all singularities . using the general result that @xmath163it is straightforward to obtain @xmath164 ^{2}\gamma \left ( \frac{3}{2}\left ( n-1\right ) \right ) } \nonumber \\ & & \times \exp \left [ -\frac{\left ( e - mc^{2}\right ) } { mc^{2}}\left\ { \frac{% \left ( 5n+3\right ) \left ( n-1\right ) + 8n\sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}% \frac{\left ( l - k\right ) } { l\left ( n - k\right ) } } { 12(n-1)}\right\ } \right ] \label{mi5}\end{aligned}\]]to leading order in @xmath1 .
similarly using ( [ mi4 ] ) in ( [ mi1 ] ) , we have @xmath165using ( [ can21 ] ) and ( [ can34 ] ) we have @xmath143 \nonumber \\ & & \mbox { \ } \times \sum_{l=1}^{n-1}\left [ \left\ { a_{l}^{n}\left ( 1+\frac{1}{% 2\beta mc^{2}}\left ( \frac{\beta ^{2}p^{4}\left ( 1+\left ( n-1\right ) ^{3}\right ) } { \left ( 2m\right ) ^{2}\left ( n-1\right ) ^{3}}+\frac{3\left ( n-2\right ) \beta p^{2}}{2m\left ( n-1\right ) ^{2}}+\frac{3\left ( n-2\right ) ^{2}}{4\left ( n-1\right ) } \right ) \right ) \right . \right .
\nonumber \\ & & -\frac{1}{\beta mc^{2}}\left [ b_{l}^{n}-a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \nonumber \\ & & + \frac{2}{\beta mc^{2}}\left ( \frac{n}{2\left ( n-1\right ) } -\frac{\beta p^{2}}{2m}\left [ \frac{n^{2}}{\left ( n-1\right ) ^{2}}\right ] \right ) \mbox{\ } \left [ c_{l}^{n}-\frac{1}{l}a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \label{mi6a } \\ & & \mbox{\ \ \ } \left .
-\frac{1}{\beta mc^{2}}\left [ d_{l}^{n}+k_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \right\ } \exp \left ( -\frac{n\beta p^{2}}{2m\left ( n-1\right ) } -2\pi gn\beta m^{2}l\left| z\right| \right ) \right ] \nonumber\end{aligned}\]]@xmath166 \nonumber \\ & & \times \sum_{l=1}^{n-1}\left\ { \left [ a_{l}^{n}+\left ( \frac{a_{l}^{n}}{% mc^{2}}\left [ \frac{3p^{2}\left ( n-2\right ) } { 4m\left ( n-1\right ) ^{2}}% \right ] -\frac{c_{l}^{n}-\frac{1}{l}a_{l}^{n}}{mc^{2}}\left [ \frac{% 2n^{2}p^{2}}{2m\left ( n-1\right ) ^{2}}\right ] \right ) \right ] \upsilon _ { + } ^{3n/2 - 4}\right .
\nonumber \\ & & + \left [ \frac{2n\left ( \pi gm^{2}\right ) \left| z\right| } { mc^{2}}\left ( a_{l}^{n}\left ( \frac{n}{\left ( n-1\right ) } -l\right ) + lk_{l}^{n}\right ) % \right ] \upsilon _ { + } ^{3n/2 - 4 } \nonumber \\ & & \mbox{\ } + \frac{\left ( \frac{3}{2}n-4\right ) } { \left ( e - mc^{2}\right ) } % \frac{a_{l}^{n}}{2mc^{2}}\left [ \frac{p^{4}n\left ( n^{2}-3n+3\right ) } { % ( 2m)^{2}\left ( n-1\right ) ^{3}}-\frac{2p^{2}}{m}\left [ \frac{2n^{3}\pi gm^{2}\left| z\right| } { \left ( n-1\right ) ^{2}}\right ] \right ] \upsilon _ { + } ^{3n/2 - 5 } \label{mi7 } \\ & & \mbox{\ } \left .
+ \frac{\left ( e - mc^{2}\right ) } { \left ( \frac{3}{2}% n-3\right ) mc^{2}}\left ( a_{l}^{n}\left [ \frac{3\left ( n-2\right ) ^{2}}{% 8\left ( n-1\right ) } + 1\right ] -b_{l}^{n}\mbox{\ } + \frac{n\left ( c_{l}^{n}-% \frac{1}{l}a_{l}^{n}\right ) } { \left ( n-1\right ) } -d_{l}^{n}-k_{l}^{n}\right ) \upsilon _ { + } ^{3n/2 - 3}\right\ } \nonumber\end{aligned}\]]as the expression for the relativistic microcanonical partition function , valid to @xmath167 , where @xmath168 * * employing the expressions @xmath169the density distribution is @xmath170 \nonumber \\ & & \times \sum_{l=1}^{n-1}\left\ { \left ( \frac{3n-5}{2}\right ) \left [ a_{l}^{n}+\frac{2\left ( \pi gm^{2}\right ) \left| z\right| } { mc^{2}}\left ( -la_{l}^{n}+lk_{l}^{n}\right ) \right ] \left ( \upsilon \left ( 0,z\right )
\right ) _ { + } ^{3n/2 - 7/2}\right . \nonumber \\ & & \mbox { \ } + \left ( \frac{\left ( e - mc^{2}\right ) } { mc^{2}}\right ) \left ( \upsilon \left ( 0,z\right ) \right ) _ { + } ^{3n/2 - 5/2}\left [ \left ( -\left ( c_{l}^{n}-\frac{1}{l}a_{l}^{n}\right ) \left [ \frac{n^{2}}{\left ( n-1\right ) } % \right ] \right ) \right .
\nonumber \\ & & \mbox { \ \ \ \ } + a_{l}^{n}\left [ \frac{3\left ( n^{2}-3n+3\right ) } { 8\left ( n-1\right ) } + \frac{3\left ( n-2\right ) } { 4\left ( n-1\right ) } + \frac{3n\left ( n-2\right ) ^{2}}{8\left ( n-1\right ) } + n\right ] \nonumber \\ & & \mbox{\ \ \ \ \ \ } \left . \mbox{\ } + \left .
\left ( -nb_{l}^{n}\mbox{\ } + % \frac{n^{2}\left ( c_{l}^{n}-\frac{1}{l}a_{l}^{n}\right ) } { \left ( n-1\right ) } % -n\left ( d_{l}^{n}+k_{l}^{n}\right ) \right ) \right ] \right\ } \nonumber \\ & = & \left ( \frac{2n\pi gm^{2}}{\left ( e - mc^{2}\right ) } \right ) \exp \left [ \frac{\left ( e - mc^{2}\right ) } { mc^{2}}\left\ { \frac{\left ( 5n+3\right ) \left ( n-1\right ) + 8n\sum_{k=1}^{n-1}\sum_{j = k+1}^{n-1}\frac{\left ( j - k\right ) } { j\left ( n - k\right ) } } { 12(n-1)}\right\ } \right ] \nonumber \\ & & \times \sum_{l=1}^{n-1}\left\ { \left ( \frac{3n-5}{2}\right ) \left [ a_{l}^{n}-\frac{2\left ( \pi gm^{2}\right ) l\left| z\right| } { mc^{2}}\left ( a_{l}^{n}-k_{l}^{n}\right ) \right ] \left ( \upsilon \left ( 0,z\right )
\right ) _ { + } ^{3n/2 - 7/2}\right . \nonumber \\ & & \left .
+ \left ( \frac{\left ( e - mc^{2}\right ) } { mc^{2}}\right ) \left ( \upsilon \left ( 0,z\right ) \right ) _ { + } ^{3n/2 - 5/2}\left ( \left ( \frac{% 3(n-1)^{2}}{8}\right ) a_{l}^{n}-n\left ( b_{l}^{n}-a_{l}^{n}\right ) -n\left ( d_{l}^{n}+k_{l}^{n}\right )
\right ) \right\ } \nonumber \\ & & \label{mi9}\end{aligned}\]]the normalization of the density @xmath171 implies @xmath172 \label{mi10}\end{aligned}\]]which , using ( [ can37 ] ) , is easily shown to be satisfied to first order in @xmath173 , where the latter quantity is the dimensionless fraction of excess energy above the total rest mass .
the momentum distribution is @xmath174 \nonumber \\ & & \times \sum_{l=1}^{n-1}\left\ { \left [ \frac{2a_{l}^{n}}{l}+\left ( \frac{% 2a_{l}^{n}}{lmc^{2}}\left [ \frac{3p^{2}\left ( n-2\right ) } { 4m\left ( n-1\right ) ^{2}}\right ] -\frac{2\left ( c_{l}^{n}-\frac{1}{l}a_{l}^{n}\right ) } { lmc^{2}}\left [ \frac{n^{2}p^{2}}{m\left ( n-1\right ) ^{2}}\right ] \right ) % \right ] \left ( \upsilon \left ( p,0\right ) \right ) _ { + } ^{3n/2 - 3}\right .
\nonumber \\ \mbox { \ \ \ \ \ } & & + \left [ \frac{4\left ( e - mc^{2}\right ) } { mc^{2}\left ( 3n-4\right ) } \left ( \frac{a_{l}^{n}}{l^{2}}\left ( \frac{n}{\left ( n-1\right ) } -l\right ) + \frac{k_{l}^{n}}{l}\right ) \right ] \left ( \upsilon \left ( p,0\right ) \right ) _ { + } ^{3n/2 - 2 } \nonumber \\ & & + \frac{2a_{l}^{n}}{l^{2}mc^{2}}\left [ -\frac{p^{2}}{m}\left [ \frac{n^{2}}{% \left ( n-1\right ) ^{2}}\right ] \right ] \left ( \upsilon \left ( p,0\right ) \right ) _ { + } ^{3n/2 - 3 } \nonumber \\ & & + \frac{\left ( \frac{3}{2}n-3\right ) } { \left ( e - mc^{2}\right ) } \frac{% a_{l}^{n}}{lmc^{2}}\left [ \frac{p^{4}n\left ( n^{2}-3n+3\right ) } { % ( 2m)^{2}\left ( n-1\right ) ^{3}}\right ] \left ( \upsilon \left ( p,0\right ) \right ) _ { + } ^{3n/2 - 4 } \nonumber \\ & & + \frac{4\left ( e - mc^{2}\right ) } { \left ( 3n-4\right ) mc^{2}}\left ( \frac{% a_{l}^{n}}{l}\left [ \frac{3\left ( n-2\right ) ^{2}}{8\left ( n-1\right ) } + 1% \right ] -\frac{b_{l}^{n}}{l}\mbox{\ } \right .
\nonumber \\ & & \mbox { \ \ \ \ } \left .
\left . + \frac{n\left ( c_{l}^{n}-\frac{1}{l}% a_{l}^{n}\right ) } { l\left ( n-1\right ) } -\frac{d_{l}^{n}+k_{l}^{n}}{l}\right ) \left ( \upsilon \left ( p,0\right )
\right ) _ { + } ^{3n/2 - 2}\right\ } \nonumber \\ & = & \left ( \frac{n}{2\pi m\left ( n-1\right ) \left ( e - mc^{2}\right ) } \right ) ^{1/2}\frac{\gamma \left ( \frac{3}{2}\left ( n-1\right ) \right ) } { \gamma \left ( \frac{3}{2}n-2\right ) } \nonumber \\ & & \times \exp \left [ \frac{\left ( e - mc^{2}\right ) } { mc^{2}}\left\ { \frac{% \left ( 5n+3\right ) \left ( n-1\right ) + 8n\sum_{k=1}^{n-1}\sum_{j = k+1}^{n-1}% \frac{\left ( j - k\right ) } { j\left ( n - k\right ) } } { 12(n-1)}\right\ } \right ] \nonumber \\ & & \times \left\ { \left [ 1-\frac{n}{mc^{2}}\left ( \frac{p^{2}}{4m}\left [ \frac{8n^{2}-11n+6}{\left ( n-1\right ) ^{2}}\right ] \right ) \right ] \left ( \upsilon \left ( p,0\right ) \right ) _ { + } ^{3n/2 - 3}\right . \nonumber \\ & & + \frac{n}{\left ( e - mc^{2}\right ) } \frac{3\left ( n-2\right ) } { 4mc^{2}}\left [ \frac{p^{4}n\left ( n^{2}-3n+3\right ) } { ( 2m)^{2}\left ( n-1\right ) ^{3}}\right ] \left ( \upsilon \left ( p,0\right ) \right ) _ { + } ^{3n/2 - 4 } \nonumber \\ & & \left .
-\frac{2n\left ( e - mc^{2}\right ) } { \left ( 3n-4\right ) mc^{2}}\left ( \frac{\left ( 5n^{2}-20n+12\right ) } { 8\left ( n-1\right ) } \mbox{\ } % + \sum_{s=1}^{n-1}\sum_{t = s+1}^{n-1}\frac{\left ( t - s\right ) } { t\left ( n - s\right ) } \right ) \left ( \upsilon \left ( p,0\right ) \right ) _ { + } ^{3n/2 - 2}\right\ } \label{mi11}\end{aligned}\]]it is straightforward to show that @xmath175 from ( [ mi8a ] [ mi8c ] ) to first order in @xmath176 .
for statisical systems ( such as those of interest in stellar dynamics ) , the large @xmath0 limit is of considerable physical interest .
this is the limit in which the total energy @xmath177 and total mass @xmath97 are fixed . in the non - relativistic case
, the single - particle distribution function approaches the isothermal solution of the vlasov equations in the large @xmath0 limit rybicki .
however the relativistic case is somewhat more subtle , since the expressions we have obtained are valid only if the speed of light is sufficiently large relative to other quantities of the same dimension , and as @xmath0 becomes large we must ensure that this approximation remains valid . in order to investigate the large @xmath0 limit it is necessary to rewrite all quantities in terms of @xmath177 , @xmath178 , and @xmath0 . as in ref .
@xcite , we adopt the dimensionless variables @xmath179 where @xmath180 are the characterisitic length and velocity scales of the system .
the scaled distributions functions are correspondingly defined @xmath181 so that @xmath182 consider first the partition function ( [ can21 ] ) , which can be rewritten as @xmath183 } { \sqrt{n}% \left ( \sqrt{2\pi } g / c^{3}\right ) ^{\left ( n-1\right ) } \left [ \left ( n-1\right ) ! \right ] ^{2}}\left ( \frac{n}{\beta mc^{2}}\right ) ^{3\left ( n-1\right ) /2 } \label{lnz}\ ] ] the approximation ( [ can8 ] ) is valid provided @xmath184 which sets an upper bound on the thermal energy @xmath185 of the system for a given value of @xmath0 . for fixed @xmath186
the relativistic corrections are valid only as the temperature becomes vanishingly small in the limit of large @xmath0 .
the exponential approximation is slightly better than the polynomial one because of the positivity of the partition function .
consider next the average energy in the canonical case as given by eq .
( [ can22 ] ) , which we rewrite as @xmath187where @xmath173 , as before .
when the thermal energy @xmath185 of the system is sufficiently small relative to its rest energy @xmath186 , the expression for the average energy @xmath188 does not differ much from the non - relativistic value given by its first term . as the thermal energy grows
( i.e. as @xmath189 decreases ) the value of @xmath190 increases more slowly than its non - relativistic counterpart , reaching a maximum when @xmath191after which the average energy decreases with decreasing @xmath189 , becoming negative when @xmath192 . as @xmath0 becomes large , @xmath193 . at @xmath194
the average energy has half the value of its non - relativistic counterpart . in figure[fig2 ] we plot the maximum value of @xmath195 as a function of @xmath0 .
the curve asymptotes to the constant value of @xmath196 as @xmath197 .
of course the relativistic expansion ( [ can5 ] ) breaks down well before @xmath198 reaches this point . in figure [ fig-1 ]
we plot the average energy @xmath190 as a function of @xmath152 for @xmath199 .
the relativistic case is clearly distinguishable from its non - relativistic counterpart once @xmath200 .
however the upper bound on the thermal energy is @xmath201 for @xmath199 . in [ fig1a ] we plot the average energy over the allowed range of @xmath152 illustrating that the distinction between the two cases is about @xmath202 at most .
for @xmath203 the maximum difference between the two cases is less than one part in a thousand over the allowed range of @xmath204 . in the canonical case
we take the energy @xmath177 to be the fixed total average energy as given by eq .
( [ can22 ] ) . solving this equation for the inverse temperature @xmath189 yields @xmath205 \zeta } { 12\left ( e - mc^{2}\right ) } = \frac{a_{n}% } { mc^{2}\zeta }
\left ( 1-\frac{b_{n}}{a_{n}^{2}}\zeta \right ) \label{ln5}\]]where @xmath206are defined for convenience .
the limit ( [ lnza ] ) implies that @xmath207where the latter limit holds for large @xmath0 . since the exponential forms a better approximation than the polynomial one , the value of @xmath208 is probably a bit larger than what is given in eq .
( [ ln6a ] ) , although it is not clear how much .
we obtain @xmath209 \nonumber \\ & & \times \sum_{l=1}^{n-1}\left [ \left\ { a_{l}^{n}-\frac{n\zeta } { a_{n}}\left [ b_{l}^{n}-a_{l}^{n}\left ( 1-\frac{4a_{n}}{3n}\left ( 1-\frac{b_{n}}{a_{n}^{2}}% \zeta \right ) l\left| \xi \right| \right ) \right ] \left ( 1-\frac{b_{n}}{% a_{n}^{2}}\zeta \right ) ^{-1}\right .
\right .
\nonumber \\ & & + a_{l}^{n}\frac{n\zeta } { 2a_{n}}\left [ a_{n}^{2}\frac{4\eta ^{4}\left ( n^{2}-3n+3\right ) } { 9n\left ( n-1\right ) ^{3}}\left ( 1-\frac{b_{n}}{a_{n}^{2}}% \zeta \right ) \right ] \nonumber
\\ & & + a_{l}^{n}\frac{n\zeta } { 2a_{n}}\left [ a_{n}\frac{\eta ^{2}\left ( n-2\right ) } { n\left ( n-1\right ) ^{2}}+\frac{3\left ( n-2\right ) ^{2}}{4\left ( n-1\right ) } \left ( 1-\frac{b_{n}}{a_{n}^{2}}\zeta \right ) ^{-1}\right ] \nonumber \\ & & + \frac{2n\zeta } { a_{n}}\left ( \frac{n}{2\left ( n-1\right ) } \left ( 1-\frac{% b_{n}}{a_{n}^{2}}\zeta \right ) -\frac{2a_{n}\eta ^{2}}{3n}\left [ \frac{n^{2}% } { \left ( n-1\right ) ^{2}}\right ] \right ) \nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } \times \left [ c_{l}^{n}-\frac{1}{l}a_{l}^{n}\left ( 1-\frac{4a_{n}}{3n}\left ( 1-\frac{b_{n}% } { a_{n}^{2}}\zeta \right ) l\left| \xi \right| \right ) \right ] \nonumber \\ & & \left .
-\frac{n\zeta } { a_{n}}\left [ d_{l}^{n}+k_{l}^{n}\left ( 1-\frac{% 4a_{n}}{3n}\left ( 1-\frac{b_{n}}{a_{n}^{2}}\zeta \right ) l\left| \xi \right| \right ) \right ] \left ( 1-\frac{b_{n}}{a_{n}^{2}}\zeta \right ) ^{-1}\right\ } \nonumber \\ & & \left .
\mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } \times \exp \left ( -\left ( 1-\frac{b_{n}}{a_{n}^{2}}\zeta \right ) \left ( \eta ^{2}+\frac{4a_{n}% } { 3n}l\left| \xi \right| \right ) \right ) \right ] \label{ln7}\end{aligned}\]]which is valid only to first order in @xmath210 .
the canonical density ( [ can35 ] ) becomes @xmath211 \nonumber \\ & & \times \sum_{l=1}^{n-1}\left\ { a_{l}^{n}\left ( 1-\frac{b_{n}}{a_{n}^{2}}% \zeta \right ) + \frac{n\zeta } { a_{n}}\left ( \frac{3}{8}\frac{\left ( n-1\right ) ^{2}}{n}a_{l}^{n}-b_{l}^{n}-d_{l}^{n}\right ) \right\ } \label{mirho } \\ & & \left .
+ \frac{n\zeta } { a_{n}}\left [ a_{l}^{n}-k_{l}^{n}\right ] \left ( 1-% \frac{4a_{n}}{3n}\left ( 1-\frac{b_{n}}{a_{n}^{2}}\zeta \right ) l\left| \xi \right| \right ) \right\ } \exp \left ( -2\left ( 1 - 1/n\right ) \left ( 1-\frac{% b_{n}}{a_{n}^{2}}\zeta \right ) l\left| \xi \right| \right ) \nonumber\end{aligned}\]]we plot in fig .
[ fig3 ] the non - relativistic canonical single - particle density function @xmath212 for various values of @xmath213 , recovering the results of rybicki@xcite . with the exception of @xmath214 ,
the central density grows with increasing @xmath0 and the distribution becomes slightly more sharply peaked . it can be shown that as @xmath215 the canonical density @xmath216h@xmath217 , and the single particle distribution function approaches the isothermal solution of the vlasov equation @xcite . in figs .
[ fig4 ] [ fig7 ] we plot the relativistic canonical function @xmath218 for differing values of @xmath0 .
as is clear from these figures , relativistic effects significantly enhance the central density by as much as 30% depending on the magnitude of @xmath210 . even for @xmath219 ,
the central density is larger than its value of @xmath220 in the non - relativistic large @xmath0 limit .
the falloff of the relativistic density functions is also more rapid than in the non - relativistic case .
unfortunately there is no closed - form expression for the terms @xmath221 and @xmath222 and so it is not possible to evaluate an explicit expression for either @xmath223 or @xmath224 in the large @xmath0 limit . instead
these quantities must be computed using symbolic algebra ; for @xmath225 this involve the factorization of thousands of terms , and computer memory limitations make this a prohibitive task
. however the large @xmath0 behaviour should not be too different from the @xmath226 case , at least for small values of @xmath210 .
figures [ fig8 ] and [ fig9 ] plot the density for different @xmath0 at two different values of @xmath210 .
the canonical momentum distribution function ( [ can36 ] ) is straightforwardly evaluated to be @xmath227 \left ( 1+\frac{n\zeta } { a_{n}}\frac{\left ( n^{2}-3n+3\right ) } { % 2n\left ( n-1\right ) } \left ( 1-\frac{b_{n}}{a_{n}^{2}}\zeta \right ) \eta ^{4}\right .
\nonumber \\ & & \times \left .
\frac{n\zeta } { a_{n}}\left ( -\frac{\eta ^{2}\left ( 4n^{2}-7n+6\right ) } { 2n\left ( n-1\right ) } + \frac{5n(n-1)+3}{8n\left ( n-1\right ) } \left ( 1-\frac{b_{n}}{a_{n}^{2}}\zeta \right ) ^{-1}\right ) \right ) \label{mitheta}\end{aligned}\]]where each of ( [ mirho],[mitheta ] ) are also valid to first order in @xmath228 . here
we can easily take the large @xmath0 limit , which is @xmath229to leading order in @xmath210 .
we plot in figs .
[ fig10 ] [ fig12 ] the behaviour of the canonical momentum density as a function of the rescaled momentum @xmath230 for differing values of @xmath231 the central momentum density increases with increasing @xmath210 , and falls off more rapidly than in the non - relativistic case .
however for @xmath232 , the momentum density grows relative to its non - relativistic counterpart , overtaking this value for large enough @xmath230 .
the relative growth is exponential , although the overall momentum density is exponentially damped for any @xmath210 . as @xmath0 increases , the differences between the non - relativistic and relativistic cases become less pronounced , although the basic features remain the same even in the limit that @xmath197 .
the microcanonical results are @xmath233 \sum_{l=1}^{n-1}\left\ { a_{l}^{n}\left ( 1-\frac{2\eta ^{2}}{3\left ( n-1\right ) } -\frac{4l\left| \xi \right| } { 3n}\right ) _ { + } ^{3n/2 - 4}\right .
\nonumber \\ & & + \zeta \eta ^{2}\left ( a_{l}^{n}\left [ \frac{\left ( n-2\right ) } { \left ( n-1\right ) ^{2}}\right ] -\left ( c_{l}^{n}-\frac{1}{l}a_{l}^{n}\right ) \left [ \frac{4n^{2}}{3\left ( n-1\right ) ^{2}}\right ]
\right ) \left ( 1-\frac{2\eta ^{2}}{3\left ( n-1\right ) } -\frac{4l\left| \xi \right| } { 3n}\right ) _ { + } ^{3n/2 - 4 } \nonumber \\ & & + \left ( \frac{3}{2}n-4\right ) \zeta a_{l}^{n}\left [ \frac{2\eta ^{4}\left ( n^{2}-3n+3\right ) } { 9\left ( n-1\right ) ^{3}}-\frac{16n\left| \xi \right| \eta ^{2}}{9\left ( n-1\right ) ^{2}}\right ] \left ( 1-\frac{2\eta ^{2}}{% 3\left ( n-1\right ) } -\frac{4l\left| \xi \right| } { 3n}\right ) _ { + } ^{3n/2 - 5 } \nonumber \\ & & + \left [ \frac{4\zeta } { 3}\left| \xi \right| \left
( a_{l}^{n}\left ( \frac{n% } { \left ( n-1\right ) } -l\right ) + lk_{l}^{n}\right ) \right ] \left ( 1-\frac{% 2\eta ^{2}}{3\left ( n-1\right ) } -\frac{4l\left| \xi \right| } { 3n}\right ) _ { + } ^{3n/2 - 4 } \nonumber \\ & & + \frac{n\zeta } { \left ( \frac{3}{2}n-3\right ) } \left ( a_{l}^{n}\left [ \frac{% 3\left ( n-2\right ) ^{2}}{8\left ( n-1\right ) } + 1\right ] -b_{l}^{n}\mbox{\ } + % \frac{n\left ( c_{l}^{n}-\frac{1}{l}a_{l}^{n}\right ) } { \left ( n-1\right ) } % -d_{l}^{n}-k_{l}^{n}\right ) \nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } \left .
\times \left ( 1-\frac{2\eta ^{2}}{3\left ( n-1\right ) } -\frac{% 4l\left| \xi \right| } { 3n}\right ) _ { + } ^{3n/2 - 3}\right\ } \label{ln10}\end{aligned}\ ] ] @xmath234 \sum_{l=1}^{n-1}\left\ { \left ( \frac{3n-5}{3n}\right ) \left [ a_{l}^{n}-\frac{4l\left| \xi \right| } { 3% } \zeta \left ( a_{l}^{n}-k_{l}^{n}\right ) \right ] \left ( 1-\frac{4l\left| \xi \right| } { 3n}\right ) _ { + } ^{3n/2 - 7/2}\right .
\nonumber \\ & & \left .
+ \frac{2}{3}\zeta \left ( 1-\frac{4l\left| \xi \right| } { 3n}\right ) _ { + } ^{3n/2 - 5/2}\left ( \left ( \frac{3(n-1)^{2}}{8n}\right ) a_{l}^{n}-\left ( b_{l}^{n}-a_{l}^{n}\right ) -\left ( d_{l}^{n}+k_{l}^{n}\right )
\right ) \right\ } \mbox { \ \ \ \ \ \ \ \ \ \ \ \ } \label{ln11}\end{aligned}\ ] ] and [ thetmic ] @xmath235 \nonumber \\ & & \times \left\ { \left [ 1-\left ( \frac{\eta ^{2}}{3}\left [ \frac{8n^{2}-11n+6% } { \left ( n-1\right ) ^{2}}\right ] \right ) \zeta \right ] \left ( 1-\frac{2\eta ^{2}}{3\left ( n-1\right ) } \right ) _ { + } ^{3n/2 - 3}\right .
\nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ } + \frac{\zeta \left ( n-2\right ) } { 3}\left [ \frac{\eta ^{4}\left ( n^{2}-3n+3\right ) } { \left ( n-1\right ) ^{3}}\right ] \left ( 1-\frac{2\eta ^{2}}{3\left ( n-1\right ) } \right ) _ { + } ^{3n/2 - 4 } \label{ln12 } \\ & & \mbox{\ } \left .
-\frac{2n\zeta } { \left ( 3n-4\right ) } \left ( \frac{\left ( 5n^{2}-20n+12\right ) } { 8\left ( n-1\right ) } \mbox{\ } + \sum_{s=1}^{n-1}% \sum_{t = s+1}^{n-1}\frac{\left ( t - s\right ) } { t\left ( n - s\right ) } \right ) \left ( 1-\frac{2\eta ^{2}}{3\left ( n-1\right ) } \right ) _ { + } ^{3n/2 - 2}\right\ } \nonumber\end{aligned}\]]where each of ( [ ln10],[ln11],[ln12 ] ) are valid to first order in @xmath210 . as @xmath215
we have @xmath236note that this differs from the canonical momentum density unless @xmath237 the microcanonical density function @xmath238 for differing values of @xmath0 is plotted in figs [ fig13][fig16 ] . for @xmath239 ,
the microcanonical density is uniform until @xmath240 , after which it falls linearly to zero .
for @xmath241 , at most two particles can contribute to the density ; in this region relativistic effects enhance their contribution .
however for @xmath242 , at most one particle can contribute , and relativistic effects suppress its contribution until @xmath243 , after which the density vanishes . for larger @xmath0 , as in the canonical case , relativistic effects significantly enhance the central density by as much as 30% , depending on the size of @xmath210 .
their falloff is more rapid , and for sufficiently large @xmath244 the non - relativistic density distribution is larger . as @xmath0 increases
, the density becomes more sharply peaked and the contrasts between the non - relativistic and relativistic cases become less pronounced . in figs .
[ fig17 ] , [ fig18 ] we plot the microcanonical density distribution for increasing values of @xmath245 and fixed @xmath210 .
we plot the microcanonical momentum distributions in figs .
[ fig19]fig21 .
the results are qualitatively similar to the canonical case , although the actual functional forms differ . for small @xmath0 ,
the relativistic approximation breaks down even for @xmath210 as small as @xmath246 , and the momentum distribution function goes negative , as shown in fig.[fig19 ] . however for larger @xmath0
the momentum distribution is positive for all @xmath210 @xmath247 , for example as in fig .
[ fig20 ] .
the relativistic densities are more sharply peaked and for sufficiently large momentum parameter @xmath230 are larger than their non - relativistic counterparts .
we have carried out the first analysis of the statistical behaviour of a rogs to leading order in @xmath1 .
the qualitative behaviour of the rogs as compared to its non - relativistic ogs counterpart @xcite is clear . at a given energy ,
the rogs temperature is smaller than the ogs temperature ; relativistic effects cool the gas down .
the one - particle distribution functions become more sharply peaked in each case with increasing @xmath0 . for a given @xmath0 ,
the rogs density functions become more sharply peaked as the relativistic parameter @xmath210 increases . for both canonical and microcanonical distribution functions , @xmath248 for sufficiently large position parameter @xmath244 .
however for the momentum densities , this is not true .
although @xmath249 for intermediate values of the momentum parameter @xmath230 , once @xmath230 becomes large enough this inequality is reversed .
this behaviour is presumably due to quadratic character of the rogs potential relative to its linear ogs counterpart in the former case , and from the @xmath250 corrections in the hamiltonian ( [ can7 ] ) in the latter situation .
this work can be extended in several directions .
it would be straightforward to extend these results to the charged and cosmological systems considered in refs .
@xcite to see what effects these impose on the distribution functions .
extensions to unequal masses would also be interesting , although considerably more difficult .
it would be very interesting to go beyond leading order in @xmath1 to investigate non - perturbative effects of the rogs .
further understanding the rogs will undoubtedly require numerical experiments for various values of @xmath0 .
the equations of motion yield quartic ( as opposed to quadratic ) time - dependence of the position variables ( to leading order in @xmath1 ) , and so can be straightforwardly ( although somewhat tediously ) integrated to investigate its equilibrium and equipartion properties .
work on this is in progress .
acknowledgements this work was supported by the natural sciences and engineering research council of canada .
we would like to thank t. ohta for interesting discussion and correspondence .
here we must compute @xmath259 \right ) \label{appl4a}\ ] ] of which the relevant part is @xmath260 and so @xmath261 \right ) \nonumber \\ & = & \left ( \frac{2\pi m}{\beta } \right ) ^{\left ( n-2\right ) /2}\frac{1}{\sqrt{% n-1}}\exp \left ( -\frac{\beta p^{2}n}{2m\left ( n-1\right ) } \right ) \left ( 1-% \frac{4\beta \lambda _ { 4}m^{3}\left ( \pi g\right ) ^{2}}{c^{2}}% \sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}\left ( n - l\right ) \left ( l - k\right ) ku_{l}u_{k}\right ) \label{appl4c}\end{aligned}\ ] ] where ( [ can18 ] ) was used .
since @xmath270 , this becomes @xmath271 -\frac{\left ( n-2\right ) } { 2\left ( n-1\right ) } \right ) \right\ } \nonumber \\ & = & -\frac{\mbox{2}\lambda _ { 2}\pi gm}{\sqrt{n-1}c^{2}}\left ( \frac{2\pi m}{% \beta } \right ) ^{\left ( n-2\right ) /2}\exp \left ( -\frac{\beta p^{2}n}{% 2m\left ( n-1\right ) } \right ) \left ( 2\sum_{l=1}^{n-1}l\left ( n - l\right ) u_{l}\left\ { \frac{\beta p^{2}}{2m\left ( n-1\right ) ^{2}}+\frac{\left ( n-2\right ) } { 2\left ( n-1\right ) } \right\ } \right .
\nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } + \left .
\left ( \sum_{s=1}^{n-1}su_{s}+\sum_{s = n}^{n-1}(n - s)u_{s}\right ) \left ( + \frac{\beta p^{2}}{2m}\left [ \frac{n\left ( n-2\right ) } { \left ( n-1\right ) ^{2}}\right ] -% \frac{\left ( n-2\right ) } { 2\left ( n-1\right ) } \right ) \right ) \label{appl2d}\end{aligned}\ ] ] now we must do @xmath273where the integrations are @xmath274giving @xmath275using again @xmath270 , this becomes @xmath276 \right ) \right\ } \nonumber \\ & = & \frac{\mbox{2}\lambda _
{ 3}\pi gm}{\sqrt{n-1}c^{2}}\left ( \frac{2\pi m}{% \beta } \right ) ^{\left ( n-2\right ) /2}\exp
\left ( -\frac{\beta p^{2}n}{% 2m\left ( n-1\right ) } \right ) \left ( 2\sum_{l=1}^{n-1}l\left ( n - l\right ) u_{l}\left ( \frac{\beta p^{2}}{2m\left ( n-1\right ) ^{2}}-\frac{1}{2\left ( n-1\right ) } \right ) \right .
\nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } \left .
+ \left ( \sum_{s=1}^{n-1}su_{s}+\sum_{s = n}^{n-1}(n - s)u_{s}\right ) \left ( \frac{1}{% \left ( n-1\right ) } -\frac{\beta p^{2}}{2m}\left [ \frac{2n}{\left ( n-1\right ) ^{2}}\right ] \right ) \right ) \nonumber \\ & & \label{appl3d}\end{aligned}\ ] ] from eqs .
( [ appl4c],[appl1c],[appl2d],[appl3d ] ) we have @xmath278 \right ) \right ] \nonumber \\ & & \mbox { \ \ } + \frac{\mbox{2}\lambda _ { 3}\pi gm}{c^{2}}\left [ 2\sum_{l=1}^{n-1}l\left ( n - l\right ) u_{l}\left ( \frac{\beta p^{2}}{2m\left ( n-1\right ) ^{2}}-\frac{1}{2\left ( n-1\right ) } \right ) \right .
\nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } \left .
+ \left ( \sum_{s=1}^{n-1}su_{s}+\sum_{s = n}^{n-1}(n - s)u_{s}\right ) \left ( \frac{1}{% \left ( n-1\right ) } -\frac{\beta p^{2}}{2m}\left [ \frac{2n}{\left ( n-1\right ) ^{2}}\right ] \right ) \right ] \nonumber \\ & & \mbox { \ \ \ } \left .
-\frac{4\beta m^{3}}{c^{2}}\lambda _ { 4}\left ( \pi g\right ) ^{2}\sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}\left ( n - l\right ) \left ( l - k\right ) ku_{l}u_{k}\right ) \label{appfula}\end{aligned}\ ] ] or alternatively @xmath279 \label{appfulb}\end{aligned}\ ] ] when the @xmath280 s are all equal to unity , this is @xmath281 \nonumber \\ & & + \left ( \frac{\mbox{2}\pi mg}{c^{2}}\right ) \left [ -\sum_{l=1}^{n-1}l% \left ( n - l\right ) u_{l}-\left ( \frac{n^{2}\beta p^{2}}{2m\left ( n-1\right ) ^{2}}-\frac{n}{2(n-1)}\right ) \left ( \sum_{s=1}^{n-1}su_{s}+\sum_{s = n}^{n-1}(n - s)u_{s}\right ) \right ] \nonumber\end{aligned}\ ] ] now consider eq .
( [ can24b ] ) , which can be rewritten as @xmath283where @xmath284the integrals are now @xmath285and we must divide by @xmath0 and sum over all values of @xmath286 to obtain the correct result . for example @xmath287
\nonumber \\ & = & \frac{1}{n\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n-1}}\sum_{n=1}^{n}% \frac{1}{\left ( n-1\right ) !
( n - n)!}\frac{\gamma \left ( n - n+1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( n+i\alpha \right ) } { % \gamma \left ( n+i\alpha \right ) } \nonumber \\ & = & \frac{1}{\gamma \left ( n - i\alpha \right ) \gamma \left ( n+i\alpha \right ) \left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n-1}}\sum_{n=1}^{n}\frac{\left ( n-1\right ) ! } { \left ( n-1\right ) ! ( n - n)!}\int_{0}^{1}dww^{n - n - i\alpha } \left ( 1-w\right ) ^{n+i\alpha -1 } \nonumber \\ & = & \frac{1}{\gamma \left ( n - i\alpha \right ) \gamma \left ( n+i\alpha \right ) \left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n-1}}\int_{0}^{1}dww^{-i\alpha } \left ( 1-w\right ) ^{i\alpha } \left ( w+1-w\right ) ^{n-1 } \nonumber \\ & = & \frac{1}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n-1}}\frac{\gamma \left ( 1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( 1+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \label{appfz6}\end{aligned}\]]this function has single poles at @xmath288 , where @xmath286 is an integer taking its values between @xmath289 and @xmath0 , except for @xmath290 .
its residues are @xmath291 _ { a = in}=\frac{in\left ( -1\right ) ^{n}}{\gamma \left ( n - n\right ) \gamma \left ( n+n\right ) } \label{appfz7}\]]yielding @xmath292 ^{2}}\sum_{l=1}^{n-1}a_{l}^{n}\exp \left ( -2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \label{appfz8}\]]for the leading non - relativistic term , where @xmath293 ^{2}}{% \gamma \left ( n - n\right ) \gamma \left ( n+n\right ) } \label{appfza}\ ] ] the remaining integrals are somewhat more difficult .
we obtain @xmath294 _ { \sigma = 1 } \nonumber \\ & = & -\frac{1}{\lambda \beta n}\frac{d}{d\sigma } \left [ \sum_{n=1}^{n}\frac{1}{% \left ( \lambda \beta \sigma \right ) ^{n-1}}\prod_{l=1}^{n-1}\frac{1}{c_{l}+i% \frac{\alpha } { \sigma } d_{nl}}\right ] _ { \sigma = 1 } \nonumber \\ & = & -\frac{d}{d\sigma } \left [ \frac{\sigma ^{1-n}}{\left ( \lambda \beta \right ) ^{n}}\frac{\gamma \left ( 1-i\alpha /\sigma \right ) } { \gamma \left ( n - i\alpha /\sigma
\right ) } \frac{\gamma \left ( 1+i\alpha /\sigma \right ) } { % \gamma \left ( n+i\alpha /\sigma \right ) } \right ] _ { \sigma = 1 } \nonumber \\ & = & \frac{1}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}}\frac{\gamma \left ( 1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( 1+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \nonumber \\ & & \times \left [ n-1+i\alpha \left ( \psi \left ( n - i\alpha \right ) -\psi \left ( 1-i\alpha \right ) -\psi \left ( n+i\alpha \right ) + \psi \left ( 1+i\alpha \right ) \right ) \right ] \label{appfz9}\end{aligned}\]]where@xmath295 is the digamma function .
this function on the right - hand side of ( [ appfz9 ] ) has a combination of single and double poles , each located at @xmath288 , where @xmath296 is an integer taking its values between @xmath289 and @xmath0 , except for @xmath290 .
writing @xmath297we have the last line of ( [ appfz9 ] ) proportional to @xmath298 \\ & = & \frac{\gamma \left ( 1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } % \frac{\gamma \left ( 1+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } % \left [ n-3 - 2\alpha ^{2}\sum_{s\neq n}^{n-1}\left ( \frac{1}{s^{2}+\alpha ^{2}}% \right ) + \left\ { \frac{2n^{2}}{\alpha ^{2}+n^{2}}\right\ } \right ] \end{aligned}\]]where the term in curly brackets contains the double - pole at @xmath288 . hence @xmath299
\nonumber\end{aligned}\]]the residue from the double pole at @xmath288 is @xmath300 _ { x = n } \nonumber \\ & = & \frac{-in\left ( -1\right ) ^{n}\exp ( n\beta \lambda nz)}{2\gamma \left ( n - n\right ) \gamma \left ( n+n\right ) } \left [ 1 + 2n\beta \lambda nz+2n\left ( \psi \left ( n - n\right ) -\psi \left ( n+n\right ) \right ) \right ] \label{appfz12}\end{aligned}\]]provided @xmath301 , and so the total residue at @xmath288 from ( [ appfz11 ] ) is @xmath302 \nonumber \\ & \equiv & \frac{-i\exp ( n\beta \lambda nz)}{\left [ \gamma ( n)\right ] ^{2}}% \left ( b_{n}^{n}-a_{n}^{n}\left ( 1+n\beta \lambda nz\right ) \right ) \label{appfz13}\end{aligned}\]]giving finally @xmath303 ^{2}}\sum_{l=1}^{n-1}\left [ b_{l}^{n}-a_{l}^{n}% \left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \exp \left ( -2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \label{appfz14}\end{aligned}\]]where @xmath304 ^{2}}{% \gamma \left ( n - n\right ) \gamma \left ( n+n\right ) } \left [ n-\frac{5}{2}% + 2n^{2}\sum_{s\neq n}^{n-1}\left ( \frac{1}{s^{2}-n^{2}}\right ) + n\left ( \psi \left ( n+n\right ) -\psi \left ( n - n\right ) \right ) \right ] \label{appfz15}\ ] ] using the relation @xmath305we find @xmath306 \left ( \prod_{l=1}^{n-1}\frac{1}{% c_{l}+i\alpha d_{nl}}\right ) \nonumber \\ & = & \frac{1}{n\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}}% \sum_{n=1}^{n}\left\ { \left [ \psi \left ( 1-n+i\alpha \right ) -\psi \left ( n - n+i\alpha \right ) + \psi \left ( n+i\alpha \right ) -\psi \left ( n+i\alpha \right ) \right ] \right .
\nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } % \left .
\times \left ( \frac{1}{\left ( n-1\right ) !
( n - n)!}\frac{\gamma \left ( n - n+1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( n+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \right ) \right\ } \nonumber \\ & = & \frac{1}{n\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}}% \sum_{n=1}^{n}\left\ { \left [ \psi \left ( n - i\alpha \right ) + \psi \left ( n+i\alpha \right ) -\psi \left ( n - n+1-i\alpha \right ) -\psi \left ( n+i\alpha \right ) \right ] \right .
\nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } % \left .
\times \left ( \frac{1}{\left ( n-1\right ) !
( n - n)!}\frac{\gamma \left ( n - n+1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( n+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \right ) \right\ } \nonumber \\ & = & \frac{1}{n\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}}\sum_{n=1}^{n}% \left [ \left [ \psi \left ( n - i\alpha \right ) + \psi \left ( n+i\alpha \right ) % \right ] \left ( \frac{1}{\left ( n-1\right ) ! ( n - n)!}\frac{\gamma \left ( n - n+1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( n+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \right ) \right .
\nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } \left . -\frac{d}{d\sigma } \left ( \frac{1}{\left ( n-1\right ) !
( n - n)!}\frac{% \gamma \left ( n - n+1-i\alpha + \sigma \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( n+i\alpha + \sigma \right ) } { \gamma \left ( n+i\alpha \right ) } \right ) \right ] _ { \sigma = 0 } \nonumber \\ & = & \frac{1}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}}\left [ \psi \left ( n - i\alpha \right ) + \psi \left ( n+i\alpha \right ) \right ] \frac{\gamma \left ( 1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( 1+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ } -% \left [ \frac{1}{\gamma \left ( n+1\right ) } \frac{d}{d\sigma } \left ( \frac{% \gamma \left ( n+1 + 2\sigma \right ) \gamma \left ( 1-i\alpha + \sigma \right ) \gamma \left ( 1+i\alpha + \sigma \right ) } { \gamma \left ( 2 + 2\sigma \right ) \gamma \left ( n - i\alpha \right ) \gamma \left ( n+i\alpha \right ) } \right ) % \right ] _ { \sigma = 0 } \nonumber \\ & = & \frac{1}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}}\frac{\gamma \left ( 1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( 1+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \nonumber \\ & & \times \left [ \psi \left ( n - i\alpha \right ) + \psi \left ( n+i\alpha \right ) -\psi \left ( 1-i\alpha \right ) -\psi \left ( 1+i\alpha \right ) -2\psi \left ( n+1\right ) + 2\psi \left ( 2\right ) \right ] \nonumber \\ & = & \frac{1}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}}\frac{\gamma \left ( 1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( 1+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \left [ \sum_{s=1}^{n-1}% \frac{1}{s - i\alpha } + \sum_{s=1}^{n-1}\frac{1}{s+i\alpha } -2\sum_{s=1}^{n-1}% \frac{1}{s+1}\right ] \nonumber \\ & = & \frac{1}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}}\frac{\gamma \left ( 1-i\alpha \right ) } { \gamma \left ( n - i\alpha \right ) } \frac{\gamma \left ( 1+i\alpha \right ) } { \gamma \left ( n+i\alpha \right ) } \left [ \sum_{s\neq n}^{n-1}\frac{1}{s - i\alpha } + \sum_{s\neq n}^{n-1}\frac{1}{s+i\alpha } % -2\sum_{s=1}^{n-1}\frac{1}{s+1}+\left\ { \frac{2n}{\alpha ^{2}+n^{2}}\right\ } % \right ] \label{appfz17}\end{aligned}\]]where the curly bracket contains the double - pole at @xmath288 , and all other poles are in the same locations as before .
so we obtain @xmath307 \nonumber\end{aligned}\]]the total residue at @xmath288 from ( [ appfz18 ] ) is @xmath308 \nonumber \\ & = & \frac{-i\exp ( n\beta \lambda nz)}{\left [ \gamma ( n)\right ] ^{2}}\left ( c_{n}^{n}-\frac{1}{n}a_{n}^{n}\left ( 1+n\beta \lambda nz\right ) \right ) \label{appfz19}\end{aligned}\]]provided @xmath301 ( otherwise it vanishes ) and so @xmath309 ^{2}}\sum_{l=1}^{n-1}\left [ c_{l}^{n}-\frac{1}{l}% a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) % \right ] \exp \left ( -2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \mbox { \ \ \ \ \ \ \ \ \ \ \ }
\label{appfz20}\end{aligned}\]]where @xmath310 ^{2}}{% \gamma \left ( n - n\right ) \gamma \left ( n+n\right ) } \left [ 2\sum_{s\neq n}^{n-1}\left ( \frac{s}{s^{2}-n^{2}}\right ) -2\sum_{s=1}^{n-1}\frac{1}{s+1}+% \frac{1}{2n}+\left ( \psi \left ( n+n\right ) -\psi \left ( n - n\right ) \right ) % \right ] \label{appfz21}\ ] ] finally , we consider the expression @xmath311interchanging the order of the sums gives @xmath312 \label{appfz23}\end{aligned}\]]when summed over @xmath286 , we obtain @xmath313which has double poles and single - pole residues at @xmath314 for every nonzero value of @xmath315 .
the coefficients of the polynomial in the numerator are calculable , but we have not found any closed - form expression for them .
the table below contains results for values up to @xmath199 .
@xmath316 ^{2}}\sum_{l=1}^{n-1}a_{l}^{n}\exp \left ( -2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \label{appfz26 } \\ & & \frac{1}{n}\sum_{n=1}^{n}\int \frac{dk}{2\pi } \int d{\bf u}\left ( \sum_{s=1}^{n-1}s\left ( n - s\right ) u_{s}\right ) { \bf \;}e^{\left ( -ikz-\lambda \beta \sum_{l=1}^{n-1}\left ( c_{l}+i\alpha d_{nl}\right ) u_{l}\right ) } \nonumber \\
\mbox { \ \ \ \ } & = & \frac{2n\beta \pi gm^{2}}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}\left [ \gamma ( n)\right ] ^{2}}\sum_{l=1}^{n-1}\left [ b_{l}^{n}-a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \exp \left ( -2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \label{appfz27 } \\ & & \frac{1}{n}\sum_{n=1}^{n}\int \frac{dk}{2\pi } \int d{\bf u}\left ( \sum_{s=1}^{n-1}si_{n , s}^{1}+\sum_{s = n}^{n-1}(n - s)i_{n , s}^{1}\right ) { \bf \;}% e^{\left ( -ikz-\lambda \beta \sum_{l=1}^{n-1}\left ( c_{l}+i\alpha d_{nl}\right ) u_{l}\right ) } \nonumber \\ \mbox { \ \ \ \ } & = & \frac{2n\beta \pi gm^{2}}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n}\left [ \gamma ( n)\right ] ^{2}}\sum_{l=1}^{n-1}\left [ c_{l}^{n}-% \frac{1}{l}a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \exp \left ( -2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \mbox { \ \ \ \ } \label{appfz28 } \\ & & \frac{1}{n}\sum_{n=1}^{n}\int \frac{dk}{2\pi } \int d{\bf u}\left ( \sum_{s=1}^{n-1}\sum_{t = s+1}^{n-1}\left ( n - t\right ) \left ( t - s\right ) su_{s}u_{t}\right ) { \bf \;}e^{\left ( -ikz-\lambda \beta \sum_{l=1}^{n-1}\left ( c_{l}+i\alpha d_{nl}\right ) u_{l}\right ) } \nonumber \\ \mbox { \ \ \ \ } & = & \frac{2n\beta \pi gm^{2}}{\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n+1}\left [ \gamma ( n)\right ] ^{2}}\sum_{l=1}^{n-1}\left [ d_{l}^{n}+k_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \exp \left ( -2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \mbox { \ \ \ \ \ } \label{appfz29}\end{aligned}\ ] ] the final expression for the one - particle distribution function is @xmath317 ^{2}\left ( \frac{2\pi m}{\beta } \right ) ^{\left ( n-2\right ) /2}\frac{2n\beta \pi gm^{2}\left ( \beta mc^{2}\right ) ^{\frac{% 3\left ( n-1\right ) } { 2}}}{\sqrt{n-1}\left ( 2\pi g\beta m^{2}\mbox { } \right ) ^{n-1}\left [ \gamma ( n)\right ] ^{2 } } \nonumber \\ & & \times \exp \left [ -\frac{1}{\beta mc^{2}}\left\ { \frac{3\left ( n-1\right ) ^{2}}{8n}\lambda _ { 1}-\frac{\left ( n-1\right ) \left ( \lambda _ { 2}\left ( n-1\right ) + \lambda _ { 3}\right ) } { n}-\lambda _ { 4}\sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}\frac{\left ( l - k\right ) } { l\left ( n - k\right ) } \right\ } \right ] \nonumber \\ & & \times \sum_{l=1}^{n-1}\left [ \left\ { a_{l}^{n}\left ( 1+\frac{\lambda _ { 1}% } { 2\beta mc^{2}}\left ( \frac{\beta ^{2}p^{4}\left ( 1+\left ( n-1\right ) ^{3}\right ) } { \left ( 2m\right ) ^{2}\left ( n-1\right ) ^{3}}+\frac{3\left ( n-2\right ) \beta p^{2}}{2m\left ( n-1\right ) ^{2}}+\frac{3\left ( n-2\right ) ^{2}}{4\left ( n-1\right ) } \right ) \right ) \right . \right .
\nonumber \\ & & \mbox { \ \ \ \ } + \frac{1}{\beta mc^{2}}\left ( \frac{\left ( \lambda _ { 3}-\lambda _ { 2}\right ) \beta p^{2}}{m\left ( n-1\right ) ^{2}}-\frac{\left ( \lambda _ { 3}+(n-2)\lambda _ { 2}\right ) } { \left ( n-1\right ) } \right ) \left [ b_{l}^{n}-a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \nonumber \\ & & + \frac{2\left ( 2\lambda _ { 3}+(n-2)\lambda _ { 2}\right ) } { \beta mc^{2}}% \left ( \frac{1}{2\left ( n-1\right ) } -\frac{n\beta p^{2}}{2m}\left [ \frac{1}{% \left ( n-1\right ) ^{2}}\right ] \right ) \mbox{\ } \left [ c_{l}^{n}-\frac{1}{l}% a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right|
\right ) % \right ] \nonumber \\ & & \mbox { \ \ \ \ \ \ \ \ \ \ \ \ } \left .
-\frac{\lambda _ { 4}}{\beta mc^{2}}\left [ d_{l}^{n}+k_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \right\ } \exp \left ( -\frac{n\beta p^{2}}{% 2m\left ( n-1\right ) } -2\pi gn\beta m^{2}l\left| z\right| \right ) \right ] \nonumber \\ & = & \frac{\left ( 2\pi gm^{2}\right ) \left ( n\beta \right ) ^{3/2}}{\sqrt{2\pi m\left ( n-1\right ) } } \nonumber \\ & & \times \exp \left [ -\frac{1}{\beta mc^{2}}\left\ { \frac{3\left ( n-1\right ) ^{2}}{8n}\lambda _ { 1}-\frac{\left ( n-1\right ) \left ( \lambda _ { 2}\left ( n-1\right ) + \lambda _ { 3}\right ) } { n}-\lambda _ { 4}\sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}\frac{\left ( l - k\right ) } { l\left ( n - k\right ) } \right\ } \right ] \nonumber \\ & & \mbox { \ } \times \sum_{l=1}^{n-1}\left [ \left\ { a_{l}^{n}\left ( 1+\frac{% \lambda _ { 1}}{2\beta mc^{2}}\left ( \frac{\beta ^{2}p^{4}\left ( 1+\left ( n-1\right ) ^{3}\right ) } { \left ( 2m\right ) ^{2}\left ( n-1\right ) ^{3}}+\frac{% 3\left ( n-2\right ) \beta p^{2}}{2m\left ( n-1\right ) ^{2}}+\frac{3\left ( n-2\right ) ^{2}}{4\left ( n-1\right ) } \right ) \right ) \right . \right .
\nonumber \\ & & \mbox{\ } + \frac{1}{\beta mc^{2}}\left ( \frac{\left ( \lambda _ { 3}-\lambda _ { 2}\right ) \beta p^{2}}{m\left ( n-1\right ) ^{2}}-\frac{\left ( \lambda _ { 3}+(n-2)\lambda _ { 2}\right ) } { \left ( n-1\right ) } \right ) \left [ b_{l}^{n}-a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \nonumber \\ & & + \frac{2\left ( 2\lambda _ { 3}+(n-2)\lambda _ { 2}\right ) } { \beta mc^{2}}% \left ( \frac{1}{2\left ( n-1\right ) } -\frac{n\beta p^{2}}{2m}\left [ \frac{1}{% \left ( n-1\right ) ^{2}}\right ] \right ) \mbox{\ } \left [ c_{l}^{n}-\frac{1}{l}% a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) % \right ] \nonumber \\ & & \mbox { \ \ } \left .
{ 4}}{\beta mc^{2}}\left [ d_{l}^{n}+k_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \right\ } \exp \left ( -\frac{n\beta p^{2}}{2m\left ( n-1\right ) } -2\pi gn\beta m^{2}l\left| z\right| \right ) \right ] \nonumber \\ & & \label{appfz30}\end{aligned}\ ] ] the canonical density distribution function is given by integration of @xmath318 over @xmath144 @xmath319 \nonumber \\ & & \mbox { \ } \times \sum_{l=1}^{n-1}\left [ \left\ { a_{l}^{n}\left ( 1+\frac{% \lambda _
{ 1}}{2\beta mc^{2}}\left ( \frac{3\left ( 1+\left ( n-1\right ) ^{3}\right ) } { 4n^{2}\left ( n-1\right ) } + \frac{3\left ( n-2\right ) } { 2n\left ( n-1\right ) } + \frac{3\left ( n-2\right ) ^{2}}{4\left ( n-1\right ) } \right ) \right ) \right
. \right .
\nonumber \\ & & + \frac{1}{\beta mc^{2}}\left ( \frac{\left ( \lambda _ { 3}-\lambda _ { 2}\right ) } { n\left ( n-1\right ) } -\frac{\left ( \lambda _ { 3}+(n-2)\lambda _ { 2}\right ) } { \left ( n-1\right ) } \right ) \left [ b_{l}^{n}-a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \nonumber \\ & & + \frac{2\left ( 2\lambda _ { 3}+(n-2)\lambda _ { 2}\right ) \pi gm}{\beta mc^{2}}% \left ( \frac{1}{2\left ( n-1\right ) } -\left [ \frac{1}{2\left ( n-1\right ) } % \right ] \right ) \mbox{\ } \left [ c_{l}^{n}-\frac{1}{l}a_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \nonumber \\ & & \mbox{\ \ } \left .
-\frac{\lambda _ { 4}}{\beta mc^{2}}\left [ d_{l}^{n}+k_{l}^{n}\left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right ] \right\ } \exp \left ( -2\pi gn\beta m^{2}l\left| z\right| \right ) \right ] \nonumber \\ & = & \left ( 2\pi gm^{2}n\beta \right ) \nonumber \\ & & \times \exp \left [ -\frac{1}{\beta mc^{2}}\left\ { \frac{3\left ( n-1\right ) ^{2}}{8n}\lambda _ { 1}-\frac{\left ( n-1\right ) \left ( \lambda _ { 2}\left ( n-1\right ) + \lambda _ { 3}\right ) } { n}-\lambda _ { 4}\sum_{k=1}^{n-1}\sum_{l = k+1}^{n-1}\frac{\left ( l - k\right ) } { l\left ( n - k\right ) } \right\ } \right ] \nonumber \\ & & \times \sum_{l=1}^{n-1}\left\ { a_{l}^{n}+\frac{1}{\beta mc^{2}}\left ( \frac{3\lambda _ { 1}}{8}\frac{\left ( n-1\right ) ^{2}}{n}a_{l}^{n}-\left ( \frac{\left ( \lambda _ { 2}\left ( n-1\right ) + \lambda _ { 3}\right ) } { n}\right ) b_{l}^{n}-\lambda _ { 4}d_{l}^{n}\right ) \right .
\nonumber \\ & & \left .
+ \frac{1}{\beta mc^{2}}\left [ \left ( \frac{\left ( \lambda _ { 2}\left ( n-1\right ) + \lambda _ { 3}\right ) } { n}\right ) a_{l}^{n}-\lambda _ { 4}k_{l}^{n}\right ] \left ( 1 - 2n\left ( \beta \pi gm^{2}\right ) l\left| z\right| \right ) \right\ } \exp \left ( -2\pi gm^{2}n\beta l\left| z\right| \right ) \nonumber \\ & & \label{appfz31}\end{aligned}\]]it is straightforward to show that the coefficients @xmath320 and @xmath321 obey the sum rules @xmath322where eq .
( [ asumrule ] ) was previously derived in the @xmath323 limit @xcite . since @xmath324 @xmath325
, we must have @xmath326 \label{appfz32}\end{aligned}\]]or alternatively , to the relevant order in @xmath2 , @xmath327each of which can be straightforwardly verified .
we also can show @xmath328 | we consider the statistical mechanics of a general relativistic one - dimensional self - gravitating system .
the system consists of @xmath0-particles coupled to lineal gravity and can be considered as a model of @xmath0 relativistically interacting sheets of uniform mass . the partition function and one - particle distitrubion functions
are computed to leading order in @xmath1 where @xmath2 is the speed of light ; as @xmath3 results for the non - relativistic one - dimensional self - gravitating system are recovered .
we find that relativistic effects generally cause both position and momentum distribution functions to become more sharply peaked , and that the temperature of a relativistic gas is smaller than its non - relativistic counterpart at the same fixed energy .
we consider the large-@xmath0 limit of our results and compare this to the non - relativistic case .
watphys - th01/01 * statistical mechanics of relativistic one - dimensional self - gravitating systems * + r.b . mann and p. chak + dept . of physics ,
university of waterloo waterloo , ont n2l 3g1 , canada + pacs numbers : 5.20.-y , 04.40.-b 5.90.+m + + |
during the past few years , there has been continuing interest @xcite in higher dimensional kinematical models for proper and unified theory of subluminal ( bradyon ) and superluminal ( tachyon ) objects @xcite .
the problem of representation and localization of extended particles and superluminal objects may be solved only by the use of higher dimensional space .
several attempts of extending special theory of relativity to superluminal realm in the usual four - space @xcite led to controversies @xcite and satisfactory theory for tachyons could not be made acceptable so far .
several experimental investigations @xcite have been shown towards the evidences for the existence of tachyons : particle moving faster than light @xmath1 ( superluminal particles ) .
still there are doubts ( due to lack of experimental verification)@xcite for the existence of tachyons and it still needs the necessity of constructing a self - consistent quantum field theory , which might yield their quantum properties relevant for their production and detection .
the continuing interest of tachyons showed that these particles are not in contradiction to the special theory of relativity .
a basic disagreement in the various models of tachyons was the spin - statistics relationship .
tanaka @xcite , dhar - sudarshan @xcite , aron - sudarshan @xcite assumed the spin for tachyons as that of bradyons . on the other hand
, feinberg @xcite , hamamoto @xcite showed that spin - statistics for tachyons is reversed to that for bradyons .
rajput and coworkers @xcite described a lorentz invariant quantum field theory of tachyons of various spins and showed that tachyons are not localized in space . the scalar and spinor field theory of free tachyons @xcite shows that tachyons are localized in time .
the localization space for the description of tachyons is @xmath0 - space where the role of space and time ( and that of momentum and energy ) are interchanged on passing from bradyons to tachyons .
the initial values of tachyons lie on the hyper plane @xmath2 while those for tachyons lie on @xmath3 .
the space @xmath4 - ( i.e. the localization space for bradyons ) and @xmath0 - ( the time representation space for tachyons ) demonstrate the structural symmetry between these two and directly lead to space - time duality between superluminal and subluminal objects .
supersymmetry , i.e. the fermi - bose symmetry @xcite , is one of the most fascinating discoveries in the history of physics .
the local extension of supersymmetric theories @xcite ( i.e. super gravity ) provides a natural framework for the unification of fundamental interactions of elementary particles .
it was believed earlier @xcite that spin@xmath5 tachyons are quantized only with anti commuting relations but their localization in space creates problem with lorentz invariance .
now it is made clear that tachyons are localized in time and their localization space is @xmath0 - space which behaves as that of bradyons do in @xmath4 - space and as such it is not possible to accelerate directly a particle from @xmath4 - space to @xmath0- space ( or subluminal to superluminal ) for which we need superluminal lorentz transformations ( slt s ) . keeping these facts in mind , in this paper
we have undertaken the study of the super partner hamiltonians in @xmath0 - space , having the identical energy spectrum , expect for the ground states , and all the potentials exhibiting the shape invariance contain the exact solutions . constructing the operations connecting the states of same energy for super partner hamiltonians @xmath6 and @xmath7 and supersymmetric partner potentials @xmath8 and @xmath9 in @xmath0 - space , all bound state wave functions have been calculated from ground state and shape invariant potentials are obtained for supersymmetric linear harmonic oscillator .
it has been shown that @xmath10 excited state energy of @xmath11 is identical to @xmath12 excited state energy of @xmath13 while any supersymmetric eigen state of one of the super partner potentials in @xmath0 - space is paired in energy with a symmetric eigen state of the other partner potential .
we have also constructed a semi - unitary transformation ( sut ) to obtain the supersymmetric partner hamiltonians for a one dimensional harmonic oscillator and it has been demonstrated that under this transformation in @xmath0 - space the supersymmetric partner @xmath11 loses its ground state but its eigen functions constitute a complete orthonormal set in a subspace of full hilbert space .
in order to over come the various problems associated with superluminal lorentz transformations ( slts ) , six - dimensional formalism @xcite of space time is adopted with the symmetric structure of space and time having three space and three time components of a six dimensional space - time . in this formalism , a subluminal observer @xmath14 in the usual @xmath15 space is surrounded by a neighbourhood in which one measures the scalar time @xmath16 and spatial vector @xmath17 out of six independent coordinates @xmath18 of the six - dimensional space @xmath19 . on passing from @xmath20 to @xmath21 ( where @xmath22 stands the tachyon ) via imaginary slt s , the usual of observer in @xmath19 will appear as @xmath23 to the observer @xmath14 in @xmath19 . the resulting space for bradyons and tachyons is thus identified as the @xmath19- or @xmath24 space where both space and time as well as energy and momentum are considered as vector quantities .
superluminal lorentz transformations ( slts ) between two frames @xmath25 and @xmath26 moving with velocity @xmath27 are defined in @xmath19- or @xmath24 space as follows ; @xmath28 where @xmath29 these transformations lead to the mixing of space and time coordinates for transcendental tachyonic objects , @xmath30 where equation ( [ eq:1 ] ) takes the following form ; @xmath31 it shows that we may have only two four dimensional slices of @xmath19- or @xmath24 space @xmath32 and @xmath33 .
when any reference frame describes bradyonic objects it is necessary to describe @xmath34\,(r^{4}-space)$ ] so that the coordinates @xmath35 and @xmath36 are not observed or couple together giving @xmath37 . on the other hand when a frame describes bradyonic object in frame @xmath25 , it will describe a tachyonic object ( with velocity @xmath38 ) in @xmath26 with @xmath39 space i.e.@xmath40=[z,\ , t_{x},\ , t_{y},\ , t_{z}](t^{4}-space).$ ] as such , we define @xmath39 space as @xmath0- space or @xmath41 space where @xmath42 and @xmath43 are not observed but coupled together giving rise to @xmath44 . as such , the spaces @xmath4 and @xmath0 are two observational slices of @xmath19- or @xmath24 space .
unfortunately the space @xmath45 is not consistent with special theory of relativity and accordingly the subluminal and superluminal lorentz transformations lose their meaning in @xmath19- or @xmath46 space in the sense that these transformations do not represent either the bradyonic or tachyonic objects in this space .
it has been shown earlier @xcite that the true localizations space for bradyons is @xmath4 - space while that for tachyons is @xmath0 - space .
so a bradyonic @xmath47 space now maps to a tachyonic @xmath48 space or vice versa i.e. @xmath49 hence @xmath4 , the suitable space for describing bradyonic phenomena , is mapped under slt s onto the space @xmath0 which is the space suitable to describe the tachyonic phenomena . in such formalism
we get the following mapping for the position four - vector @xmath50 , four - differential @xmath51 and four - potential @xmath52 under slt s as @xmath53 @xmath54 where @xmath55 are @xmath56 differential operators in terms of three spatial and three temporal coordinates , respectively . applying these mappings on the equations for generalized electric field , magnetic field and lorentz force associated with dyons ,
we get the following expressions respectively for generalized electric field , magnetic field and lorentz force on dyons under superluminal transformations in @xmath0 - space i.e. ; @xmath57 where @xmath58 is introdued @xcite as magnetic four - potential beside the electric four - potential @xmath59 and @xmath60 is described as generalized electromagnetic vector field in subluminal frame of reference ; and @xmath61 is inverse velocity .
these equations show the superluminal fields associated with tachyonic dyons containing both longitudinal and transverse parts .
equation @xmath62 describes the lorentz force acting on dyon in @xmath0-space .
it may be recalled as the force of tachyonic dyon which is similar to that of dyonic force in @xmath4- space if we replace the particle velocity @xmath63 to that of the inverse velocity @xmath61 alon with the role of space variables is interchanged with time variables .
so , the theories of tachyons be better understood in @xmath0- space where they behave as bradyons do in @xmath4- space . as such in view of their localizability and other quantum properties , we can formulate the supersymmetric theories of tachyons in @xmath0- space in the parallel ground of bradyonic supersymmetric theories in @xmath4- space .
similarly , the total hamiltonian @xmath64 of a tachyonic supersymmetric system may be decomposed in to its bosonic part @xmath65 and the fermionic part @xmath66 such that @xmath67 with @xmath68 and @xmath69=-iw''(t)y_{t}\label{eq:8}\end{aligned}\ ] ] where@xmath70is the rest mass of tachyon , @xmath71 is super potential while @xmath72 and @xmath73 are the fermionic variables in @xmath0- space describing spin degree of freedom and the operator @xmath74 is defined as @xmath75 in ( [ eq:8 ] ) , prime and double prime denote first and second order derivatives .
supersymmetric hamiltonian may also be constructed in terms of non - hermitian supercharge operators @xmath76 and @xmath77 as @xmath78\label{eq:10}\end{aligned}\ ] ] such that @xmath79 & = & 0,\,\,\,\ , q^{2}=q^{+^{2}}=0\label{eq:11}\end{aligned}\ ] ] and @xmath80 & = \left[h^{t},q^{+}\right]= & 0;\nonumber \\
\left[y_{t}\,,q\right ] & = & -iq;\nonumber \\
\left[y_{t},\ , q^{+}\right ] & = & iq^{+}.\label{eq:12}\end{aligned}\ ] ] from equations ( [ eq:6 ] ) and ( [ eq:10 ] ) , the supercharges may readily be constructed accordingly in the following form ; @xmath81\overline{\psi_{t}};\,\,\,\ , q=\left[e+iw'(t)\right]\psi_{t}.\label{eq:13}\end{aligned}\ ] ] any state @xmath82 satisfies the conditions @xmath83 is denoted as the bosonic state for which we have @xmath84 similarly , fermionic state @xmath85 satisfies the conditions @xmath86 which gives @xmath87 using these relations , it may readily be demonstrated that the operator @xmath76 transforms states @xmath85 in to states @xmath82 of the same eigen energy @xmath88 and the operator @xmath77 transforms the states @xmath82 into states @xmath85 , i.e. @xmath89 which directly shows @xmath90=0,\,\,\,\ , q^{2}=q^{+^{2}}=0 $ ] ; the supersymmetry between tachyonic fermions and tachyonic bosons with positive momentum eigen values while the negative momentum value corresponds to singularity problem associated with super potential in this supersymmetric model .
alternatively , supersymmetric quantum mechanics may be worked out in terms of a pair of bosonic hamiltonians @xmath13 and @xmath11 which are supersymmetric partners of supersymmetric hamiltonian i.e. , @xmath91 in order to construct these super partner hamiltonians for the system described by equation ( [ eq:6 ] ) , let us introduce the potential @xmath9 whose ground state energy has been adjusted to zero with the corresponding ground state wave function given by ; @xmath92.\label{eq:18}\end{aligned}\ ] ] substituting it in the schrdinger equation ( in the units of @xmath93 ) , we get @xmath94 hence we have the following hamiltonian corresponding to the potential , @xmath95 if the ground state wave function @xmath96 is square integrable then the supersymmetry will be broken .
this hamiltonian may also be written in the following form in terms of bosonic operator @xmath97 and @xmath98 ; @xmath99 where @xmath100 let us introduce the hamiltonian @xmath101 where @xmath102 the potentials @xmath103 and @xmath104 are called supersymmetric partner potentials and @xmath105 is the hamiltonian corresponding to the potential @xmath8 . from equations ( [ eq:13 ] ) and ( [ eq:18 ] ) , we get @xmath106 & = & w^{2}(t);\nonumber \\
\left[\hat{b},\hat{b^{+}}\right ] & = & 2w'(t)\label{eq:25}\end{aligned}\ ] ] showing that @xmath107 is the average of the potential @xmath8 and @xmath9 , whereas @xmath108 is proportional to the commutator of @xmath97 and @xmath98 .
the supersymmetric charges have been defined in terms of operators @xmath98and @xmath97 .
hence , the hamiltonians @xmath11 and @xmath13 are denoted as supersymmetric partners .
it can be visualized by introducing supersymmetric charges @xmath76 and @xmath77 as @xmath109;\,\,\ , q^{+}=\sqrt{2}\left[\begin{array}{cc } 0 & b^{+}\\ 0 & 0\end{array}\right].\label{eq:26}\end{aligned}\ ] ] then we have @xmath110\end{aligned}\ ] ] or @xmath111 & = \left[h_{susy}^{t},q^{+}\right ] & = 0\end{aligned}\ ] ] or @xmath112 for any eigen function @xmath113 of @xmath13 with the corresponding eigen value @xmath88 , we have @xmath114 or @xmath115 & = & e^{t}\hat{[b}\psi^{t(-)}]\label{eq:28}\end{aligned}\ ] ] which shows that @xmath116 is the eigen function of @xmath11 , with the same eigen value @xmath88 .
similarly it can be demonstrated that if @xmath117 is an eigen function of @xmath11 with the eigen value @xmath118 then @xmath119 is an eigen function of @xmath13 with the same eigen value .
thus @xmath13 and @xmath11 have the identical energy spectrum except for the ground state of @xmath13 .
it is also obvious that the operators @xmath97 and @xmath98 connect the states of the same energy for two different supersymmetric partner potentials @xmath8 and @xmath9 .
specific relations between spectra of @xmath120 and @xmath105 are based on the symmetry between the solvable potential and its supersymmetric partner potential .
for supersymmetry one - dimensional harmonic oscillator we have the super potential @xmath121 then the supersymmetic hamiltonian is described as @xmath122 and equations ( [ eq:19 ] ) and ( [ eq:20 ] ) give @xmath123 similarly , equation ( [ eq:28 ] ) leads to @xmath124 from equation ( [ eq:22 ] ) , we may then readily get @xmath125 which gives @xmath126 substituting relation ( [ eq:22 ] ) into equation ( [ eq:18 ] ) , we get @xmath127\label{eq:35}\end{aligned}\ ] ] where the constant @xmath128 is the result of the orthonormality of ground state wave function . from equation ( [ eq:30 ] ) and ( [ eq:34 ] ) , it is obvious that @xmath129 i.e. @xmath130 using relations ( [ eq:34 ] ) and ( [ eq:35 ] ) , we get @xmath131 where @xmath132 is the classical frequency of the oscillator .
this equation shows that ground state of @xmath11 is not the zero energy state i.e. @xmath133 .
first excited state @xmath134 of @xmath13 may be obtained by the raising operator @xmath135 i.e. @xmath136 & \left(\frac{\omega}{\pi}\right)^{\frac{1}{4}}\exp\left[-\frac{\omega t^{2}}{2}\right]\nonumber \\ = & 2\left(\frac{\omega}{\pi}\right)^{\frac{1}{4}}\omega t\ , & \exp\left[-\frac{\omega t^{2}}{2}\right].\label{eq:38}\end{aligned}\ ] ] then we get @xmath137 i.e. @xmath138 equations ( [ eq:38 ] ) and ( [ eq:39 ] ) yield @xmath139 i.e. ground state energy of @xmath11is the first excited state energy of @xmath13 .
similarly , we get @xmath140 i.e. @xmath141 second excited state of @xmath13 may be constructed as follows ; @xmath142\exp[-\frac{\omega t^{2}}{2}]\label{eq:42}\end{aligned}\ ] ] for which we have @xmath143 showing that first excited state energy of @xmath11 is state @xmath134 is identical to the second excited energy of @xmath13 in the state @xmath144 .
generalizing these results , it may be inferred that with respect to the states @xmath145 , the @xmath10 excited state energy of @xmath11 is identical to the @xmath12 excited state energy of @xmath13 . in other words ,
any state @xmath145 of the super partner hamiltonian @xmath13 corresponding to the eigen value @xmath146 of harmonic oscillator is also the eigen state of @xmath11 with the corresponding eigen value @xmath147 .
let us construct the eigen state @xmath148 of @xmath11 from the states @xmath149 of @xmath13 in the following form ; @xmath150 which may be written in the following form by using equation ( [ eq:18 ] ) ; @xmath151.\label{eq:45}\end{aligned}\ ] ] then we get , @xmath152 similarly , we get @xmath153;\nonumber \\
\psi_{2}^{t(+)}(t ) & = & 2\left(\frac{\omega}{\pi}\right)^{\frac{1}{4}}(\omega)^{\frac{3}{2}}[-1 + 2\omega t^{2}]\exp[-\frac{\omega t^{2}}{2}];\label{eq:47}\end{aligned}\ ] ] which gives rise to @xmath154 and @xmath155 generalizing these results , we may infer that@xmath156 which shows that @xmath10 excited state energy of @xmath11 is identical to @xmath12 excited state energy of @xmath13 .
in other words , we may infer that if @xmath145 is the @xmath10 eigen state of @xmath13 with the corresponding eigen value @xmath157 then @xmath158 is the @xmath159 eigen state of @xmath11 with the energy @xmath157 .
it is obvious from equations ( [ eq:34 ] ) and ( [ eq:45 ] ) ; ( [ eq:37 ] ) , ( [ eq:38 ] ) and ( [ eq:42 ] ) ; and ( [ eq:47 ] ) that for supersymmetry harmonic oscillator we have @xmath160 with the corresponding energy eigen values satisfying condition ( [ eq:50 ] ) . from equations ( [ eq:31 ] ) and ( [ eq:32 ] )
we get the following conditions for the supersymmetry partner potentials for one - dimensional harmonic oscillator @xmath161 any anti - symmetric eigen state of one of these potentials is paired in energy with the symmetric eigen state of the other equations ( [ eq:34 ] ) , ( [ eq:42 ] ) , ( [ eq:45 ] ) and ( [ eq:47 ] ) shows that all eigen functions of @xmath11 and @xmath13 will go to non - zero constant as @xmath162 and hence these are unacceptable as physical states of @xmath163
. these unacceptable solutions of both these super partner potentials are paired in energy and hence the degeneracy theorem holds for @xmath163 maintaining the supersymmetry of the system .
equation ( [ eq:5 ] ) for lorentz force acting on dyon interacting with generalized superluminal electromagnetic fields in @xmath0 - space is similar to that of a dyon interacting with generalized subluminal electromagnetic fields in @xmath4- space except that the role of velocity @xmath164 has been changed with the inverse velocity @xmath165 and consequently the role of space and time variables are interchanged on passing from subluminal to superluminal realm via slt s and thus suggests that tachyonic dyons are localized in time with their fields having rotational symmetry in temporal planes and their motion is bi - dimensional in time .
it is emphasized that there are two observations slices , @xmath4- or @xmath166 ( one time and three space ) and @xmath0- or @xmath167 ( three time and one space ) , in the existing @xmath19-space , @xmath46 ( three time and three space ) .
the transformations for transcendent velocity @xmath168 reveal that the tachyons and the bradyons can not be described individually in the @xmath19 - space . as such ,
the equivalence of space and time and the necessity of a four - dimensional minkowski space to describe any physical event consistent with the special theory of relativity , the @xmath0 - space have been constructed as the natural space for the description of the tachyons , whereas a bradyon is described in the usual @xmath4- space .
it has been stressed that a bradyonic @xmath4- space maps under slt s to a tachyonic @xmath0 - space and on passing from a bradyon to a tachyon via slt s , the @xmath0 - space becomes the representation space of the observables while the @xmath4- space represents the internal space , where only internal degree for bradyons are represented .
it means that the usual pseudo - euclidean group , @xmath169 of space - time representation of the poincare group will be visualized as the usual rotation group in the three time planes in the @xmath0- space .
further an object moving with velocity , @xmath170 ( bradyonic ) in subspace @xmath4 ( forward velocity @xmath171 ) appears to be moving with an inverse velocity @xmath172 with @xmath173 .
it may also be emphasized that for all - purpose the superluminal fields ( tachyonic ) in the time - energy representation ( @xmath0- space ) behave as subluminal ( bradyonic ) fields in space - momentum representation ( @xmath4- space ) .
equation ( [ eq:6 ] - [ eq:8 ] ) shows the total supersymmetric hamiltonian in @xmath0- space associated with tachyons in terms of energy eigen values and super potentials .
equation ( [ eq:16 ] ) shows that the supersymmetry between tachyonic fermions and tachyonic bosons with positive momentum eigen values while the negative momentum values corresponds to singularity problem associated with super potential in the supersymmetric model .
constructing the potential @xmath9 in the form given by equation ( [ eq:19 ] ) with the ground state energy vanishing and the ground state wave function given by equation ( [ eq:18 ] ) the super - partner hamiltonian @xmath13 has been obtained in the form of equation ( [ eq:20 ] ) and it has been found that its ground state energy identical to ground state momentum of a bradyon is vanishing provided that the wave - function defined by equation ( [ eq:19 ] ) is square integrable .
the hamiltonian @xmath11 ( super - partner potential of @xmath13 ) has been constructed in the form of equation ( [ eq:23 ] ) in terms of potential @xmath8 ( super - partner potential of @xmath9 ) given by equation ( [ eq:24 ] ) .
the supersymmetric charge operators are introduced by equation ( [ eq:26 ] ) in @xmath0- space .
it has been shown that @xmath11 and @xmath13 have identical energy ( momentum ) spectrum in @xmath174 - space except for the ground state of a hamiltonian .
equation ( [ eq:41 ] ) and ( [ eq:43 ] ) shows that the first excited state energy of @xmath11 in state @xmath134 is identical to the second excited energy of @xmath13 in the state @xmath144 .
generalization of these results leads to the inference that with respect to the states @xmath145 the @xmath10 excited state energy of @xmath11 is identical to the @xmath12 excited state energy of @xmath13 .
in other words any state @xmath145 of the super partner hamiltonian @xmath13 corresponding to the eigen value @xmath146 of harmonic oscillator is also the eigen state of @xmath11with the corresponding eigen value @xmath147 .
equations ( [ eq:34 ] ) , ( [ eq:42 ] ) , ( [ eq:45 ] ) and ( [ eq:47 ] ) provide that all eigen functions of @xmath11 and @xmath13 will go to non - zero constant as @xmath162 and hence these are unacceptable as physical states of @xmath163 .
these unacceptable solutions of both these super partner potentials are paired in energy and hence the degeneracy theorem holds for @xmath163 maintaining the supersymmetry of the system in @xmath0 - space .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ * acknowledgement : * - this work is carried out under uttarakhand state council for science and technology , dehradun research scheme . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ | constructing the operators connecting the state of energy associated with super partner hamiltonians and super partner potentials for a linear harmonic oscillator has been discussed and it is shown that any super symmetric eigen state of one of the super partner potentials in @xmath0- space is paired in energy with a symmetric eigen state of the other partner potential .
department of physics + kumaun university + s. s. j. campus + almora - 263601 ( india ) email : - [email protected] [email protected] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ key words : - supersymmetry , superluminal transformation pacs no : 14.80ly _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ |
dendrimers have been the subject of intensive investigations ever since their synthesis@xcite in the late 1970s .
they are characterized by a high degree of monodispersity and a well - defined , highly branched internal structure ; efficient dendrimer assembly has been boosted by recent progress in synthetic techniques.@xcite a great deal of research activity has focused on the issue of whether they possess an open , _ dense - shell _ or a collapsed , _ dense - core _ configuration , the motivation arising by the potential to employ them as hollow , carrier - type molecules in the former case . for neutral dendrimers , a large number of simulation studies,@xcite careful self - consistent field calculations@xcite and not least scattering experiments@xcite have revealed that the dense - shell conformation is _ not _ the real one . due to back - folding of the end - groups , caused by entropic considerations , a dense - core calculation results instead , leading even to compact , hard - sphere - like conformations at high generation numbers.@xcite from the point of view of applications , this may sound like a disappointing result , as one would like to have dense - shell molecules .
however , seen from the angle of fundamental research , the growing compactness of dendrimers with increasing generation number is very welcome , since it allows to use them as model colloidal / nano particles with tunable stiffness,@xcite bridging the gap between flexible polymers and rigid spheres .
the issue of dendrimer conformations is less clear when _ charged _ or _ polyelectrolyte _ dendrimers are considered .
charge on the building blocks of , e.g. , poly(amidoamine ) ( pamam ) dendrimers can be manipulated by changing the ph of the solution@xcite and their conformations can be further influenced by added salt .
the expectation that charged dendrimers may achieve stretching is based on experience with other branched polyelectrolytes ( pe ) , such as pe - stars , for instance.@xcite however , the sans - study of nisato _
et al._@xcite has led to a negative result : the size of dendrimers is insensitive to ph changes and thus to charge .
this experimental fact is at odds with the earlier work by welch and muthukumar,@xcite who predicted , by means of simulation and theory , an ` opening up ' of pe dendrimers upon increase of the number of charged units ; similar conclusions were reached in the brownian dynamics simulations of lyulin _
et al._@xcite however , both simulational works quoted above employed a debye - hckel ( screened coulomb ) interaction potential acting between charged units , treating thereby the counterions as a continuum . giupponi _
, on the other hand , pointed out that a more realistic treatment of the molecules should employ the bare coulomb interactions among all involved species ( monomers and counterions ) , which is the method they employed in their own , recent simulations.@xcite interestingly enough , they found that the dendrimer size is indeed very weakly dependent on charge , due to local charge neutrality , a condition that is masked when one employs the debye - hckel approximation .
though the work of giupponi _ et al._@xcite has illuminated a number of issues pertaining to monovalently charged monomer units accompanied by monovalent counterions , the question of the influence of valency on dendrimer conformations has not been studied so far .
the purpose of this work is to examine precisely this issue , which appears relevant on the grounds that counterion valency is known to bring about drastic changes in , e.g. , the size of spherical pe brushes.@xcite we focus thereby on dendrimers of the fourth generation ( @xmath0 ) and examine separately the conformations of neutral dendrimers , as a reference point , as well as the combinations @xmath1 and @xmath2 , where @xmath3 stands for the valency of the monomers and @xmath4 for the valency of the counterions .
we find that the case of divalent monomers and monovalent counterions brings about a substantial change of the dendrimer size , accompanied by a strong stretching of the chemical bonds , whereas the same phenomena are less pronounced for the other two cases . after describing our simulation model in sec .
ii , we present and discuss the results in sec .
iii , whereas in sec .
iv we draw our conclusions .
the dendrimers we use within our simulations are built from a central pair of joined monomers , the so - called generation 0 . a successive generation @xmath5 of dendrimer is formed by connecting two additional monomers , the functionality of the dendrimer is therefore three , to each outer monomer of the dendrimer of generation @xmath6 . in doing so
, the number of monomers @xmath7 of a given generation @xmath6 in a dendrimer follows a simple power law , i.e. @xmath8 .
there are in general three types of interactions between monomers .
the first type of interaction prevents the collapse of monomers onto each other and is a short - range repulsive interaction given by a simple , shifted and purely repulsive lenard
jones potential @xmath9 & r \leq r_c \\ 0 & r >
r_c \end{array } \right.\ ] ] where @xmath10 and @xmath11 are the unity of length and energy respectively , and @xmath12 is the range of the interaction .
the second type of interaction is of an attractive nature and describes the bonds between joined monomers in order to prevent the molecule from flying apart .
this interaction is described by a fene potential@xcite @xmath13 & r \leq r_0 \\ 0 & r >
r_0 \end{array } \right.\ ] ] where @xmath14 is the maximum allowed distance between two connected monomers .
the last type of interaction to be included in the model is the coulomb potential between charged monomers @xmath15 with @xmath16 and @xmath17 the charge numbers , @xmath18 the boltzmann constant , @xmath19 the temperature , and @xmath20 the bjerrum length given by @xmath21 we have performed molecular dynamics simulations at constant density and temperature using the nos - hoover thermostat@xcite on dendrimers of generation 4 , i.e. , a dendrimer formed by 62 identical monomers , where the monomers with charge number @xmath3 are either neutral , monovalent , or divalent . in order to guarantee charge neutrality the charged dendrimers need to be balanced , for which we in either case used both monovalent and divalent counterions ( charge number @xmath4 ) .
apart from the coulomb interaction between monomers and counterions , as well as between the counterions themselves , we use for simplicity the same short range repulsion ( [ e : v_lj ] ) as is used for the monomers .
simulation parameters are chosen such that our unit of length @xmath22 and we fixed the temperature to @xmath23 . using an implicit solvent that mimics the behavior of water , i.e. , a temperature of 300k and a relative permittivity @xmath24 we arrive at @xmath25 .
periodic boundary conditions have been applied in combination with the ewald summation method to include the long - range electrostatic interactions .
the volume of the simulation box was chosen such that effectively the dendrimers can be considered to be independent , indicated by the independence of the results on larger volume sizes .
the initial configurations were equilibrated over times long enough for the counterions to diffuse in to the core of the dendrimer , and to reach a steady state for a in- and out - flux . the simulation time is chosen long enough for the individual counter - ions to explore the full dendrimer , i.e. , inner and outer ranges .
in order to characterize the size of the dendrimer , the radius of gyration @xmath26 is measured , which is defined by @xmath27 where the summation runs over the @xmath28 positions @xmath29 of the monomers of the dendrimer and @xmath30 is the center of mass of the dendrimer .
the resulting values for the neutral and charged dendrimers are listed in table [ t : rg ] .
the first observation we can make , is that by charging the dendrimer the radius of gyration increases .
this is actually not so surprising , because the monomers will repel each other due to their charges . what is more interesting is the fact that the behavior on charging is not monotonic , i.e. , the radius of gyration of the divalently charged monomers and counterions is smaller with respect to that of the monovalent case
this suggests that the swelling of the dendrimer due to the increasing charge is counteracted by the transport of counterions into the deeper regions of the dendrimer , which results in screening effects and reduces the swelling .
@xcite .the radius of gyration for the neutral dendrimer and the dendrimers with charged monomers and counterions , with charge numbers @xmath3 and @xmath4 respectively .
[ cols="^,<,<",options="header " , ] it also appears that divalent counterions reduce the size of the dendrimer with respect to that of monovalent counterions .
not only do they result in a better screening of the monomers , but also less of them are required which also leads to less steric hindrance . using a similar argument
, the monovalent monomers reduce the size of the dendrimer with respect of that of divalent monomers , because in the latter case more counterions are required within the dendrimer , preventing its shrinking .
note that for a neutral dendrimer there is no counterpart for the counterions present within the dendrimer , hence it is only the short - range repulsion between the monomers that prevents the dendrimer from collapsing .
the radius of gyration is also experimentally observable from the small wave - vector limit of the form factor @xmath31 defined by @xmath32 by expanding in small wave - vectors @xmath33 and averaging over the directions it is easy to show that this results in @xmath34.\ ] ] in the present case this expansion is valid up to @xmath35 as can be seen in fig .
[ f : fq ] where we compare the measured form factor , eq .
( [ e : fq ] ) , from the simulation with the small wave - vector limit , eq .
( [ e : fq - small ] ) . ) from the simulation ( sim ) , with the small wave - vector limit , eq .
( [ e : fq - small ] ) , denoted ( th).,width=302 ] in fig .
[ f : rho ] the radial density profiles of the monomers and counterions are shown , both measured with respect to the center of mass of the dendrimer .
note that at small distances from the center of mass the noisy behavior is purely due a poor statistical sampling caused by the lack of particles in that region .
it is immediately clear that the presence of charge on the monomers results in a much more structured density profile , in which the monomers are mostly found in a shell - like structure .
the mutual repulsion of the monomers due to their charge not only leads to a larger size of the dendrimer but it enables the counterions to diffuse into the dendrimer as well ; this can be seen from their density profiles .
their distribution , however , shows less structure and has a wider range .
the latter can also easily be understood , since the presence of counterions within the dendrimer will also lead to steric hindrance and there is an obvious entropic gain in surrounding the dendrimer and move in the region lying outside its extent . and
divalently [ ( c),(d ) ] charged monomers , and monovalently [ ( a),(c ) ] and divalently [ ( b),(d ) ] counterions.,width=302 ] to clarify the ordering in the charged dendrimer cases , the same data are shown in fig .
[ f : rho_m_c ] where the radial density profiles of the monomers and counterions are directly compared . although the highest density for the counterions is reached roughly in between the first and second peak of the monomer distributions ( except for the case @xmath36 and @xmath37 ) , there is no layered structure present for the counterions .
this is also confirmed by the absence of plateaus in the cumulative counterion density profile ( not shown ) .
the absence of charges on the monomers in the neutral dendrimer does not only affect its size , i.e. , it is more compact , but it also modifies its internal structure as is illustrated in fig . [
f : rhog_00 ] , where the density profiles for the different generations are plotted versus the distance to the center of mass .
this reveals that the density of generation 4 monomers at smaller distances is larger than that of those from generation 3 . in other words ,
the dendrimer starts to fold in to itself .
whereas the repulsive coulomb interactions between charged monomers tend to stretch the dendrimer causing a loss in entropy , the neutral monomers just feel the short - range lennard - jones repulsion and can exploit the open space in the core of the dendrimeric structure .
a not unimportant other reason is the absence of the equivalent of counterions within the dendrimer .
this effect is only weakly visible in the cases of the divalently charged counterions , but it is to be expected that it will be more pronounced for increasing value of the generation of the dendrimer . a more detailed description of the internal structure of the dendrimer can be obtained by analyzing the the bond - lengths @xmath38 , shown in figs .
[ f : bonds_00 ] and [ f : bonds_xx ] .
the bond - length probability distributions , @xmath39 , are decomposed per generation , whereby a bond of generation @xmath40 is formed by a monomer of generation @xmath6 with its parent of generation @xmath41 ( with the exception of generation 0 ) . for the neutral dendrimer
only a small shift in the distribution of lengths is found towards shorter bond - lengths for higher generations , i.e. , bonds near the center of the dendrimer tend to be more stretched than those near the border .
this implies that there is a collective behavior in which the mutual repulsion of monomers in higher generations that prefers to expand the dendrimer , forcing the fewer central bonds to stretch .
[ f : bonds_xx ] shows the same distributions but now for the charged dendrimer cases .
the first thing one can observe is that the bond - lengths in the charged cases are more stretched .
this is not so surprising , since the bond - length is directly affected by the mutual charge repulsion of the monomers and even more so for the divalent monomers .
also the tendency for the central bonds to be more stretched than those in the outer regions is apparent .
the most interesting case , however , is that of the divalent monomers and monovalent counterions , in which case the stretching of the bond - lengths shifts significantly with the generation . in fig .
[ f : bonds_x ] this is illustrated even more clearly by comparing the bonds of generation 0 and 4 for the various models .
this suggests that the abundance of monovalent counterions required for an effective screening of the divalent monomer bonds can not be packed within the core of the dendrimer . and
divalently [ ( c),(d ) ] charged monomers , and monovalently [ ( a),(c ) ] and divalently [ ( b),(d ) ] charged counterions.,width=302 ] and @xmath42 angles mentioned in the text.,width=151 ] in the final method we used to examine the internal structure of the dendrimer , we consider the angles between bonds in the dendrimer .
the angle @xmath42 as illustrated in fig .
[ f : angle ] is the angle between the bonds of a monomer of generation @xmath6 and the two monomers bounded to it of generation @xmath5 ( a bond generation @xmath5 ) .
the angle @xmath43 is the angle between a bond of generation @xmath6 and a bond of generation @xmath5 . [ ( a ) ] and @xmath42 [ ( b ) ] for the different dendrimer cases averaged over all nodes in the dendrimer.,width=302 ] the results of the angle analysis are shown in fig .
[ f : angles_m ] , where we have averaged over all 30 nodes in the dendrimer . in the case of the neutral dendrimer the probability distributions are broader than in the case of the charged dendrimers .
this is partially due to the fact that there is no penalty for smaller angles , whereas in the charged dendrimer cases the angles will be more focused in the equal angles area due to the coulomb repulsion . ;
( b ) : angle @xmath42.,width=302 ] also these angle distributions will , just like in the case of the bond - lengths , depend on the position of the monomer within the dendrimer , i.e. its generation , as can be seen in fig .
[ f : angles_mm ] in the case of monovalent monomers and dendrimers . for nodes near the core
the @xmath42 angles are more or less centered about 120 , but for nodes near the boundary of the dendrimer the peaks shifts to smaller angles .
this is specially clear for the angle @xmath42 for the outer most nodes lies between 70 and 80 for the various cases presented here . at the same time
the @xmath43 angles show a similar behavior , except that they become larger , albeit that the shift is not as significant as is the case for @xmath42 .
this is not the only information that can be extracted from the angles @xmath43 and @xmath42 . in general
the three bonds at a given node will not lie within a single plane , but span a space angle , also called spherical excess,@xcite @xmath44 , which ranges from zero to @xmath45 .
both extremes can only be obtained in the limiting case of three bonds lying within the same plane .
the fact that for such a planar configuration the spherical excess can take two values can be understood by realizing that two cases need to be distinguished . if the sum of the three angles between the pairs of bonds @xmath46 , the space angle reaches it maximum @xmath45 and spans a half sphere . in the other case
the largest of the three bond angles equals the sum of the smaller ones , e.g. @xmath47 and the spherical excess reaches it minimum 0 . in other words , when the three bonds completely open up and stretch away from one another , one has the maximum value , @xmath48 ; on the other hand , when there is complete backfolding of the two bond connecting the @xmath6-monomer with the two @xmath5-monomers towards the bond between the @xmath6-monomer and the @xmath41-monomer in fig . 8 , then @xmath49 .
in this fashion , the probability distribution @xmath50 of the spherical excess offers valuable information on the presence of back - folding within the dendrimer .
a distribution with strongly suppressed values around @xmath49 and high values around @xmath48 implies a stretched dendrimer with planar three - bond junctions at the branching points .
relatively flat distributions with non - negligible values around @xmath49 point rather to strongly back - folded dendrimers , akin to neutral ones .
the probability distribution of the spherical excess for the neutral dendrimer and for the case of a charged dendrimer with both monomers and counterions monovalently charged , is shown in fig .
[ f : sph_exc ] .
it reveals that the spherical excess decreases for nodes that lie further away from the core of the dendrimer .
the preference for spherical excess to be large near the center of the dendrimer can be understood by realizing that such behavior will maximize stretching in the dendrimer and enable to exploit the so available space for the more distant monomers more effectively .
the decrease of the spherical excess is particular striking in the case of the neutral dendrimer and is a necessary consequence from our earlier observation that the monomers from the last generation can fold back into the dendrimer.@xcite as can be seen in fig . [
f : sph_exc](b ) , however , and in conjunction with fig .
[ f : sph_exc](a ) , the @xmath51-dendrimer shows a considerably higher degree of stretching than the neutral one ; its distribution of the spherical excess is shifted towards higher values and only at the third generation do we observe a nonvanishing contribution around @xmath49 .
[ f : sph_exc_g ] shows the spherical excess of the nodes of generation 0 and 3 for the various models discussed . in full agreement with all our previous measures , we find that charged dendrimers carrying divalent monomers and monovalent counterions show the most dramatic degree of stretching , all the way from the inner to the outermost generations .
physically , this is caused by the strong attraction between the monomers and the counterions . in this case
, there are twice as many counterions as in the cases @xmath52 and @xmath53 that have to be accommodated in the inner region of the dendrimer , giving rise to stretching to create space .
two counterions per monomer are required to achieve local charge neutrality,@xcite rendering the composite , monomer - counterion groups too big to allow for the backfolding seen for neutral dendrimers . and
monovalently charged monomers and counterions [ ( b)].,width=302 ] and generation 3 [ ( b ) ] nodes for the different cases monomers and counterions.,width=302 ]
we have presented simulation data of simple , fourth generation dendrimers and investigated the effect that chargeable monomers will have on their size and internal structure . in particular ,
we have allowed for the possibility of divalent monomer units and/or counterions and monitored the conformational changes induced by the various scenarios . for the simplest case , @xmath51 , our findings are in full agreement with the recent ones of guipponi _
et al._,@xcite in the sense that no dramatic changes in the dendrimer conformation and size have been detected .
this is also in agreement with the experimental results of nisato _
et al._@xcite and at odds with the prediction of welch and muthukumar.@xcite at this point , it appears that the assumption of linear counterion screening , which has been employed in ref . in the form of a debye - hckel approximation ,
is not valid when strong counterion absorption and condensation effects are present .
more so at the nanoscale - distances involved in the inner of a dendrimer . at the same time , we could establish that a remarkable change of the size of the dendrimer , accompanied by a stiffening and stretching of its bonds , takes place when the former carries divalent chargeable groups , which release two monovalent counterions per site . here ,
although the resulting configuration is not yet of the dense - shell type , the dendrimer size grows by almost 50% in comparison to a neutral dendrimer and strong correlations between the monomer- and counterion profiles appear .
it is now reasonable to assume that , given a guest molecule that carries the same type of charge as the counterions and being small enough to ` fit ' inside the stretched dendrimer , absorption will take place , as in this fashion a number of counterions will be released from the inner part of the dendrimer , resulting in a concomitant entropy gain.@xcite another possibility is to endow the dendrimers with a spacer length @xmath54 , which allows for creation for more space in their interior but at the same time it also increases the possibilities to backfolding .
work along these lines is currently in progress . | we study the effect of chargeable monomers on the conformation of dendrimers of low generation by computer simulations , employing bare coulomb interactions .
the presence of the latter leads to an increase in size of the dendrimer due to a combined effect of electrostatic repulsion and the presence of counterions within the dendrimer , and also enhances a shell - like structure for the monomers of different generations . in the resulting structures the bond - length between monomers , especially near the center ,
will increase to facilitate a more effective usage of space in the outer - regions of the dendrimer .
+ + * keywords : * dendrimers , molecular dynamics simulations , polyelectrolytes , scattering |
a germ @xmath0 of complex analytic variety equidimensional of dimension @xmath1 is _ quasi - ordinary _ ( q.o . ) if there exists a finite map @xmath2 which is unramified outside a normal crossing divisor in @xmath3 .
quasi - ordinary singularities admit fractional power series parametrizations , which generalize newton - puiseux expansions of plane curves ( see @xcite ) .
quasi - ordinary surface singularities appear classically in the jung s method to parametrize and resolve surface singularities ( see @xcite ) and are related to the classification of singularities by zariski s dimensionality type ( see @xcite ) .
classical examples are plane curve singularities , hirzebruch - jung surfaces and simplicial toric varieties .
in addition to the applications in equisingularity problems , the class of q.o .
singularities is also of interest to test and study various open questions and conjectures for singularities in general , particularly in the hypersurface case ( see @xcite ) . in many cases the results passed by using the fractional power series parametrizations of these singularities , which allow explicit computations combining analytic , topological and combinatorial arguments
( see for instance @xcite ) .
it is natural to investigate new invariants of singularities , such as those arising in the development of motivic integration , on this class of singularities with the hope to extend the methods or results to wider classes ( for instance by passing through jung s approach ) .
we recall the definition of the _ geometric motivic poincar series _ of a germ @xmath4 of complex algebraic variety ( or complex analytic or algebroid ) equidimensional of dimension @xmath1 .
the set @xmath5 of formal arcs of the form , @xmath6 \rightarrow ( z,0)$ ] has a scheme structure over @xmath7 ( not necessarily of finite type ) .
let @xmath8 , the set @xmath9 of @xmath10-jets of @xmath11 of the form @xmath12 / ( t^{s+1 } ) \rightarrow ( z,0)$ ] , has the structure of algebraic variety over @xmath7 . by a theorem of greenberg @xcite , the image of @xmath5 by the natural morphism of schemes @xmath13 , which maps an arc to its @xmath10-jet ,
is a constructible subset of @xmath14 . the _ grothendieck ring _
@xmath15 of @xmath7-_varieties _ is generated by the symbols @xmath16 $ ] for @xmath17 an algebraic variety , subject to relations : @xmath16 = [ x']$ ] if @xmath17 is isomorphic to @xmath18 , @xmath16 = [ x- x ' ] + [ x ' ] $ ] if @xmath18 is closed in @xmath17 and @xmath16[x ' ] = [ x \times x']$ ] .
since a constructible set @xmath19 has an image @xmath20 $ ] in the _ grothendieck ring _
@xmath21 of varieties , it is natural to consider the _
geometric motivic poincar series _ of the germ @xmath4 : @xmath22
t^s \in k_0 ( \var_\c ) [ [ t]].\ ] ] this series , which was introduced more generally by denef and loeser in @xcite , is inspired by related poincar series in arithmetic geometry @xcite .
it has a rational form : ( see @xcite theorem 1.1 ) set @xmath23 $ ] .
there exist @xmath24 and @xmath25 , for @xmath26 and @xmath27 [ t]$ ] such that @xmath28 [ [ t]].\ ] ] there is not a general formula for this invariant in terms of a resolution of singularities of @xmath4 ( notice that these kind of formulas exists for other motivic invariants as the _ motivic zeta function _ , see @xcite ) .
there is no conjecture on the meaning of the exponents @xmath29 , @xmath30 which may appear in the denominator of a rational form of @xmath31 .
it is not known if such a rational expression for this series holds in the ring @xmath32 . if @xmath11 is an analytically irreducible germ of plane curve the series @xmath33 determines and it is determined by the multiplicity of @xmath4 ( see @xcite ) .
nicaise has given formulas for @xmath34 for germs @xmath4 with a _ very special _ embedded resolution of singularities @xcite , for instance if @xmath4 is the cone over a smooth hypersurface @xmath35 then @xmath36 is determined by @xmath37 $ ] , @xmath1 and the multiplicity of @xmath4 .
lejeune - jalabert and reguera have studied the motivic poincar series of a germ @xmath4 of normal toric surface at its distinguished point .
they have given a formula for the rational form of this series in terms of the hirzebruch - jung continued fraction describing the resolution of singularities of @xmath38 ( see @xcite and also @xcite for a different approach and comparison with other motivic series ) . if @xmath4 is a germ of affine toric variety of dimension @xmath1 we prove in @xcite that @xmath36 is determined by the newton polyhedra of the _ logarithmic jacobian ideals _ , which are defined in terms of the modules of differential forms with logarithmic poles outside the torus of @xmath38 .
rond studied the coefficients of the series @xmath39 associated to a germ of q.o .
hypersurface and computed the sum of this series in some particular cases ( see @xcite ) . in this paper
we describe the rational form of the geometric motivic poincar series of a germ @xmath40 of q.o . hypersurface singularity .
our approach is independent of rond s . a q.o .
hypersurface singularity @xmath40 has a fractional power series parametrization , which possesses a finite set of _ characteristic monomials _ ( generalizing the characteristic exponents of plane branches ) and which classify the embedded topological class of @xmath41 ( see @xcite ) .
since the normalization @xmath42 of the germ @xmath0 is a toric singularity ( see @xcite ) it is natural to extend the approach in @xcite to the q.o . case . for this reason
we introduce a sequence of monomial ideals @xmath43 of the analytic algebra of the normalization @xmath44 .
these ideals , called the _ logarithmic jacobian ideals _ , are defined first in a combinatorial manner in terms of the characteristic monomials .
inspired by the toric case @xcite , we prove that these ideals can be defined in terms of the composite of canonical maps @xmath45 , where @xmath46 ( resp . @xmath47 and @xmath48 ) denotes the module of holomorphic differential forms of @xmath0 ( resp . over @xmath44 and the module of forms with logarithmic poles at the complement @xmath49 of the torus in @xmath50 ) ,
see section [ torqo ] . in the toric case
the blow up of the ideal @xmath51 is the nash modification ( see @xcite ) . in the q.o .
case the normalized nash modification of @xmath0 is equal to the normalization of @xmath0 followed by the normalized blow - up of the ideal @xmath51 ( see @xcite ) . we study the arc space @xmath52 of @xmath0 by relating it with the arc space @xmath53 of the normalization @xmath44 .
we denote by @xmath54 ( resp . by @xmath55 ) the set of arcs with generic point in the torus of @xmath56 ( resp .
arcs in @xmath57 which lift to an arc in @xmath58 ) .
for @xmath8 we have that @xmath59 ( see @xcite ) but @xmath60 in general if @xmath61 is not normal ( the simplest example is whitney umbrella ) . to avoid this difficulty we consider a finite set of q.o .
coordinate sections @xmath62 of @xmath0 , with @xmath63 if @xmath64 , which are compatible with the toric structure of the normalization .
we define an auxiliary motivic series @xmath65 whose coefficients are the image in the grothendieck ring of the @xmath10-jets of arcs in @xmath66 which are not jets of arcs through any proper q.o .
coordinate section @xmath67 , @xmath68 .
it follows from this that @xmath69 .
this reduces the study of @xmath70 to the study of @xmath65 . since @xmath44 is a toric singularity , the arc space of the torus of @xmath56 acts on @xmath53 .
we use the orbit decomposition of @xmath53 under this action studied by ishii in @xcite .
the set @xmath71 decomposes as a disjoint union of orbits @xmath72 , parametrized by the arc space of the torus , where @xmath73 runs through certain subset of the lattice @xmath74 of one - parametric subgroups of the torus .
we prove that the @xmath10-jets of these orbits are locally closed subsets which are either disjoint or equal and we characterize the equality in combinatorial terms .
then we prove that the coefficient of @xmath75 in the series @xmath65 expand as a sum of classes @xmath76 $ ] , where @xmath73 runs through a finite subset of the lattice @xmath74 .
the description of the rational form of the geometric motivic poincar series @xmath77 , is done by using the methods and combinatorial results of @xcite .
we determine first a formula for the class @xmath76 $ ] in terms of the newton polyhedra of the logarithmic jacobian ideals in theorem [ keyqo ] .
these ideals satisfy similar combinatorial properties in the toric case and in the q.o .
case , though the proofs are more difficult in the second case .
in addition , new combinatorial features are needed to relate the parametrization of @xmath0 with fractional power series with the parametrization of @xmath78 by using the coordinates of the arc space of the torus .
we prove then that the rational form of the series @xmath65 is determined by the newton polyhedra of the logarithmic jacobian ideals and the lattice @xmath74 ( see theorem [ plambdaracqo ] ) .
we deduce a formula for the rational form of @xmath77 , which holds in the ring @xmath79(t)\subset k_0 ( { \var}_\c ) ( t)$ ] , see corollary [ p - geomqo ] .
in particular a finite set of _ candidate poles _ of @xmath70 is obtained .
our result implies that the embedded topological type of the germ @xmath80 determines the series @xmath77 .
the converse is not true even if @xmath81 . as an application we deduce a formula for the _ motivic volume _ of the arc space of a q.o .
hypersurface germ @xmath0 in terms of the logarithmic jacobian ideal @xmath82 ( see corollary [ vol - mot2qo ] ) .
notice that the series @xmath77 and @xmath83 may be quite different ( for instance if the normalization of @xmath0 is smooth ) .
in section [ qoexample ] we give an example of q.o .
surface @xmath0 such that @xmath84 , where @xmath85 is the _
monomial variety _ associated to the q.o .
hypersurface ( see section [ secqo ] ) .
notice that these two series coincide in the one dimensional case .
the paper is organized as follows .
the first three sections introduce the basic notions we need on arc and jets spaces , toric geometry and q.o . singularities .
we describe the orbit decomposition of the arc space of a q.o .
hypersurface in section [ arcsandjetsqo ] . in section [ mainqo ] we state the main results . in section [ conv - newtonqo ]
we give some combinatorial convexity properties of the newton polyhedra of the logarithmic jacobian ideals . in section [ tor - jetqo ]
we prove theorem [ keyqo ] . in section [ genptorbitsqo ]
we describe the series @xmath65 and prove the rationality results . in section [ torqo ]
we define the logarithmic jacobian ideals in terms of differential forms .
the results and proofs in this paper hold if the field @xmath7 is replaced by any algebraically closed field of characteristic zero .
we refer to @xcite for expository papers on arc and jet schemes and/or motivic integration .
we introduce arc and jet spaces on an equidimensional germ of complex analytic variety @xmath4 ( or complex algebraic variety or algebroid ) .
arc and jet spaces can be defined on any algebraic variety ( without fixing the origin of the arcs ) .
we have an embedding @xmath86 in such a way that the germ @xmath87 is defined by the ideal @xmath88 .
an arc @xmath89 \rightarrow ( \a^n_\c , 0)$ ] ( resp .
a @xmath90-jet @xmath91 /(t^{m+1 } ) \rightarrow ( z , 0)$ ] ) is defined by @xmath92 formal power series @xmath93 ( resp .
@xmath92-polynomial expressions ( [ arc - expqo ] ) @xmath94 ) . for any @xmath95 ,
the coefficient of @xmath96 in the series @xmath97 is a polynomial expression @xmath98 , where @xmath99 , for @xmath100 .
this arc ( resp .
@xmath90-jet ) factors through @xmath11 if for any @xmath95 we have @xmath101 .
@xmath102 ) . the arc space @xmath5 ( resp . @xmath90-jet space @xmath103 ) is the reduced scheme underlying the affine scheme @xmath104 , where @xmath105 / ( \a_f^{(k ) } ( \underline{a}^{(1 ) } , \dots , \underline{a}^{(k)}))_{k \geqslant 1 , f \in i}$ ] ( resp .
@xmath106 , where @xmath107 / ( \a_f^{(k ) } ( \underline{a}^{(1 ) } , \dots , \underline{a}^{(k ) } ) ) _ { k = 1 , \dots , m , f \in i } ) $ ] .
we have morphisms of schemes @xmath108 and @xmath109 induced by truncating arcs or jets @xmath110 , for every @xmath111 .
we have that @xmath112 . if @xmath113 is a formal power series and @xmath111 we set @xmath114 .
see @xcite for general references on toric geometry . if @xmath115 is a lattice we denote by @xmath116 ( resp .
@xmath117 ) the vector space spanned by @xmath74 over the field @xmath118 ( resp . over @xmath119 ) . if @xmath120 we denote by @xmath121 the linear subspace spanned by the @xmath122 over @xmath119 . in what follows a _ cone _ mean a _ rational convex polyhedral cone _ : the set of non negative linear combinations of vectors @xmath123 .
the cone @xmath124 is _ strictly convex _ if it contains no lines , in that case we denote by @xmath125 the @xmath125-dimensional face of @xmath124 .
the _ dual cone _
@xmath126 ( resp .
_ orthogonal cone _
@xmath127 ) of @xmath124 is the set @xmath128 ( resp .
@xmath129 ) @xmath130 .
we denote by @xmath131 or by @xmath132 the relative interior of the cone @xmath124 .
@xmath133 is a family of strictly convex cones in @xmath116 such that any face of such a cone is in the family and the intersection of any two of them is a face of each .
the relation @xmath134 ( resp .
@xmath135 ) denotes that @xmath136 is a face of @xmath124 ( resp .
@xmath137 is a face of @xmath124 ) .
the _ support _
( resp . the @xmath138-_skeleton _ ) of the fan @xmath133 is the set @xmath139 ( resp .
@xmath140 ) .
we say that a fan @xmath141 is a _ subdivision _ of the fan @xmath133 if both fans have the same support and if every cone of @xmath141 is contained in a cone of @xmath133 . if @xmath142 for @xmath143 are fans with the same support their intersection @xmath144 is also a fan .
let @xmath124 be a strictly convex cone rational for the lattice @xmath74 .
the toric variety @xmath145 $ ] , denoted also by @xmath146 or @xmath147 , is normal .
the torus @xmath148 is an open dense subset of @xmath149 , which acts on @xmath149 and the action extends the action of the torus on itself by multiplication .
there is a one to one correspondence between the faces @xmath150 of @xmath124 and the orbits @xmath151 of the torus action on @xmath149 , which reverses the inclusions of their closures .
the closure of @xmath151 is the toric variety @xmath152 for @xmath153 .
the orbit @xmath154 is reduced to a closed point called the origin @xmath125 of the toric variety @xmath149 .
the ring @xmath155 of germs of holomorphic functions at @xmath156 is a subring of the ring @xmath157 $ ] of formal power series with exponents in @xmath158 .
if @xmath159 , the _ newton polyhedron _ @xmath160 of the monomial ideal defined by @xmath161 is the minkowski sum of sets @xmath162 . the _ support function _
@xmath163 of the polyhedron @xmath164 is defined by @xmath165 , @xmath166 .
a vector @xmath167 defines the face @xmath168 of the polyhedron @xmath160 .
all faces of @xmath160 are of this form , the compact faces are defined by vectors @xmath169 .
the _ dual fan _
@xmath170 associated to an integral polyhedron @xmath171 is a fan supported on @xmath124 which is formed by the cones @xmath172 , for @xmath173 running through the faces of @xmath174 .
notice that if @xmath175 and if @xmath176 then we have that @xmath177 and we denote this face of @xmath171 also by @xmath178 .
the affine varieties @xmath179 corresponding to cones in a fan @xmath133 glue up to define a _
toric variety _ @xmath180 .
the subdivision @xmath141 of a fan @xmath133 defines a _ toric modification _ @xmath181 .
if @xmath182 is a monomial ideal of @xmath149 @xmath183 , the toric modification @xmath184 is the _
normalized blowing up _
of @xmath149 centered at @xmath161 ( see @xcite for instance ) .
a germ @xmath40 of complex analytic variety equidimensional of dimension @xmath1 is _ quasi - ordinary _ ( q.o . ) if there exists a finite projection @xmath185 which is a local isomorphism outside a normal crossing divisor .
if @xmath40 is a hypersurface there is an embedding @xmath186 , defined by an equation @xmath187 , where @xmath188 $ ] is a _
. polynomial _ : a weierstrass polynomial with discriminant @xmath189 of the form @xmath190 for a unit @xmath191 in the ring @xmath192 of convergent power series in the variables @xmath193 and @xmath194 .
we suppose that @xmath0 is analytically irreducible , that is @xmath195 $ ] is irreducible .
the jung - abhyankar theorem guarantees that the roots of a q.o .
polynomial @xmath196 , called _
branches _ , are fractional power series in @xmath197 , for @xmath198 ( see @xcite ) .
[ expo ] ( see @xcite , prop .
let @xmath199 $ ] be an irreducible q.o .
let @xmath200 be a root of @xmath196 with expansion : @xmath201 there exists @xmath202 such that if @xmath203 and @xmath204 for @xmath205 , then : 1 .
@xmath206 and if @xmath207 then @xmath208 where @xmath209 is the unique integer such that @xmath210 and @xmath211 ( where @xmath212 means coordinate - wise and we convey that @xmath213 ) .
2 . for @xmath205
, we have @xmath214 , hence the index @xmath215 $ ] is @xmath216 . the exponents @xmath217 in lemma [ expo ] above ( resp .
the monomials @xmath218 ) are called _ characteristic _ of the q.o . branch @xmath200 .
we denote by @xmath219 the lattice @xmath220 and we call it the lattice associated to the q.o . branch @xmath200 .
we denote its dual lattice by @xmath74 . for convenience
we denote @xmath221 . without loss of generality
we relabel the variables @xmath222 in such a way that if @xmath223 for @xmath205 , then we have : @xmath224 where @xmath225 is lexicographic order . the q.o .
branch @xmath200 is normalized if @xmath226 is not of the form @xmath227 with @xmath228 .
lipman proved that the germ @xmath0 can be parametrized by a normalized q.o .
branch ( see @xcite , appendix ) .
we assume from now on that the q.o .
branch @xmath200 is normalized .
the semigroup @xmath229 has a minimal set of generators @xmath230 which is a basis of the lattice @xmath231 .
the dual basis of the dual lattice @xmath232 spans a regular cone @xmath233 in @xmath234 .
it follows that @xmath235 , where @xmath236 is the dual cone of @xmath233 .
the @xmath7-algebra @xmath237 is isomorphic to @xmath238 .
this isomorphism sends the monomial @xmath239 in the monomial @xmath240 if @xmath241 . the local algebra @xmath242/(f)$ ] of the singularity @xmath0 is isomorphic to @xmath243 $ ] . by lemma [ expo ]
the series @xmath244 can be viewed as an element @xmath245 of the algebra @xmath246 .
[ normal ] ( see @xcite ) .
the homomorphism @xmath247 is the inclusion of @xmath248 in its integral closure in its field of fractions .
the previous lemma shows that the normalization @xmath42 of a q.o .
hypersurface singularity @xmath40 is the germ of the toric variety @xmath249 at the distinguished point . more generally , the normalization of a q.o .
singularity , non necessarily hypersurface , is a toric singularity ( see @xcite and @xcite ) .
if @xmath250 is the normalization map the composite @xmath251 is a q.o .
projection , since it is the toric map defined by the inclusion of algebras , @xmath252 , induced by the finite index lattice extension @xmath253 ( see @xcite ) .
if @xmath254 , the toric map between the orbit closures @xmath255 is the composite of @xmath256 where @xmath257 and @xmath258 denote respectively the restrictions of @xmath92 and @xmath259 and @xmath260 is the coordinate section of @xmath40 given by : @xmath261 . it follows that @xmath67 is q.o .
( see @xcite ) .
the germ @xmath67 is a q.o .
hypersurface of dimension @xmath262 parametrized by the series @xmath263 . for @xmath264 , we call @xmath265 the q.o .
coordinate section associated to @xmath136 , we denote by @xmath266 the lattice associated to the q.o .
branch @xmath267 and by @xmath268 its dual lattice .
[ notlattqo ] notice that the dual cone of @xmath269 is equal to the image @xmath270 of the cone @xmath233 in the quotient vector space @xmath271 .
we have finite index lattice extensions @xmath272 .
the map @xmath257 in ( [ tor - sec ] ) is a ramified covering with @xmath273 $ ] sheets , which is unramified over the torus .
the index @xmath274 $ ] is equal to the number of irreducible components of the germ of @xmath61 at a generic point of @xmath265 ( cf .
proposition 4.5.4 in @xcite ) .
the elements of @xmath219 defined by : @xmath275 and @xmath276 for @xmath277 , span the semigroup @xmath278 .
the semigroup @xmath279 defines an analytic invariant of the germ @xmath0 ( see @xcite ) .
[ zs ] the _ monomial variety _ associated to @xmath40 is the toric variety @xmath280 $ ] . following teissier s approach @xcite , @xcite ,
the singularity @xmath0 can be presented , after re - embedding in a suitable affine space of larger dimension , as the generic fiber of a @xmath281-parametrical deformation with special fiber equal to the monomial variety @xmath85 ( see @xcite ) .
this family is equisingular in the sense that one toric morphism of the affine space provides a simultaneous embedded resolution of singularities of the family ( see @xcite for the one dimensional case , @xcite in the q.o .
case and @xcite for related results in a more general context ) .
the normalizations of @xmath0 and of @xmath282 coincide .
see @xcite for the properties of the equisingular deformation of @xmath85 with generic fiber @xmath283 .
in this section we study the arcs in the q.o . hypersurface @xmath0 by using the toric structure of @xmath284 , following the approach of @xcite .
we keep notations of section [ secqo ] .
recall that the normalization @xmath56 of @xmath61 is equal to the toric variety @xmath285 .
the set @xmath54 of arcs of @xmath53 with generic point in the torus is @xmath286 any arc @xmath287 defines two group homomorphisms @xmath288 and @xmath289^ * $ ] by @xmath290 if @xmath291 then we have that @xmath292 hence @xmath293 belongs to @xmath294 .
it follows that @xmath295 defines an arc in the torus @xmath296 ( see @xcite ) .
we define similarly the set @xmath297 : ( see @xcite ) [ estrella ] @xmath298 .
the set @xmath299 consists of those arcs @xmath300 such that there exists @xmath301 such that @xmath302 .
notice that if @xmath303 then the generic point of @xmath304 is not contained in the singular locus @xmath305 of @xmath61 , since @xmath306 by definition of @xmath259 .
by the valuative criterion of properness applied to the normalization map , there exists a unique arc @xmath307 such that @xmath308 . since ( [ composite ] ) defines a q.o
. projection it follows that @xmath301 and @xmath309 defines a bijective correspondence between the sets @xmath310 and @xmath311 . for any @xmath312
we define the sets : @xmath313 ishii noticed that the space of arcs in the torus acts on the arc space of a toric variety ( see @xcite ) .
the set @xmath314 is an orbit of the action of the arc space of the torus of @xmath315 ( see @xcite ) .
the map @xmath316 defines a bijective correspondence between the sets @xmath317 and @xmath318 .
we usually denote the set @xmath319 by @xmath320 .
the set @xmath58 is invariant for this action of the arc space of the torus and we obtain the partitions : @xmath321 and @xmath322 .
notice that @xmath323 can be defined also as the set of arcs @xmath324 such that the arc @xmath325 has generic point in the torus , where we consider the affine space @xmath326 with the structure of toric variety @xmath327 . if @xmath328 then @xmath329 factors through a unique minimal orbit closure , of the form @xmath330 , in such a way that @xmath329 has generic point in the torus of @xmath331 . in this case , it follows from the properties of ( [ composite ] ) that the arc @xmath304 factors through @xmath67 and belongs to the set @xmath332 .
we deduce from this observations the following partition of the arc space @xmath333 .
we state in this section the main results of the paper .
we keep notations of the previous sections .
[ dplusg ] we denote by @xmath334 the elements of the canonical basis of @xmath335 . we also denote the characteristic exponent @xmath336 of the q.o . branch @xmath200 by @xmath337 , for @xmath205 .
we set @xmath338 and @xmath339 .
we introduce the following subsets of @xmath340 associated to the q.o .
branch @xmath200 parametrizing the germ @xmath0 .
@xmath341 we call the monomial ideal of @xmath342 defined by ( [ j - kqo ] ) the _ @xmath343-logarithmic jacobian ideal _ of @xmath0 relative to the q.o . projection @xmath259 , for @xmath344 .
we abuse slightly of notation by denoting this ideal also by @xmath345 .
we denote by @xmath346 ( resp . by @xmath347 ) the dual subdivision of @xmath233 ( resp .
the support function ) associated to the newton polyhedron of the @xmath343-logarithmic jacobian ideal @xmath348 , for @xmath349 .
[ idealesjac ] these ideals can be defined in terms of holomorphic differential forms with logarithmic poles outside the torus of @xmath315 ( see section [ torqo ] ) .
the terminology is inspired by @xcite and @xcite .
now we introduce an auxiliary series to study @xmath350 . by greenberg s theorem , for any @xmath254 we have that @xmath351 is a constructible subset of the @xmath10-jet space of @xmath0 .
it follows that @xmath352 is also a constructible subset .
taking the images of theses sets in the grothendieck ring it is natural to consider the auxiliary poincar series , @xmath353t^s \in k_0 ( { \var}_\c ) [ [ t]],\ ] ] which measures the jets of arcs through @xmath0 which do not factor through any proper q.o .
coordinate section @xmath354 , @xmath355 .
we will show that the coefficient of @xmath356 in the series @xmath65 is a finite sum of classes @xmath357 $ ] .
notice that if @xmath358 the series @xmath359 is defined similarly by formula ( [ auxiliarpqo ] ) .
4.1 @xcite ) [ descomppgeomqo ]
we have the equality @xmath360 .
@xmath361 [ p - codim1qo ] 1 .
the series @xmath362 takes into account those jets of arcs in @xmath57 which truncate to @xmath125 .
we have that @xmath363 .
if @xmath364 is of codimension one then @xmath365 is a plane curve .
we have that @xmath366 , where @xmath90 is the multiplicity of @xmath265 ( see proposition 10.2.1 in @xcite ) . recall that if @xmath133 is a fan then @xmath367 denotes the set of one dimensional cones of @xmath133 ( see section [ sec - tor ] ) .
notation 4.7 @xcite ) [ jotak - bisqo ] the following maps are piece - wise linear functions on the cone @xmath233 : @xmath368 we define @xmath369 and @xmath370 by convenience . if @xmath371 is a cone of dimension one , we denote by @xmath372 the generator of the semigroup @xmath373 .
we define the finite set : @xmath374 this definition applies for the q.o .
sections @xmath265 , for @xmath375 .
for @xmath376 we set @xmath377 .
[ ak ] ( cf .
@xcite ) for @xmath378 we set @xmath379 [ keyqo ] ( cf .
7.1 @xcite ) if @xmath380 then the jet space @xmath381 is a locally closed subset of @xmath382 isomorphic to @xmath383 if @xmath384 or to @xmath385 if @xmath386 .
theorem [ keyqo ] is essential to prove the main results of the paper : ( cf
4.9 @xcite ) [ plambdaracqo ] let @xmath0 be an irreducible germ of q.o . hypersurface .
then there exists a polynomial @xmath387 $ ] determined by the lattice @xmath219 and the newton polyhedra of the logarithmic jacobian ideals of @xmath0 such that @xmath388 combining theorem [ plambdaracqo ] with proposition [ descomppgeomqo ] we deduce the following corollary : ( cf .
cor . 4.10 @xcite ) [ p - geomqo ]
we have that @xmath389 the series @xmath390 depends only on the logarithmic jacobian ideals and lattices @xmath391 associated to the q.o .
sections @xmath265 for @xmath392 ( with respect to the q.o .
projection @xmath259 ) . as an application
we give a formula for the _ motivic volume _ of the arc space of a q.o .
we refer to @xcite for the definition of _ measurable sets _ and properties of the _
motivic volume_. let @xmath124 be a strictly convex cone rational for the lattice @xmath74 .
the _ generating series _ @xmath393 has a rational form @xmath394\ ] ] ( see @xcite for instance ) . if @xmath395 we denote by @xmath396\rightarrow\z[\l^{\pm 1}]$ ] the toric map given by @xmath397 .
10.1 @xcite ) the motivic volume of the arc space @xmath57 of the q.o .
hypersurface @xmath40 is equal to @xmath398 [ vol - mot2qo ] in section [ qoexample ] we give an example of q.o .
surface singularity such that all the candidate poles in corollary [ p - geomqo ] are actual poles .
we give a geometrical interpretation of the set of candidate poles of the series @xmath399 . for @xmath400
we denote by @xmath401 the composite of the normalization map @xmath402 with the toric modification of @xmath403 defined by the subdivision @xmath404 of @xmath233 .
the modification @xmath405 is the minimal modification with normal source , which factors through the normalization of @xmath61 and the normalized blowing up of @xmath56 with center @xmath406 , for @xmath407 .
the rays @xmath408 in the fan @xmath409 correspond bijectively to orbit closures of @xmath410 which are of codimension one . if @xmath372 is the generator of the semigroup @xmath373 we denote by @xmath411 the irreducible component corresponding to @xmath408 .
we denote by @xmath412 the divisorial valuation of the field of fractions of @xmath413 , which is associated to the divisor @xmath411 .
if @xmath414 then we have that @xmath415 we have that @xmath416 if and only if @xmath411 is a codimension one irreducible component of the exceptional fiber of @xmath417 . if @xmath418 the pull - back @xmath419 of @xmath406 by @xmath405 is a locally principal monomial ideal sheaf on the toric variety @xmath410 and by ( [ val1 ] ) we have that @xmath420 [ val3 ] for @xmath400 we have that @xmath421 define locally principal monomial ideal sheaves on @xmath410 such that @xmath422
in this section we give a series of results on the properties of the support functions of the logarithmic jacobian ideals associated to a q.o .
hypersurface germ @xmath0 , which are used in sections [ tor - jetqo ] and [ genptorbitsqo ] .
if @xmath423 we denote by @xmath424 the partial order on @xmath219 defined by @xmath425 if @xmath426 .
[ notationorder ] the logarithmic jacobian ideals @xmath427 of the normalization @xmath44 , which are studied in @xcite , are different than those of @xmath0 in general .
recall that we have @xmath249 .
if @xmath428 are the minimal sequence of generators of the semigroup @xmath340 then we have that @xmath429 .
the combinatorial convexity properties of the support functions , @xmath430 for @xmath431 , are simpler in the toric case .
given @xmath432 , up to relabeling , we can assume for simplicity that @xmath433 .
we define inductively @xmath434 and @xmath435 for @xmath436 .
then we have that @xmath437 and @xmath438 .
the following example shows that the q.o . case is not as simple as the toric case .
[ examalg ] consider a q.o .
branch with characteristic exponents @xmath439 and @xmath440 . the points @xmath441 and @xmath442 belong to the lattice @xmath74 .
it is easy to check that @xmath443 , @xmath444 , @xmath445 and @xmath446 .
then we get @xmath447 while @xmath448 .
we fix a partial order on the set @xmath449 by @xmath450 in such a way that if @xmath451 for @xmath452 and @xmath453 , then @xmath454 . by lemma [ expo ]
we have the inequalities @xmath455 in the following proposition we use the convention @xmath456 , for @xmath457 ( see notations [ dplusg ] ) .
5.1 @xcite ) let @xmath458 .
with respect to the order ( [ ordenqo ] ) , the vector @xmath73 defines a sequence @xmath459 such that @xmath460 and @xmath461 .
we set @xmath462 .
suppose that @xmath463 is already defined .
set : 1 .
@xmath464 2 .
@xmath465 3 .
@xmath467 5 .
@xmath468 , for @xmath469 .
set @xmath470 and @xmath471 .
then we have that : @xmath472 [ pruebaalgqo ] we prove it by induction on @xmath138 . for @xmath473
the assertion holds , since @xmath474 .
we suppose that the statement is true for @xmath138 and we prove it for @xmath475 . by induction hypothesis
we have @xmath476 and @xmath477 . if @xmath478 the argument coincides with the proof of proposition 5.1 in @xcite .
we assume that @xmath479 and that there exists @xmath480 different from @xmath481 and such that @xmath482 if the vector @xmath483 with @xmath484 and @xmath485 appears in the expansion of @xmath486 as sum of @xmath475 linearly independent elements of @xmath449 , then @xmath483 appears in the expansion of @xmath481 .
_ proof of the assertion_. if @xmath483 does not appear in the expansion of @xmath481 , then @xmath487 belongs to @xmath345 and we deduce @xmath488 , where the first inequality follows by the induction hypothesis and the second by ( [ hipalgqo ] ) .
then we get @xmath489 and since @xmath485 the vector @xmath490 belongs to @xmath491 , but this is in contradiction with the choice of @xmath481 in the algorithm , hence the assertion holds .
now we distinguish various cases : 1 . if @xmath492 , then we obtain @xmath493 by definition . by the assertion there is an integer @xmath494 such that @xmath495 . by ( [ ch - order2 ] ) and ( [ hipalgqo ] ) we deduce that @xmath496 , but then @xmath497 and @xmath498 , this is a contradiction with the induction hypothesis .
2 . if @xmath499 and @xmath493 and if @xmath500 does not appear in the expansion of @xmath486 we apply the argument of case ( i ) to get a contradiction .
3 . if @xmath499 , @xmath493 and if @xmath500 appears in the expansion of @xmath486 then by the assertion we have that @xmath501 and @xmath502 , where @xmath503 . by definition @xmath504 , hence @xmath505 does not belong to @xmath506 since @xmath480 . by definition of @xmath507
we deduce that @xmath508 .
then it follows that @xmath509 , which contradicts ( [ hipalgqo ] ) .
4 . if @xmath499 and @xmath510 the assertion implies that @xmath500 appears in the expansion of @xmath511 . we deduce that @xmath512 and @xmath513 . formula ( [ hipalgqo ] ) implies that @xmath514 which is a contradiction with the definition of @xmath507 .
@xmath515 the two choices appearing in ( [ rithm ] ) occur ( see example [ examalg ] ) .
[ op2 ] notice that @xmath516 implies @xmath517 and @xmath518 ( otherwise @xmath519 would verify that @xmath520 ) .
[ c1 ] ( cf .
lemma 5.3 @xcite ) for all @xmath458 we have that @xmath521 the assertion is equivalent to the inequality @xmath522 for @xmath523 , where @xmath524 and @xmath481 are defined by the algorithm and @xmath525 . we distinguish the cases : 1 . if @xmath526 and @xmath527 the result follows by definition of @xmath528 in proposition [ pruebaalgqo ] .
2 . if @xmath526 and @xmath529 , we get @xmath530 and @xmath531 .
we distinguish two subcases below : 1 .
if @xmath532 we get @xmath533 by the definition of the algorithm hence we deduce @xmath534 since @xmath535 .
2 . if @xmath536 we obtain @xmath537 and @xmath538 .
3 . if @xmath539 and @xmath527 then we get @xmath540 and @xmath541 . since the vector @xmath542 is not a characteristic exponent
we deduce that @xmath543 .
then we get the inequalities @xmath544 , and it follows that : @xmath545 .
4 . if @xmath539 and @xmath529 we obtain @xmath540 and @xmath510 . since the vector @xmath546 belongs to @xmath345 we deduce @xmath547 .
we get @xmath548 .
[ ele ] if @xmath549 there is a unique integer @xmath378 such that @xmath380 .
we set @xmath550 ( cf . definition [ ak ] , proposition [ pruebaalgqo ] and lemma [ c1 ] ) .
[ crucial ] ( cf .
lemma 5.7 @xcite ) if @xmath380 and @xmath551 is a vector such that @xmath552 , then we have @xmath553 in addition , if @xmath554 and @xmath555 then @xmath556 implies that @xmath557 .
let @xmath558 be the vector defined by proposition [ pruebaalgqo ] .
we denote by @xmath559 ( resp .
by @xmath560 ) the linear subspace of @xmath561 spanned by the vectors in the expansion of @xmath562 ( resp .
@xmath563 ) .
we prove first that @xmath564 .
if @xmath560 and @xmath565 are distinct we verify that @xmath566 .
suppose that there exists a vector @xmath567 appearing in the expansion of @xmath563 and such that @xmath568 .
we distinguish two cases : 1 .
if @xmath569 then the vector @xmath570 belongs to @xmath491 .
we get @xmath571 hence @xmath572 .
if @xmath573 then the vector @xmath574 belongs to @xmath575 and we find the contradiction @xmath576 .
hence we obtain @xmath577 thus @xmath566 holds by lemma [ c1 ] .
2 . if for any @xmath484 , with @xmath483 appearing in the expansion of @xmath563 then @xmath578 , then we get @xmath579 and @xmath580 , for @xmath581 and @xmath582 .
then @xmath583 and by ( [ ch - order2 ] ) the vector @xmath584 can not be a characteristic exponent .
thus @xmath478 and we get @xmath585 .
we obtain @xmath586 thus @xmath566 holds by lemma [ c1 ] .
we obtain that @xmath587 by checking that @xmath588 for @xmath589 . if @xmath473 the inequality is trivial
. if @xmath590 and @xmath591 is a vector appearing in the expansion of @xmath562 then there are two possibilities . 1 . if there is @xmath592 such that @xmath593 then we get @xmath594 . by lemma [ c1 ]
we deduce @xmath595 .
2 . if there is @xmath592 such that @xmath596 , with either @xmath597 or @xmath598 then in both cases we get @xmath599 , since @xmath600 and @xmath601 . we prove that @xmath602 . if @xmath603 for @xmath604 and if @xmath605 then @xmath606 belongs to @xmath491 ; we deduce that @xmath607 contradicting the hypothesis .
we have also shown that @xmath608 hence if @xmath609 then @xmath610 by definition of @xmath507 . if @xmath554 then @xmath611 by ( [ ch - order2 ] ) and it is easy to see from the algorithm in proposition [ pruebaalgqo ] that @xmath612 belongs to @xmath613 .
it follows that @xmath602 . for the last assertion notice that if @xmath614 and if @xmath615 then we get a contradiction since @xmath616 belongs to @xmath345 and verifies that @xmath617 .
@xmath618 [ lema - obs ] if @xmath380 we have the following inequalities 1 .
@xmath619 if @xmath620 or if @xmath621 and @xmath622 .
2 . @xmath623 if @xmath624 .
_ if @xmath620 then the vector @xmath625 belongs to @xmath491 hence @xmath626 .
we deduce from this that @xmath627 . if @xmath628 then we get @xmath629 by proof of lemma [ c1 ] .
this implies that @xmath630 since @xmath631 by remark [ op2 ] .
if @xmath380 , @xmath632 it may happen that @xmath633 .
for instance , consider a q.o .
branch with characteristic exponents @xmath634 and @xmath635 .
if @xmath636 then we get @xmath637 , @xmath638 and @xmath639 .
if @xmath640 then we obtain @xmath641 , @xmath642 , @xmath643 , @xmath644 and @xmath645 . 1
. if @xmath646 we denote by @xmath647 the smallest coordinate subspace containing @xmath648 .
2 . with respect to the fixed vector @xmath649 , we set @xmath650 .
[ deflandaq ] [ nk ] it is easy to see from the algorithm in proposition [ pruebaalgqo ] that if @xmath651 then @xmath652 and if @xmath653 then @xmath654 . let @xmath655 for some @xmath656 . with the notations of proposition [ pruebaalgqo ] we define the integer @xmath657 by @xmath658 [ defpk ] notice that @xmath659 .
[ ap3crucial ]
if @xmath655 for @xmath400 and @xmath660 the following inequality holds @xmath661 if in addition @xmath662 and @xmath663 we have @xmath664 suppose first that @xmath665 for some @xmath666 . it follows that ( [ cla2 ] ) holds by remark [ op2 ] .
we deduce also that ( [ cla ] ) holds since @xmath667 hence @xmath668 .
we deal first with the case @xmath669 . if @xmath670 then the set @xmath671 is non empty and @xmath517 .
we introduce the terms @xmath672 , which are the characteristic exponents appearing in the expansion of the terms @xmath673 defined in proposition [ pruebaalgqo ] by the vector @xmath73 for suitable integers @xmath674 .
first we set @xmath675 and then inductively @xmath676 and @xmath677 .
after finitely many steps we have an integer @xmath304 such that @xmath678 and @xmath679 , and the process stops .
it is easy to check that @xmath680 @xmath681 .
notice that @xmath682 for @xmath683 and if @xmath663 we have also that @xmath684 .
remark that the definition of @xmath685 in proposition [ pruebaalgqo ] involves the choice @xmath686 in ( [ rithm ] ) , i.e. , we have : @xmath687 and similarly by lemma [ c1 ] and definition [ ak ] , @xmath688 summing the inequalities in ( [ eccrucialc1 ] ) for @xmath689 with ( [ ecu - h ] ) provides the inequality @xmath690 now we distinguish two cases : \(i ) if @xmath691 then @xmath692 by definition
. then the inequality ( [ cla2 ] ) hold by ( [ eccrucialc1prima ] ) since @xmath693 . adding ( [ eccrucialc1prima ] ) with the inequality @xmath694 shows that ( [ cla ] ) holds .
\(ii ) otherwise @xmath695 and we have that @xmath696 for @xmath697 .
the characteristic exponent @xmath612 is not chosen at any step of the algorithm in proposition [ pruebaalgqo ] , that is , @xmath698 then we check that ( [ cla2 ] ) holds by adding the inequalities ( [ eccrucialc1primados ] ) and ( [ eccrucialc1prima ] ) .
the same happens for ( [ cla ] ) by adding ( [ eccrucialc1primados ] ) , ( [ eccrucialc1prima ] ) and ( [ ecu - p ] ) .
it remains to prove that ( [ cla ] ) and ( [ cla2 ] ) hold for @xmath699 .
in this case @xmath700 and the inequalities hold trivially by the definition of @xmath657 since @xmath701 .
[ nota ] in this section we fix @xmath380 for some @xmath703 and we simplify our notations .
let @xmath704 be the vector defined by @xmath73 in proposition [ pruebaalgqo ] .
we relabel the vectors in @xmath705 in such a way that @xmath706 if @xmath478 , and @xmath707 if @xmath662 .
we denote the integers @xmath708 and @xmath657 , defined in the section [ conv - newtonqo ] in terms of @xmath73 , @xmath138 and @xmath10 , simply by @xmath709 , @xmath710 , and @xmath711 , respectively .
we begin by recalling some definitions and results from @xcite .
let @xmath712 be a basis of the lattice @xmath219 .
we consider a set of variables @xmath713 to define the @xmath7-algebra @xmath714\otimes_\c \c [ u_j(m_i)]_{i=1,\dots , d}^{j\geqslant 1}.\ ] ] since @xmath712 form a basis of @xmath219 there is a unique homomorphism of semigroups @xmath715^*\ ] ] given by @xmath716 where @xmath717 for @xmath718 .
we associate to @xmath719 the terms @xmath720 and @xmath721 , for @xmath722 , in the ring @xmath723 , by @xmath724 in ( [ family ] ) , where @xmath720 is the constant term of the series and @xmath725 is of the form @xmath726 .
the following result show some relations among the elements @xmath727 , when we vary @xmath728 and @xmath729 , in terms of linear dependency relations among the @xmath729 .
[ vectorqo ] ( see lemma 6.2 in @xcite ) .
let @xmath730 be linearly independent vectors in the lattice @xmath219 spanning the linear subspace @xmath565 of @xmath731 .
for any @xmath732 and @xmath733 the term @xmath734 belongs to @xmath735_{j=1}^k.$ ] .
we use in an essential manner the parametrization of the sets @xmath318 by the arc space of the torus @xmath296 .
we have an homomorphism of semigroups : @xmath736 , \cdot ) \mbox { given by } m \mapsto t^{\langle\nu , m\rangle}c(m)u(m ) \mbox { , for } m\in\s^\vee\cap m.\ ] ] it defines a parametrization of @xmath737 by the arc space of the torus @xmath296 .
recall that by lemma [ normal ] we have that the analytic algebra of the germ @xmath0 is @xmath738 $ ] .
the maximal ideal of @xmath739 is @xmath740 where @xmath741 for @xmath718 and @xmath742 ( see the notations of section [ secqo ] ) .
the restriction of the homomorphism @xmath743 $ ] defined by ( [ fam ] ) to the local algebra @xmath744 parametrizes the set @xmath745 .
this homomorphism verifies that @xmath746 where @xmath747 is a series of the form @xmath748 .
for @xmath749 we have that @xmath750 is the image of @xmath751 by the map ( [ family ] ) , in particular @xmath752 .
we use the expansion ( [ expan ] ) of @xmath753 as a power series in @xmath342 to describe expansions of the terms @xmath754 in terms of @xmath755 , for @xmath756 and @xmath757 . for @xmath758
we denote by @xmath759 the @xmath7-algebra of @xmath723 generated by @xmath760 for @xmath761 .
we denote @xmath762 simply by @xmath763 .
[ defcj ] [ notr_j ] if @xmath764 and @xmath765 are fixed we set the sequence @xmath766 by : @xmath767 [ salva ] notice that if @xmath768 then we have that @xmath769 for every @xmath758 .
[ ur ] let @xmath770 be an integer .
in @xmath723 we have the expansion : @xmath771 we use notation [ notr_j ] .
if @xmath772 and @xmath773 then we set @xmath774 the following properties hold : 1 . @xmath775 .
2 . the coefficient @xmath776 of a term @xmath777 in @xmath778 belongs to @xmath759 .
if the term @xmath777 appears in @xmath778 then @xmath779 unless @xmath780 and @xmath781 , for @xmath782 .
[ prop7 - 1 ] we deduce that @xmath783 by comparing the definition of @xmath784 with the expansion ( [ expan ] ) of @xmath785 in @xmath342 .
the equality follows by collecting the terms in @xmath786 in ( [ exp(d+1 ) ] ) .
the sum in ( [ zc2 ] ) is finite because there is a finite number of lattice points @xmath787 verifying that @xmath788 and @xmath789 , since @xmath790 . notice that @xmath791 by definition and the term obtained for @xmath792 ( resp .
@xmath793 ) in the sum ( [ zc2 ] ) is @xmath794 ( resp .
@xmath795 ) .
the other assertions are obtained by collecting the indices in the sum ( [ zc2 ] ) according to ( [ rsequence ] ) .
for simplicity we convey that the term @xmath778 equals zero whenever @xmath796 .
hence we can write @xmath797 for any @xmath798 [ alg - indqo ] ( see proposition 6.3 in @xcite ) .
if the vectors @xmath799 in @xmath219 are linearly independent then the terms , @xmath800 for @xmath801 , are algebraically independent over the field of fractions of @xmath763 , and in addition if @xmath802 they generate @xmath803 as a @xmath763-algebra .
[ atn ] let us denote by @xmath804 the @xmath7-algebra @xmath805_{i=1 , \dots d+1}.\ ] ] we set @xmath806 for @xmath807 .
suppose that for @xmath808 the algebras @xmath809 have been defined by induction .
then we have that @xmath810 is a multiplicative subset of @xmath811 and the localization @xmath812 $ ] is a subalgebra of @xmath803 which is equal to @xmath803 if @xmath813 .
notice that @xmath814[f_j^{-1}]=c_j$ ] .
by definition we have that @xmath815 for @xmath816 hence it follows that @xmath817 is a multiplicative subset
. then we apply proposition [ ur ] for @xmath818 .
we deduce that @xmath819 where @xmath820 belongs to @xmath821 .
it follows that @xmath822 belongs to @xmath821 and this implies that @xmath823 is a multiplicative subset of @xmath821 .
suppose that the assertion is true for @xmath824 .
then we apply proposition [ ur ] for @xmath825 and we deduce similarly that @xmath826 is a multiplicative subset of @xmath811 .
finally , if @xmath827 we deduce that @xmath828 is a @xmath763-subalgebra of @xmath803 ( see notation [ defcj ] ) which contains the set @xmath829 .
this set generate @xmath803 as a @xmath763-algebra by proposition [ alg - indqo ] , since the vectors @xmath334 are linearly independent . the idea of the proof of theorem [ keyqo ] is rather similar to that of proposition [ atn ] , though we have to precise the form of the terms @xmath830 which remains when taking a @xmath10-jet .
this is controlled by the following proposition : [ ur2 ] suppose that @xmath380 .
we consider an integer @xmath831 such that @xmath832 the terms @xmath833 in the expansion @xmath834 vanish for @xmath835 .
if in addition , the term @xmath836 appears with non zero coefficient in the expansion of @xmath837 , for some @xmath838 , then @xmath839 by proposition [ ur ] we have that a term @xmath840 may be non - zero in the expansion of @xmath841 if @xmath842 ( see notation [ notr_j ] ) . by remark
[ salva ] this happens if @xmath843 this holds if and only if @xmath844 ( see notation [ nota ] ) .
suppose that @xmath836 appears with non zero coefficient in @xmath840 for some @xmath845 .
this implies that @xmath846 appears in the expansion ( [ expan ] ) of @xmath200 , @xmath847 and @xmath848 , thus @xmath208 and @xmath849
. then we obtain @xmath850 we deduce from these inequalities , using ( [ cla ] ) , @xmath851 and @xmath852 that @xmath853 since @xmath208 is @xmath854 we have a unique expansion of the form @xmath855 , where @xmath856 belongs to @xmath857 and @xmath858 are integers . since @xmath859 it follows from proposition [ crucial ] that @xmath860 belongs to @xmath861 .
we deduce that @xmath862 and @xmath863 if @xmath864 .
this implies that @xmath865 if @xmath866 and together with ( [ cla3 ] ) proves that @xmath867 .
we denote by @xmath868 the @xmath7-algebra of @xmath723 generated by : @xmath869 we set @xmath870 the @xmath7-algebra generated by @xmath871 [ algebrao ] [ pro - aux ] we use notations [ nota ] and [ algebrao ] .
the @xmath870-algebra generated by @xmath868 coincides with the @xmath870-algebra of @xmath803 generated by the following elements : @xmath872 together with @xmath873 we consider the case @xmath874 ( resp .
@xmath875 and @xmath876 ) .
by lemma [ crucial ] we have that if @xmath877 and @xmath878 then @xmath879 .
@xmath880 ) .
it is sufficient to prove that the elements @xmath881 for @xmath882 belong to the @xmath870-algebra generated by ( [ one ] ) and ( [ two ] ) . by propositions [ ur ] and [ ur2 ]
we get @xmath883 where @xmath884 , @xmath885 and @xmath886 .
this implies that @xmath887 ( resp .
@xmath888 ) and @xmath865 for @xmath889 ( resp . for @xmath890 ) thus by lemma [ vectorqo ] we deduce that @xmath891_{i= 1\ldots k ( \mbox{{\rm \small resp . } } k-1)}$ ] .
we deal then with the case @xmath875 and @xmath892 .
it is sufficient to prove that the elements @xmath881 for @xmath893 belong to the @xmath870-algebra generated by ( [ one ] ) and ( [ two ] ) .
the conclusion follows by the same arguments , by using propositions [ ur ] , [ ur2 ] and lemma [ vectorqo ] . [ atn2 ] we set @xmath894 for @xmath895 .
suppose that for some @xmath896 the algebras @xmath897 have been defined by induction .
then @xmath898 is a multiplicative subset of @xmath899 and the localization @xmath900 $ ] is a subalgebra of @xmath803 .
then the ring @xmath901 is generated by @xmath868 as a @xmath870-algebra .
the ring @xmath902 is the coordinate ring of @xmath903 . _
proof_. set @xmath904 [ ( \bar{f}_j)^{-1 } ] $ ] . by definition
we have that @xmath815 for @xmath905 , hence it follows that @xmath906 is a multiplicative subset .
we apply propositions [ ur ] and [ ur2 ] for @xmath907 .
we deduce that @xmath819 where @xmath820 belongs to @xmath908 .
it follows that @xmath822 belongs to @xmath908 and this implies that @xmath909 is a multiplicative subset of @xmath908 .
suppose that the assertion is true for @xmath910 .
then we apply proposition [ ur2 ] for @xmath911 and we deduce similarly that @xmath898 is a multiplicative subset of @xmath899 .
we can iterate this procedure for @xmath912 . by proposition
[ pro - aux ] it follows that @xmath913 is the @xmath870-algebra generated by @xmath914 . by construction @xmath913 is the coordinate ring of @xmath903 ( compare with proposition [ atn ] ) .
@xmath515 the generators ( [ one ] ) and ( [ two ] ) of the @xmath870-algebra @xmath913 are algebraically independent over the field of fractions of @xmath915 .
[ lemaprekey ] if @xmath916 and if @xmath917 then we have that @xmath754 is of the form : @xmath918 ( see notation [ notr_j ] ) . by proposition [ ur ]
we have that @xmath919 and if @xmath920 appears in the expansion ( [ alfa - r - j ] ) of @xmath921 with non zero coefficient then @xmath922 and @xmath208 verifies that @xmath923
if @xmath924 then we have the inequality @xmath925 .
the equality @xmath926 implies that @xmath927 , thus @xmath787 belongs to @xmath928 since @xmath929 .
if @xmath930 then we have that @xmath931 by ( [ flu ] ) .
now we prove the proposition by distinguishing the following cases : 1 . if @xmath874 then the assertion is consequence of proposition [ alg - indqo ] .
2 . if @xmath875 and @xmath876 the assertion is consequence of ( [ flu2 ] ) and proposition [ alg - indqo ] .
3 . if @xmath875 and @xmath892 then by proposition [ vectorqo ] and the definition of @xmath932 we have that the @xmath870 algebra generated by @xmath933 coincides with the @xmath870 algebra generated by @xmath934 .
we deduce the statement from this , the equation ( [ flu2 ] ) and proposition [ alg - indqo ] .
_ proof of theorem [ keyqo ] . _
the statement is clear if @xmath935 .
suppose that @xmath936 for some @xmath937 and @xmath938 .
by lemma [ lemaprekey ] we have @xmath939 generators of @xmath913 as a @xmath870 algebra .
these generators , indicated in ( [ one ] ) and ( [ two ] ) , are algebraically independent over the field of fractions of @xmath870 .
notice that @xmath870 is the @xmath7 algebra of the rank @xmath138 sublattice of @xmath940 hence @xmath941 is a @xmath138-dimensional torus .
it follows from this and proposition [ atn2 ] that we can define an isomorphic parametrization @xmath942
first we study the relations between the sets @xmath903 when @xmath73 varies .
[ eqqo ] define an equivalence relation in the set @xmath943 for @xmath944 : @xmath945 we denote by @xmath946 $ ] the equivalence class of @xmath655 by this relation .
[ remkqo ] 1 . the set @xmath947\ |\ ( \nu , s_0)\in { a}_k\}$ ] is finite for @xmath948 and @xmath937 .
2 . if @xmath949 the equivalence relation is the equality .
8.3 @xcite ) let @xmath950
. the following are equivalent : 1 .
@xmath951 2 .
@xmath952 and @xmath953 3 .
4 . @xmath955 .
[ prop - equivqo ] the equivalence between ( i ) and ( ii ) follows by lemma [ crucial ] , definition [ ak ] and the definition of the equivalence relation . if ( ii ) holds , then the vectors obtained in the algorithm in proposition [ pruebaalgqo ] coincides for both @xmath73 and @xmath956 .
it follows that the indices @xmath957 , @xmath958 and @xmath657 are the same for @xmath73 and @xmath956 .
these facts imply that the isomorphic parametrizations ( [ etapar ] ) corresponding to @xmath959 coincide .
hence @xmath954 and ( iii ) holds .
if ( iv ) holds we prove that ( ii ) holds .
there exists @xmath960 and @xmath961 such that @xmath962 .
then for any @xmath963 the inequality @xmath964 implies that @xmath965 .
we denote by @xmath958 and @xmath957 ( resp . @xmath966 and @xmath967 ) the integers associated to @xmath968 ( resp .
@xmath969 ) in proposition [ pruebaalgqo ] . if @xmath620 then the assertion follows . if @xmath662 we set @xmath970 and @xmath971 .
we have that @xmath972 .
if @xmath973 appears in @xmath974 and @xmath975 then @xmath976 whenever @xmath977 ( this is consequence of lemma [ crucial ] ) .
in addition , if @xmath978 then it follows by lemma [ vectorqo ] that @xmath979_{1\leqslant i\leqslant d}^{\langle\nu , e_i\rangle\leqslant s}$ ] for @xmath980 .
then it is easy to see that the initial coefficients of @xmath981 and @xmath982 coincide for @xmath983 .
we deduce that @xmath984 hence @xmath985 is of order @xmath986 .
since @xmath987 we deduce that @xmath988 hence @xmath989 and ( ii ) holds .
[ auxiqo ] the cone @xmath990 is rational for the lattice @xmath991 . 1 .
if @xmath992 and @xmath993 we set @xmath994 2 . if @xmath995 then we set @xmath996 .
[ pwlqo ] if @xmath124 is a cone contained in a cone of the fan @xmath997 then we have that if @xmath998 , then the closure of @xmath999 in @xmath1000 is a convex polyhedral cone , rational for the lattice @xmath1001 ( since in this case the functions @xmath1002 , hence also @xmath1003 , are linear on @xmath124 and the function @xmath1004 , hence also @xmath1005 , is piece - wise linear and convex on @xmath124 ) .
the set @xmath1006 may be empty , for instance , if @xmath124 is contained in the boundary of @xmath233 or if for all @xmath73 in the interior of @xmath124 we have that @xmath1007 .
[ 86qo ] @xmath361 1 .
for @xmath400 we have that @xmath1008 .
the vectors @xmath1009 are equivalent by relation @xmath1010 in ( [ equivalence - bis ] ) if and only if there exists a cone @xmath1011 such that @xmath73 and @xmath956 belong to the relative interior of @xmath124 and we have that @xmath1012 , for @xmath1013 .
it follows that @xmath1014 , where @xmath1015 is the set of equivalent classes of elements in the set @xmath1006 by relation ( [ equivalence - bis ] ) .
8.7 @xcite ) if @xmath1016 , @xmath733 and @xmath1017 then the following relations are equivalent : 1 .
@xmath1018 , 2 .
@xmath1019 , 3 .
@xmath1020 .
[ thetapropositionqo ] if ( i ) holds , there is an arc @xmath1021 with @xmath1022 . then if @xmath1023 the vector @xmath751 belongs to @xmath1024 . by lemma [ crucial ]
those vectors generate the subspace @xmath1025 and hence ( iii ) holds .
suppose that ( iii ) holds .
the vector @xmath1026 belongs to the lattice @xmath268 and it is in the interior of the cone @xmath270 ( see notation [ notlattqo ] ) .
by definition we have that @xmath1027 and @xmath1028
. then we continue as in the proof of ( iv ) @xmath1029 ( ii ) in proposition [ prop - equivqo ] .
[ newqo ] if @xmath1030 and @xmath380 then the following assertions are equivalent : 1 .
the intersection @xmath1031 is empty .
the face @xmath1032 of the polyhedron @xmath1033 determined by @xmath73 is contained in the interior of @xmath1034 .
_ the proof coincides with that of proposition 8.8 @xcite . @xmath1035
[ skqo ] ( cf .
@xcite ) if @xmath400 we define the set @xmath1036 as the subset of cones @xmath1037 such that the face @xmath1038 of @xmath1039 is contained in the interior of @xmath1034 .
[ ddqo ] notice that @xmath1040 . if @xmath1041 , the set @xmath1042 is non - empty if and only if @xmath1043 . as a consequence of the results of this section we have the following propositions : [ 29qo ] ( cf .
8.11 @xcite ) let us fix an integer @xmath1044 .
we have the following partition as union of locally closed subsets : @xmath1045 \in a_{k , \t } /_{\sim } } j_{s_0 } ( h_{\nu}^*).\ ] ] if @xmath1046 the coefficient of @xmath1047 in the auxiliary series @xmath65 is obtained by taking classes in the grothendieck ring in ( [ trunc - sqo ] ) , and then using theorem [ keyqo ] .
prop . 8.12 @xcite ) [ mk ] if @xmath1048 we set @xmath1049 \in a_{k , \t } /_{\sim } } \l^{sk-\ord_{\j _ k}(\nu)}t^s.\ ] ] we have that @xmath1050 the main results on the geometric motivic poincar series are based on theorem [ keyqo ] and proposition [ 29qo ] .
the proofs of the main results of this section follow by applying the method introduced in the toric case @xcite . recall that if @xmath1051 is of dimension one we denote by @xmath372 the integral primitive vector in @xmath74 .
9.5 @xcite ) if @xmath1052 then the rational form of the series @xmath1053 is of the form : @xmath1054 for some @xmath1055 $ ] . if @xmath949 then equation ( [ p - jqo ] ) holds by replacing in the denominator the term @xmath1056 by @xmath1057 .
both numerator and denominator of ( [ p - jqo ] ) are determined by the lattice @xmath219 and the newton polyhedra of the logarithmic jacobian ideals .
[ ratp_jqo ] the proof follows by using the same argument of proposition 9.5 @xcite .
the factor @xmath1057 does not appear in the denominator of @xmath1053 for @xmath1052 .
_ proof of theorem [ plambdaracqo]_. it is consequence of propositions [ ratp_jqo ] and [ mk ] . _ proof of corollary [ p - geomqo]_. it follows from proposition [ descomppgeomqo ] and theorem [ plambdaracqo ] .
_ proof of corollary [ vol - mot2qo]_. the proof is the same as proposition 10.1 @xcite .
we introduce a distinguished sequence of monomial ideals of the coordinate ring of the normalization of the q.o . hypersurface .
let @xmath1058 be a germ of complex analytic variety .
its analytic algebra @xmath1059 is of the form @xmath1060 , for @xmath1061 .
we denote by @xmath1062 the @xmath1063-module of _ ( k " ahler ) holomorphic differential forms _ and by @xmath1064 its canonical derivation .
we denote by @xmath1065 the @xmath1063-module @xmath1066 .
see section 1.10 of @xcite for instance .
first , we review the normal toric case following @xcite chapter 3 and @xcite appendix .
we consider the toric singularity @xmath38 with analytic algebra of the form @xmath1067 .
we denote by @xmath49 the equivariant weil divisor defined by the sum of orbit closures of codimension one in the toric variety @xmath38 .
the @xmath1068-module @xmath1069 of @xmath281-forms of @xmath38 with logarithmic poles along @xmath49 is identified with @xmath1070 .
we have a map of @xmath1071-modules @xmath1072 , determined by @xmath1073 , for @xmath1074 .
notice that if @xmath1075 generate the semigroup @xmath1076 then @xmath1077 generate the @xmath1071-module @xmath1078 . if @xmath1079 then @xmath1080 .
notice that @xmath1081 , where @xmath1082 denote the coordinates of @xmath1083 in terms of a basis @xmath1084 of the lattice @xmath219 .
for @xmath1085 we have the following homomorphism of @xmath1068-modules @xmath1086 @xmath1087 notice that fixing a basis @xmath1088 of the rank @xmath1 lattice @xmath219 defines an isomorphism @xmath1089 , given by @xmath1090 .
this provides an homomorphism of @xmath1071-modules @xmath1091 the image of @xmath1092 by the composite @xmath1093 is an ideal of @xmath1071 , which is independent of the basis of @xmath219 chosen .
this ideal is called the _ logarithmic jacobian ideal _ of @xmath38 in @xcite . if @xmath40 is the germ of q.o .
singularity then its normalization is a toric singularity and we have a canonical map @xmath1094 for @xmath937 , which induced by the normalization map @xmath1095 .
we denote also by @xmath1096 and @xmath1097 the maps ( [ f ] ) and ( [ w ] ) if @xmath1098 .
11.1 @xcite ) the _ @xmath343-logarithmic jacobian ideal _ of @xmath0 is the @xmath1099-module generated by the set @xmath1100 .
[ jotak ] ( generalizing theorem 3.3 of @xcite , cf . prop .
11.2 @xcite ) for @xmath1101 the @xmath343-logarithmic jacobian ideal of a q.o .
hypersurface @xmath0 is the monomial ideal @xmath345 of @xmath1102 of definition [ idealesjac ] .
we denote by @xmath1103 the ideal generated by @xmath1100 .
for @xmath1104 we analyze the images of the elements @xmath1105 , which generate @xmath1106 as a @xmath744-algebra , by the homomorphism @xmath1097 .
we have two possibilities . 1 .
if @xmath1107 then @xmath1108
. then taking wedges with elements of @xmath1109 and applying the homomorphism @xmath1096 we obtain a generator of @xmath1103 equal to @xmath1110 if and only if @xmath1111 .
2 . if one of the @xmath1112 , say @xmath1113 , is equal to @xmath1114
, then we set @xmath1115 .
the image of @xmath785 in @xmath1116 is of the form @xmath1117 where @xmath1118 and @xmath1119 .
then we have that @xmath1120 in @xmath47 hence @xmath1121 and @xmath1122 then taking wedges with elements of @xmath1109 and applying the homomorphism @xmath1096 we obtain a generator of @xmath1103 of the form : @xmath1123 where @xmath1124 and @xmath1125 are units in @xmath1126 , @xmath1127 and @xmath1128 .
the elements we obtain in ( i ) and ( ii ) are generators of @xmath1103 .
since @xmath1129 the terms @xmath1130 belong to @xmath1103 hence we deduce that @xmath1131 .
[ e1 ] we compute the series @xmath77 for a q.o .
surface parametrized by a q.o .
branch @xmath200 with characteristic exponents @xmath1132 .
we have that @xmath1133 and @xmath1134 .
it follows that @xmath0 has smooth normalization thus @xmath1135 .
we denote by @xmath1136 and @xmath1137 the one - dimensional faces of @xmath233 .
the plane curves @xmath1138 and @xmath1139 have multiplicities two and four respectively which determine the terms @xmath1140 and @xmath1141 by remark [ p - codim1qo ] . notice that @xmath1142 thus @xmath1143 we determine this sum by computing the generating functions of the sets @xmath1144 and applying the method of proposition 9.5 and section 12 of @xcite .
we get : the semigroup @xmath279 is determined by the characteristic exponents . in this example
the semigroup is generated by @xmath1150 and @xmath1151 .
the lattice @xmath219 coincides with the one of @xmath56 .
the dual cone of @xmath1152 coincides with @xmath233 . since the monomial curves @xmath1153 and @xmath1154 have multiplicities @xmath1155 and @xmath1156 respectively , we get @xmath1157 for @xmath1158 . since @xmath1142 we obtain that @xmath1159 is obtained by a formula analogous to the right hand - side of ( [ aaa ] ) , in terms of the subdivision associated to the logarithmic jacobian ideals of @xmath1160 ( see @xcite ) .
the subdivision obtained for the monomial variety is different that the one obtained for the q.o .
we get : , on some rational generating series occuring in arithmetic geometry , in _ geometric aspects of dwork theory _ , edited by a. adolphson , f. baldassarri , p. berthelot , n. katz and f. loeser , volume 1 , de gruyter , 509 - 526 ( 2004 ) . ; the zeta - function of a quasi - ordinary singularity .
ii ; _ topics in algebraic and noncommutative geometry ( luminy / annapolis , md , 2001 ) _
, 109122 , contemp .
, 324 , amer .
soc . , providence , ri , 2003 .
darstellung der funktionen eines algebraischen krpers zweier unabhangigen vernderlichen @xmath1166 , @xmath1167 in der umgebung einer stelle @xmath1168 , @xmath1169 , _
j.reine angew .
_ , * 133 * ( 1908 ) , 289 - 314 .
the denef - loeser series for toric surface singularities .
proceedings of the international conference on algebraic geometry and singularities ( spanish ) ( sevilla , 2001 ) . _ rev . mat .
* 19 * ( 2003 ) , no 2 , 581 - 612 . ; some conjectures about quasi - ordinary singularities ; _ topics in algebraic and noncommutative geometry ( luminy / annapolis , md , 2001 ) _ , 185193 ,
, 324 , amer .
math.soc . , providence , ri , 2003 .
arc spaces , motivic integration and stringy invariants , advanced studies in pure mathematics 43 , proceedings of `` singularity theory and its applications , sapporo ( japan ) , 16 - 25
september 2003 '' ( 2006 ) , 529 - 572 . | the _ geometric motivic poincar series _ of a germ @xmath0 of complex algebraic variety takes into account the classes in the grothendieck ring of the jets of arcs through @xmath0 .
denef and loeser proved that this series has a rational form .
we give an explicit description of this invariant when @xmath0 is an irreducible germ of _ quasi - ordinary hypersurface singularity _ in terms of the newton polyhedra of the _ logarithmic jacobian ideals_. these ideals are determined by the _
characteristic monomials _ of a quasi - ordinary branch parametrizing @xmath0 . |
the deuteron is the simplest and best understood nuclear system bound by strong interactions . as such
it has received an extraordinary amount of attention .
the binding energy of the deuteron is rather small compared to the nuclear potential and to the binding energy of heavier nuclei and the average distance between the nucleons in the deuteron is larger than the typical distance in nuclei . at first glance these facts would seem to discourage one from using the deuteron for studying short range correlations ( src ) between nucleons . however , a conventional theory of the deuteron based on a phenomenological potential of the nucleon - nucleon interaction with a repulsive core at small distances predicts a fairly substantial contribution of high - momentum nucleon components in the wave function ( wf ) @xcite .
note that the fourier transform of the deuteron wf implies that high momentum nucleon components originate predominantly from src .
the recent measurements of elastic magnetic , electric and quadrupole form factors at @xmath11 @xcite as well as previous data for elastic @xcite and nearthreshold @xmath12 @xcite scattering at @xmath13 up to @xmath14 , elastic @xmath1 scattering @xcite and the backward production of secondary @xmath15 , @xmath16 on the deuteron @xcite are all consistent with a conventional nonrelativistic theory of the deuteron which has a core in the nn force ( see e.g. @xcite ) .
all these data show that the probability to find a nucleon with momentum above @xmath17 in the deuteron wave function is at the level of @xmath18 and furthermore , that the @xmath19 wave dominates for the high momentum components .
although the above experiments establish the presence of src in the deuteron any detailed knowledge of the high momentum components in the deuteron wf is quite limited .
the elastic and inclusive @xmath20 cross sections depend on an integral over nucleon momenta in the deuteron .
also , the inclusive character of the backward particle production does not allow the unambiguous reconstruction of the space - time picture of these processes .
high energy wide angle exclusive reactions belong to the special class of processes which can probe rather directly the structure of src in the deuteron .
the ability to measure simultaneously the spectator and the interacting nucleons opens up a unique opportunity to cross check various interpretations of src dominated phenomena .
this is the reason why we consider the high energy exclusive reactions to be the most straightforward method for investigating src .
it is a timely venture to investigate theoretically these processes on the deuteron since the high accuracy facilities for studying these processes experimentally with high intensity proton @xcite and electron @xcite beams will be available very soon . in this paper
we analyze the capacity for studying src in the deuteron by a complete kinematical measurement of the hard proton - deuteron quasielastic scattering .
the salient feature of this reaction ( discussed in section 2 ) is that within the framework of the plane wave impulse approximation ( pwia ) the strong energy dependence of the hard @xmath21 elastic cross section boosts the contribution of the high momentum components of the wave function@xcite .
we show ( section 3 ) that initial and final state interactions do not spoil this conclusion . in section 4
we show that there is a significant difference between two varying approaches to the description of deuteron structure which enables us to demonstrate the importance of relativistic effects .
the fact that the processes are sensitive to high momentum components of the deuteron wf can be also used to investigate models of the emc effect . in order to demonstrate that
, we calculate the effect of the suppression of point - like configurations due to src @xcite . in the appendix we present some details of the calculation of the initial and final state interactions .
when calculating the cross sections of hard processes it is necessary to include relativistic kinematics and the fact that high energy processes are developed along the light cone ( lc ) .
we calculate the cross section for the quasielastic @xmath22 reaction within the light - cone impulse approximation . in section 4
we show that for small spectator momenta in the deuteron the predictions based on lc mechanics are indistinguishable from those based on a nonrelativistic theory . within the lc impulse approximation
the cross section is a convolution of the elementary @xmath21 cross section and the deuteron light cone density function @xcite : @xmath23 where @xmath24 where @xmath25 , @xmath26 , @xmath27 , @xmath28 , @xmath29 are the four - momenta of the target nucleus , incoming , scattered , produced and spectator nucleons , respectively ( see e.g. fig.4c ) . the indices @xmath30 and @xmath31 denote the transverse and longitudinal direction with respect to the incoming proton momentum @xmath32 .
the @xmath33 and @xmath34 denote the energy and longitudinal components of the four - momenta in the light cone reference frame : @xmath35 .
the factor @xmath36 accounts for the effects of initial and final state interactions . in the plane wave approximation @xmath37 .
the light - cone density function of deuteron expressed in terms of @xmath38 and @xmath19 components of the wave function is as follows @xcite : @xmath39 note that in eq .
( [ rho ] ) we ignore possible admixtures of non - nucleonic degrees of freedom in the wf .
this is justified by the analysis of high energy processes on the deuteron which indicate that the nucleon - nucleon component of deuteron wf dominates in the region of momenta below @xmath40 ( for a review see @xcite ) .
we write @xmath41 for the invariant cross section for elastic @xmath21 scattering .
we use phenomenological @xmath21 differential cross section parameterized by : @xmath42\cdot f(s,\theta_{c.m . } ) \label{x_pp}\end{aligned}\ ] ] where @xmath43 , @xmath44 and @xmath45)^{-2}$ ] . at @xmath46
eq .([x_pp ] ) reproduces the parameterization of ralston and pire @xcite for @xmath47 while at fixed @xmath48 it reproduces the parameterization of brodsky @xcite .
the function @xmath49 is used to further adjust the phenomenologically motivated parameterization of the experimental data in the range @xmath50 .
we present the function @xmath49 for four typical @xmath51 scattering angles ( fig.1 ) . at the limit of large @xmath48 and large @xmath51 scattering angle
the phenomenological motivated approximation is the best and @xmath49 approach unity .
as far as we get from the ideal case the larger @xmath49 is . for cross sections at
lower @xmath48 we use the phase - shifts generated by the said code @xcite .
we draw attention to the strong dependence on @xmath48 of the elementary cross section . according to eq .
( [ x_pp ] ) , for a fixed value of @xmath52 ( @xmath53 ) , the @xmath21 cross section behaves as @xmath54 which according to eq .
( [ alpha ] ) will introduce a @xmath55 dependence in the pwia cross section ( see eq .
( [ x_lc ] ) ) .
thus , in pwia the cross section for proton wide angle scattering on the deuteron is sensitive to the low @xmath56 values of the wave function @xcite i.e. to that high fermi momentum of the struck nucleon which is predominantly parallel to the incoming beam .
this is the key feature that makes the process a useful tool to magnify src effects .
indeed it complements other probes such as electrons which are usually more sensitive to the @xmath57 region .
we calculated the cross sections using eqs.([x_lc])-([x_pp ] ) .
they are shown in fig.2 as a function of the momentum ( @xmath58 ) and polar angle ( @xmath59 ) of the spectator neutron with respect to the incoming proton .
one can see a significant enhancement of the cross section for large momenta of the spectator nucleon at large projectile momenta ( @xmath60 ) and when the target proton momentum ( @xmath61 ) tends to be parallel to the projectile .
thus the discussed process is rather sensitive to the high momentum part of the deuteron wave function which in turn is the result of the strong @xmath48 dependence of the elementary @xmath21 cross section . because of this strong @xmath48 dependence it is necessary to pay special attention to off - shell effects , which are proportional to the difference between the invariant energies at the intermediate ( @xmath62 ) and final ( @xmath63 ) states . since in lc mechanics
only the `` - '' and `` t '' components of momenta are conserved in an intermediate state , this difference is : @xmath64 where @xmath65 .
the difference between @xmath62 and @xmath63 does not increase with incoming energy since @xmath66 .
we estimate the size of the effect by calculating the cross section with eq.([x_lc ] ) for two values of the invariant energy @xmath63 and @xmath62 .
the two calculations were done for the kinematics where the projectile is parallel to the struck proton and the results are shown in fig.3 . note that the off - shell effects depend only weakly on the direction of the struck nucleon momentum ( see eq.([off_sh ] ) ) .
we see also that the off - shell effects cause at most @xmath67 change in the cross section values for spectator momenta up to @xmath68 .
since the kinematics of quasielastic scattering on a deuteron is fully determined by the momenta of the two ejected protons one can reconstruct the geometry of the elementary @xmath21 hard scattering .
we calculate the cross sections for three extreme geometries where the momentum of the struck nucleon is parallel , antiparallel and perpendicular to the direction of the incident proton and we refer to them accordingly as parallel , antiparallel and perpendicular .
in order to extract usefull information from the data one has to account for the initial state interaction ( isi ) of the incoming proton and the final state interactions ( fsi ) between the outgoing protons and the spectator neutron .
we will choose kinematical conditions such that the isi and the fsi corrections will be small or , at least , will not introduce any additional energy dependence in the spectra of scattered nucleons . we will use the eikonal approximation to calculate the sum of the diagrams shown in fig.4 which describe the hard pwia scattering ( fig.4a ) and the lowest order soft @xmath0 rescatterings ( fig.4b - d ) .
the soft interactions between the incident and target protons as well as the interactions between the scattered and produced protons are included in the phenomenological parameterization of the @xmath21 hard scattering amplitude . since at high energies the soft rescattering amplitude depends almost linearly on the energy of the collision @xmath48 , the contribution of the isi diagram ( fig.4b ) is equal to the contribution of any one of the fsi ( fig.4c , d ) diagrams .
therefore , just one of the three diagrams ( b , c , d ) has to be evaluated and the final result is given by the pwia diagram plus three times the contribution from any one of the rescattering diagrams ( see details in the appendix ) . since the soft scattering decreases exponentially with transfered momentum while the hard scattering has a power law dependence it is convenient to factorize the effect of isi and fsi into a scale factor @xmath69 : @xmath70 \over u^{2}(k ) + w^{2}(k ) } \left ( { \sqrt{m^2+k'^2 } \over \sqrt{m^2 + k^2 } } \right ) ^{{1\over 2 } } \cdot f(q_{t } ) \cdot{d^2q_{t } \over ( 2\pi)^2 } \ \ \ + \nonumber \\
\nonumber \\ & & { 9 \over 16 } \int { \left ( u(k_1)u(k_2 ) + w(k_1)w(k_2)[{3\over 2}{(k_1k_2)^{2 } \over k_1 ^ 2 k_2 ^ 2 } - { 1\over 2 } ] \right ) \left ( \sqrt{(m^2 + k_1 ^
2)(m^2 + k_2 ^ 2)}\right ) ^{{1\over 2 } } \over \left ( u^{2}(k ) + w^{2}(k ) \right ) \sqrt{m^2+k^2}}\cdot \nonumber \\ & & \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \
\cdot f(q_{1t})f(q_{2t}){d^2q_{1 t } \over ( 2\pi)^2 } { d^2q_{2 t } \over ( 2\pi)^2 } \label{kappa}\end{aligned}\ ] ] where @xmath71 , @xmath72 , @xmath73 , @xmath74 are given by eq .
( [ rho ] ) and @xmath75 is the imaginary part of the elastic @xmath0 soft scattering amplitude . for simplification we assumed that @xmath76 and neglected nucleon spin rotations in the lc which may add uncertainties of the order of 1% ( see @xcite for estimates ) .
the eq.([kappa ] ) contains the distinctive feature of soft high energy processes that the @xmath56 component of the target nucleon momentum is conserved in soft rescatterings . in fig.5
we present the value of @xmath36 as a function of @xmath56 for different values of the spectator transverse momenta ( @xmath77 ) .
it shows that the isi and fsi are significant for @xmath78 for any @xmath77 and at large @xmath77 ( @xmath79 ) , for a large range of @xmath56 . in these kinematical conditions
the contribution from higher order rescatterings is not negligible and consequently our results could not be trustworthy .
the important feature of isi and fsi is that at small @xmath77 the @xmath36 is a smooth function of @xmath56 for @xmath80 , which corresponds to spectator momenta @xmath81 in parallel geometry and @xmath82 in antiparallel geometry . in these cases
the difference of @xmath83 from @xmath84 is typically @xmath85 and the effects of initial and final state interactions can be accounted for by rescaling the pwia cross section .
thus , quasielastic processes with small @xmath77 and @xmath56 within the above discussed kinematical range are well suited for investigating high momentum nucleon component in the deuteron .
it means that under these conditions the deuteron is a good testing ground for the studying the basic ingredients of the nucleon - nucleon interaction .
note that in this paper we do not consider any possible effects due to color transparency ( see e.g. @xcite and references therein ) which could suppress considerably the effects of isi and fsi especially at larger projectile energies .
we will elaborate on the physics of color transparency in forthcoming publications .
an interesting question is to what extent the effects due to the nucleon relativistic motion in the deuteron are important and whether they can be investigated experimentally . in fig.6
we show two time - ordered diagrams which represent the standard impulse approximation contribution ( a ) with on - energy shell @xmath0 amplitude and the relativistic contribution ( b ) . to examine to
what extent the relativistic contribution ( diagram b ) is important we compare the results obtained in the light - cone impulse approximation with those of the virtual nucleon formalism ( vn ) ( see e.g. @xcite ) : @xmath86 where @xmath87 .
the factor @xmath88 accounts for the conservation of baryon charge @xcite . in the lc impulse approximation we know ( see e.g. @xcite , @xcite ) that the contribution of diagram b is negligible or included into the definition of the deuteron wave function . by comparing the lc impulse aproximation to the vn approximation , where diagram b is included in different way , we actually can learn about the importance of diagram b. it is interesting to note that we account for isi and fsi in the virtual nucleon impulse approximation by an equation which is very similar to eq.([kappa ] ) .
the only major difference with the lc approximation is the different argument ( momentum ) of the deuteron wave function .
the ratio of the cross sections calculated in the lc impulse approximation ( eq . ( [ x_lc ] ) ) and in the virtual nucleon approach ( eq . ( [ x_vn ] ) ) is shown in fig.7 .
the ratio is displayed as a function of the spectator ( neutron ) momentum for the above defined parallel ( antiparallel ) and perpendicular geometries .
we use deuteron wave functions calculated with the bonn@xcite and paris @xcite potentials .
formally , the main difference between the lc and virtual nucleon formalisms is in the value of the momentum at which the deuteron wave function is calculated by eq .
( [ rho ] ) . for the perpendicular geometry as well as for small spectator momenta
@xmath89 the internal momentum @xmath90 and the spectator momentum in the laboratory system are rather close to each other . under those conditions there is no substantial difference between the relativistic and non - relativistic calculations .
however , there is a considerable ( @xmath91 ) difference between these approaches even at relatively low spectator momenta of about @xmath92 where the deuteron wave function is known with an accuracy of a few percent and where the isi and fsi can be replaced by a rescaling factor ( see previous section ) .
another way to confront various approaches that account for the relativistic motion of nucleons in the deuteron is to measure the asymmetry between the cross sections for parallel and antiparallel geometries at light - cone momentum fractions @xmath56 and @xmath93 .
we define the asymmetry parameter @xmath94 as @xcite : @xmath95/2 } \label{asm}\ ] ] where @xmath96 ( see eq .
( [ x_lc ] ) ) and @xmath97 , @xmath98 . since the light - cone formulae ( eq . ( [ x_lc ] ) and eq . ( [ kappa ] ) )
are invariant under exchange of @xmath56 and @xmath99 we get that @xmath100 .
this is because the two nucleons in the deuteron are treated symmetrically in light - cone quantum mechanics .
in contrast , in the virtual nucleon description the interacting nucleon is off - shell while the spectator is on the mass shell . as a result
the symmetry with respect to the @xmath101 transposition is lost .
the significance of this effect can be seen in fig.8 which shows the asymmetry calculated in the vn approximation according to eq .
( [ asm ] ) . a significant difference between the lc prediction ( @xmath102 ) and the vn one @xmath103 reveals itself at comparatively low spectator momenta @xmath104 for parallel and @xmath105 for antiparallel geometries .
the initial and final state interactions plays insignificant role at low spectator momenta ( cf . above discussion ) and will even enhance the difference at higher momenta .
one can also investigate the difference between the two approaches by looking at the symmetry of the cross sections under the transformation which changes the sign of the spectator momenta @xmath106 .
such a symmetry is expected in the vn approximation eq.([x_vn ] ) but not in the light - cone description eq.([x_lc ] ) .
it is worth mentioning that at the large energies of the experiment where one measures the two emerging protons will yield a better resolution in @xmath56 than in @xmath58 @xcite .
therefore , from a practical point of view there is an advantage in testing the asymmetry in @xmath56 rather than in @xmath58 .
the isi and fsi effects suppress the sensitivity to the various deuteron wave functions as can be seen in figs .
the suppression occurs because the integrals in eq .
( [ kappa ] ) are sensitive mainly to the lower momenta in the deuteron wave function .
the fact that the target nucleon is bound in a deuteron can significantly alter the picture of high - energy hard scattering especially for the large momenta .
one consequence of nucleon binding is the suppression of point - like configurations for bound nucleons @xcite . at distances where the two - nucleon interaction is dominated by the attractive part of the nucleon - nucleon potential the probability of small size quark - gluon configurations ( point - like configurations ( plc ) )
is suppressed ( color screening phenomenon @xcite ) . for the discussion how this effect reveals itself in the interaction of two nucleons
see @xcite . the amount of suppression
can be estimated by multiplying the deuteron lc wave function by the factor @xcite : @xmath107 where @xmath108 is the deuteron binding energy and @xmath109 ( @xmath110 ) is a parameter which characterizes the bound nucleon excitation in the deuteron . the additional @xmath111 dependence ( @xmath112 @xcite accounts for the fact that point - like configurations dominate in the bound nucleon wave function for reactions with sufficiently large transfered momentum @xmath111 .
we used the value @xmath113 .
in fig.9 we show the suppression of plc as a function of the spectator momentum for the parallel , perpendicular and antiparallel geometries . except for very large spectator momenta
the effect is the same for the various geometries .
the suppression is relatively small in the parallel geometry for @xmath114 due to the small energy transfer in this kinematical condition .
this is a consequence of the @xmath111 dependence in eq .
( [ asm ] ) .
the plc suppression decreases the cross section and increases the difference between the lc and vn approximations for parallel kinematics ( solid line in fig.9 ) .
note also that for @xmath115 the @xmath111 dependence becomes weaker and that the symmetry under @xmath101 transformation is restored .
we summarize our results in fig.10 where we display the cross section for hard quasi - elastic proton scattering on the deuteron by taking into account the various src effects discussed in this paper .
we see that isi and fsi and nuclear binding effects do not mask the predicted enhancement of high momentum components for parallel kinematics .
we reiterate that a fully kinematical measurement at the appropriate kinematical conditions makes it possible to separate the various effects we discussed .
for example , at neutron spectator momenta of about @xmath116 , where the deuteron wave functions are well known , any difference between a bound and a free nucleon in parallel kinematics tends to increase the differences between the light - cone and virtual nucleon results . also the contributions from isi and fsi are small and have a relatively weak momentum dependence .
we have shown that there is considerable difference for the asymmetry of cross sections in parallel and antiparallel geometries when calculated in the lc and vn approximations in a kinematical region where isi and fsi effects and effects of emc are well understood .
these differences allow us to determine the importance of the vacuum diagram contribution .
eventually , after we achieve some understanding of hard scattering , nucleon binding , isi and fsi effects , any new data will be able to teach us something about the deuteron wf at momenta above the region studied thus far .
* acknowledgment * we wish to thank jonas alster , steve hepelmann , yael mardor and israel mardor for helpful discussions .
this work was supported in part by the basic research foundation administrated by the israel academy of science and humanities , by the u.s.a .
- israel binational science foundation grant no .
9200126 and by the u.s .
department of energy under contract no .
de - fg02 - 93er40771 .
to acount for the intial state interaction of the incoming proton and of the final state interaction of the scattered and knocked out protons with the spectator neutron we need to calculate the sum of the diagrams in fig.4 : @xmath117 where @xmath118 , @xmath119 and @xmath120 .
@xmath121 is the amplitude for the hard @xmath21 elastic scatterings , @xmath122 is the amplitude for soft @xmath123 rescatterings and @xmath124 is the deuteron light cone wave function .
spin effects are not included . the labels `` b '' , `` c '' and `` d '' correspond to the diagrams in fig.4 .
it is well known that the sum of these diagrams leads to lowest order formulae of the eikonal approximation where the intermediate particles are on the mass shell .
we calculate only the diagram `` b '' since the calculation for the other two rescattering diagrams is practicaly identical . in terms of light cone variables
@xmath125 the invariant phase volume is @xmath126 and the denominator of integrand `` b '' is given by : @xmath127\cdot & \nonumber \\
( p_s - k)_-\left [ p_{s+ } - k_+ - { m^2+(p_1-k)_t^2\over p_{s-}-k_- } + i{\varepsilon\over p_{s-}-k_- } \right ] & \label{a2}\end{aligned}\ ] ] to evaluate the integral we take residues over @xmath128 and @xmath129 which correspond to on - shell intermediate nucleons ( namely the incident proton and spectator neutron ) . finally , for diagram `` b '' we obtain : @xmath130 where @xmath131 is the center of mass energy for the incident proton and the spectator neutron rescattering . as follows from eq .
( [ a3 ] ) at sufficiently high energies and soft rescatterings @xmath132 and @xmath133 and therefore the longitudinal ( @xmath56 ) component of wave function is not changed by the soft rescattering .
similarly , for the diagrams `` d '' and `` c '' one gets the following expressions : @xmath134 where the @xmath135 and @xmath136 are the invariant energies corresponding to rescattering of the scattered and produced protons with the spectator neutron . for soft rescattering @xmath137 is practically independent to @xmath48 .
since the dependence on @xmath111 of the hard scattering is slower than the exponential dependence of soft rescattering , one can neglect the difference between @xmath111 and @xmath138 in the hard scattering amplitude . as a result
we find that the diagrams of `` b '' , `` c '' and `` d '' contribute equally .
therefore , the scattering amplitude can be expressed by the sum of the born diagram plus three times the contribution of any one of the diagrams `` b '' , `` c '' , `` d '' .
factorizing the hard scattering amplitude from this sum , including the spin dependence of the deuteron wave function and taking the square of the amplitude we obtain the final expression for the effect of initial and final state interaction which is presented in eq .
( [ kappa ] ) . for simplicity
we neglect the effects of spin rotation in the light cone spin density function of the deuteron ( see @xcite ) which is a small effect ( 1% ) in our case .
g. e. brown and a. d. jackson , the nucleon - nucleon interaction , , 1976 .
i. the et al .
, , * 67 * , 173 ( 1991 ) . r. arnold et al .
, , * 35 * , 776 , ( 1975 ) .
r. arnold et al .
, , * 58 * , 1723 , ( 1987 ) . s. rock et al . , , * 49 * , 1139 , ( 1982 ) . g. alberi and g. goggi , , * 74 * , 1 , ( 1981 ) . a. m. baldin et al . , preprint jinr , dubna , ( 1977 ) . l. m. anderson jr . , preprint lbl , 6769 , berkley , ( 1979 ) .
frankfurt and m. i. strikman , phys.rep .
, * 76 * , 214 , ( 1981 ) .
l. l. frankfurt and m. i. strikman , phys.rep .
, * 160 * , 235 ( 1988 ) .
l. l. frankfurt and m. i. strikman , nucl.phys .
, * b148 * , 107 , ( 1979 ) . a. s. caroll et al . , bnl experiment , 850 . conceptual design report ,
cebaf 1990 . g. r. farrar , l. l. frankfurt and m. i. strikman , phys . rev .
lett . , * 62 * , 1095 , ( 1989 )
. l. l. frankfurt and m. i. strikman , nucl .
, * 250 * , 143 ( 1985 ) .
ralston and b. pire , phys.rev.lett .
, * 61 * , 1823 , ( 1988 ) . s. j. brodsky , high energy collisions - 1973 ( stony brook ) conf.proc
. r. a. arndt , phys.rev .
* d 35 * , 212 , ( 1987 ) . l. l. frankfurt and m. i. strikman , nucl .
phys . , * a405 * , 557 , ( 1983 ) .
w. b. atwood and g. b. west , phys .
, * d 7 * , 773 , ( 1973 ) .
l. l. frankfurt and m. i. strikman , phys.lett .
, * 64b * , 435 , ( 1976 ) . j. b. kogut and d. e. soper , phys.rev .
, * d 1 * , 2901 , ( 1970 ) .
r. machleidt , k. holinde and c. elster , phys.rep . ,
* 149 * , 1 , ( 1987 ) .
m. lacombe et al .
, phys.rev .
, * c 21 * , 861 , ( 1980 ) .
l. l. frankfurt , g. miller and m. i. strikman , phys.rev.lett . * 68 * , 17 , ( 1992 ) .
jin - yuan wu , thesis , pennsylvania state university 1992 .
* @xmath139 is the function used to best fit the world data for @xmath21 elastic scattering to the phenomenological motiviated parametrization of @xcite and @xcite .
the parametrization is ideal for large ( @xmath140 ) @xmath48 and @xmath141 and under this condition @xmath142 . * the lc pwia cross sections for hard quasielastic proton scattering of the deuteron as a function of the momentum ( @xmath58 ) and polar angle ( @xmath59 ) of the spectator neutron .
( a ) - @xmath143 , ( b ) - @xmath144 . * the ratio of lc pwia cross sections . in the nominator the cross section for @xmath21 hard scattering was calculated with a final state invariant energy @xmath48 and the denominator cross section was calculated for the invariant energy of the intermediate state .
the ratio is presented as a function of the spectator momenta ( @xmath145 ) .
the dashed line is for @xmath143 and the solid line is for @xmath144 . * the diagrams show the pwia and the lowest order soft @xmath123 rescattering processes .
the broken line represents the soft interaction .
the empty circles represent the hard @xmath21 scattering vertices .
* the factor @xmath69 represents the effect of isi and fsi as calculated in eikonal approximation according to the diagrams of fig.4 .
the projectile momentum is @xmath144 .
the different lines ( from top to bottom ) correspond to spectator transverse momenta of @xmath146 . * time - ordered diagrams representing the impulse approximation contribution ( a ) and the relativistic ( vacuum ) contribution ( b ) . *
the ratio of the cross sections as a function of @xmath58 calculated in lc over the cross section obtained in the vn approximation for @xmath144 .
( a ) parallel and antiparallel geometries , ( b ) perpendicular geometries ( see definitions in section 2 ) . the solid line is for the pwia calculation with the `` paris '' deuteron wave function .
the dashed line is for the pwia calculation with the `` bonn '' wave function . the curves with `` @xmath147 '' correspond to the calculation with isi and fsi .
* the @xmath148 dependence of the asymmetry @xmath149 defined in eq.([asm ] ) and calculated in the vn approximation .
the lines and symbols are defined in the previous figure .
note that @xmath150 in lc .
the bottom scale represents the corresponding values of longitudinal component of the spectator momenta @xmath151 calculated for @xmath152 ( negative @xmath151 ) and @xmath153 ( positive @xmath151 ) at @xmath154 and @xmath155 . *
the ratio of cross sections calculated in the lc approximation with and without taking into account the plc suppression .
the ratio is shown as a function of @xmath58 for parallel ( solid ) , antiparallel ( dashed ) and perpendicular ( dot - dashed ) geometries .
the calculations are presented for @xmath144 with the `` paris '' wave function for the deuteron . * the cross section for @xmath156 = @xmath157 of hard quasielastic @xmath1 scatterings as a function of spectator momentum at ( anti)parallel geometries .
dotted line - vn pwia , dash - dotted - lc pwia , dashed - lc with the effect of plc suppression and solid line - lc with plc suppression and with isi , fsi .
the `` paris '' wave function for the deuteron was used . | we show that the strong dependence of the amplitude for @xmath0 hard scattering on the collision energy can be used to magnify the effects of short range nucleon correlations in quasielastic @xmath1 scattering . under specific kinematical conditions
the effect of initial and final state interactions can be accounted for by rescaling the cross section calculated within the plane wave impulse approximation .
the feasibility to investigate the role of relativistic effects in the deuteron wave function is demonstrated by comparing the predictions of different formalisms .
binding effects due to short range correlations in deuteron are discussed as well . * probing short range nucleon correlations in high energy hard quasielastic @xmath1 reactions + * l. frankfurt@xmath2 , e. piasetsky@xmath3 , m. sargsyan@xmath4 , m. strikman@xmath5 _ @xmath6 raymond and beverly sackler faculty of exact science , + school of physics and astronomy , tel aviv university , + tel aviv , 69978 , israel + @xmath7 on leave of absence from st .
petersburg nuclear physics institute , + st .
petersburg , russia + @xmath8 also at the yerevan physics institute , yerevan , 375036 , armenia + @xmath9 pennsylvania state university , university park , pa , 16802 , usa + @xmath10 also at the st .
petersburg nuclear physics institute , st .
petersburg , russia _ |
the exceptional piezoelectric properties of the lead - oxide class of relaxor ferroelectrics has fueled an intense amount of scientific research on these materials over the past several years .
@xcite numerous studies have been performed in an attempt to understand the fundamental mechanism(s ) responsible for the ultrahigh piezoelectric response of these compounds .
@xcite during the course of these studies , additional anomalous behavior has been documented including the truly remarkable situation in which the structure of the bulk of a single crystal relaxor specimen can differ from that of the near - surface region , or `` skin , '' which spans tens of microns .
the x - ray scattering studies of xu _ et al . _
have revealed a striking discrepancy between the low - temperature crystal structure reported by earlier x - ray scattering studies on the parent compound pb(zn@xmath3nb@xmath4)o@xmath5 ( pzn ) and that observed using high - energy x - rays .
@xcite by varying the energy of the incident x - rays , and thus the penetration depth , distinct crystal structures were observed .
subsequent neutron scattering studies by gehring
_ et al . _ have found that the same discrepancy exists in solid solutions of pb(mg@xmath3nb@xmath4)o@xmath5 ( pmn ) mixed with 10% pbtio@xmath6 ( pt ) .
@xcite more recent neutron residual stress measurements on a large single crystal of pmn at room temperature show that the cubic unit cell lattice parameter varies with the depth normal to the crystal surface , with a corresponding large surface strain .
@xcite in this paper we review and expand upon the current understanding of the skin effect in the pzn-@xmath0pt and pmn-@xmath0pt relaxor systems .
new data are reported that demonstrate the high variability in crystal structure that can be obtained from x - ray measurements made on the same single crystal specimen , but at different locations over the crystal surface .
a detailed analysis of high resolution neutron data obtained on pmn-10%pt single crystal reveals the presence of strong internal strain that does not manifest itself as a true structural transition , but is nonetheless correlated with the cubic - to - rhombohedral phase transition temperature determined with low - energy x - rays .
new data is also presented that demonstrates that the depth dependence of the lattice constant measured in pure pmn varies significantly with temperature
. while these findings are of obvious significance from a fundamental research point of view , they are also important because the skin effect has misled researchers into drawing incorrect conclusions about the pzn-@xmath0pt and pmn-@xmath0pt phase diagrams , as well as several other erroneous assumptions about the true relaxor state .
the phase diagrams of the pzn-@xmath0pt and pmn-@xmath0pt systems are striking in their similarity ; both exhibit cubic - to - rhombohedral phase transitions at low pt content , cubic - to - tetragonal phase transitions at high pt content , and a steeply - sloped morphotropic phase boundary ( mpb ) separating monoclinic @xcite and tetragonal phases over a narrow range of intermediate pt concentrations .
figure 1 shows the phase diagram for pmn-@xmath0pt .
in addition , all compounds below the mpb exhibit exceptionally large piezoelectric coefficients as well as other properties characteristic of relaxors , namely a strongly frequency dependent dielectric susceptibility and a very large room temperature dielectric constant . given these numerous similarities , the @xmath7 case of
pure pmn stands out as an oddity because several studies have shown that the crystal structure retains an average cubic structure down to 5 k , @xcite whereas the x - ray measurements by lebon _
_ show that pure pzn transforms to a rhombohedral phase at a transition temperature @xmath8 k. @xcite the unique status of pmn was made all the more curious after an intriguing study by ohwada _
on a zero - field cooled single crystal of pzn-8%pt did not find the expected rhombohedral phase below @xmath9 .
@xcite these results motivated the decision by xu _ et al .
_ to re - examine the zero - field structure of pure pzn , especially because previous studies of the structure of pzn-@xmath0pt had been done on poled single crystals .
@xcite another motivating factor were the measurements by noheda _
et al . _ that showed that x - ray diffraction results on pzn-@xmath0pt single crystals depended strongly on surface structures .
@xcite zero - field phase diagram of pmn-@xmath0pt.,width=226 ] single crystals of pzn were grown via spontaneous nucleation from a high temperature solution using pbo as flux .
@xcite two triangular - shaped plates 1 mm thick and with edges roughly 3 mm in length were cut with large surfaces parallel to the cubic [ 111 ] axis from the same as - grown single crystal .
the faces of both crystals were polished using a series of diamond pastes down to 3 @xmath1 m , and gold was sputtered onto the large faces of one crystal , which was then poled at 200 c and cooled to room temperature in a field of 20 kv / cm .
the unsputtered sample was neither poled nor thermally treated , thereby leaving a potentially larger mechanical strain within the crystal .
( our measurements have shown that the structures of the skin / bulk do not vary after thermal cycling up to 800 k. ) the x - ray measurements on pzn reported here were all performed at the brookhaven national laboratory national synchrotron light source ( nsls ) , primarily on beamline x-17b1 , which can provide x - rays with an incident energy of 67 kev and an energy resolution @xmath10 .
the spatial extent of the focused x - ray beam on the sample was @xmath11 mm@xmath12 .
additional measurements were made on beamline x-22a , which provided access to lower incident x - ray energies of 32 and 10.7 kev .
radial scans of the x - ray scattering intensity measured across the pzn ( 111 ) bragg peak .
panels on the left ( right ) correspond to measurements made with 10.7 kev ( 32 kev ) incident energy x - rays .
different panels correspond to different locations on the crystal surface.,width=226 ] figure 2 shows measurements on the unpoled single crystal of pzn at room temperature ( which is well below the reported value of @xmath8 k ) using two different incident energies of 10.7 and 32 kev , corresponding to penetration depths of 13 @xmath1 m and 59 @xmath1 m , respectively .
the different panels in each figure column correspond to measurements made on different locations of the crystal surface .
the data correspond to radial ( @xmath13 ) scans of the scattering intensity measured across the ( 111 ) bragg peak , which must split into two peaks in a rhombohedral phase .
it is immediately apparent that the different measurements do not confirm a single crystal structure .
the top - left panel shows a cubic - like single peak at one location ( panel ( a ) ) , whereas a two - peaked structure , suggestive of rhombohedral symmetry , is observed at two others ( panels ( b ) and ( c ) ) .
the same discrepancy holds at the higher incident x - ray energy of 32 kev ( panels ( d ) and ( e ) ) even though these data correspond to a deeper penetration depth .
top panel : radial scans of pzn ( 111 ) bragg peak measured at 300 k ( open circles ) and 650 k ( closed circles ) using 67 kev incident energy x - rays .
bottom panel : radial scans of pzn ( 111 ) bragg peak measured at 300 k ( open circles ) and 650 k ( closed circles ) using 10.2 kev incident energy x - rays .
inset shows temperature dependence of the rhombohedral angle @xmath14 .
@xcite , width=226 ] in order to probe the structure to still greater depths within the crystal , x - ray measurements were made of the same ( 111 ) reflection at an incident energy of 67 kev , which corresponds to a penetration depth of 412 @xmath1 m .
the upper panel of fig . 3 shows the resulting ( 111 ) bragg peak profile , using the same radial scans shown in fig . 2 , at temperatures above ( 650 k ) and below ( 300 k )
@xmath8 k. @xcite remarkably , a single , resolution - limited peak is observed at both temperatures thereby indicating the absence of the rhombohedral phase previously reported by lebon _
et al_. at all different locations measured on the crystal at this energy , only a single resolution - limited ( 111 ) bragg peak was observed .
an effort to reproduce the lower energy ( 8.9 kev from cu k@xmath15 radiation ) results of lebon _ et al .
_ were made by repeating the same measurements at 10.7 kev .
these data are shown in the lower panel of fig . 3 where it is seen that a cubic structure ( single peak )
is observed at 650 k as was the case at 67 kev . at 300 k , however , a clear split ( 111 ) bragg peak profile is seen , consistent with a rhombohedral distortion .
the temperature dependence of the rhombohedral angle @xmath14 is plotted in the inset to fig . 3 , and gives a transition temperature of 410 k , also consistent with previous reports .
radial scans through the ( 111 ) bragg peak at 300 k for pmn ( top panel ) , pmn-4.5%pt ( middle panel ) , and pmn-8%pt ( bottom panel ) .
open ( closed ) circles correspond to an incident energy of 67 kev ( 10.2 kev ) .
@xcite , width=226 ] this combination of x - ray measurements , which probed varying depths of the pzn near - surface region , provides conclusive evidence for the existence of a skin effect whereby the bulk and skin of a single crystal exhibit different structures .
additional evidence of this effect was reported in a recent neutron scattering study on pzn by stock _
et al . _ , who found that the bragg peaks broaden significantly below the x - ray determined value of @xmath16 k along both the radial and transverse directions , but saw no sign of a rhombohedral distortion .
@xcite assuming that the phase diagrams of pzn-@xmath0pt and pmn-@xmath0pt should be `` universal , '' the high - energy x - ray data on pzn reconcile the puzzling discrepancy at @xmath7 in that they demonstrate that both pmn and pzn retain an average cubic structure down to low temperatures .
this leads naturally to the question of how the skin is affected by the addition of pbtio@xmath6 ( pt ) . to answer this , three single crystals of pzn-@xmath0pt with pt concentrations @xmath7 , 4.5% , and 8% were studied at room temperature by xu _
_ in the same manner as described above , i. e. using incident x - ray energies of 10.2 kev and 67 kev .
the data are shown in fig .
@xcite the top panel summarizes the situation just described for pzn ( @xmath7 ) , where skin and bulk exhibit different crystal structures at 300 k. at a pt concentration of 4.5% , however , both the skin region ( 10.2 kev data ) and the crystal bulk ( 67 kev data ) exhibit a rhombohedral phase .
the difference here is that the bulk rhombohedral phase is characterized by a smaller rhombohedral angle @xmath14 than is that of the skin .
similar results were obtained at a pt concentration of 8% , just below the mpb , where again both skin and bulk exhibit a rhombohedral distortion , but with a smaller value of @xmath14 measured in the bulk .
given the penetration depth of the 10.2 kev x - rays , the thickness of the skin was estimated at roughly 10 @xmath1m@xmath1750 @xmath1 m for all samples .
composition dependence of the rhombohedral angle @xmath14 .
closed circles correspond to the skin , or outer - layer , measured at 10.2 kev .
open circles correspond to the crystal bulk measured at 67 kev .
@xcite , width=226 ] the pt dependence of @xmath14 for both the outer - layer skin region and the crystal bulk is given in fig .
@xcite these data demonstrate that the addition of ferroelectric pt stabilizes a low - temperature rhombohedral phase that is otherwise absent in pure pzn , and they provide new insight into the way the structure of the relaxor pzn evolves as one approaches the mpb where the greatest piezoelectric response is observed .
the x - ray studies of pzn reported here have exploited the variability of the penetration depth by appropriate tuning of the incident x - ray energy .
neutrons , by contrast , have penetration depths of order 1 cm , and thus provide bulk information .
motivated by the x - ray work on pzn , a program of neutron scattering measurements on pmn-@xmath0pt crystals was undertaken to determine how the rhombohedral phase evolves when adding pt into pure pmn .
the accepted phase diagram for pmn-@xmath0pt , shown previously in fig . 1
, indicates that a cubic - to - rhombohedral phase transition takes place for @xmath0 less than about 32% .
@xcite the solid data points are values of the transition temperature @xmath9 measured by different authors , @xcite and the solid line is a linear fit to these data .
the solitary open circle at @xmath7 corresponds to the first - order phase transition that has been reported to occur in pmn near 213k , but only after first cooling in an electric field @xmath18kv / cm .
@xcite the notation @xmath19 shown by the arrow is used to denote the fact that pmn retains an average cubic structure below this temperature when zero - field cooled ( zfc ) . a monoclinic ( m@xmath20 )
phase was discovered by noheda _
et al . _
@xcite at higher pt concentrations next to the well - known morphotropic phase boundary ( mpb ) separating the m@xmath20 phase from the tetragonal phase .
temperature dependence of the rhombohedral angle @xmath14 and diffuse scattering measured with x - rays by dkhil _
et al . _ on pmn-10%pt .
@xcite lines are guides to the eye.,width=226 ] an interesting and detailed comparative study between pmn and pmn-10%pt was carried out by dkhil and co - workers using both x - ray and neutron scattering techniques on powder and on single crystal samples .
@xcite their results point to the existence of competing rhombohedral and tetragonal order that never produces a ferroelectric distortion in pmn , but does in pmn-10%pt below a critical temperature @xmath21 k. in addition to a rhombohedral phase , critical - like diffuse scattering was observed that peaked at @xmath9
. the temperature dependence of both this diffuse scattering , as well as the rhombohedral angle @xmath14 is given in fig .
@xcite subsequent x - ray scattering experiments by ye _
et al_. on a series of pmn-@xmath0pt samples detected a low - temperature rhombohedral distortion for pt concentrations as small as 5% .
@xcite because all of the published x - ray and neutron structural data agree in the case of pmn ( both indicate a persistent cubic phase down to 5 k ) it was believed that the skin effect in the pmn-@xmath0pt system , if it exists at all , must be significantly smaller than that in pzn-@xmath0pt .
to test this idea , a single crystal of pmn-10%pt was obtained for the purpose of performing a high resolution neutron structural characterization of the material over a wide range of temperatures .
@xcite the dashed vertical arrow in fig . 1
locates the structural transition at 10% pt .
the neutron scattering study was performed on the bt9 triple - axis spectrometer located at the nist center for neutron research .
diffuse scattering measurements were performed using fixed incident and final neutron energies @xmath2214.7mev ( @xmath23 ) , which were obtained via bragg diffraction from the ( 002 ) reflection of highly - oriented pyrolytic graphite ( hopg ) crystals as monochromator and analyzer , and relatively relaxed horizontal beam collimations of 40@xmath24 - 46@xmath24-s-40@xmath24 - 80@xmath24 ( s = sample ) . to maximize the instrumental @xmath25-resolution , and thus the sensitivity to subtle structural distortions and changes in lattice spacing
, the spectrometer was operated with a perfect ge crystal as analyzer .
more importantly , the @xmath26-spacings of both the analyzer and the pmn-10%pt bragg reflections were specifically chosen to match one another as closely as possible .
this latter condition yields a minimum in the instrumental longitudinal @xmath25-resolution .
@xcite for the perovskite pmn-10%pt sample the @xmath26-spacings associated with the ( 111 ) and ( 022 ) bragg peak ( 2.333 and 1.4287 , respectively ) are well - matched with those for ge ( 220 ) and ( 004 ) ( 2.000 and 1.4143 ) .
our choice of horizontal beam collimations of 10@xmath24 - 46@xmath24-s-20@xmath24 - 40@xmath24 was based on monte carlo intensity simulations to improve the @xmath25-resolution without an unreasonable reduction in intensity .
@xcite an incident neutron energy of 8.5mev was used for measurements of the ( 111 ) bragg peak , which was provided via a highly oriented pyrolytic graphite monochromator , and resulted in a sharp @xmath25-resolution of 0.0018rlu fwhm ( 1rlu = 2@xmath27a = 1.5576@xmath28 ) .
the measurements of the ( 022 ) bragg peak yielded an even better @xmath25-resolution approaching 0.0010rlu fwhm , thanks to a more closely matched @xmath26-spacing condition between sample and analyzer . a 2.65 g single crystal of pmn-10%pt , with approximate dimensions of @xmath29 mm@xmath30 ,
was produced using a top - seeded solution growth technique .
@xcite the crystal mosaic full - width at half - maximum ( fwhm ) measured at the ( 200 ) bragg peak at 500 k is less than 0.06@xmath31 indicating an excellent quality sample .
access to bragg reflections of the form @xmath32 and @xmath33 were needed to study the rhombohedral phase ; therefore the crystal was mounted with the [ 01@xmath34 axis aligned vertically .
the sample was then mounted inside a high - temperature closed - cycle @xmath30he refrigerator and fixed to the goniometer of the bt9 triple - axis spectrometer .
neutron diffuse scattering intensity measured as a function of temperature at @xmath35 on a pmn-10%pt single crystal .
the inset shows the full diffuse scattering profile measured along the [ 011 ] direction ( transverse to @xmath36 ) at 200k and 450k .
@xcite , width=226 ] the neutron diffuse scattering data on pmn-10%pt were measured near the ( 300 ) bragg peak ( along the [ 011 ] direction ) where it is known that such scattering is very strong .
@xcite the temperature dependence of the diffuse scattering is shown in fig . 7 at several selected values of the reduced wavevector @xmath25 .
@xcite an immediate discrepancy is apparent between these and the x - ray data of dkhil _
et al_. shown in fig
. 6 inasmuch as the neutron diffuse scattering data do not exhibit a peak at any temperature down to 50 k. instead , a slight suggestion of a change in slope is observed in the vicinity of @xmath37 k for @xmath38 rlu .
the inset to fig .
7 shows two full scans of the ( 300 ) peak at 200 k and 450 k showing the strong growth of the diffuse scattering intensity with decreasing temperature .
the presence of strong diffuse scattering below @xmath9 directly contradicts the x - ray results , which were obtained using an incident x - ray energy of 8.9 kev .
this finding strongly suggests that a skin effect is also present in the pmn-@xmath0pt system as it provides a natural explanation with which to reconcile the neutron and x - ray measurements .
comparison of x - ray measurements by dkhil _
et al . _ at ( 222 ) ( top ) @xcite and our neutron measurements at ( 111 ) ( bottom ) on pmn-10%pt .
the horizontal scale of the x - ray data have been reduced by a factor of two to allow direct comparison with the neutron ( 111 ) data .
solid bars indicate the bt9 instrumental @xmath25-resolution fwhm .
@xcite , width=226 ] neutron scattering measurements were made at both the ( 111 ) and ( 022 ) bragg reflections of the pmn-10%pt sample using the high @xmath25-resolution configuration of the bt9 spectrometer described earlier .
all reflections of the form @xmath32 and @xmath33 must split into two in the presence of a rhombohedral distortion because the unit cell is elongated along one diagonal and compressed along the others , thus giving rise to two distinct @xmath26-spacings .
hence ( 111 ) and @xmath39 bragg peaks will no longer be identically situated in reciprocal space , and their relative intensities will give the relative domain populations for each reflection .
radial scans of the neutron scattering intensity across the ( 111 ) bragg peak are presented in fig .
8 both above and below @xmath40 k , and these are compared to equivalent scans across the ( 222 ) bragg peak measured with x - rays .
@xcite the horizontal scales of the x - ray data have been adjusted to allow for a direct scan by scan comparison . at high temperatures , both neutron and x - ray data
show a single resolution - limited bragg peak consistent with a cubic structure .
but below @xmath9 the x - ray measurements reveal a definitive splitting of the ( 222 ) bragg peak whereas the neutron data do not . instead
, the neutron measurements continue to show a single peak .
moreover , the neutron peak width is less than the splitting observed in the x - ray scan , thus providing even stronger confirmation of the absence of any rhombohedral distortion in the bulk of the pmn-10%pt single crystal .
some broadening is apparent in the neutron case at 100 k. this could either be the result of internal strain , which would manifest as a radial broadening , and/or it could be due to the contribution from the skin .
spatially resolved neutron diffraction measurements of the near - surface strain by conlon _
_ in a single crystal of pure pmn , where both a dramatic change in the lattice constant and the bragg peak intensities is observed over a length scale of order 100 @xmath1 m , @xcite suggest that strain is the most likely source of this broadening , and will be discussed in the following section .
temperature dependence of the two component fitting analysis for the ( 022 ) bragg peak measured with high @xmath25-resolution .
top and middle panels compare the intensity and , width=226 ] the thermal expansion of the pmn-10%pt sample was measured from 10 k to 600 k by carefully measuring the ( 022 ) bragg peak radial profile with high @xmath25-resolution .
it was noted that these data were significantly better fit using the sum of two gaussian functions of @xmath25 , one sharp and the other broad .
the temperature dependences of the resulting peak intensity and lattice spacing for these two components are shown in the top and middle panels of fig .
the notable feature here is that the sharp component intensity and associated lattice constant show no significant feature that correlates with @xmath9 .
by contrast , the same parameters associated with the broad component correlate quite well with the x - ray determined value of @xmath9
. in particular , the thermal expansion shown in the middle panel behaves as expected for a normal ferroelectric .
@xcite we emphasize here that the broad component is a single peak , and still quite narrow , and thus does not represent a rhombohedral distortion .
rather the broad peak is believed to represent a highly strained region of the crystal bulk .
the lattice constant associated with the sharp component is intriguing in that it exhibits an almost temperature independent behavior from 10 k to 400 k , with the @xmath26-spacing changing by less than 0.001 over this temperature range .
however at 400 k both the lattice constant and the peak intensity exhibit abrupt changes .
indeed , the rate of thermal expansion becomes extremely large , being roughly @xmath41 k@xmath28 , which is considerably larger than the value @xmath42 k@xmath28 , which is typical of other oxides .
it is suggestive that this temperature is very close to those observed in pure pmn and pmn-20%pt , where both the maximum broadening of the ta mode occurs and the minimum value of the to soft mode is obtained . @xcite the bottom panel of fig .
9 shows the variation of the peak widths ( fwhm ) for the two gaussian components . in this case
it is interesting that the sharp component fwhm appears to correlate more obviously with @xmath9 than does that of the broad component .
the x - ray and neutron results obtained on pmn-10%pt , when viewed together , point strongly to the presence of a skin effect like that documented by xu _
the concept of a near - surface region having a structure distinct from that of the bulk would also provide a natural explanation of the results from x - ray and neutron powder diffraction studies of the low - temperature structure of both pmn and pzn , for which a two - phase model , one cubic and the other rhombohedral , is required to describe the observed powder peak intensity profiles .
@xcite because neutrons interact weakly with matter ( typical penetration depths being of order 1 cm ) any observable scattering signal from the skin would imply a substantial skin volume
. this could be the result of a `` thick '' skin of order many microns in single crystals , or a large effective surface area , as would be the case in powdered samples .
schematic diagram of the neutron residual stress instrument used for spatially resolved measurements of strain .
the gauge volume is that region of the crystal bounded by the intersection of the incident and scattered neutron beams , which are given by the solid lines the slits on the incident and scattered beams were 5 mm tall and 300 @xmath1 m wide .
@xcite , width=226 ] to examine this idea , neutron residual stress measurements were performed by conlon _
_ on the l3 double - axis spectrometer , which is located at the national research universal ( nru ) reactor at chalk river laboratories , in order to study neutrons diffracted from a very large 9.3 cc single crystal of pmn as a function of depth .
@xcite the pmn crystal was grown by the modified bridgeman technique , the details of which are described elsewhere .
@xcite a schematic diagram of the l3 instrument is shown in fig .
10 , where a ge(113 ) crystal was used to monochromate the incident beam to an energy of @xmath43 mev , and horizontal beam collimations of 60@xmath24 - 30@xmath24-s-30@xmath24 were used in conjunction with cadmium ( strong neutron absorbers ) slits placed both before and after the sample .
an hopg filter was placed in the incident beam to remove higher order harmonics ( i. e. neutrons with wavelengths @xmath44 , @xmath45 , etc . ) .
the crystal was cut in order to provide a large and well - defined 100 surface , and neutrons scattered from the ( 200 ) bragg peak from this surface at room temperature were recorded in a reflection geometry as indicated in the schematic .
the residual stress technique allows a measurement of lattice strain as a function of depth normal to the crystal surface through the use of narrow absorbing slits , or masks , that provide a high degree of angular collimation of both the incident and scattered beams .
the result is that only neutrons scattered from a small `` gauge '' volume , represented by the diamond - shaped region in fig .
10 , are recorded by the detector . by translating the large single crystal along the scattering vector @xmath36 , in this case the ( 200 ) bragg peak , and then measuring radial scans of the bragg peak intensity at each value of translation , the lattice constant of the sample can be probed as a function of depth .
radial ( @xmath46 scans of the ( 200 ) bragg peak scattering intensity are shown for three values of translation of the pmn crystal along the [ 100 ] axis dashed lines are fits to a gaussian lineshape .
@xcite , width=226 ] data from @xmath47 scans are shown in fig .
11 for three different translational settings .
@xcite two features are readily apparent from this figure : the first is the shift of the center of the ( 200 ) bragg peak in @xmath48 , thus indicating a measurable change of the lattice parameter over a depth of 0.3 mm ( 300 @xmath1 m ) , while the second is the dramatic change in the peak intensity .
this latter effect is due to the extinction effect , which has been observed previously in analogous neutron scattering measurements of the second length scale phenomenon in tb and srtio@xmath6 for which a residual stress configuration was used .
@xcite as a cross check of these data , a perfect germanium single crystal was studied in exactly the same way using the ge ( 220 ) reflection .
the @xmath26-spacing of the ge ( 220 ) bragg peak is almost identical to that of the pmn ( 200 ) bragg peak .
this allows the instrument scattering angles to remain nearly the same , and thus provides a direct test of the observed change in lattice parameter with depth ( translation ) .
unlike the case of pmn , no significant strain or change in lattice spacing was observed in the near surface region of the perfect germanium single crystal over a depth of more than 1000 @xmath1 m .
an additional check was made by measuring rocking curves ( @xmath49 scans ) to determine the crystal mosaic as a function of depth .
the mosaics obtained near the surface and deep within the bulk were both measured to be @xmath50 , thus demonstrating that the change in lattice constant is not an artifact of a variation in mosaic spread or the presence of another crystal domain near the surface .
top panel : ( 200 ) bragg peak intensity as a function of translation ( depth ) .
bottom panel : normalized lattice spacing ( strain ) @xmath51 as a function of translation , where @xmath52 is the bulk value of the lattice spacing @xmath26 .
the location of the crystal surface is indicated by the vertical dashed line .
@xcite , width=226 ] figure 12 shows the depth dependence of the ( 200 ) bragg peak intensity ( top panel ) and the crystal lattice spacing ( bottom panel ) over a range exceeding 500 @xmath1 m .
@xcite the large variation in the peak intensity is due to the extinction effect mentioned above , and can not be ascribed solely to the linear attenuation of the incident and scattered neutron beams as they traverse through the pmn crystal .
the variation of bragg intensity is important , however , as it can be used to locate the surface of the sample on the translation axis .
the crystal surface should correspond to the translation that gives the maximum intensity as this will be when the gauge volume lies completely below the crystal surface . at this point ,
the scattering volume is a maximum , while the effects from extinction are minimal .
the bottom panel shows a strong variation in the lattice spacing spanning several hundred microns ( see fig .
11 ) , and thus confirms the existence of a skin effect in pure pmn
. it thus appears to be the case that the skin is not `` thin '' in pmn ; rather the crystal structures of the bulk and skin are both cubic , and there is a substantial near surface strain .
normalized lattice spacing , or strain , @xmath53 k@xmath54 as a function of translation , measured at 400 k , 500 k , and 700 k. here @xmath55 k ) is the bulk value of the lattice spacing @xmath26 at room temperature .
the location of the crystal surface is indicated by the vertical dashed line.,width=226 ] all of the residual stress measurements presented so far were performed at room temperature .
figure 13 shows data obtained in a manner identical to that in the bottom panel of fig . 12 , but at three higher temperatures , namely 400 k , 500 k , and 700 k. the striking aspect of these data is the reversal of the strain minimum apparent at 300 k in fig . 12 , which slowly evolves into a maximum at 700 k in the bottom panel of fig . 13 .
while the number of temperatures reported here are insufficient to draw any definitive conclusions about this effect , it is very interesting to note that the burns temperature @xmath56 , below which local regions of randomly oriented polar order on a nanometer scale first appear , @xcite lies between 600 k and 700 k. it is tempting to speculate that these polar nanoregions might have some relationship to the skin , and that they may also play an important role in the invar - like thermal expansion observed from 10 k to 400 k.
the combination of x - ray and neutron measurements presented here provide conclusive proof of a significant surface layer in single crystal specimens of both pzn-@xmath0pt and pmn-@xmath0pt systems .
the presence of this skin has led to incorrect phase assignments in the phase diagrams of both systems inasmuch as no rhombohedral phase is present at sufficiently low pt concentrations in either case , the low - temperature phase remaining cubic .
a crossover to a low - temperature rhombohedral phase does take place in pzn-@xmath0pt between @xmath57 and @xmath58% whereas a similar crossover occurs in the pmn-@xmath0pt system between @xmath59% and @xmath60% .
@xcite that the skin effect persists over such a large temperature range , and that the skin thickness spans many tens of microns , are extremely unusual . in the case of the second length scale phenomena observed in tb and srtio@xmath6 , the anomalous extra component in the critical scattering associated with a phase transition
was also observed over a large depth of order 100 @xmath1 m , but it persisted only over a very limited temperature range .
@xcite in some of these relaxor compounds , the skin and crystal bulk could exhibit different crystal structures ( pzn and pmn-10%pt ) , while in others they possess the same structure ( cubic or rhombohedral ) , but with either different lattice spacings ( pmn ) or different sized rhombohedral distortions ( pzn-4.5%pt and pzn-8%pt ) .
some logical questions for future study are whether or not the skin varies with electric field ?
if so does it get bigger or smaller ?
such studies are in progress , including one to examine the field - induced transition in pmn with high-@xmath25 resolution , as well as the effects on the strong diffuse scattering .
@xcite it is evident that future theoretical models of these relaxor systems will need to account for the observed anomalous skin effect .
the authors wish to acknowledge first and foremost the friendship and support provided by dr .
gen shirane of brookhaven national laboratory , who passed away suddenly in january of 2005 . his guidance ,
genius , and boundless enthusiasm formed the foundation for all of the relaxor studies presented here .
gen s unceasing scientific curiosity and his tireless drive to discover new physical phenomenon were a continuous source of inspiration to all of us
. he shall be sorely missed .
we dedicate this paper to his memory , and we express our deep gratitude for all that he has taught us .
the authors also wish to acknowledge stimulating discussions with dwight viehland and jiefang li .
finally , financial support from the u.s .
department of energy under contract de - ac02 - 98ch10886 is also gratefully acknowledged .
p. bonneau , p. garnier , e. husson , and a. morell , mater .
* 24 * , 201 ( 1989 ) ; p. bonneau , p. garnier , g. calvarin , e. husson , j. r. gavarri , a. w. hewat , and a. morell , j. solid state chem . * 91 * , 350 ( 1991 ) . | x - ray and neutron scattering studies of the lead - based family of perovskite relaxors pzn-@xmath0pt and pmn-@xmath0pt have documented a highly unusual situation in which the near - surface region of a single crystal can exhibit a structure that is different from that of the bulk when cooled to low temperatures .
the near - surface region , or `` skin '' can also display critical behavior that is absent in the crystal interior , as well as a significantly different lattice spacing . by varying the incident photon energy , and thus the effective penetration depth ,
x - ray measurements indicate a skin thickness of order 10 @xmath1 m to 50 @xmath1 m for pzn-@xmath0pt samples with @xmath2% .
neutron residual stress measurements on a large pmn single crystal reveal a uniform lattice spacing within the bulk , but an increased strain near the surface .
the presence of this skin effect has led to incorrect phase diagrams for both the pzn-@xmath0pt and pmn-@xmath0pt systems and erroneous characterizations of the nature of the relaxor state . |
a wide variety of mass transport systems ranging from traffic flow to polymer gels @xcite exhibit nonequilibrium condensation phenomena .
these include basic microscopic dynamics ubiquitous in nature such as aggregation , fragmentation and diffusion .
the nonequilibrium steady states of these systems are classified into two types of phases , so - called fluid phase and condensed phase .
a finite fraction of total particles condenses on a single site in the condensed phase . in the fluid phase
, the particle number on each site fluctuates around the density of total particles ( @xmath10 ) without the condensation .
as the rates of these processes vary , the condensation phase transitions between the two phases may take place at a certain critical density @xmath22 .
one of the simplest mass transport models exhibiting the condensation transitions is a conserved - mass aggregation ( ca ) model @xcite .
ca model evolves via diffusion , chipping and aggregation upon contact which arise in a variety of phenomena such as polymer gels @xcite , the formation of colloidal suspensions @xcite , river networks @xcite and clouds @xcite . in one - dimensional ca model ,
the mass @xmath4 of site @xmath1 moves either to site @xmath23 or to site @xmath24 with unit rate , and then @xmath25 and @xmath26 . with rate @xmath11 , unit mass chips off from site @xmath1 and moves to one of the nearest neighboring sites ; @xmath27 and @xmath28 .
the generalization to higher dimensions is straightforward . as total masses are conserved , the conserved density @xmath10 and the rate @xmath11 determine the phase of ca model .
the @xmath29 case corresponds to the well - known zero range process ( zrp ) with a constant hopping rate @xcite .
the existence of the condensation transitions in ca model depends on the symmetry of movement , the constraints of diffusion rate and the underlying network structure @xcite . in the symmetric ca ( sca )
model @xcite in which diffusion and chipping direction are unbiased , the condensation transitions take place at a certain @xmath22 .
the steady state properties of sca model is exactly described by the mean field theory @xcite .
the single site mass distribution @xmath30 was shown to undergo phase transitions on regular lattices @xcite . for a fixed @xmath11 ,
as @xmath10 is varied across the critical density @xmath31 , the behavior of @xmath30 for large @xmath32 was found to be @xcite @xmath33 mean field theory predicts @xmath34 and @xmath35 @xcite .
recently , ca model on unweighted scale - free networks ( sfns ) with degree distribution @xmath36 was studied to investigate the effect of underlying network structure on the condensation transitions @xcite .
we call networks with equal weight on all links unweighted networks . on unweighted sfns ,
the same type of the condensation transitions as those of sca in regular lattice take place for @xmath37 .
however for @xmath38 , the condensation always occurs for any density @xmath10(@xmath39 ) .
it was shown that the existence of the transitions is directly related to the diffusive capture process on unweighted sfns @xcite .
on the other hand , most real - world networks exhibit not only a heterogeneous distribution of degree , but also heterogeneous distribution of weights @xcite .
weights assigned on links characterize the interaction strengths between nodes .
there have been various attempts to understand the underlying mechanism and scale - free behaviors of empirically observed weighted networks @xcite .
also there have been attempts to understand the effect of heterogeneous weights on various dynamics such as synchronization , dynamics of random walks , transport and percolation , and condensation of zero - range process @xcite .
these studies showed that dynamical properties are modified and exhibit non - trivial dependence on the strength of weight . in this paper , as the generalization of our study on ca model on complex networks , we investigate the effect of both heterogeneous degree and weight on the condensation phenomena of ca model on weighted networks .
the weight @xmath40 represents the weight to a link from the node @xmath1 to @xmath2 . in general
, the strength @xmath41 of the node @xmath1 scales with the degree @xmath42 as @xmath43 .
the exponent @xmath44 varies with network structures @xcite .
thus it is natural to take the weight @xmath40 as @xmath45 . in this paper
, we study the condensation transitions of ca on the wsfns with degree distribution @xmath46 and the symmetric weight @xmath47 . as in one dimension ,
the diffusion of the whole masses and the fragmentation of unit mass occur with the unit rate and the rate @xmath11 , respectively .
in addition , masses move from a node @xmath1 to @xmath2 with hopping rate proportional to @xmath48 .
we found that a certain critical @xmath8 exists below which the condensation transitions take place .
however for @xmath9 , the condensation always occurs for any density @xmath49 . to find @xmath8 as a function of the degree exponent @xmath13 ,
one needs the steady state distribution @xmath14 of finding a walker at nodes with degree @xmath15 on the wsfns .
@xmath14 gives the capture probability @xmath16 with which two walkers meet at a node with degree @xmath15 .
we analytically derived @xmath50 and @xmath16 , and finally obtained @xmath51 .
this paper is organized as follows . in sec .
ii , we discuss the condensation transitions of ca model on the wsfns . to verify the existence @xmath8 , we investigate the steady state property of a single walker and the diffusive capture process on the wsfns in sec .
iii and iv .
we discuss the behavior of an average mass @xmath19 of a node with degree @xmath15 in sec .
v and finally summarize our results in sec .
we consider ca model on wsfns with the weight @xmath0 from node @xmath2 to @xmath1 defined as @xmath52 . for the construction of wsfn
, we first construct an unweighted static sfn with @xmath53 nodes and @xmath54 links @xcite .
the degree @xmath42 of a node @xmath1 is defined as the number of its links connected to other nodes .
the average degree of a node @xmath55 is given as @xmath56 .
the degree distribution @xmath57 of sfns is a power - law distribution @xmath36 . in the static model @xcite , it is desired to use large @xmath55 to construct fully connected networks . in simulations ,
we use @xmath58 .
after then , we assign a weight @xmath59 to the link between node @xmath1 and @xmath2 .
thus the hopping probability of masses from node @xmath1 to an @xmath60s linked neighbor @xmath2 is @xmath61 .
@xmath62 denotes the sum over the linked neighbors of node @xmath1 .
each node has an integer number of particles , and the mass on a node is defined as the number of particles on the node .
initially @xmath63 particles are randomly distributed on @xmath53 nodes with given conserved density @xmath64 .
next a node @xmath1 is chosen at random and one of the following events occurs : \(i ) diffusion : with the unit rate , the whole mass @xmath4 of node @xmath65 moves to one of the linked neighbors @xmath2 with probability @xmath5
. then the aggregation takes place ; @xmath66 and @xmath67 .
\(ii ) chipping : with the rate @xmath11 , unit mass moves to a linked neighbor @xmath2 with the probability @xmath5 , and then the aggregation takes place , i.e. @xmath68 , @xmath69 .
the @xmath29 case corresponds to zrp with constant chipping rate on wsfns @xcite .
we perform monte carlo simulations with random initial mass distribution on the wsfns with @xmath70 and @xmath71 .
we set @xmath72 and the network size @xmath73 with @xmath58 .
we measure the single node mass distribution @xmath30 in the steady states . in fig . 1
, we plot @xmath30 for @xmath74 with two different @xmath44 , @xmath75 and @xmath76 .
@xmath30 exhibits quite different behavior according to the value of @xmath44 . for @xmath75 ( fig .
1(a ) ) , @xmath30 decays exponentially without aggregates for sufficiently low density @xmath77 . on the other hand , for sufficiently high density , @xmath78 , an aggregate forms with the power - law decaying background mass distribution .
it means that the condensation transition takes place at a certain critical density @xmath79 .
hence @xmath30 follows eq .
( [ p ] ) . since in unweighted sfns , i.e
@xmath21 , the condensation phase transitions take place for @xmath80 @xcite , one may expect the condensation transitions for very small @xmath44 .
based on the following steps , we estimate @xmath22 and the exponent @xmath81 . in the condensed phase , the total density @xmath10 is written as @xmath82 , where @xmath83 is the density of an aggregate .
since @xmath10 is given as @xmath84 , one can estimate @xmath22 from @xmath85 , where the upper bound @xmath86 is the cut - off mass at which the background distribution terminates . using this method
, we estimate @xmath87 .
we estimate the exponent @xmath81 from the scaling plot @xmath88 using @xmath30 of @xmath78 ( inset of fig.1(a ) ) . since the background distribution does not change for @xmath89 , we use @xmath30 of @xmath90 for the scaling plot .
we estimate @xmath91 . on the other hand , for @xmath92 ( fig.1(b ) ) , @xmath30 shows the complete different behavior .
the condensation takes place with an exponentially decaying background distribution for both sufficiently low and high density , @xmath93 and @xmath94 .
therefore we conclude that the condensation always occurs for any nonzero density so a system is always in the condensed phase without any transitions for @xmath92 .
the two different behaviors of @xmath30 for @xmath75 and @xmath76 indicate that a crossover @xmath44 ( @xmath8 ) should exist in the range @xmath95 for @xmath74 .
a system undergoes the condensation transition for @xmath96 , while the condensation always occurs without the transition for @xmath97 . for @xmath74 with @xmath75 ( a ) and @xmath92 ( b ) .
the inset of ( a ) shows the scaling plot @xmath98 with @xmath99 when @xmath78.,title="fig:",width=302][fig1 ] for @xmath70 with @xmath100 ( a ) and @xmath101 ( b ) .
the inset of ( a ) shows the scaling plot @xmath98 with @xmath102 when @xmath78 .
, title="fig:",width=302][fig2 ] similarly , for @xmath70 , @xmath30 exhibits the same different behavior according to the value of @xmath44 .
the difference from the @xmath74 case is that the condensation transitions are observed for a negative @xmath44 .
we observe the condensation transitions for @xmath103 ( fig .
2(a ) ) . with the same method used in the @xmath74 case ,
we estimate @xmath104 and @xmath105 respectively . however , for @xmath101 , the condensation is observed even for very low density @xmath93 , which means that a system is always in the condensed phase for @xmath101 ( fig .
therefore , the crossover @xmath8 also exists for @xmath70 , but its value is negative unlike the @xmath74 case .
together with the results of @xmath74 , we conclude that the crossover @xmath8 exists for any @xmath106 and @xmath8 varies with @xmath13 . in what follows
, we discuss the existence of @xmath8 and next the condensation nature for @xmath96 .
first , the condensation phenomena of ca model on wsfns is similar to that on unweighted sfns . on unweighted sfns ,
the condensation transitions exist for @xmath37 , while the condensation always occurs for @xmath38 @xcite .
hence the crossover @xmath13 is @xmath107 .
intriguingly , it was shown that the existence of transitions is determined by the survival probability of a diffusing prey chased by a diffusing predator , so - called the lamb - lion problem @xcite .
the reason is as what follows . in the limit @xmath108 ,
let us assume only an infinite aggregate exists . with the rate @xmath11 , the unit mass is chipped off and moves around with the unit rate . if the chipped mass meets again with the infinite aggregate within a finite time interval , then the infinite aggregate is stable against the chipping process . on the other hand ,
if the chipped mass and the infinite aggregate does not meet again within a finite time interval , then the infinite aggregate would disappear by repeated chipping processes .
therefore , the stability of the infinite aggregate is physically related to the capture process in which a diffusing lion ( infinite aggregate ) chases a diffusing lamb ( chipped mass ) . for the unweighted sfns with @xmath38 , it was shown that the survival probability @xmath109 of a lamb decays exponentially with finite life time @xmath110 @xcite . however , for @xmath111 , @xmath109 is finite in the thermodynamic limit .
the behavior of @xmath109 implies that the condensation transition exist for @xmath37 due to the stable fluid phase in the limit @xmath108 , but only the condensation exist for @xmath38 . as a result ,
the asymptotic behavior of the survival probability of a lamb in the lamb - lion capture process determines the existence of the condensation transitions on unweighted sfns . similarly , on the wsfns , the existence of the condensation transitions is also expected to depend on the survival probability @xmath109 of a lamb . to see this ,
let us consider two limits , @xmath112 and @xmath113 for a given @xmath13 . in the limit @xmath114
, a walker always moves to a node with the larger degree .
once a walker reach the hub node with the maximal degree , the walker is trapped at the hub node forever . as a result , a lion always captures a lamb at the hub node within a finite time interval .
hence , @xmath109 should decay exponentially with a finite life time . on the other hand , in the limit @xmath115 ,
a walker is forced to reach nodes with the minimal degree . due to the inhomogeneous structure ,
the nodes with the minial degree are connected by nodes with larger degree .
hence , a walker can not escape from one of the nodes with the minimum degree in this limit .
it means that a lion can not always capture a lamb at some other node , so that @xmath109 is finite . from the behavior of @xmath109 in the two opposite limits
, there should be a crossover @xmath8 .
@xmath109 is finite for @xmath116 and decays to zero for @xmath97 . for the condensation phenomena ,
one expects no condensation transitions ( @xmath117 ) for @xmath97 due to finite life time of a lamb .
instead , the condensation always occurs . on the other hand ,
the condensation transitions occur for @xmath116 .
we analytically find @xmath12 for a given @xmath13 in sec .
iv . from @xmath12
, one reads @xmath118 for @xmath70 and @xmath119 for @xmath120 respectively .
our simulation results for @xmath120 and @xmath121 confirm the existence of @xmath8 and also the sign of @xmath8 for each @xmath13 .
next , we discuss the critical behavior of ca model on wsfns
. the ca model on any dimensional regular lattice and unweighted sfns with @xmath37 is well described by mean - field theory @xcite .
on wsfns , interestingly , the transitions take place even for @xmath122 , which means that the transition nature is not affected by the inhomogeneity of network structure .
since @xmath8 diverges for @xmath123 , the critical behavior of ca model on sfns with @xmath116 should be the same as that on random networks where @xmath44 , i.e. weight , does not have no special meaning due to the uniform degree distribution . as a result
, one expects the mean - field critical behavior of sca model in regular lattice .
our numerical estimates of @xmath81 , @xmath91 for @xmath70 and @xmath105 for @xmath71 , well agree with the mean - field value @xmath35 . therefore , we conclude that the critical behavior of ca model for @xmath124 on the wsfns belongs to the universality class of sca model in regular lattice . in summary , for a fixed @xmath13
, there is a crossover weight exponent @xmath8 .
ca model undergoes the same type of condensation transitions as those of sca model in regular lattice for @xmath96 , while the condensation always takes place for nonzero density for @xmath97 . to find @xmath8 as a function of the degree exponent @xmath13 ,
one needs the steady state distribution @xmath14 of finding a walker at nodes with degree @xmath15 on the wsfns . in the next section ,
we derive @xmath14 on the wsfns . in sec .
iv , we study lamb - lion capture process on the wsfns and finally find @xmath8 using @xmath14 .
we consider a single walker on weighted networks with the weight @xmath0 .
the connectivity of the network is represented by the adjacency matrix @xmath125 whose element @xmath126 if there is a link from a node @xmath2 to @xmath1 .
otherwise , @xmath127 .
we set @xmath128 conventionally .
the degree @xmath42 of a node @xmath1 is given as @xmath129 .
since we consider weighted networks with weight @xmath0 , we define the weighted adjacency matrix @xmath130 as @xmath131 .
the motion of a walker on the weighted networks defined by the matrix @xmath130 is a stochastic process in the discrete time .
we derive the stationary distribution @xmath132 of a walker being at node @xmath1 following the method of ref .
@xcite . to set up the equation
, we define the transition probability as follows .
a walker at node @xmath1 at time @xmath133 selects one of its @xmath42 linked nodes with hopping probability @xmath5 .
then , at time @xmath134 , the walker moves to the selected node .
the hopping probability @xmath5 from node @xmath1 to @xmath2 is then given as @xmath135 , where @xmath136 is the strength of node @xmath1 .
as an initial condition , assume that the walker starts at the node @xmath137 at time @xmath138 .
then the recurrence relation of the transition probability @xmath139 of finding the walker at node @xmath1 at time @xmath133 is @xmath140 then the transition probability @xmath141 is written by iterating as @xmath142 for a symmetric @xmath130 with @xmath143 , one finds @xmath144 by comparing @xmath145 and @xmath139 . in the stationary state ,
the probability @xmath132 of finding a walker at node @xmath1 should be independent of initial starting nodes , which gives @xmath146 . summing up over @xmath137 ,
one finds @xmath147 where @xmath148 .
in weighted networks with symmetric weights , @xmath132 is proportional to the strength of node @xmath1 , i.e. the sum of the weights of the nearest neighboring nodes .
the same result was found in the recent study on the dynamics of random walks on growing weighted networks @xcite . in this paper
, we consider the symmetric weight @xmath0 , @xmath149 for the weight ( [ w ] ) , @xmath132 is not given as a simple form .
hence it is better to handle the distribution @xmath14 of finding a walker at nodes with degree @xmath15 . using eq .
( [ pi ] ) , one can see that @xmath150 to express the sum in eq .
( [ pk1 ] ) in terms of degree @xmath15 , we arrange the sum as follows .
only terms with @xmath151 contributes nontrivially to the sum @xmath152 and thus @xmath153 nodes with the degree @xmath15 in a network have the nontrivial contributions to the sum .
the node with the degree @xmath15 has @xmath15 linked neighbors whose degree ranges from @xmath154 to the maximal degree of the network @xmath155 . hence
, the number of nontrivial terms in the sum @xmath156 is @xmath157 , which can be arranged in the order of increasing degree .
then , the double sum of eq .
( [ pk1 ] ) is written as @xmath158 , where @xmath159 is the degree distribution of the node involved in such @xmath157 terms . for large @xmath53 , we approximate @xmath159 to @xmath160 .
then @xmath14 is approximately given as @xmath161 on sfns with degree distribution @xmath36 , the integral in the second line is finite for @xmath162 .
hence we finally obtain @xmath14 on wsfns as @xmath163 the exponent @xmath164 varies with @xmath44 and @xmath13 , and also changes its sign . for @xmath165 ,
i.e. @xmath166 , @xmath167 is independent of degree @xmath15 so a walker does not feel the inhomogeneity of the underlying network structure .
while a walker performs biased walks to nodes with the larger degree for @xmath168 , the direction of the bias is reversed for @xmath169 .
since the exponent @xmath44 is a free parameter , one can controls the direction of the bias for a given @xmath13 . and
@xmath164 for @xmath70 ( a ) and @xmath170 ( b ) .
insets show the relation ( [ pk ] ) ( solid line ) and numerical estimates of @xmath164 ( symbols ) .
, title="fig:",width=302][fig3 ] to check the scaling relation ( [ pk ] ) , we perform monte carlo simulations on the wsfns with @xmath73 and the average degree @xmath58 . in the steady states , we measure @xmath14 for various @xmath44 up to @xmath171 for @xmath70 and @xmath94 for @xmath170 . fig .
3 shows the plot of @xmath14 against @xmath15 for several values of @xmath44 .
as shown , @xmath14 scales in power - law with @xmath15 .
the inset in each panel shows the plot of @xmath164 against @xmath44 .
the simulation results agree well with the analytical prediction ( [ pk ] ) .
in this section , we consider the capture process or the lamb - lion problem on wsfns with the symmetric weights ( [ w ] ) . a lamb and a lion initially locate separately on randomly selected two nodes .
then the probability @xmath16 of finding two walkers at the same node with degree @xmath15 at the same time is proportional to @xmath172 . from eq .
( [ pk ] ) , one gets @xmath173 with @xmath174 then the probability @xmath175 of finding two walkers on the same node with any degree is given as @xmath176 since the upper bound @xmath155 diverges with @xmath53 , the integral @xmath177 diverges for @xmath178 .
hence there exists a crossover value @xmath8 given as @xmath179 for @xmath116 , the lamb survives indefinitely with a finite probability .
however , for @xmath97 , the lion captures the lamb with the unit probability . to check the scaling relation ( [ dk ] ) , we measure @xmath16 on the wsfns with @xmath73 and @xmath58 . for @xmath180 trials , we count the number @xmath181 of capture events on nodes with the degree @xmath15 .
we obtain @xmath16 to divide @xmath181 by total trials(@xmath180 ) .
4 shows the plot of @xmath16 for several values of @xmath44 , which scales well with @xmath15 in power - law . as shown in the insets of fig .
4 , numerical estimates for @xmath182 satisfy the relation ( [ dk ] ) very well . and @xmath182 for @xmath70 ( a ) and @xmath170 ( b ) .
insets show the relation of ( [ dk ] ) ( solid line ) and numerical estimates of @xmath182 ( symbols ) . ,
title="fig:",width=302][fig4 ] of a lamb for @xmath170 ( a ) and @xmath70 ( b ) .
the solid line is a guide to the eye .
insets show the semi - logarithmic plots of @xmath109 for @xmath73 . from top to bottom , each line corresponds to the @xmath109 of @xmath96 , @xmath8 and @xmath183 respectively.,title="fig:",width=294][fig5 ] to verify the existence of @xmath8 by another method , we now consider the survival probability @xmath109 of a lamb .
@xmath109 always satisfies @xmath184 on random and scale - free networks due to the small world nature @xcite .
as @xmath184 in sfns with any @xmath13 , we are interested in the average life time @xmath185 of a lamb rather than @xmath109 itself . from @xmath186\ ; dt$ ] and @xmath187 , we have @xmath188 .
hence @xmath189 is infinite for @xmath116 and finite for @xmath97 in the limit @xmath190 .
however , for the finite - sized networks , a lamb is eventually captured within @xmath53 time steps for any @xmath44 . for @xmath96 , the maximum life time
should be the order of @xmath53 to guarantee the finite survival probability in the limit @xmath190 .
hence @xmath185 is expected to scale as @xmath191 for @xmath96 .
we measure @xmath185 on wsfns of @xmath70 and @xmath192 with network size @xmath53 up to @xmath193 . from eq .
( [ ac ] ) , one reads @xmath194 for @xmath70 and @xmath195 for @xmath170 . in fig . 5 , we plot @xmath185 against @xmath53 .
as shown in each inset , @xmath109 exponentially decays for any @xmath44 . for @xmath170(fig .
5(a ) ) , @xmath185 increases with @xmath53 as @xmath196 with @xmath197 for @xmath198 ( @xmath199 ) and @xmath200 for @xmath201 .
we estimate @xmath202 by measuring successive slopes from the log - log data in fig .
5(a ) . for @xmath92 ( @xmath203 )
, @xmath185 tends to saturate to the asymptotic value @xmath204 with decreasing successive slopes .
the exponent @xmath202 of @xmath75 is close to the expected value @xmath205 . for @xmath206
, @xmath185 seems to diverge with @xmath207 .
however , since @xmath185 of @xmath92 already tends to saturate , it is expected that @xmath185 of @xmath97 would saturate to a finite value in the network with @xmath208 , where @xmath209 is the characteristic size for given @xmath44 .
for example , for @xmath92 , @xmath185 does not get into the saturation region even after @xmath210 , which implies @xmath211 for @xmath92 .
since @xmath212 should increase as @xmath213 , it is empirically impossible to see the saturation of @xmath185 via simulations .
therefore , the initial slope @xmath202 at @xmath8 may have no special meaning as that of @xmath183 .
the same behavior for @xmath185 was observed for @xmath21 case @xcite , where @xmath185 initially algebraically increases with continuously varying @xmath202(@xmath214 ) as @xmath215 from below . for @xmath70 ( fig .
5(b ) ) , we estimate @xmath216 for @xmath100 ( @xmath199 ) as expected . however , for @xmath101 ( @xmath217 ) , @xmath185 algebraically increases with @xmath218 .
since for @xmath21 @xcite , @xmath212 is already larger than @xmath193 for @xmath219 , it is difficult to see the saturation of @xmath185 . for @xmath220 , we estimate @xmath200 . as in @xmath170 ,
the initial slope for @xmath97 has no special meaning . based on our numerical results
, we are convinced that @xmath221 approaches a finite value @xmath110 for @xmath97 and becomes infinite for @xmath96 in the limit @xmath190 .
hence in the limit @xmath190 , we have @xmath222 with the characteristic time @xmath223 .
another interesting quantity in condensation phenomena on networks is the average mass @xmath19 at a node with degree @xmath15 in the steady state @xcite . in zrp with chipping rate @xmath224 ,
the complete condensation takes place for @xmath225 , where @xmath226 for unweighted sfns @xcite and @xmath227 for wsfns with the weight ( [ w ] ) @xcite . for @xmath228
, @xmath19 increases as @xmath20 for @xmath229 , and as @xmath230 for @xmath231 on the wsfns . especially for @xmath232 ,
@xmath19 increases @xmath20 until @xmath233 and jumps to the value @xmath234 at @xmath155 .
hence the condensation takes place at the node with @xmath155 degree in zrp . for @xmath170 .
the solid and the dashed line correspond to @xmath92 and @xmath235 respectively .
the inset shows the scaling plot @xmath236 with @xmath237 for @xmath75 ( dashed line ) and @xmath238 for @xmath92 ( solid line).,title="fig:",width=340][fig6 ] the recent study on ca model on unweighted sfns showed that @xmath19 linearly increases up to @xmath155 without the jump at @xmath155 unlike in zrp with constant chipping rate @xcite .
the linearity of @xmath19 comes from the fact that all masses can diffuse .
the mass @xmath239 formed on the node with the degree @xmath155 can diffuse throughout network to make the steady sate distribution @xmath132 . by taking average over all nodes
, the @xmath239 soaks into the average mass @xmath19 unlike in zrp where all samples have @xmath239 at @xmath155 .
the linearity of @xmath19 on unweighted sfns results from @xmath240 @xcite .
therefore , from @xmath241 on the wsfns , we expect @xmath242 up to @xmath243 . to see this explicitly , we derive the relation @xmath244 as follows .
we consider the average total mass @xmath245 of nodes with degree @xmath15 defined as @xmath246 where @xmath247 is the probability of finding a walker with mass @xmath32 at nodes with degree @xmath15 in the steady state .
since the mass distribution @xmath30 in the steady state is independent of @xmath15 , we write @xmath248 . from ( [ pk2 ] ) and ( [ mk ] ) , one reads @xmath249 where we drop the normalization constant of @xmath14 .
since the number of nodes with degree @xmath15 is @xmath153 , @xmath19 is given as @xmath250 to confirm the scaling behavior of @xmath19 , we measure @xmath19 in the condensed phase on the wsfns of @xmath170 and @xmath73 . in fig . 6 , we plot @xmath19 against @xmath15 for @xmath75 and @xmath76 . with @xmath72 , we set @xmath78 which corresponds to the condensed phase for both @xmath44 values .
assuming @xmath251 , one expects @xmath252 for @xmath75 and @xmath253 for @xmath92 respectively . from the scaling plot @xmath254 ( inset of fig .
6 ) , we estimate @xmath255 for @xmath75 and @xmath256 for @xmath92 which agree well with the predictions .
in summary , we investigate the properties of conserved - mass aggregation ( ca ) model on weighted scale - free networks ( wsfns ) . in wsfns , the weight @xmath0 is assigned to the link between node @xmath1 and @xmath2 .
we consider the symmetric weight given as @xmath3 .
in ca model , masses diffuse with unit rate and unit mass chips off from mass with rate @xmath11 . in addition , the hopping probability @xmath5 from node @xmath1 to @xmath2 is given as @xmath257 . on the wsfns
, a walker finally reaches the hub node with @xmath155 degree for @xmath258 , while it is trapped forever at nodes with the minimal degree for @xmath259 . in the lamb - lion capture process , it means that the lion captures the lamb at the hub node within finite time interval for @xmath260 . on the other hand ,
a lamb survives indefinitely with finite probability for @xmath259 , because the lion can not escape from a node with the minimal degree to capture a lamb at some other node . in - between the two limits
, one expects a crossover @xmath8 below which the life time @xmath221 of a lamb is infinite .
however , for @xmath97 , @xmath221 is finite .
the dependence of @xmath261 on @xmath44 is similar to that on unweighted sfns of @xmath21 where @xmath221 is infinite for @xmath80 and finite otherwise @xcite . to verify the existence of @xmath8
, we need the stationary distribution @xmath14 of finding a walker at nodes with degree @xmath15 . from the equation for the transition probability @xmath262 to go from node @xmath1 to @xmath2 in @xmath133 time steps , we analytically find @xmath17 .
next , we consider the so - called lamb - lion capture process . with @xmath14 , we find the probability @xmath16 of finding two walkers at the same node with degree @xmath15 at the same time to scale @xmath263
. finally , integrating out @xmath16 , we find the death probability @xmath175 of a lamb .
a lamb survives indefinitely with the finite survival probability for @xmath116 , while it is eventually captured by a lion for @xmath97 .
we analytically find @xmath264 .
therefore , in the limit @xmath190 , the life time @xmath221 of a lamb is finite for @xmath9 , while it is infinite for @xmath265 .
we numerically confirm the all analytical results .
the existence of the condensation transitions is known to depend on @xmath221 of a lamb @xcite . for @xmath9 ,
@xmath221 is finite so the condensation always occurs for any nonzero density . on the other hand , for @xmath266 , the infinite @xmath221 ensures the condensation transitions at a certain critical density @xmath22 . for @xmath97 , we numerically confirm that the condensation always takes place at very low density .
we also numerically confirm that for @xmath116 , ca model on the wsfns undergoes the same type of the condensation transitions as those of sca model in regular lattice . finally , we investigate the behavior of the average mass @xmath19 of a node with degree @xmath15 . in zrp with constant chopping rate on networks
@xcite , @xmath19 increases as @xmath20 , and jumps to the total mass of the system at @xmath155 .
however , in the sca model on unweighted sfns , it was shown that @xmath19 linearly increases with @xmath15 up to @xmath155 without any jumps @xcite .
furthermore the linearity of @xmath19 is valid for any @xmath267 , which comes from the fact that the diffusion is only the relevant physical factor to decide the distribution @xmath19 .
similarly , on the wsfns , we analytically find and numerically confirm that @xmath19 algebraically increases as @xmath20 for any @xmath268 without any jumps .
this work was supported by the korea science and engineering foundation(kosef ) grant funded by the korea government(most ) ( no .
r01 - 2007 - 000 - 10910 - 0 ) and by the korea research foundation grant funded by the korean government ( moehrd , basic research promotion fund ) ( krf-2007 - 313-c00279 ) .
m. r. evans , europhys . lett .
36 , 13 ( 1996 ) .
o. j. oloan , m. r. evans , and m. e. cates , phys .
e 58 , 1404 ( 1998 ) .
d. van der meer et al .
, j. stat .
theor . exp .
04 , p04004 ( 2004 ) ; j. torok , cond - mat/0407567 .
r. m. ziff , j. stat .
23 , 241 ( 1980 ) .
w. h. white , j. colloid interface sci .
87 , 204 ( 1982 ) .
a. e. scheidegger , bull . i.a.s.h .
12 , 15 ( 1967 ) .
a. maritan , a. rinaldo , r. rigon , a. giacometti , and i. r. iturbe , phys .
e 53 , 1510 ( 1996 ) ; m. cieplak , a. giacometti , a. maritan , a. rinaldo , i. r. iturbe , and j. r. banavar , j. stat . phys .
91 , 1 ( 1998 ) .
s. k. friedlander , smoke , dust and haze ( wiley interscience , new york , 1977 ) .
s. n. majumdar , s. krishnamurthy , and m. barma , phys .
81 , 3691 ( 1998 ) ; s. n. majumdar , s. krishnamurthy , and m.barma , j. stat .
99 , 1 ( 2000 ) .
r. rajesh and s. n. majumdar , phys .
e * 63 * , 036114 ( 2001 ) .
r. rajesh and s. krishnamurthy , rhys .
e 66 , 046132 ( 2002 ) .
r. rajesh , d. das , b. chakraborty , and m. barma , phys .
e * 66 * , 056104(2002 ) .
s. kwon , s. lee , and yup kim , phys .
e * 73 * , 056102 ( 2006 ) .
m. r. evans , braz.j.phys .
30 , 42 ( 2000 ) ; m. r. evans and t. hanney , j. phys .
a * 38 * , r195 ( 2005 ) .
j. d. noh , g. m. shim , and h. lee , phys .
94 , 198701 ( 2005 ) .
m. tang , z. liu , and j. zhou , phys .
e * 74 * , 036101 ( 2006 ) .
s. l. pimm , _ food webs _ ( university of chicago press .
; a. e. krause et al , nature ( london ) * 426 * , 282 ( 2003 ) .
m. e. j. newman , phys .
e * 64 * , 016132 ( 2001 ) .
a. barrat , m .
barthlemy , r. pastor - santorras , and a. vespignani , proc .
natl . acad .
* 101 * , 3747 ( 2004 ) .
s. h. yook , h. jeong , a .-
barabasi , and y. tu , phys . rev . lett .
bf 85 , 5835 ( 2001 ) ; a. barrat , m. barthelemy , and a. vespignani , _ ibid _ , * 92 * , 228701 ( 2004 ) ; k .-
goh , b. kahng , and d. kim , phys .
e * 72 * , 017103 ( 2006 ) .
j. d. noh and h. rieger , phys .
lett . * 92 * , 118701 ( 2004 ) .
s. lee , s. -h .
yook , and yup kim , phys .
e * 74 * , 046118 ( 2006 ) .
goh , b. kahng , and d. kim , phys .
rev . lett . * 87 * , 278701 ( 2001 ) . p. l. krapivsky and s.redner , j. phys .
a 29 , 5347 ( 1996 ) . | we investigate the condensation phase transitions of conserved - mass aggregation ( ca ) model on weighted scale - free networks ( wsfns ) . in wsfns
, the weight @xmath0 is assigned to the link between the nodes @xmath1 and @xmath2 .
we consider the symmetric weight given as @xmath3 .
in ca model , the mass @xmath4 on the randomly chosen node @xmath1 diffuses to a linked neighbor of @xmath1,@xmath2 , with the rate @xmath5 or an unit mass chips off from the node @xmath1 to @xmath2 with the rate @xmath6 . the hopping probability @xmath5 is given as @xmath7 , where the sum runs over the linked neighbors of the node @xmath1 .
on the wsfns , we numerically show that a certain critical @xmath8 exists below which ca model undergoes the same type of the condensation transitions as those of ca model on regular lattices .
however for @xmath9 , the condensation always occurs for any density @xmath10 and @xmath11 .
we analytically find @xmath12 on the wsfn with the degree exponent @xmath13 .
to obtain @xmath8 , we analytically derive the scaling behavior of the stationary distribution @xmath14 of finding a walker at nodes with degree @xmath15 , and the probability @xmath16 of finding two walkers simultaneously at the same node with degree @xmath15 .
we find @xmath17 and @xmath18 respectively . with @xmath14
, we also show analytically and numerically that the average mass @xmath19 on a node with degree @xmath15 scales as @xmath20 without any jumps at the maximal degree of the network for any @xmath10 as in the sfns with @xmath21 . |
understanding the phase structure of quantum chromodynamics ( qcd ) is one of the key issues in current high - energy physics .
thorough phenomenological knowledge of properties of the hadron spectrum as well as nuclear matter is now being complemented by increasingly precise first - principle numerical studies of qcd at nonzero temperature .
however , the application of lattice techniques to matter at high baryon chemical potential @xmath1 remains a major challenge due to the infamous sign problem @xcite .
the difficulties encountered in simulations of qcd triggered interest in similar theories which are free of the sign problem .
these include simulations at imaginary chemical potential @xcite , qcd at nonzero isospin density @xcite , a qcd - like theory with adjoint quarks @xcite , and two - color qcd @xcite .
the latter , two - color qcd , will be the subject of the present article .
two - color qcd differs in several aspects from the world we live in .
the most notable difference perhaps is that the colorless baryons are formed from two quarks , and hence are bosons .
dense matter is then not realized as an interacting fermi sea of nucleons , but rather as a bose gas of diquarks which undergoes bose
einstein condensation ( bec ) at sufficiently low temperature .
therefore the ground state of cold and dense two - color quark matter forms a superfluid @xcite .
another noteworthy feature of two - color qcd , stemming from the fact that the @xmath2 gauge group has only pseudo - real representations , is the pauli grsey symmetry @xcite connecting quarks with antiquarks . as a consequence , the spectrum of nambu
goldstone ( ng ) bosons of the spontaneously broken chiral symmetry contains diquark states in addition to the pseudoscalar mesons .
the presence of light particles carrying baryon number ( i.e. baryonic pions ) is quite peculiar to two - color qcd ( and qcd with adjoint quarks ) and crucial for understanding the phase structure at @xmath3 and @xmath4 by means of the chiral lagrangian approach @xcite .
the model - independent chiral lagrangian arguments have been complemented by investigations in various models such as the linear sigma model @xcite , the random matrix theory @xcite , the strong - coupling expansion @xcite .
the nambu jona - lasinio ( njl ) model was first applied to two - color qcd in ref .
@xcite .
in fact there has been a tight communication between a number of effective model studies and the monte - carlo simulations on the lattice .
recent progress in this direction led to first attempts to probe bec of diquarks and the region of moderate baryon density @xcite .
thus , based on the knowledge achieved in these preceding papers , the present paper aims to make a proposal to use the two - color qcd model as a controllable test setting to clarify the following controversial issues on the real - qcd phase diagram .
* in - medium @xmath0 symmetry restoration * + the qcd lagrangian has global @xmath0 symmetry at the classical level .
the quantum anomaly , however , leads to a nonconserving contribution to the axial current , which breaks @xmath0 symmetry explicitly . from the point of view of quantum field theory
the anomaly comes from highly ultraviolet modes , and thus , the anomaly should be insensitive to any infrared scales such as @xmath5 , @xmath1 , and the quark mass @xmath6 . in this sense
the @xmath0 anomaly is never restored at any @xmath5 nor @xmath1 . in the effective model description
the @xmath0 anomaly manifests itself in the form of a @xmath0-breaking interaction which is microscopically induced by instantons @xcite .
because instantons are suppressed at high @xmath5 or @xmath1 @xcite , the @xmath0-breaking interaction is anticipated to weaken in a medium , leading to `` effective restoration '' of the @xmath0 symmetry @xcite .
in fact , a quantitative estimate of the reduction of the @xmath0 effect is crucial for locating the qcd critical point in the @xmath7 phase diagram @xcite .
the numerical study by the lattice monte - carlo simulation is in principle possible in two - color qcd .
the clear signal for @xmath0 restoration is degeneracy in the spectra of mesons connected by a @xmath0 rotation @xcite .
that is , in the two - flavor case , the @xmath0 partners are @xmath8 the masses of these multiplets become degenerate when the chiral symmetry is also restored .
since the @xmath9 meson involves the so - called disconnected diagrams , the lattice simulation is too noisy to see the degeneracy with @xmath10 ( flavor - singlet @xmath11 ) which is also in a noisy channel .
it is , however , feasible to check the degeneracy between @xmath12 and @xmath13 in the finite-@xmath5 and finite-@xmath1 lattice simulation of two - color qcd . in this paper
we will give a quantitative guide from an effective model study . * superfluidity and quarkyonic matter * + a new state of matter at high density has been recognized and it is now referred to as _ quarkyonic matter _ @xcite . because the definition of quarkyonic matter is clear only in the large @xmath14 limit , in which color non - singlet interactions are subleading , there seem to be some confusing arguments like that quarkyonic matter overwhelms color superconductivity .
this is not quite true especially when @xmath14 is finite .
intuitively , quarkyonic matter is characterized by the properties that the thermodynamic quantities ( pressure , baryon density , etc . )
should be almost saturated by the degenerated fermi liquid of quarks and the collective excitations on top of the fermi surface should be colorless ( i.e. color confined ) .
this is actually one of the known properties fulfilled by a certain color - superconducting phase , that is , the color - flavor locked ( cfl ) phase @xcite . to form a colorless object in the cfl phase , however , we need to treat a meson composed of four quarks , which is technically complicated . instead of real qcd , here ,
we shall make use of the two - color system ; this exhibits superfluidity at high density which is reminiscent of color superconductivity in qcd , and besides , we need not treat four quarks , since diquarks ( that is , baryons in the two - color world ) can make a color singlet
. we will demonstrate that the realization of quarkyonic matter is not so exclusive to disfavor superfluidity and diquark condensation .
the goal of our discussions in this part is to make convincing the phase diagram as drawn in fig .
[ fig : phase ] .
normalized by the `` stefan boltzmann '' ( see sec .
[ sec : phase ] for detailed explanations ) value ranges from 0.4 to 0.6 . ]
let us now specify the model that we will use .
physics of cooper pairing of quarks ( dense fermionic matter in general ) is well described by njl - type effective models .
however , the njl - type models suffer from a serious drawback : the lack of confinement , which may bring about artificial model predictions . in order to capture the essential features of confinement physics and
yet maintain the technical simplicity of the njl model , it was augmented with the polyakov loop , equipped with a phenomenological potential designed to reproduce selected lattice data at finite @xmath5 and zero @xmath1 @xcite .
( see also , ref .
@xcite for example , for a related approach of the polyakov loop extended quark - meson model . ) this extended model is now called the pnjl model . in spite of its simplicity
, the pnjl model has shown remarkable agreement with thermodynamic quantities measured in the lattice qcd simulations .
so far , among available finite-@xmath5 and finite-@xmath1 model options at least , the pnjl model is one of the best tools to unveil the qcd phase structure . in this paper , in a sense , we _ downgrade _ the pnjl model into the two - color setting , hoping that our model predictions would guide the future two - color simulations .
figure [ fig : phase ] is actually the phase diagram of two - color two - flavor qcd predicted by means of the pnjl model .
we note that the phase diagram is divided into three regimes : one at high @xmath5 consists of deconfined and chiral symmetric particles , which is to be identified as a quark - gluon plasma .
the other one at low @xmath5 and low @xmath1 is of course the hadronic phase .
the last one at low @xmath5 and high @xmath1 is commonly referred to as a superfluid state .
we will later discuss that this state can be regarded as quarkyonic matter
so to speak , quarkyonic superfluid of two - color quark matter . keeping in mind this phase structure
, we will look at the pole and screening masses of mesons to extract the information on @xmath0 symmetry .
the meson spectrum in the pnjl model was first analyzed in ref .
@xcite . along the same line
, we will perform the calculations and examine the dependence of meson spectrum on the @xmath0-breaking interaction strength .
the paper is organized as follows . in sec .
[ sec : model ] we introduce the model lagrangian and derive some basic analytic formulas . before going into the numerical study
, we describe in sec .
[ sec : parfix ] in detail the way we fix the parameters of our model .
most of the results have been obtained numerically and are presented in sec .
[ sec : results ] . finally , in sec .
[ sec : conclusions ] we summarize and conclude .
two - color qcd with @xmath15 massless quark flavors has a global @xmath16 flavor invariance at the classical level owing to the pauli grsey symmetry @xcite .
the axial anomaly explicitly breaks @xmath16 to @xmath17 . in the vacuum ,
the @xmath17 symmetry is spontaneously broken by the standard chiral condensate down to its @xmath18 subgroup . in the @xmath19 case , which is the subject of the present paper
, the symmetry - breaking pattern can be equivalently cast as @xmath20 @xcite .
the spectrum of ng modes therefore consists of a single 5-plet , including three pions and two diquarks ( a diquark and an antidiquark a baryonic and an antibaryonic pion ) . within the njl model , this degeneracy
is reflected by the equality of couplings in the meson and diquark channels @xcite .
let us begin the njl analysis with the @xmath0-invariant interaction lagrangian , @xmath21 \,,\end{gathered}\ ] ] where @xmath22 and @xmath23 denote pauli matrices in color and flavor spaces , and @xmath24 the charge conjugation of the dirac spinor @xmath25 with @xmath26 .
the interaction @xmath27 is minimal in the sense that it only involves the scalar and pseudoscalar channels with isospin zero and unity . while the above @xmath28 is invariant under a @xmath0 rotation , we further need an interaction which breaks the @xmath0 symmetry . to that end
we consider the analogous interaction as follows : @xmath29 \,.\end{gathered}\ ] ] the general interaction lagrangian is thus a sum of these two ; @xmath30 . at @xmath31
only @xmath28 remains and the interaction preserves @xmath0 , whereas at @xmath32 the remaining piece @xmath33 breaks @xmath0 maximally , being equivalent to the two - flavor instanton - induced interaction @xcite . indeed @xmath34 is a usually used @xmath0-violating parameter @xcite but we will also use @xmath35 for notation simplicity . in previous works , @xmath36 ( @xmath37 )
was used to discuss the phase structure @xcite . performing the hubbard
stratonovich transformation in all six channels , we arrive at the total lagrangian as @xmath38 the covariant derivative @xmath39 involves coupling of the quarks to the background gauge field @xmath40 which translates to the polyakov loop in the end .
in addition , we note that @xmath6 and @xmath41 denote the current quark mass and the quark chemical potential .
later we will introduce @xmath1 to denote the baryon chemical potential ; @xmath42 where 2 comes from the number of colors .
the collective fields @xmath43 represent in order the mesons in the scalar - isoscalar , pseudoscalar - isovector , pseudoscalar - isoscalar , and scalar - isovector channels , and the scalar and pseudoscalar diquarks .
( we hereafter omit the subscript `` 0 '' out of @xmath10 and @xmath12 for simplicity . ) in the absence of isospin chemical potential the isovector modes do not develop a vacuum expectation value . moreover , the vafa witten theorem @xcite guarantees that the chiral condensate in the vacuum has positive parity . in our model approach , this requires @xmath44 , that is , @xmath45 .
( at the same time , @xmath46 , i.e. , @xmath47 is needed for our mean - field treatment using the hubbard
stratonovich transformation @xcite . )
consequently , with the exception of the @xmath0-conserving limit @xmath31 , the scalar chiral and diquark condensates will always be preferred to the pseudoscalar ones .
we will therefore take into account only the @xmath9 and @xmath48 condensates .
the mean - field thermodynamics of the system is then independent of the parameter @xmath34 , which will only affect the propagation of collective modes , to be discussed in sec . [
sec : collective ] . in the pnjl model ,
one introduces a constant temporal gauge field which couples to the quarks via the covariant derivative . in the polyakov gauge this gauge field
is diagonal in the color space , and for the color @xmath2 group it has a form , @xmath49 , where @xmath50 is a real `` phase '' .
the order parameter for deconfinement is then the traced polyakov loop given by @xmath51 where @xmath52 is the inverse temperature .
in the mean - field approximation , the thermodynamic potential is given by a sum of the gauge and quark parts , @xmath53 in the following , we will refer to the two quark colors for simplicity as the `` red '' and `` green '' . combining the red quark and the green antiquark into the nambu gorkov spinor , @xmath54 , the background gauge field becomes proportional to the unit matrix in this doubled space and the quark thermodynamic potential can be expressed as @xmath55 where the nambu
gorkov hamiltonian reads @xmath56 here @xmath57 is the constituent quark mass and the @xmath9 and @xmath48 now stand for the condensates .
the four eigenvalues of the hamiltonian are easily found as @xmath58 and @xmath59 corresponding to the gapped quasiparticle dispersion relations , where @xmath60 the total thermodynamic potential thus becomes @xmath61 + \frac{\sigma^2+\delta^2}{4 g } \\ & - 4\sum_{i=\pm}\int\!\frac{d^3{{\boldsymbol{k}}}}{(2\pi)^3 } \biggl [ e^i_{{{\boldsymbol{k } } } } \ ! + \ !
t\log\left ( 1 \ ! + \ !
2\phi e^{-\beta e^i_{{{\boldsymbol{k } } } } } \ ! + \ !
e^{-2\beta e^i_{{{\boldsymbol{k } } } } } \right)\biggr ] \ , .
\end{split } \label{tdpot}\ ] ] the first term is the gauge part @xmath62 having two model parameters @xmath63 and @xmath64 .
we assume the simple form motivated by lattice strong - coupling expansion @xcite .
it differs from the three - color expression by a simpler logarithmic term due to the @xmath2 haar measure , and by the rescaled prefactor of the exponential term , which is in general proportional to @xmath65 .
note that , as usual in the pnjl model literature , we simulate the effects of gauge dynamics by a constant background temporal gauge field .
we then adopt a phenomenological ansatz for the gauge contribution to the mean - field thermodynamic potential , chosen to reproduce selected features of the pure gauge theory .
therefore , the parameters @xmath66 only enter the thermodynamic potential since there are no dynamical gauge degrees of freedom in our model lagrangian .
it is interesting to recall that in the three - color case the thermodynamic potential in general can not be written in terms of the polyakov loop variable @xmath67 ( and the conjugate @xmath68 ) only , and one has to use two phases analogous to our @xmath50 to parameterize it . on the contrary , in the two - color pnjl model
the thermodynamic potential depends just on @xmath67 even in the presence of a diquark condensate .
this is because diquarks are colorless .
this considerably simplifies the discussion and also avoids technical ambiguities stemming from generally complex effective actions involving the diquark condensate @xcite .
the values of the condensates in thermodynamic equilibrium are determined by minimizing the thermodynamic potential with respect to the variables @xmath9 , @xmath48 , and @xmath67 . in the quark sector , the effect of the polyakov loop as compared to the simple njl model is to make the replacement @xmath69 in the quasiparticle contribution to the thermodynamic potential .
similarly , in the gap equations as well as collective mode propagators , one generalizes @xmath70 , where @xmath71 is the fermi
dirac distribution , to @xmath72 \label{eq : mod_f}\\ & = \left(1-\frac{\phi}{\cosh(\beta e)+\phi}\right ) \bigl [ 1 - 2f(2e ) \bigr ] .
\label{eq : mod_f2}\end{aligned}\ ] ] the latter two forms ( [ eq : mod_f ] ) and ( [ eq : mod_f2 ] ) of the function @xmath73 illustrate that for @xmath74 thermal excitations are dominated by quark modes with baryon number @xmath75 ( i.e. deconfinement ) , while for @xmath76 by baryon modes with baryon number @xmath77 ( i.e. confinement ) . with these replacements , one can readily generalize the results of the pure njl model to the polyakov loop - extended one , as observed in ref .
@xcite for the three - color case . at zero temperature
the polyakov loop expectation value is zero for all values of the chemical potential .
this led to the suggestion that the pnjl model can naturally describe quarkyonic matter @xcite at high chemical potential , that is , a phase where chiral symmetry is restored but confinement persists @xcite . a simple glance at eq
. shows that when @xmath78 , thermal excitations are indeed governed by the term @xmath79 , i.e. , they correspond to colorless baryons .
let us take a closer look at how @xmath80 arises as @xmath81 .
we consider the gap equation for @xmath67 following from eq .
, that is , @xmath82 at high enough @xmath41 and low @xmath5 the system is in the bardeen cooper schrieffer ( bcs ) regime where cooper pairing of quarks occurs close to the fermi sea @xcite .
the right - hand side of this equation is then dominated by the particle ( @xmath83 ) part .
we can further simplify the calculation by using the high - density approximation , in which we expand the dispersion relation around the fermi surface , @xmath84 , and replace the measure @xmath85 by @xmath86 , where @xmath87 is the density of states at the fermi surface .
the integral thus becomes gaussian near @xmath88 and we arrive at the asymptotic result as @xmath89 the polyakov loop itself is hence suppressed exponentially at low temperature .
the propagators of the collective modes are easily obtained by a second variation of the effective action that follows from eq . after integrating the quarks
fields out .
denoting a set of collective fields symbolically as @xmath90 , the inverse propagator at imaginary ( bosonic matsubara ) frequency @xmath91 is given by @xmath92 , \end{gathered}\ ] ] where @xmath93 and @xmath94 stands for the fermionic matsubara frequencies .
the constant @xmath95 is equal to @xmath96 for @xmath97 , to @xmath98 for @xmath99 , to @xmath100 for @xmath101 , and to @xmath102 for @xmath103 .
the trace is taken in dirac , flavor , as well as nambu
gorkov space .
when calculating the partial derivatives of the hamiltonian , that determine the yukawa couplings of quarks to the bosonic modes , we demand that all fields have to be kept as in eq . .
in the diquark condensation phase , the baryon number is spontaneously broken . as a result some of the modes mix and we have to find their dispersion relations by diagonalizing a matrix propagator .
this applies to the scalar - isoscalar modes , @xmath104 , as well as pseudoscalar - isoscalar modes , @xmath105 .
the only modes that do not mix are @xmath106 and @xmath107 thanks to conservation of isospin and parity .
their inverse propagators are trivial in isospin space , namely , @xmath108 . after analytic continuation to real frequencies they acquire the following forms : @xmath109,\\ d^{-1}_a(\omega,{{\boldsymbol{p}}})=&\frac{\zeta}{2g}-\zeta^2\sum_{i , j=\pm } \sum_{k , l=\pm } \int\!\frac{d^3{{\boldsymbol{q}}}}{(2\pi)^3 } \frac1{2e^i_{{{\boldsymbol{q}}}+\frac{{{\boldsymbol{p}}}}2 } e^{-j}_{{{\boldsymbol{q}}}-\frac{{{\boldsymbol{p}}}}2 } } \left(1-ij\frac{m^2-{{\boldsymbol{q}}}^2+\frac{{{\boldsymbol{p}}}^2}{4 } } { \epsilon_{{{\boldsymbol{q}}}+ \frac{{{\boldsymbol{p}}}}{2 } } \ , \epsilon_{{{\boldsymbol{q}}}-\frac{{{\boldsymbol{p}}}}{2}}}\right)\\ & \times\frac{(e^i_{{{\boldsymbol{q}}}+\frac{{{\boldsymbol{p}}}}2}+ik\ , \xi^i_{{{\boldsymbol{q}}}+\frac{{{\boldsymbol{p}}}}2})(e^{-j}_{{{\boldsymbol{q}}}-\frac{{{\boldsymbol{p}}}}2 } + jl\,\xi^{-j}_{{{\boldsymbol{q}}}-\frac{{{\boldsymbol{p}}}}2})-kl\delta^2 } { \omega+k e^i_{{{\boldsymbol{q}}}+\frac{{{\boldsymbol{p}}}}2 } + l e^{-j}_{{{\boldsymbol{q}}}-\frac{{{\boldsymbol{p}}}}2 } } \bigl[\varphi(k e^i_{{{\boldsymbol{q}}}+\frac{{{\boldsymbol{p}}}}2 } ) + \varphi(l e^{-j}_{{{\boldsymbol{q}}}-\frac{{{\boldsymbol{p}}}}2})\bigr ] \ , .
\end{split } \label{medium_props}\ ] ] here , we have changed the momentum notation from @xmath110 to @xmath111 to reserve @xmath112 to take a summation over @xmath113 .
it should be noted that in the limit @xmath114 , only the @xmath115 terms survive and the expressions somewhat simplify .
the first of the formulas can be used together with the gap equation for @xmath48 to show that in the diquark condensation phase the pion ( pole ) mass is exactly equal to @xmath116 at zero temperature .
the proof is straightforward , hence we omit the details . in the normal phase ( where @xmath117 ) all propagators can be evaluated easily .
we provide here the list of expressions that we later in sec .
[ sec : num_collective ] use to calculate the masses numerically . to recall that these are the propagators in normal matter , let us write a superscript @xmath118 : @xmath119 propagators of the antidiquarks , @xmath120 and @xmath121 , are obtained from the diquark ones by charge conjugation , i.e. changing the sign of the chemical potential .
armed with these analytic expressions we are now ready to proceed in the numerical calculations .
first we will explain our choice of the model parameters .
next we will investigate the phase diagram , in which we particularly look into the nature of the superfluid phase .
finally we will come to the discussion of the meson spectrum dependence on the @xmath0-breaking parameter @xmath34 .
our model has five parameters whose values have to be chosen appropriately : @xmath63 and @xmath64 in the polyakov loop sector , and @xmath122 , @xmath6 , and the sharp three - momentum cutoff @xmath123 in the njl sector .
( the parameter @xmath34 or @xmath124 will be treated as an undetermined free parameter . )
let us first concentrate on the latter .
in the three - color njl model one normally fixes the values of @xmath122 , @xmath6 , and @xmath123 from the physical pion mass @xmath125 , decay constant @xmath126 , which are both measured experimentally , and the chiral condensate , which is calculated on the lattice or estimated from the qcd sum rules @xcite . the authors of ref .
@xcite chose to set up the two - color njl model using roughly the same input values , while in ref .
@xcite the pion mass , decay constant and the ( three - color ) constituent quark mass were used , resulting in a rather different parameter set .
we shall argue here that both these fits may be actually overdetermined and one should be very careful especially when comparing the model outputs with lattice data .
the chiral limit qcd , as a yang
mills theory coupled to massless quarks in the color fundamental representation , has a single mass scale , namely @xmath127 .
all physical quantities should be expressed as an appropriate power of @xmath127 times a dimensionless number . in the njl model
one can therefore choose freely a single input quantity , basically just to set the energy unit .
dimensionless combinations of observables are a pure prediction of qcd , and shall thus not be tuned to arbitrary values . in particular , either the pion decay constant , the chiral condensate , or the constituent quark mass may serve for this purpose , but not two of them independently . with the fact in mind that the current quark mass is a free parameter in the lattice simulation
, we note that the pion mass can in principle acquire any value and represents a new physical scale in addition to @xmath127 . .
the thick lines show the pnjl result ( black solid : chiral condensate in units of the vacuum value @xmath128 ; red dashed : polyakov loop @xmath67 ) , while the thin lines indicate the pure chiral ( njl ) and pure gauge transitions .
as expected , the deconfinement transition in the absence of quarks is of second order . ]
unfortunately , we are not aware of any lattice data which would provide us with the input needed to fix the njl model parameters unambiguously . in order to make at least an educated guess
, we use an argument based on the @xmath14 scaling of physical quantities . using the fact that @xmath126 is proportional to @xmath129 and the chiral condensate to @xmath14 , we rescale the three - color values by factors @xmath130 and @xmath131 , respectively .
regarding the polyakov loop sector parameters , the constant @xmath63 is related to the critical temperature @xmath132 for deconfinement in the pure gauge theory by @xmath133 . since @xmath132 in the first approximation does not depend on @xmath14 @xcite , we use the value @xmath134 .
the parameter @xmath64 can in principle be adjusted in order to make the chiral and deconfinement crossovers happen at about the same temperature @xcite . here
we use the estimate based on lattice strong - coupling expansion , @xmath135 , where @xmath136 is the physical string tension .
the input values as well as the fitted parameter set are summarized in table [ tab : parameters ] . in order to check that our parameter set is reasonable , we plot in fig .
[ fig : mu0xovers ] the expectation values of @xmath9 and @xmath67 as a function of @xmath5 at @xmath137 .
the positions of the two crossovers move very close to each other when the coupling between the quark and polyakov loop sectors is switched on .
we note that this observation of simultaneous crossovers is less clear if we use the unrescaled input parameters .
1em 1em in the vacuum the diquark is degenerate with the pions and its mass is therefore @xmath125 .
thus , when @xmath138 exceeds @xmath125 , the diquark condenses and one enters the bec phase which forms a superfluid component .
at this moment the constituent quark mass @xmath139 is still rather large .
however , as @xmath1 further increases , @xmath139 drops .
once @xmath140 , a fermi sea of quarks appears .
here we note that the baryon number density , @xmath141 , acquires a contribution from the diquark condensate and becomes nonzero as soon as @xmath142 or @xmath143 .
because of the binding energy , naturally , @xmath144 is smaller than @xmath139 , and thus the onset of @xmath145 emerges first and then a quark fermi sea shows up with increasing @xmath41 . in the presence of the fermi sea the diquark condensation is closer to the bcs pairing of quarks sitting near the fermi surface rather than to bec of bound bosonic molecules .
there is no phase transition associated with this qualitative change of behavior , so one speaks of a bcs bec crossover . even though it is not particularly sharp , it can be conveniently defined by the condition @xmath146 @xcite .
we will draw the phase diagram of our model in a `` conventional '' way . in the phase diagram the deconfinement crossover
is conveniently defined by the condition @xmath147 .
the deconfinement temperature is then almost independent of the chemical potential .
in fact , the value of the polyakov loop at all temperatures depends on @xmath1 very weakly , as can be seen from fig .
[ fig:3dcondensates ] , where all condensates are plotted as a function of @xmath148 and @xmath149 where @xmath125 is fixed to be @xmath150 .
this behavior of the polyakov loop can be traced back to the fact that the two - color diquark is a color singlet , and therefore does not break center symmetry to induce nonzero polyakov loop directly . even at nonzero quark mass
the diquark condensate exhibits a clear second - order phase transition as shown in the plot for @xmath151 ( where @xmath128 is the vacuum value of the chiral condensate ) in fig .
[ fig:3dcondensates ] . unlike the deconfinement crossover
, we can draw a well - defined phase boundary in the phase diagram separating the normal and superfluid phases . as we have already mentioned above
, we can confirm that nonzero @xmath48 certainly appears at @xmath142 at @xmath152 . plane .
we indicate the chiral - symmetry breaking ( @xmath153sb ) , bec , and bcs phases .
solid line : critical onset of diquark condensation where @xmath48 starts to be nonzero . dashed line : deconfinement crossover ( defined by @xmath147 )
. dash - dotted line : bcs bec crossover ( defined by @xmath146 ) . ]
a compilation of these data leads to the phase diagram we present in fig .
[ fig : phase_diagram ] .
the solid line represents a second - order phase transition between the normal and superfluid phases . in the superfluid region
we have added a dash - dotted line which indicates @xmath146 and can be interpreted as a bcs bec - type crossover .
the dashed line is the deconfinement crossover defined by @xmath147 .
figure [ fig : phase_diagram ] is a basis to understand a more `` advocative '' way of presenting the phase diagram as shown in fig . [ fig : phase ] .
now we are well prepared to discuss the physical meaning of each phase labeled in fig .
[ fig : phase ] .
the red band which spreads almost straight along the horizontal axis represents the deconfinement crossover .
because a crossover has a width and does not have a unique definition , it should be much more reasonable to express the transition region not by a line but a band .
the band width in fact tells us how rapid or slow the crossover is .
we drew the deconfinement band by the condition that @xmath67 ranges from 0.4 to 0.6 .
the blue band showing a sudden decrease around @xmath154 is the chiral crossover defined similarly by the condition that @xmath155 ranges from 0.4 to 0.6 .
we see that the two crossovers of deconfinement and chiral restoration take place simultaneously at zero density , and this coincidence persists until around @xmath154 .
the pink dotted line represents the superfluid onset , which is a well - defined phase transition .
finally the green band which shows behavior similar to the chiral crossover is drawn by the baryon number density @xmath141 .
we can compute @xmath141 in the pnjl model and normalize it by the `` stefan boltzmann '' value .
since the pnjl model is a cutoff theory , even the free ( non - interacting ) limit suffers from the cutoff artifact and deviates from the standard formula , which is also the case in the lattice simulation @xcite .
therefore we evaluate the baryon number density in the stefan boltzmann limit @xmath156 using the pnjl model with @xmath157 and @xmath158 imposed by hand . in this way , we indicate by the green band in fig . [
fig : phase ] the region in which @xmath159 ranges from 0.4 to 0.6 .
because the chiral phase transition controls the dynamical quark mass , it is conceivable that the bcs bec crossover in fig .
[ fig : phase_diagram ] is associated with the chiral crossover .
this is indeed the case ; the dash - dotted line in fig .
[ fig : phase_diagram ] is covered by the chiral crossover band in fig .
[ fig : phase ] .
hence , as labeled in fig .
[ fig : phase ] , the right - bottom region is characterized by small @xmath67 ( confinement ) and small @xmath155 ( chiral symmetric ) .
this is in fact in accord with the identification of quarkyonic matter in refs .
@xcite . in this case the blue band is interpreted as the quarkyonic transition .
if we use @xmath159 to define the quarkyonic transition according to refs .
@xcite , the green band , instead of the blue band , plays the role of the quarkyonic boundary .
we would emphasize here that the former criterion makes more physical sense at least in the present case of @xmath160 .
actually , our consideration of the bcs bec crossover provides us with a clear view point on this issue .
as we have already discussed , in the chemical potential window @xmath161 , finite baryon number density grows .
the carriers of the baryon number are , however , not quarks but baryons ( baryonic pions ) .
this is so because @xmath162 in this region .
therefore , in the phase region we called bec in fig . [
fig : phase_diagram ] , the more appropriate physical interpretation should be `` superfluid nuclear matter '' rather than a quarkyonic state in which the pressure is mostly given by fermi - degenerated quarks .
this difference in the interpretation makes a contrast to the large-@xmath14 arguments @xcite , and it still remains a question to which the real world with @xmath163 is closer , infinite @xmath14 or @xmath160 ?
answering this question goes beyond our current scope .
if the quark - hadron continuity scenario driven by the cfl state is realistic , we can say that the situation at @xmath160 is more relevant .
before closing this section we mention the previous studies . in preceding works
there was some controversy regarding the order of the phase transition from the diquark condensation phase to the normal phase . in
@xcite it was concluded that there is a tricritical point at @xmath149 somewhere in the range @xmath164 , and for higher @xmath1 the transition becomes first order . on the other hand
, the authors of @xcite used the thouless criterion to calculate the critical temperature , which _ assumes _ that the transition is second order .
our numerical results for the diquark condensate as a function of @xmath148 and @xmath149 ( see fig . [ fig:3dcondensates ] ) suggest that the transition is second order everywhere .
we have further confirmed the statement that there is no first - order phase transition by looking at the quartic ginzburg
landau coefficient in our numerical calculation .
detailed computations and arguments are given in the appendix .
-breaking parameter @xmath34 .
the black solid is @xmath125 , the red dashed @xmath165 , the blue dotted @xmath166 , and the green dash - dotted @xmath167 .
we see that @xmath168 and @xmath169 in the @xmath0 symmetric ( @xmath31 ) case . ]
we are going to investigate the mass spectrum of collective modes as a function of @xmath5 and @xmath1 for different values of the @xmath0-breaking parameter @xmath34 .
a useful starting point therefore is the @xmath34-dependence of the masses in the vacuum .
this is shown in fig .
[ fig : vacmass ] .
it is worth noting that we only display the meson spectrum ; thanks to the unbroken @xmath170 symmetry at @xmath137 , the scalar diquark @xmath48 is degenerate with @xmath171 mesons and the pseudoscalar diquark @xmath172 is degenerate with the @xmath173 mesons .
we should perhaps emphasize that whenever we speak of a ( pole ) mass , we have in mind the zero of the real part of the inverse propagator @xcite .
we will therefore sometimes refer to it as the real - part mass .
this coincides with the position of the pole if the zero appears below the threshold for decay into quark pairs .
another prescription for the mass would be to take the real part of the complex pole of the meson propagator , yielding a somewhat different result , or to compute the spectral function whose peak position and broadness indicate the physical mass and decay width . in the @xmath0-symmetric limit ( @xmath31 ) , in our calculations , the masses of @xmath10 and @xmath173 are equal to those of @xmath171 and @xmath9 , respectively .
the degeneracy of @xmath10 and @xmath171 is generally exact only when the quark mass is strictly zero , so that they are both massless ng bosons . in the ( p)njl model in the mean - field approximation ,
this exact degeneracy holds regardless of finite quark masses , which is an artifact of the approximation .
in fact , @xmath10 and @xmath171 belong to different irreducible representations of the unbroken @xmath170 ; @xmath10 is a singlet and @xmath171 ( with @xmath48 ) form a quintet .
the degeneracy is only approximate for small quark mass once higher - order meson loops are taken into account . off the limit of @xmath31 , both @xmath166 and @xmath167
increase steeply .
the typical value of @xmath34 in the three - color njl model with two light quark flavors , that one can obtain either by fitting the physical @xmath174 mass or by a reduction of the three - flavor model , is in the range @xmath175 @xcite . in order to be able to investigate the convergence to the @xmath0-symmetric limit
, we consider in the following two particular values , @xmath176 and @xmath177 .
that is , if we assume that the vacuum value is @xmath178 , we consider two examples of no reduction at all and @xmath179 reduction of @xmath0 effects .
it is quite unlikely that @xmath34 goes to zero in the @xmath5 and @xmath1 range of our interest . as a function of @xmath148 .
the notation for the lines is the same as in fig .
[ fig : vacmass ] : the black solid is @xmath125 , the red dashed @xmath165 , the blue dotted @xmath166 , and the green dash - dotted @xmath167 .
the thick lines correspond to @xmath166 and @xmath167 at @xmath176 , while the thin lines to @xmath177 . ] in fig .
[ fig : mu0mass ] the meson real - part masses are plotted as a function of @xmath148 while keeping @xmath137 .
the reason why some curves exhibit a cusp structure is that the corresponding modes cross the quark
antiquark ( or quark quark in the case of diquarks ) threshold as @xmath5 increases ( and @xmath139 decreases accordingly ) .
of course , such a decay into quarks is unphysical and is just an artifact , following from the lack of confinement in the njl model . even in the pnjl model
these decay processes are not sufficiently suppressed @xcite , since the coupling to the polyakov loop only imitates confinement in a statistical sense . as a function of @xmath148 .
the notation for the lines is the same as in fig .
[ fig : vacmass ] .
the black solid is @xmath125 , the red dashed @xmath165 , the blue dotted @xmath166 , and the green dash - dotted @xmath167 .
the thick lines correspond to @xmath166 and @xmath167 at @xmath176 , while the thin lines to @xmath177 . ]
the real - part masses are difficult to measure on the lattice because the number of lattice sites in the temporal direction is severely limited .
instead , one can straightforwardly determine the damping of correlators of the fermion - bilinear interpolating fields for the mesons at large _
spatial _ separations .
the exponential decay of the correlations is related to the screening mass of the lightest mode in the selected channel . in the ( p)njl model , this
can be found as the pole of the static propagator , @xmath180 , in the complex - momentum plane .
the results are shown in fig .
[ fig : mu0screen ] . at @xmath152
the real - part and screening masses should coincide thanks to the lorentz invariance . in the model calculation , however , this nice feature is slightly breached by the three - momentum cutoff , but the difference of the two masses in the vacuum turns out to be about a few percent at most . hence the apparent cutoff artifacts are reasonably small .
the masses of the chiral partners become degenerate at high temperature , signaling the restoration of chiral symmetry . on the other hand , at any ( fixed ) @xmath181 the masses of the parity partners , connected by a @xmath0 rotation , do not converge even at the highest temperatures considered . from the technical point of view , this is a consequence of the simple structure of the inverse propagators : when chiral symmetry is restored , the self - energies of the parity partners become equal up to a simple rescaling by @xmath182 . physically , in reality , one should expect the coupling @xmath34 to vary with @xmath5 since it is induced by instantons whose density is exponentially suppressed at high @xmath5 @xcite . if we consider that the @xmath0-breaking interaction strength is proportional to the ( pure ) topological susceptibility , we can infer the @xmath5-dependence from the lattice data in the pure gauge simulation . instead of doing so , in this work , we pick up several values of @xmath34 . a proper way to understand fig .
[ fig : mu0screen ] is thus as follows . at @xmath152
naturally @xmath34 is nonzero , and if precise two - color simulation data is available for @xmath166 and @xmath167 , in principle , @xmath34 can be fixed by the data .
we can perform the pnjl model calculations using the determined @xmath34 to go to the higher temperature . if we see a reduction of @xmath166 and @xmath167 toward degenerated @xmath125 and @xmath165 , it is a signal for the effective @xmath0 restoration .
we can deduce how far @xmath34 decreases by adjusting @xmath34 to fit @xmath166 and @xmath167 at each temperature . hence , fig .
[ fig : mu0screen ] is a demonstration for all these possible investigations once the two - color simulation successfully measures the screening masses in good precision . and
@xmath171 as a function of @xmath149 and @xmath148 in the extreme case of @xmath31 .
this quantity indicates the progressive @xmath0 restoration in the hot and/or dense medium .
the masses are denoted by @xmath183 with asterisk to make clear that these are in - medium quantities and different from the vacuum @xmath125 which is a fixed parameter of the model ; see table [ tab : parameters ] . ] finally , in fig .
[ fig:3dmass ] , we plot the difference of screening masses of @xmath173 and @xmath171 as a function of @xmath149 and @xmath148 . since these two modes do not mix even in presence of the diquark condensate , the masses can be calculated straightforwardly also in the bec phase using the in - medium propagators .
the choice of the screening masses instead of the real - part ones here is motivated by the lattice measurement , and also technically favored : while the real - part masses are obscured by landau damping in the diquark condensation phase at @xmath3 , the screening masses remain well defined .
the results indicate effective restoration of @xmath0 symmetry at high temperature and/or chemical potential . while in the @xmath5-direction the degeneracy more - or - less copies the restoration of chiral symmetry which is another source of @xmath0 breaking , in the @xmath1-direction the convergence of @xmath184 and @xmath185 is slower as a result of additional @xmath0 breaking by the diquark condensate .
we note that fig .
[ fig:3dmass ] is the extreme example of @xmath31 , which is not likely near the phase boundaries , but could be the case in the quark - gluon plasma region in view of the lattice data of the topological susceptibility @xcite . in the future , by combining the two - color lattice outputs and the pnjl model analysis , it would be possible to make a 3d plot of @xmath34 which should approach zero at high @xmath5 and/or high @xmath1 .
we have adopted the pnjl model as an effective approach to two - color qcd and applied it to the case of two light quark flavors .
this is the simplest case which exhibits nontrivial low - energy spectrum due to spontaneous chiral symmetry breaking , and at the same time can be simulated by lattice monte - carlo techniques .
we argued that one can fit the parameters of the njl part of the model using physical ( three - color ) observables , but their values have to be rescaled appropriately . once this is done and the quark sector is coupled to the polyakov loop , the model yields locking behavior of chiral and deconfinement crossovers as long as @xmath1 is zero .
we checked older results on the phase structure of the two - color njl model in the plane of @xmath148 and @xmath149 , and analyzed its modification induced by the coupling to the polyakov loop .
the phase transition between the normal and superfluid phases is second order for all values of the chemical potential considered . in a large part of the diquark condensation phase the expectation value of the polyakov loop is small , which resembles
quarkyonic matter predicted using large-@xmath14 arguments .
we carefully clarified the realization of quarkyonic matter in the two - color system .
the baryon number density @xmath141 appears finite as soon as @xmath1 exceeds the mass of the baryonic pion , but still the quark contribution to @xmath141 is not substantial until @xmath1 surpasses the twice of the dynamical quark mass .
after then quark degrees of freedom supersede baryons , meaning a transition from `` superfluid nuclear matter '' into `` quarkyonic superfluid . ''
our model analysis of the phase diagram is based on two simplifying assumptions .
the first one is the mean - field approximation which treats the system as a gas of noninteracting fermionic quasiparticles .
this may not be quantitatively accurate in some regions of the phase diagram such as for @xmath186 at nonzero temperature where the system behaves rather as a dilute bose gas .
on the other hand , in the theory of strongly - interacting fermi gases the mean - field approximation is known to be reliable at zero temperature . moreover , the structure of the phase diagram concerning diquark condensation and chiral symmetry restoration is robust , being a direct consequence of the symmetry of two - color qcd .
the second assumption is the extrapolation of the gauge part of the thermodynamic potential to nonzero baryon chemical potential .
this has been justified for three - color qcd and low baryon chemical potential by a comparison with available lattice data and , in fact , is the source of the predictive power of the pnjl model .
therefore , we are confident about the existence of the quarkyonic superfluid phase as depicted in fig . [
fig : phase ] . on the other hand
, the most recent lattice data @xcite suggest that the polyakov loop at a fixed low value of temperature starts to rise at @xmath187 , signalling possible deconfinement .
this certainly presents a challenge to the pnjl model and determines the direction of our future research efforts . finally , we studied the dependence of the spectrum of collective excitations ( scalar and pseudoscalar mesons and diquarks ) on the strength of the axial anomaly . for that sake
we introduced a njl - type interaction with a tunable @xmath0-breaking parameter .
for all modes we calculated both the real - part mass and the screening mass , which governs the decay of spatial correlators , in order to facilitate a direct comparison with lattice simulations . above the chiral restoration / deconfinement temperature the masses of the chiral partners
become degenerate , as expected . at the same time , our results indicate that the restoration of @xmath0 symmetry in terms of the masses of parity partners can not be hidden by chiral restoration unlike the full topological susceptibility , which is good .
this would naturally incorporate in the model study the suppression of instanton effects in a hot and/or dense matter . to make this whole argument into a quantitative predictive framework ,
further physical input from the lattice simulation would be indispensable , e.g. precise measurement of @xmath167 as a function of @xmath5 and @xmath1
. it would be also interesting that , since the topological susceptibility in the two - color pure gauge theory does not cost much , we can compare the inferred @xmath34 behavior and the suppression of the pure topological susceptibility .
the authors are grateful to h. abuki , j. o. andersen , x. huang , l. kyllingstad , and w. weise for useful discussions , and to l. he for pointing out an error in an earlier version of the manuscript .
the work of t. b. was supported in part by the alexander von humboldt foundation , and by the extreme matter institute emmi in the framework of the helmholtz alliance program of the helmholtz association ( ha216/emmi ) .
k. f. is supported by japanese mext grant no.20740134 and also supported in part by yukawa international program for quark hadron sciences .
y. h. was supported by the grant - in - aid for the global coe program `` the next generation of physics , spun from universality and emergence '' from the mext of japan . .
the solid and dashed lines are the pnjl and njl model results , respectively , with the same input parameters . ]
in order to make clear whether the phase transition of diquark superfluidity becomes ( weakly ) first order at high @xmath1 , we performed the ginzburg
landau expansion of the thermodynamic potential near the second - order transition line and calculated the coefficient of the quartic term with respect to @xmath48 . in the presence of a tricritical point and the onset of a first - order phase transition , this coefficient would go to zero and then change its sign to negative .
the thermodynamic potential @xmath188 depends only on the square of the diquark condensate . in general
, @xmath188 depends on other condensates , symbolically denoted by @xmath189 , as well . in the pnjl model
, we have @xmath190 , while in the standard njl model the only other condensate would be @xmath9 .
to study the behavior of the diquark condensate near the critical temperature , it is most convenient to solve the gap equations for the other condensates , i.e. , @xmath191 .
these define @xmath189 implicitly as a function of @xmath192 , and the thermodynamic potential is then a function of @xmath192 solely , that is , @xmath193 .
the coefficients of the quadratic and quartic terms in the ginzburg
landau functional are now determined by the first and second _ total _ derivatives with respect to @xmath192 , evaluated at @xmath117 .
the first derivative vanishes at the transition point by means of the gap equation .
the second derivative defines the effective ginzburg
landau quartic coupling and is in general expressed as @xmath194 the inverse in the second term is assumed in the matrix sense .
the sign of @xmath195 decides whether the transition is of first or second order .
in particular in the njl model , this expression acquires a simple form ( with an obvious notation for the partial derivatives ) @xmath196 in the pnjl model , one has to calculate the @xmath197 matrix of second partial derivatives . given the formula for the thermodynamic potential , these are easily evaluated explicitly at @xmath117 as , @xmath198 \notag\\ & + 4t\sum_{i=\pm}\int\frac{d^3{{\boldsymbol{k}}}}{(2\pi)^3 } \frac1{(\cosh\beta\xi^i_{{{\boldsymbol{k}}}}+\phi)^2 } \,.\end{aligned}\ ] ] the njl limit is recovered by setting @xmath199 .
we plot the numerical results in fig .
[ fig : glcoupling ] , from which we conclude that the ginzburg
landau coupling is always positive , and that the phase transition hence is always second order , within a reasonable chemical potential range ( below the cutoff @xmath123 ) , and in both njl and pnjl models . | we discuss the phase structure of quantum chromodynamics ( qcd ) with two colors and two flavors of light quarks .
this is motivated by the increasing interest in the qcd phase diagram as follows : ( 1 ) the qcd critical point search has been under intensive dispute and its location and existence suffer from uncertainty of effective @xmath0 symmetry restoration .
( 2 ) a new phase called quarkyonic matter is drawing theoretical and experimental attention but it is not clear whether it can coexist with diquark condensation . we point out that two - color qcd is nontrivial enough to contain essential ingredients for ( 1 ) and ( 2 ) both , and most importantly , is a system without the sign problem in numerical simulations on the lattice .
we adopt the two - flavor nambu jona - lasinio model extended with the two - color polyakov loop and make quantitative predictions which can be tested by lattice simulations . |
much attention has been focused on the vortex pinning in type - ii superconductors .
the vortex pinning in the superconductors under magnetic fields plays an important role on various vortex - related quantities and phenomena such as the critical current , hysteresis of the magnetization , and quantum vortex tunneling .
the problem of the vortex pinning is categorized in two aspects .
one is the elementary vortex pinning force , which is the interaction between a vortex and a single defect .
the other is the summation problem , which is how the elementary vortex pinning forces add up to a holding force on the elastic vortex lattice under a distribution of defects @xcite .
a study of the elementary vortex pinning is the step necessary for the understanding of the phenomena related to the vortex pinning in the superconductors under magnetic fields .
traditionally , the mechanism of the elementary vortex pinning is considered to be that a defect locally prohibits superconducting condensation so that it attracts a normal region of a vortex core in order to avoid the loss of the condensation energy .
@xcite advanced the understanding of the mechanism of the elementary vortex pinning , taking account of a nonlocal effect that a defect as a scattering center scatters the quasiparticles .
this mechanism of the vortex pinning is that the quasiparticle scattering by the defect inside a vortex core helps to avoid suppressions of the order parameter around the vortex center up to distances of the order of the coherence length . in this paper
, we investigate the elementary vortex pinning potential in unconventional superconductors , following the analysis started by thuneberg _
@xcite for conventional @xmath1-wave superconductors .
the unconventional superconductivity is recently proposed for many superconductors such as high-@xmath3 cuprates , organic conductors , and heavy - fermion compounds .
for example , the @xmath0-wave superconductivity is believed to be realized in most of the high-@xmath3 cuprates , and the chiral @xmath2-wave one is expected in a non - copper - layered perovskite superconductor sr@xmath4ruo@xmath5 @xcite ( although it seems to be open to further discussion @xcite ) .
we calculate numerically the elementary vortex pinning potential for @xmath0- , anisotropic @xmath1- , and isotropic @xmath1-wave pairings , and compare them each other to clarify what could be one of the characteristic properties specific to the superconductors with the unconventional order parameter .
we also present an analytical consideration of a novel chirality - dependent vortex pinning in a chiral @xmath2-wave superconductor .
to investigate the vortex pinning , we use the quasiclassical theory of superconductivity @xcite .
we start with the eilenberger equation for the quasiclassical green function in the absence of the pinning , @xmath6 namely , @xmath7 = 0 , \label{eq : eilen}\end{aligned}\ ] ] where @xmath8 is the pauli matrix and the order parameter @xmath9 the eilenberger equation ( [ eq : eilen ] ) is supplemented by the normalization condition @xmath10 , and the commutator is @xmath11={\hat a}{\hat b}-{\hat b}{\hat a}$ ] .
the symbol @xmath12 denotes the center of mass coordinate of the cooper pairs and @xmath13 denotes the relative coordinate of them
. the cylindrical fermi surface is assumed .
we use units in which @xmath14 .
following thuneberg _
@xcite , the effect of the pinning is introduced as follows to the quasiclassical theory of superconductivity .
the quasiclassical green function @xmath15 in the presence of a point - like non - magnetic defect situated at @xmath16 is obtained from the eilenberger equation @xmath17 = \bigl [ { \hat t } , { \hat g}_{\rm imt } \bigr ] \delta ( { \bf r } ' ) , \label{eq : eilen - pin}\end{aligned}\ ] ] and the @xmath18 matrix due to the defect @xmath19 , \label{eq : t - matrix}\end{aligned}\ ] ] where @xmath20 , the denominator @xmath21 $ ] , the symbol @xmath22 , the normal - state density of states on the fermi surface is @xmath23 , and we assume the @xmath1-wave scattering @xmath24 to obtain eq .
( [ eq : t - matrix ] ) .
the cross section of the defect is given by @xmath25 with the scattering phase shift @xmath26 , and @xmath27 .
the free energy in the presence of the defect is , at the temperature @xmath28 , given as @xcite @xmath29 , \label{eq : free - ene}\end{aligned}\ ] ] where @xmath30 is evaluated at @xmath31 , and @xmath32 is the order parameter in the absence of the defect .
equation ( [ eq : free - ene ] ) represents the difference in the free energy between the states with and without the defect , and then gives the vortex pinning potential @xmath33 . between the vortex center and the defect .
the defect is situated on the @xmath34 axis ( in the direction of @xmath35 ) .
, title="fig:",width=302 ] between the vortex center and the defect .
the defect is situated on the @xmath34 axis ( in the direction of @xmath35 ) .
, title="fig:",width=302 ] between the vortex center and the defect .
the defect is situated on the @xmath34 axis ( in the direction of @xmath35 ) .
, title="fig:",width=302 ] between the vortex center and the defect .
the defect is situated on the @xmath34 axis ( in the direction of @xmath35 ) .
, title="fig:",width=302 ]
we calculate the vortex pinning potential @xmath33 numerically with the order parameter around a vortex , @xmath36 \exp(i\phi)$ ] , ( note that the vortex center is situated at @xmath37 ) . here , the @xmath38-dependent function @xmath39 for isotropic @xmath1-wave , @xmath40 for anisotropic @xmath1-wave , and @xmath41 for @xmath0-wave pairings .
@xmath42 is the bcs order parameter amplitude and @xmath43 the coherence length .
$ ] , @xmath45 and @xmath46 in fig .
[ fig:1 ] , we show the numerical results for the born limit ( @xmath47 ) and the unitary limit ( @xmath48 ) . the defect is situated on the @xmath34 axis ( in the direction of @xmath35 ) . in the case of the isotropic @xmath1-wave pairing , when the defect is situated in bulk , i.e. , far away from the vortex center ( @xmath49 ) , the difference in the free energy between the states with and without the defect , @xmath50 , is equal to zero .
this is because the anderson s theorem @xcite is valid for homogeneous systems in isotropic @xmath1-wave superconductors and the non - magnetic defect has no influence on the free energy . on the other hand , in the anisotropic pairing cases , @xmath51 is finite and positive as seen in fig .
[ fig:1 ] , which is the loss of the condensation energy in bulk due to the presence of the defect and to the breakdown of the anderson s theorem .
the condensation energy loss in bulk @xmath51 contributes to the depth of the vortex pinning potential @xmath50 , i.e. , the vortex pinning energy . while the contribution of @xmath51 to the vortex pinning energy was first pointed out by thuneberg _
_ @xcite in the context of an ion in @xmath2-wave superfluid @xmath52he - b , we performed here a numerical calculation explicitly for the vortex pinning potential in the @xmath0-wave superconductors .
it is noted from figs .
[ fig:1](c ) and [ fig:1](d ) that , at @xmath53 ( high temperature ) , the vortex pinning energy in the @xmath0-wave case is dominantly determined by the condensation energy loss in bulk @xmath51 , and is 4 13 times larger than those in the @xmath1-wave cases . for the chiral @xmath2-wave pairing @xmath54 @xcite , we analytically investigate the vortex pinning potential . in the case of that chiral @xmath2-wave pairing , there
exist the following two possible forms for the order parameter @xmath55 @xmath56 of axisymmetric vortex @xcite : ( i ) @xmath57 + \delta_{-}(r ) \exp\bigl[i(-\theta+\phi)\bigr]$ ] , where the chirality and vorticity are antiparallel ; ( ii ) @xmath58 + \delta_{-}(r ) \exp\bigl[i(-\theta+3\phi)\bigr]$ ] , where the chirality and vorticity are parallel . here , @xmath59 are real , @xmath60 , and @xmath61 .
we focus on @xmath62 , where both the defect and the vortex center are situated just at the origin @xmath37 .
the quasiparticles inside the vortex core , subject to the andreev reflection , run along straight lines , namely the quasiclassical trajectories @xcite .
we consider the quasiclassical trajectories which go through the origin @xmath37 . on those trajectories with zero impact parameter , the position vector is parallel to the direction of the trajectory ( i.e. , @xmath63 ) , and therefore @xmath64 .
in this situation the order parameter is @xmath65 in the case of ( i ) , and @xmath66 in the case of ( ii ) . of importance
is the difference in the phase factor of these order parameters of ( i ) and ( ii ) . on the basis of an analysis of the so - called zero - core vortex model in ref .
@xcite , the matrix elements of @xmath67 at the vortex center are approximately obtained as @xmath68 , @xmath69 , and @xmath70 , where @xmath71 .
because of the difference in the phase factor of the order parameters mentioned above , @xmath72 and @xmath73 in the case of ( i ) , while @xmath74 and @xmath75 in the case of ( ii ) . therefore , in the case of ( i ) , @xmath76 , and then @xmath77=0 $ ] from eq .
( [ eq : t - matrix ] ) . in the case of
( ii ) , @xmath78 , and then @xmath77 \neq 0 $ ] . when @xmath77=0 $ ] , the eilenberger equation ( [ eq : eilen - pin ] ) in the presence of the defect is identical to eq . ( [ eq : eilen ] ) ( the equation in the absence of the defect ) , i.e. , the defect has no influence on the green function and the free energy . from it and the above analysis , @xmath79 in the case of ( i ) when the chirality is antiparallel to the vorticity , and @xmath80 in the case of ( ii ) when the sense of the chirality is the same as that of the vorticity .
it means that the vortex pinning depends on the chirality in the chiral @xmath2-wave superconductor .
this analytical result is certainly confirmed by a numerical calculation @xcite with self - consistently obtained order parameters @xmath81 @xcite .
we investigated the elementary vortex pinning potential @xmath50 on the basis of the quasiclassical theory of superconductivity .
the numerical results were presented in fig .
[ fig:1 ] for the @xmath0-wave and @xmath1-wave pairings .
the vortex pinning energy in the @xmath0-wave pairing case is about 10 times larger than those in the @xmath1-wave pairing cases , at a high temperature , because of the breakdown of the anderson s theorem in unconventional superconductors .
we also investigated analytically the quasiparticles inside the vortex core in the chiral @xmath2-wave superconductor , and found the chirality - dependent vortex pinning .
we expect that the chirality - dependent vortex pinning revealed in this paper shares the same physics with impurity effects on the quasiparticle relaxation time @xcite and on the energy spectrum @xcite inside the vortex core in chiral @xmath2-wave superconductors .
one of the authors ( n.h . )
thanks m. ichioka , k. machida , m. takigawa , n. nakai , and m. matsumoto for useful discussions .
y.k . thanks m. sigrist for useful discussions .
this work is partly supported by grant - in - aid for scientific research on priority areas ( a ) of `` novel quantum phenomena in transition metal oxides " ( 12046225 ) from the ministry of education , science , sports and culture and grant - in - aid for encouragement of young scientists from japan society for the promotion of science ( 12740203 ) . | the elementary vortex pinning potential is studied in unconventional superconductors within the framework of the quasiclassical theory of superconductivity .
numerical results are presented for @xmath0- , anisotropic @xmath1- , and isotropic @xmath1-wave superconductors to show explicitly that in unconventional superconductors the vortex pinning potential is determined mainly by the loss of the condensation energy in bulk due to the presence of the pinning center , i.e. , by the breakdown of the anderson s theorem .
it is found that the vortex pinning energy in the @xmath0-wave pairing case is 4 13 times larger than those in the @xmath1-wave pairing cases .
this means that an enhancement of pinning effect in unconventional superconductors occurs due to the breakdown of the anderson s theorem .
the case of a chiral @xmath2-wave superconductor is also investigated in terms of the vortex core states subject to the andreev reflection , where important is whether the vorticity and chirality are parallel or antiparallel . and vortex pinning , unconventional superconductor , @xmath0-wave pairing , chiral @xmath2-wave pairing 74.60.ge |
higher - curvature corrections to the einstein - hilbert ( eh ) action appear in any sensible theory of quantum gravity as next - to - leading orders in the effective action .
quadratic terms , for instance , such as the lanczos - gauss - bonnet ( lgb ) action @xcite , appear in _ bona fide _ realizations of string theory @xcite and m - theory @xcite .
interesting implications of these terms within the context of the ads / cft correspondence has been recently the focus of thorough investigation ( see , _
e.g. _ , @xcite ) .
they allow for the holographic description of a broad family of quantum field theories ( like 4d superconformal field theories with unequal central charges @xcite ) , as well as for the study of the fluid / gravity correspondence beyond the eh gravitational sector .
a remarkable example of higher - curvature gravity is lovelock theory @xcite , the natural extension of general relativity to higher dimensions . more precisely , it is the most general theory of gravity yielding second - order field equations .
general relativity and chern - simons gravity @xcite are particular cases of lovelock theories .
although quantum corrections are not generally of the lovelock type , these theories provide a tractable playground that captures many important features that are rather generic . among them , the existence of new branches of black hole solutions corresponding to vacua with different cosmological constants that pop out as soon as higher - curvature terms are brought into place .
the main motivation of the present work is that of studying transitions between the different vacua of the theory .
this is important to investigate whether a new type of instability involving non - perturbative solutions occurs in the theory .
studying such transitions may uncover a mechanism that eventually leads to a built - in mechanism to reduce the ( absolute ) value of the effective cosmological constant .
it is a well - known fact that black holes in ads display the so - called hawking - page ( hp ) transition @xcite , _
i.e. _ , the semiclassical phase transition between thermal ads space at finite temperature and a black hole in ads with the same hawking temperature
. this has been further interpreted as a confinement / deconfinement phase transition in the dual cft @xcite .
a natural question arises regarding the role of the different black hole branches and would be transitions among them . at first glance
it might seem that such phase transitions are forbidden since distinct branches exhibit a different asymptotic behavior . in this article , however , we report on a new type of phase transition taking place in higher - curvature gravities , which can be thought of as a sophisticated version of the hp transition involving different branches .
we work out the simplest example given by lgb gravity whose ordinary hp transition was studied in @xcite , since it is enough to realize that these phase transitions occur in lovelock theory as well @xcite , and possibly in a larger class of higher - curvature theories .
we consider the theory at finite temperature and observe that , above a critical temperature , @xmath0 , the higher - curvature vacuum decays producing a bubble which hosts a black hole in its interior .
this phase transition is analogous to the _ thermalon _ transition discussed in @xcite , where the materialization of an electrically charged bubble induces the decay of a de sitter vacuum into another with a smaller cosmological constant , but containing a black hole . in the present case ,
however , the bubble is not made from matter , but from the gravitational field itself .
it separates two regions having a different effective cosmological constant and ultimately changes its value throughout the whole space , thus changing the asymptotics . in the canonical ensemble ,
whether or not the transition takes place can be decided by evaluating the euclidean action on two well - defined classical configurations . according to our free energy computation ,
the phase transition actually occurs , the preferable static configuration above @xmath0 being a black hole surrounded by a bubble . in lorentzian
signature we can analyze its subsequent dynamics . we see that the bubble eventually expand in such a way that it swallows up the whole spacetime in finite proper time , thus changing the asymptotics . the lovelock action can be written in terms of the vierbein 1-form , @xmath1 , and the curvature 2-form , @xmath2 , @xmath3 being the ( torsion - free ) spin connection , @xmath4 with the _ bulk _ contributions , @xmath5 where @xmath6 , @xmath7 being the antisymmetric symbol , while @xmath8 is a positive integer , @xmath9 $ ] , and @xmath10 refers to boundary terms to be discussed below . the coefficients @xmath11 are coupling constants with length dimensions @xmath12 .
the @xmath13 term in the lagrangian corresponds to the extension of the euler characteristic in @xmath14 dimensions .
the zero@xmath15 contribution is the cosmological constant term ( we set @xmath16 , _
i.e. _ , @xmath17 ) , the first term is the eh action ( we normalize the newton constant to @xmath18 , _
i.e. _ , @xmath19 ) , and the second term , quadratic in the riemann curvature , is the lgb action ( we take @xmath20 , and call @xmath21 the lgb _ coupling _ ) .
although lovelock theory yields second order equations of motion , they are non - linear in the curvature . as a result ,
the theory admits more than one maximally symmetric solution ; it has up to @xmath8 different ( a)ds vacua with effective cosmological constants @xmath22 ,
they are the solutions of the @xmath24 order polynomial @xcite @xmath25 \equiv \sum_{k=0}^{k}c_{k}\,\lambda^{k } = c_{k}\prod_{i=1}^{k}\left ( \lambda -\lambda _ { i}\right ) = 0 ~. \label{cc - algebraic}\ ] ] the theory exhibits degenerate behavior whenever two or more of these effective cosmological constants coincide .
we are interested in the non - degenerate case .
the first ingredient in our discussion is the black hole solution of the theory .
consider the ansatz @xmath26 where @xmath27 is the round metric on a @xmath28-dimensional sphere .
the equations of motion reduce to a single first order differential equation for @xmath29 , that can be easily solved in terms of @xmath30 as @xcite @xmath31=\sum_{k=0}^{k}c_{k}\,g^{k}=\frac{\kappa } { r^{d-1 } } ~ , \label{eqg}\ ] ] an implicit polynomial solution with up to @xmath8 branches where @xmath32 is an integration constant related to the mass of the black hole @xcite , @xmath33 ~ , \label{mass}\ ] ] @xmath34 being the location of the horizon .
we focus on the quadratic theory which is enough to illustrate the phenomenon reported in this article .
the solutions take the form @xcite @xmath35 each of the two branches in ( [ gbbranches ] ) is associated with a different value of the effective cosmological constant , @xmath36 notice that @xmath37 is needed in order to have real non - degenerate values . while @xmath38 has a well defined horizon
, @xmath39 displays a naked singularity at the origin provided @xmath40 .
this latter branch is also unstable @xcite ; the graviton propagator is proportional to @xmath41<0 $ ] , thus having the wrong sign with respect to the eh case . in the dual cft
this amounts to non - unitarity @xcite .
the stable solution @xmath42 also happens to be the one that is continuously connected to the schwarzschild - tangherlini solution of general relativity in the @xmath43 limit ; namely @xmath44 these features are also manifest for generic black holes in lovelock theory @xcite .
the existence of a second vacuum of effective cosmological constant @xmath45 in the higher - curvature theory was referred to , in @xcite , as the theory having `` its own cosmological constant problem '' . moreover , for small lgb coupling , the curvature of such vacuum becomes very large , @xmath46 .
one may thereby argue that considering such a background in the quadratic theory does not make sense because higher - curvature terms can not be neglected .
however , far from removing this second vacuum , adding higher - order terms further produce a plethora of highly - curved vacua . with the purpose of understanding the implications of considering one such vacuum in the theory
, we consider spaces that asymptote ads of typical radius @xmath47 .
let us now discuss the role of the boundary terms , @xmath10 .
their contribution is necessary for the variational principle to be well defined .
this is analogous to the gibbons - hawking term in general relativity @xcite , which in the first order formalism can be written as @xmath48 where @xmath49 is the second fundamental form associated to the extrinsic curvature .
similarly , the boundary term associated to the lgb contribution reads @xcite @xmath50 in the same way as bulk terms are the dimensional extension of the @xmath14-dimensional euler characteristic for closed manifolds , the corresponding boundary terms are needed in the extension of the gauss - bonnet theorem to manifolds with boundaries .
the boundary action is then given by @xmath51 .
we want to explore configurations consisting of a spherical _ bubble _ dividing the spacetime in two regions , the outer being taken to asymptote ads with effective cosmological constant @xmath52 .
solutions consisting of a spherically symmetric surface separating regions with different vacua are known to exist @xcite .
they have been recently considered to explore instanton transitions , @xmath53 , via bubble nucleation @xcite . in the present case
we are interested in thermodynamic phase transitions .
it is convenient to break the action in three pieces , @xmath54 the first term is integrated inside the bubble , while the last term is integrated outside and includes all the boundary terms at infinity necessary to both have a well defined variational principle and to regularize the infinities .
the term in the middle is integrated on a small region around the bubble .
we consider the limit when its width goes to zero .
proceeding in this way , we deal with terms at the boundaries of each region , which give rise to a finite @xmath55 in the thick - less limit @xcite .
the variation of @xmath55 with respect to the vierbein gives the junction conditions on the bubble @xcite .
these generalize the israel conditions of general relativity .
the configuration we will be concerned with is a _ bubble _ , whose outer region asymptotes ads with a cosmological constant @xmath52 , while the inner region hosts a black hole with mass @xmath56 , and an effective cosmological constant @xmath57 .
the opposite situation does not possess a smooth euclidean section . across the junction ,
the vierbein has to be continuous .
sticking to spherical symmetry , the two bulk regions are described by a solution of the form ( [ gbbranches ] ) with euclidean signature , @xmath58 where the @xmath59 signs denote the outer / inner regions .
the notation is consistent with the respective branches .
we start by considering arbitrary branch solutions with undetermined mass parameters on each region .
the junction conditions then fully constrain the allowed possibilities . the junction is conveniently described by the parametric equations @xmath60 with an induced metric of the form @xmath61 , which has to be the same from both sides .
this yields @xmath62 where the dot stands for derivatives with respect to the proper time @xmath63 .
the function @xmath64 appears explicitly in the induced metric , so the radius has to be continuous across the surface .
this condition allows us to write all the expressions in terms of @xmath65 and its derivatives and eventually find a dynamical equation for the bubble itself @xcite .
we are interested in static configurations which , in view of ( [ rtcond ] ) , translates into @xmath66 this means that the physical length of the euclidean time circle is the same as seen from both sides of the junction .
this matching condition will let us determine the temperature .
once the periodicity of the inner solution is fixed by demanding regularity at the black hole horizon , that of the outer solution gets fully determined , @xmath67 such that there is a _ unique _
free parameter , the temperature .
while @xmath68 is the inverse of the hawking temperature of the inner black hole solution , @xmath69 is the inverse of the temperature measured by an observer at infinity . by resorting to the usual euclidean time formalism
, we can compute the free energy associated to the bubble configuration that hosts a black hole , and then compare it with that of thermal ads at the same temperature . as we show in what follows , the computation indicates that the phase transition occurs above a critical temperature , @xmath70 .
the canonical ensemble at temperature @xmath71 is defined by the path integral over all metrics which asymptote ads identified in euclidean time with period @xmath72 , @xmath73 } } ~ , \label{path}\ ] ] where @xmath74 .
the dominant contributions come from the saddle points .
we have then to evaluate the euclidean action on a classical solution , @xmath75 which therefore gives the free energy , @xmath76 . in the present case ,
this basically amounts to computing the difference between the euclidean action of the bubble configuration and that of ads space identified with the same period in imaginary time .
the euclidean action is in general divergent due to the infinite volume of ads ; nevertheless , it can be suitably regularized by background subtraction , meaning that the free energy is actually measured with respect to the maximally symmetric solution .
the periodicity at infinity is fixed by demanding regularity of the black hole solution at the horizon , @xmath77 , supplemented by the gluing condition ( [ tglue ] ) , that determines in turn the outer periodicity . in order to simplify the discussion
, we calculate the on - shell action in terms of two parameters , the position of the bubble , @xmath65 , and the temperature , even though they are not independent from each other . unlike the computation of the hp effect in general relativity
, here we have to consider the contribution @xmath78 arising on the bubble , when writing the euclidean action in the form ( [ pieces ] ) . the boundary term in @xmath79 regularizes its divergence by subtracting the background @xmath80 with the same periodicity at infinity .
the term @xmath81 , in turn , is integrated from the horizon to the location of the bubble .
finally , @xmath78 is given by @xmath82 where the periodicity in euclidean time is inherited from the bulk regions , @xmath83 .
we can collect all contributions depending upon the location of the bubble , @xmath84 , the rest being consequently called @xmath85 . remarkably enough , a neat result comes out after a quite lengthy calculation that nicely carries on to the generic lovelock theory @xcite , once the junction conditions are imposed , @xmath86 which is the exact value needed to correct the on - shell bulk action such that the thermodynamic interpretation is safely preserved .
in fact , due to this contribution , the total action takes the form @xmath87 that is , the bubble contributes as mass carrying the mass difference between the two solutions but does not contribute to the entropy . from the hamiltonian point of view
this is naturally understood as follows .
the canonical action vanishes in this case , the only possible contributions coming from boundary terms both at infinity and at the horizon , yielding respectively @xmath88 and the entropy , which are nothing but the total charges of the solution .
the junction conditions simply imply the continuity of canonical momenta @xcite .
equation ( [ treintaytres ] ) shows that the junction conditions are important to guarantee the consistency of the thermodynamic picture .
they also imply @xmath89 so that the first law of thermodynamics holds both for the whole configuration ( @xmath90 and @xmath91 ) as well as for the black hole ( @xmath92 and @xmath56 ) .
having proven the consistency of the thermodynamic picture in the case of the bubble configuration by deriving ( [ treintaytres ] ) and ( [ new1law ] ) , we are ready to address the question of global thermodynamic stability .
this amounts to analyzing the free energy associated to the bubble configuration .
the free energy , @xmath76 , as a function of the temperature @xmath93 displays a critical temperature above which it becomes negative and , thus , the phase transition occurs ( see fig.[ceggfig2 ] ) .
( positivity bound ) , @xmath94 ( maximal @xmath95 ) , and @xmath96 ( from right to left ) .
the @xmath21 dependence of the critical temperature is displayed in a separate box.,scaledwidth=39.0% ] if the free energy is positive , however , the system is metastable .
it decays by nucleating bubbles with a probability given , in the semiclassical approximation by @xmath97 .
therefore , after enough time , the system will alway end up in the stable , black hole solution .
this is reminiscent of the hp transition , except for the fact that , here , the thermal ads vacuum decays into a black hole _ belonging to a different branch_. @xmath70 is monotonically decreasing , the phase transition becoming increasingly unlikely the more we come closer to the eh _ classical _ limit . in this sense , it is a quantum mechanical phenomenon . these high temperature configurations , or _
thermalons _ , correspond to a black hole and a bubble in equilibrium , connecting inner and outer solutions of different branches @xmath42 and @xmath98 .
such solutions exist just for positive values of @xmath21 .
the infinite temperature limit corresponds to planar black holes , in which the junction conditions lead to a remarkably simple relation that is valid also for general lovelock gravity @xcite , @xmath99 .
the corresponding free energy is always negative , @xmath100 this implying that the preferable classical solution is always the thermalon and the transition always occurs for high enough temperature .
the junction conditions considered above determine not only the equilibrium configuration but also , in lorentzian signature , the effective potential felt by the bubble and , consequently , its subsequent dynamics .
the scalar field @xmath64 specifying the location of the bubble sits at the top of a potential barrier .
it will therefore either collapse or eventually expand in such a way that it engulfs the whole spacetime in finite proper time , thus changing its asymptotic behavior @xcite .
we presented a novel mechanism for phase transitions that is a distinctive feature of higher curvature theories of gravity .
these theories have several branches of asymptotically ( a)ds solutions that might admit an interpretation as different phases of the dual field theory .
phase transitions among these are driven by the mechanism described in the present article . mimicking the thermalon configuration @xcite
, a bubble separating two regions of different cosmological constants pops out , generically hosting a black hole .
this configuration is thermodynamically preferred above some critical temperature .
the corresponding phase transition can be interpreted as a generalized hp transition for the high - curvature branches , driving the system towards the eh branch . from the holographic point of view
this looks like a confinement - deconfinement phase transition in a dual cft , involving an effective change in the t hooft coupling , both phases being strongly coupled . whether a phenomenon like this takes place in a 4d cft , particularly within the framework of the fluid / gravity correspondence where both phases might be characterized by different transport coefficients , or it is overtaken by higher curvature corrections , is an open question at this point . the bubble configuration , being unstable , dynamically changes the asymptotic cosmological constant , transitioning towards the stable horizonful branch of solutions , the only one usually considered as relevant .
this is then a natural mechanism for the system to select the general relativistic vacuum among all possible ones .
we are aware of the fact that the vacuum @xmath45 in the lgb theory exhibits ghosts .
the phenomenon presented in this article , however , takes place in the lovelock theory as well , where there are further healthy vacua than the one connected to the eh action @xcite .
we wish to thank stanley deser , juan maldacena , carlos nez and josep m. pons for their interesting comments .
thanks ceci garraffo , elias gravanis , and steve willison for previous collaborations in the subject . the work of x.o.c . and j.d.e .
was supported in part by micinn and feder ( grant fpa2011 - 22594 ) , by xunta de galicia ( consellera de educacin and grant pgidit10pxib206075pr ) , and by the spanish consolider - ingenio 2010 programme cpan ( csd2007 - 00042 ) .
the work of g.g . was supported by nsf - conicet , pip , and pict grants from conicet and anpcyt .
the work of a.g . was partially supported by fondecyt ( chile ) grant @xmath1011090753 .
x.o.c . and j.d.e .
would like to thank the fcen - uba and unab for hospitality during part of this project .
is thankful to the front of galician - speaking scientists for encouragement . c. lanczos , annals math . *
39 * , 842 ( 1938 ) .
b. zwiebach , phys .
b * 156 * , 315 ( 1985 ) .
e. s. fradkin and a. a. tseytlin , nucl .
b * 261 * , 1 ( 1985 ) .
d. j. gross and e. witten , nucl .
b * 277 * , 1 ( 1986 ) .
d. j. gross and j. h. sloan , nucl .
b * 291 * , 41 ( 1987 ) .
s. ferrara , r. r. khuri and r. minasian , phys .
b * 375 * , 81 ( 1996 ) .
i. antoniadis , s. ferrara , r. minasian and k. narain , nucl .
b * 507 * , 571 ( 1997 ) .
m. brigante , h. liu , r. c. myers , s. shenker and s. yaida , phys .
d * 77 * , 126006 ( 2008 ) .
m. brigante , h. liu , r.c .
myers , s. shenker , and s. yaida , phys .
* 100 * ( 2008 ) 191601 .
a. buchel and r. c. myers , jhep * 0908 * , 016 ( 2009 ) .
d. m. hofman , nucl .
b * 823 * , 174 ( 2009 ) .
j. de boer , m. kulaxizi and a. parnachev , jhep * 1006 * , 008 ( 2010 ) .
x. o. camanho and j. d. edelstein , jhep * 1006 * , 099 ( 2010 ) .
m. f. paulos , jhep * 1105 * , 043 ( 2011 ) .
x. o. camanho , j. d. edelstein and m. f. paulos , jhep * 1105 * , 127 ( 2011 )
. j. de boer , m. kulaxizi and a. parnachev , jhep * 1107 * , 109 ( 2011 ) .
a. buchel , r. c. myers and a. sinha , jhep * 0903 * , 084 ( 2009 ) .
d. lovelock , j. math . phys .
* 12 * , 498 ( 1971 ) .
j. zanelli , class .
* 29 * , 133001 ( 2012 ) .
s. w. hawking and d. n. page , commun .
phys . * 87 * , 577 ( 1983 ) .
e. witten , adv .
* 2 * , 505 ( 1998 ) .
r. g. cai , phys .
d * 65 * , 084014 ( 2002 ) .
x. o. camanho _ et al_. , to appear .
a. gomberoff , m. henneaux , c. teitelboim and f. wilczek , phys . rev .
d * 69 * , 083520 ( 2004 ) .
d. boulware and s. deser , phys .
lett . *
55 * , 2656 ( 1985 ) .
j. t. wheeler , nucl .
b * 268 * , 737 ( 1986 ) .
j. t. wheeler , nucl .
b * 273 * , 732 ( 1986 ) .
d. kastor , s. ray and j. traschen , class .
* 27 * , 235014 ( 2010 ) .
x. o. camanho and j. d. edelstein , arxiv:1103.3669 .
g. w. gibbons and s. w. hawking , phys .
d * 15 * , 2752 ( 1977 ) .
r. c. myers , phys .
d * 36 * , 392 ( 1987 ) .
e. gravanis and s. willison , phys .
d * 75 * , 084025 ( 2007 ) . c. garraffo , g. giribet , e. gravanis and s. willison , j. math .
phys . * 49 * , 042502 ( 2008 ) . c. charmousis and a. padilla , jhep
* 0812 * , 038 ( 2008 )
. s. c. davis , phys .
d * 67 * , 024030 ( 2003 ) .
e. gravanis and s. willison , phys .
b * 562 * , 118 ( 2003 ) .
m. baados , c. teitelboim and j. zanelli , phys .
lett . * 72 * , 957 ( 1994 ) . | we set forth a new type of phase transition that might take place in gravitational theories whenever higher - curvature corrections are considered .
it can be regarded as a sophisticated version of the hawking - page transition , mediated by the nucleation of a bubble in anti - de sitter ( ads ) space .
the bubble hosts a black hole in its interior , and separates two spacetime regions with different effective cosmological constants .
we compute the free energy of this configuration and compare it with that of thermal ads .
the result suggests that a phase transition actually occurs above certain critical temperature , ultimately changing the value of the cosmological constant .
we discuss the consistency of the thermodynamic picture and its possible relevance in the context of ads / cft . |
an anticipated highlight of diffraction at lhc is the study of new particles in central exclusive production , @xmath1 .
detailed investigations have been made for the case where @xmath2 is a higgs boson in the standard model or its supersymmetric extension ( see e.g.@xcite ) , but other systems with a strong coupling to two gluons , like a gluino pair @xmath3 @xcite , can be equally interesting . provided that rates are sufficiently high
, central exclusive production enables us to study the system @xmath2 in a clean environment , with a signal - to - background ratio often much larger than in conventional , inclusive production channels .
measurement of the outgoing proton momenta in forward detectors gives the possibility of a precise determination of the mass and possibly the width of @xmath2 , and the exclusive production mechanism strongly favours systems @xmath2 with quantum numbers @xmath4 .
if the effective two - gluon luminosity for @xmath5 can be determined from standard - model channels such as @xmath6 or @xmath7 , the cross section measurement for a new particle @xmath2 decaying into a final state @xmath8 yields the combination @xmath9 of widths .
more detail is given in several presentations at this meeting @xcite .
r0.5 hera data provides essential non - perturbative input needed to calculate central exclusive production at lhc .
this is important not only to estimate event rates before lhc measurements start , but also to control backgrounds and to help optimise triggers and selection cuts when lhc data will be available .
an example are events where the central system @xmath2 contains not only a higgs but also relatively soft gluons .
diffractive parton densities from hera are crucial to estimate the importance of this channel , which would spoil the accuracy of a higgs mass determination from the scattered protons alone @xcite .
the theory description of central exclusive production involves a number of nontrivial issues , some of which were discussed at this meeting @xcite . the simplest leading - order graph for the production mechanism
is shown in fig .
[ fig : exc - higgs ] .
there is no all - order factorisation theorem for this process , and indeed factorisation is broken by rescattering between spectator partons .
radiative corrections to the leading - order graph in fig .
[ fig : exc - higgs ] are known to be important and can in part be grouped into sudakov factors .
a full next - to - leading order calculation is , however , not available , and it has recently been argued that there may be important higher - order corrections which have not been evaluated so far @xcite . the non - perturbative input to the graph in fig .
[ fig : exc - higgs ] is given by the generalised gluon distribution .
our present knowledge of this distribution derives essentially from hera data on vector meson production and on deeply virtual compton scattering @xcite . from a theory point of view , @xmath10 photoproduction is a particularly clean channel for this purpose , but the rate at hera is too limited for detailed measurements . along with @xmath11 photoproduction
this channel could , however , be further studied in @xmath12 and @xmath13 collisions at lhc @xcite .
it is important to note that the calculation of central exclusive production involves generalised distributions @xmath14 that depend explicitly on the transverse momentum @xmath15 of the emitted gluon . put in a simplified way , the graph in fig .
[ fig : exc - higgs ] involves an integral @xmath16 where the distributions in the two colliding protons are entangled .
this can not readily be reduced to the @xmath15 integrated , collinear distributions @xmath17 that appear in @xmath18 scattering processes . while the understanding and evaluation of higher - order corrections is fairly advanced for collinear generalised parton distributions , the same is unfortunately not the case for their @xmath15 dependent counterparts . as already mentioned , the simple mechanism shown in fig .
[ fig : exc - higgs ] receives important corrections from the rescattering of spectator partons in the two colliding protons .
most calculations assume that the dominant rescattering effects can be described by elastic or quasi - elastic proton - proton interactions .
this leads to a simple representation @xmath19 for the physical cross section , where the convolution is in transverse - momentum space , @xmath20 describes the hard - scattering mechanism sketched in fig . [
fig : exc - higgs ] , and the rapidity gap survival factor @xmath21 can be inferred from @xmath12 and @xmath22 scattering data .
more complicated rescattering mechanisms have , however , been studied @xcite .
the underlying physics is related to multiple scattering in inclusive @xmath12 collisions , which by itself is of high importance for understanding final states at lhc .
the description of such rescattering effects involves considerable difficulties , and a reliable evaluation of their importance has not been achieved as yet .
it is all the more important to test the phenomenological models currently used for the description of central exclusive production .
fortunately , this is already possible by using data from the tevatron , either for inclusive diffractive channels such as @xmath23 or for exclusive reactions like @xmath24 and @xmath25 @xcite .
possible tests using early data from lhc are discussed in @xcite .
hera provides crucial input to these tests in the form of diffractive parton densities and of the generalised gluon distribution , which are needed to calculate the hard part @xmath20 of the cross section .
the importance of rescattering can also be probed in diffractive photoproduction at hera , given the hadronic component of a real photon .
this has turned out to be more complicated than initially thought , both from the experimental and the theoretical sides @xcite .
it is probably too early to draw final conclusions here , but ultimately these studies might teach us more about the double nature of the photon as a pointlike particle and a hadron than about rescattering dynamics in diffraction .
parton saturation , caused by nonlinear dynamics that sets in when parton densities become very large , has become a central topic in high - energy qcd .
it allows us to study a field theory in a strongly coupled regime , but with a small coupling constant , which makes it possible to use perturbative methods . beyond this intrinsic interest
, parton saturation entails the breakdown of a description based on collinear factorisation and dglap evolution . to quantify nonlinear effects at small @xmath26 and moderate @xmath27 at hera is hence relevant for assessing the limits of precision in dglap based extractions of parton densities . for generic reasons ,
one expects such nonlinear , higher - twist corrections to be stronger in @xmath28 than in @xmath29 .
quantitative estimates based on the colour dipole model have been given several years ago @xcite , and work is underway to update these estimates taking into account the progress in dipole phenomenology @xcite .
a convenient framework to describe parton saturation in @xmath18 collisions is the colour dipole formalism .
it permits the calculation of many inclusive and diffractive processes from the inclusive structure functions @xmath30 , @xmath31 , @xmath32 and their diffractive counterparts to exclusive vector meson production and deeply virtual compton scattering with the _ same _ non - perturbative input , namely the scattering amplitude for a dipole on a proton target .
the associated phenomenology is very successful , as presented in @xcite at this meeting . on the theoretical side ,
this formalism is , however , still largely restricted to leading order in @xmath33 .
the fluctuation of a @xmath34 into @xmath35 is readily taken into account , but the next highest fock state @xmath36 has so far only been taken in certain approximations , which limits a reliable description of inclusive diffraction .
the evolution of the dipole scattering amplitude with energy is described by the bfkl or the balitsky - kovchegov equations , whose leading - order solutions can not account for the energy dependence seen in experiment . in practice one
therefore typically takes a functional form of the dipole scattering amplitude motivated by theory , but fits the relevant parameters to data . note
that this is different from dglap type fits in collinear factorisation , where the relevant parton distributions are parametrised at a starting scale but evolved using the perturbative evolution equations .
it remains an outstanding task to formulate bfkl and saturation dynamics in a dipole framework at nlo , in a manner that would allow one to pursue phenomenological analyses .
to which extent nonlinear dynamics is seen in hera data remains rather controversial .
current saturation models find that the virtualities where saturation effects become important are below a gev at hera energies , which limits the possibilities of controlled perturbative calculations .
a prominent experimental observation is the very flat energy dependence for the ratio @xmath37 of diffractive and total structure functions at given @xmath27 .
this is explained in a natural way by the saturation mechanism @xcite .
it must , however , be noted that many dipole models , including versions without saturation , provide a good description of @xmath38 and @xmath30 in a wide kinematic range @xcite .
to better assess the situation , it may be helpful to compare models with data specifically for the structure function ratio .
a striking feature observed in hera small-@xmath0 data is geometric scaling , i.e. , the dependence of the total @xmath39 cross section ( and also of diffractive cross sections ) on a single scaling variable @xmath40 @xcite .
this is sometimes presented as evidence for saturation dynamics .
however , several investigations have shown that , both analytically and numerically , one finds approximate geometric scaling also from the dglap and the bfkl equations @xcite . to infer from geometric scaling on the underlying dynamics
, one may have to focus on the _ deviations _ from exact scaling , which should differ among the various dynamical mechanisms .
how well this can be quantified theoretically , and whether the precision and kinematic lever arm of the hera data are sufficient for such a study , remains to be seen .
there are prospects to pursue the hera studies of saturation effects for the much smaller parton momentum fractions achievable at lhc .
this will require detection of particles at very forward rapidities @xcite .
forward drell - yan pair production is of particular interest from a theory point of view . on the one hand ,
it permits a description in the dipole picture @xcite and can thus be closely related with the studies performed at hera . on the other hand
this process is very well understood in collinear factorisation , with full next - to - next - to - leading order results @xcite allowing for precise `` non - saturated '' calculations .
it is a pleasure to acknowledge discussions with d.yu .
ivanov and l. motyka , and to thank r. devenish , m. wing , and their co - organisers for hosting a great conference .
s. heinemeyer _ et al .
_ , eur .
j. c * 53 * ( 2008 ) 231 [ arxiv:0708.3052 [ hep - ph ] ] ; + j. r. ellis , j. s. lee and a. pilaftsis , phys . rev .
d * 71 * ( 2005 ) 075007 [ hep - ph/0502251 ] .
p. j. bussey _ et al .
_ , jhep * 0611 * ( 2006 ) 027 [ hep - ph/0607264 ] .
talks at this conference by b. cox , j. forshaw and m. taevski .
m. klasen and g. kramer , arxiv:0806.2269 [ hep - ph ] .
j. bartels , k. j. golec - biernat and k. peters , eur .
j. c * 17 * ( 2000 ) 121 [ hep - ph/0003042 ] .
l. motyka , talk at the workshop on hera and the lhc , cern , 2630 may 2008 , + ` http://indico.cern.ch/contributiondisplay.py?contribid=101&sessionid=6&confid=27458 ` j. kwieciski and a. m. stato , phys . rev .
d * 66 * ( 2002 ) 014013 [ hep - ph/0203030 ] ; + e. iancu , k. itakura and l. mclerran , nucl .
a * 708 * ( 2002 ) 327 [ hep - ph/0203137 ] ; + e. avsar and g. gustafson , jhep * 0704 * ( 2007 ) 067 [ hep - ph/0702087 ] ; + f. caola and s. forte , arxiv:0802.1878 [ hep - ph ] . s. j. brodsky , a. hebecker and e. quack , phys . rev .
d * 55 * ( 1997 ) 2584 [ hep - ph/9609384 ] ; + b. z. kopeliovich , a. v. tarasov and a. schfer , phys .
c * 59 * ( 1999 ) 1609 [ hep - ph/9808378 ] ; + f. gelis and j. jalilian - marian , phys .
d * 67 * ( 2003 ) 074019 [ hep - ph/0211363 ] .
see e.g. s. moch , j. phys .
g * 35 * ( 2008 ) 073001 [ 0803.0457 [ hep - ph ] ] . | i discuss and connect a number of topics in small-@xmath0 physics at hera and at lhc , pointing out recent progress and open questions in theory and phenomenology . |
non - zero neutrino masses and their flavor mixings require physics beyond the standard model ( sm ) . one of the attractive mechanisms for generating neutrino masses and mixings is the so - called radiative seesaw in which the smallness of neutrino mass is explained by the suppression from the loop factor . in this class of radiative neutrino mass models ,
dark matter ( dm ) candidate often appears naturally if we assign dark @xmath7 parity to stabilize the dm candidates @xcite .
the predictive neutrino mass model can be achieved by applying some symmetry which distinguishes fermion flavor .
flavor dependent @xmath8 gauge symmetry is one of the interesting candidates which is discussed in the case of tree level neutrino mass generation @xcite .
furthermore , flavor dependent @xmath8 gauge symmetries including the quark sector have been motivated in order to explain various anomalies gauge theories with additional higgs doublets carrying nonzero @xmath9 charges were discussed in refs .
@xcite in order to accommodate the top forward - backward asymmetry ( fba ) at the tevatron .
since this anomaly has been less significant now , we do not consider this case further .
but the model building issues addressed in refs .
@xcite still remain valid and relevant in other flavor dependent @xmath9 models for @xmath10 physics anomalies . ] in @xmath11 @xcite ; @xmath12 anomaly in lepton - universality in the ratio @xmath13 by the lhcb @xcite , and sizable deviation measured in angular distributions of @xmath14 @xcite .
these anomalies can be accounted for a shift of the wilson coefficient @xmath15 of the semileptonic operator @xmath16 @xcite which can be induced by flavor dependent @xmath6 interaction in down quark sector . in this paper
, we propose a radiative seesaw model based on flavor dependent and anomaly free @xmath17 gauge symmetry and extra discrete @xmath7 symmetry to ensure dm stability .
the active neutrino mass matrix is induced at one loop level where @xmath7 odd particles propagate inside the loop including the dm candidate which is the lightest @xmath7 odd sm singlet majorana fermion with nonzero @xmath9 charge
. then structure of the mass matrix for the majorana fermion is restricted and determined by the flavor dependent @xmath9 charge assignments .
we also study phenomenology associated with @xmath6 boson , such as collider constraints , signatures at the lhc and the wilson coefficients contributing to @xmath11 obtained from flavor dependent @xmath6 interaction .
then we show predictions in the neutrino mass matrix by carrying out numerical analysis taking into account constraints from lepton flavor violation , thermal relic density of dm and various constraints on @xmath6 interaction . in sec .
ii , we introduce our model lagrangian and discuss particle properties and their interactions . in sec.iii
we discuss phenomenology including neutrino mass matrix , charged lepton flavor violation , relic density of dm , and some processes related to @xmath6 gauge boson including the lhcb anomalies .
the numerical analysis is carried out in sec .
iv to find out the parameter region satisfying experimental constraints and to obtain some prediction for neutrino physics .
finally we summarize the results in sec .
in this section , we introduce our model and discuss some properties for our analysis in the following sections . in the fermion sector , we introduce @xmath18 singlet majorana fermions @xmath19 , and impose a flavor dependent gauge symmetry @xmath20 as summarized in table [ tab:1 ] , where @xmath21 is an arbitrary number , which originates from the structure of the right - handed neutrino mass matrix ( see eq .
( ii.6 ) below ) .
it follows from the fact that one can not distinguish @xmath22 from @xmath23
. then we would lose predictability on the neutrino sector .
therefore we shall choose @xmath24 in this paper and keep predictability on the neutrino sector . ] .
this combination of @xmath9 is known as anomaly free of the gauge symmetry @xcite .
note here that we ignore the kinetic mixing between @xmath9 and @xmath25 assuming it is negligibly small .
in addition , @xmath7-odd parity is assumed for the new fermion @xmath26 s in order to forbid the tree level neutrino masses or(and ) to assure the stability of dark matter ( dm ) .
.field contents of fermions and their charge assignments under @xmath27 , where @xmath20 ( @xmath28 ) , and each of the flavor index is defined as @xmath29 and @xmath30 . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] in the bosonic sector , we introduce an @xmath18 doublet inert boson @xmath31 , new higgs doublet @xmath32 with extra @xmath8 charge , and three @xmath18 singlet bosons @xmath33 , where the lower indices represent their charges of @xmath9 as summarized in table [ tab:2 ] .
we assume that two higgs doublets @xmath34 and @xmath35 singlet fields @xmath36 respectively break electroweak and @xmath9 gauge symmetries spontaneously due their nonzero vacuum expectation values ( vevs ) , which are denoted by @xmath37 , @xmath38 , @xmath39 , @xmath40 and @xmath41 .
we impose @xmath7 odd quantum number for the inert boson @xmath31 due to the same reason as the fermion sector .
the new higgs doublet @xmath32 is introduced in order to induce quark mass matrix element which mix the 3rd generation with first and second generations .
the @xmath42 will allow a dim-3 operator , @xmath43 , with which we can avoid a massless goldstone boson .
note also that scalar potential of @xmath44 and @xmath45 has global symmetries which would induce a massless goldstone boson since the potential is given by @xmath46 and @xmath47 due to the @xmath9 symmetry .
such global symmetries can be avoided by introducing @xmath9-charged scalar ; for example @xmath48 with @xmath9 charge @xmath49 provides a term @xmath50 which violate dangerous global symmetries . in this paper , we assume all scalar bosons have non - zero masses and we abbreviate the complete analysis of the scalar potential .
0_anomaly cancellation _ : the @xmath51 gauge symmetry is anomaly free where the anomaly is canceled within each generation of fermions @xcite .
we then assign @xmath51 charges to first and second generation fermions but charges for third generation fermions are required to be zero .
the triangle anomaly within one generation cancels as follows : @xmath52 ^ 2u(1)_r:\quad 3\left(\frac49x+\frac19(-x)\right)-x=0,\\ & [ u(1)_r]^2u(1)_y:\quad 3\left(\frac23x^2-\frac13 x^2\right)-x^2=0,\\ & [ u(1)_r]^3:\quad 3\left(x^3-x^3\right)-x^3+x^3=0,\\ & u(1)_r:\quad 3\left(x - x\right)-x+x=0.\end{aligned}\ ] ] _ yukawa interactions _ :
under these fields and symmetries , the renormalizable lagrangians for quark and lepton sector are given by @xmath53 where @xmath54 , and @xmath55 is the second pauli matrix .
after two higgs doublet develops nonzero vevs , we obtain the quark mass matrix such that @xmath56 which is the same structure as discussed in ref . @xcite . as in the sm ,
the quark mass matrices are diagonalized by unitary matrices @xmath57 and @xmath58 which change quark fields from interaction basis to mass basis : @xmath59
. then the ckm matrix is given by @xmath60 .
the ckm matrix can be approximated as @xmath61 since @xmath62 correspond to right - handed rotations , and @xmath63 is obtained @xcite .
the details of quark yukawa couplings with two higgs doublet are discussed in ref .
@xcite , and we omit the further discussion here . _
@xmath6 couplings to sm fermions _ : after the aforementioned fields rotations into the mass basis , the @xmath6 couplings to the sm fermions are written as @xmath64 where @xmath65 is the gauge coupling constant associated with the @xmath9 .
the coupling matrices @xmath66 and @xmath67 for down - type quarks are given approximately by @xmath68 where @xmath69s are the elements of ckm matrix and we applied the relation @xmath70 as we discussed above .
+ _ exotic majorana fermion mass matrix _ is defined in the basis @xmath71^t$ ] as follows : @xmath72,\label{eq : mn}\end{aligned}\ ] ] where we simply assume to be positive real for each component and define @xmath73 , @xmath74 , and @xmath75
. then @xmath76 is diagonalized by orthogonal mixing matrix @xmath77 as @xmath78,\quad n_{r_{e,\mu,\tau}}=v n_{r_{1,2,3}},\label{eq : n - mix}\end{aligned}\ ] ] where @xmath79 is the mass eigenstate .
_ the active neutrino mass matrix _ is then given at one - loop level by @xcite @xmath80 - \frac{m_i^2}{m_r^2- d_{n_k}^2 } \ln\left[\frac{m_i^2}{d_{n_k}^2}\right ] \right){\nonumber}\\ & \approx \frac1{8\pi^2 } \frac{\lambda_5 v^2}{m_r^2+m_i^2 } \sum_{k=1}^3 y_{n_i } ( v_{ik } d_{n_k}v_{kj}^t ) y_{n_j } = \frac1{8\pi^2 } \frac{\lambda_5 v^2}{m_r^2+m_i^2 } y_{n_i } ( m_n)_{ij } y_{n_j } , \label{eq : neut - mass}\end{aligned}\ ] ] where @xmath81 is the quartic coupling of @xmath82 , @xmath83 is the mass eigenstate of real(imaginary ) part of neutral component of @xmath31 , and we have used eq.([eq : n - mix ] ) in the last equation .
here we assume to be @xmath84 , which could be natural if we consider the fermion dm case .
since @xmath85 is diagonal , the form of active neutrino mass matrix is proportional to the one of @xmath76 in eq.([eq : mn ] ) , therefore we have some predictions of type @xmath1 through the texture analysis @xcite
. then @xmath76 can be rewritten in terms of pmns matrix @xmath86 and mass eigenvalues of active neutrino @xmath87 by @xmath88 , where we define @xmath89 .
combining eq .
( [ eq : neut - mass ] ) , @xmath90 can be rewritten in terms of neutrino observables and some input parameters such as @xmath85 by @xmath91 where @xmath92 . in the section of numerical analysis
, we will show some predictions combined with the other phenomenologies such as lfvs and dm , adapting the recent global data @xcite up to 3@xmath93 confidential level .
+ _ lepton flavor violations(lfvs ) _ are induced from the term @xmath85 at one - loop level , and its branching ratio is given by @xmath94 where @xmath95 is the singly charged component of @xmath31 , @xmath96[gev]@xmath97 is the fermi constant , @xmath98 is the fine structure constant , @xmath99 , @xmath100 , and @xmath101 .
experimental upper bounds are respectively given by @xmath102 @xcite , @xmath103 , and @xmath104 @xcite .
+ _ muon anomalous magnetic moment ( muon g-2 : @xmath105 ) _ can be induced via @xmath85 with negative contribution , which is in conflict with the current experiment @xmath106
however another source via the additional @xmath6 gauge boson can also be induced by @xmath107 where @xmath108 , and @xmath6 is the new gauge vector boson .
thus we could explain the sizable muon @xmath109 if we can satisfy the constraint from the neutrino trident process : @xmath1100.4 gev with @xmath111 @xcite .
this can be realized by the limit @xmath112 .
however this is nothing but a typical gauged @xmath113 symmetry @xcite .
thus we discuss another region ( heavier mass region of @xmath6 ) apart from the muon @xmath109 . here
we consider the lightest majorana fermions @xmath114 as our dm , and assume @xmath115 to forbid the mode of @xmath116 for simplicity .. ] also we neglect mixings among neutral component of @xmath117 to simply suppress higgs portal interaction for avoiding the constraint from direct detection searches .
therefore the dominant contribution to dm annihilation in estimating the relic density arises from yukawa coupling .
then the relevant lagrangian in terms of mass eigenstates is given by @xmath118+{\rm c.c . } \\
& + \frac{g'}{2}(-x|v_{11}|^2-|v_{21}|^2+|v_{31}|^2)\bar x\gamma^\mu x z'_\mu+ { \cal l}_{z'ff},{\nonumber}\end{aligned}\ ] ] where the last term is given in eq .
( [ eq : int_z ] ) .
we have three relevant precesses to explain the relic density : @xmath119 , @xmath120 , and @xmath121 via the yukawa terms @xmath85 and the gauge interaction with @xmath6 involving @xmath65 , where we have @xmath122 channels only for @xmath123 , while @xmath124 channels for @xmath125 .
we apply the @xmath126 expansion approximation @xcite to estimate the relic density of dm , taking up to the @xmath127- and @xmath128-wave contributions in the annihilation amplitudes . then the formula for thermal relic density @xmath129
is approximately given by @xcite @xmath130},\end{aligned}\ ] ] where @xmath131[gev ] is the planck mass , @xmath132 is the total number of effective relativistic degrees of freedom at the time of freeze - out , and @xmath133 is defined by @xmath134 at the freeze out temperature ( @xmath135 ) , @xmath136 is the total contributions to the @xmath127-wave , and @xmath137 is the total contributions to the @xmath128-wave , respectively . the observed relic density reported by planck suggests that @xmath138 @xcite .
but in our numerical analysis below , we will use more relaxed value @xmath139 .
+ here we discuss phenomenology of @xmath6 boson such as the constraints on interactions , the contribution to @xmath140 , and the direct @xmath6 production cross section at the lhc .
_ lep constraint : _ the @xmath6 couplings to leptons induce the following effective interactions ; @xmath141 where @xmath142 , @xmath143 and @xmath144 in our charge assignments . in this case ,
the strongest constraint comes from the @xmath145 measurement at lep @xcite : @xmath146 we will impose this constraint in the following numerical analysis . _ the constraint from neutrino trident production _ :
the couplings of @xmath6 to the second generation of lepton is constrained by the neutrino trident process @xmath147 where @xmath148 denotes a nucleon .
taking into account the ccfr data , this constraint is roughly approximated as @xmath149 gev at the 95@xmath150 c.l . for a heavy @xmath6 boson case @xcite .
when we take @xmath151 , the mass of @xmath6 should satisfy @xmath152 gev . _
@xmath6 contribution to the @xmath153 decay _ : the anomalies in the angular observable @xmath154 associated with full angular distribution of @xmath155 ( with @xmath156 ) and in the lepton - universality violation @xmath157 can be accounted by the shift in the wilson coefficient @xmath158 , which is defined by @xmath159 effective hamiltonian as @xmath160 we have suppressed other operators for simplicity , since they do not play any important role regarding those two @xmath10 physics anomalies considered here as long as the wilson coefficients of those operators do not receive new physics contributions
. the global fit of the value for @xmath158 is obtained in ref .
@xcite based on lhcb data as follows ; @xmath161\;({\rm at } \
1\sigma ) , \qquad [ -0.32,-0.08 ] \ ; ( { \rm at } \ 2\sigma ) .
\label{eq : c9_fit}\ ] ] where @xmath162 indicates new physics contribution and @xmath163 at @xmath164 gev .
note that the sm contribution @xmath165 is lepton flavor universal , unlike to @xmath162 .
in the model proposed in sec .
ii , the flavor - dependent @xmath6 interaction shall induce the following effective hamiltonian : @xmath166 where @xmath167 and @xmath168 .
thus the shift of @xmath158 relative to its sm value would be given by @xmath169 therefore , applying the lep constraint eq .
( [ eq : lep ] ) , we find the range of @xmath162 such that @xmath170 the magnitude of @xmath171 is smaller than best fit value ( @xmath172 ) but it is within @xmath173 range as shown in eq .
( [ eq : c9_fit ] ) .
note that @xmath174 is suppressed by an extra factor of @xmath175 in our model .
thus it is possible to explain the anomaly in lepton - universality in @xmath176 : @xmath177 measured by lhcb , which shows a @xmath178 deviation from the sm prediction . here
the @xmath179 can be rewritten in terms of @xmath180 ( @xmath181 ) where @xmath182 in our model , and its allowed region is found to be @xcite ; @xmath183 \leq 1.5 $ ] , applying the @xmath179 data with @xmath184 errors .
this condition can be interpreted as @xmath185 where @xmath186 is used . therefore our value of @xmath158 in eq .
( [ eq : c9 mm ] ) can be accommodated with the range . _
@xmath6 production at the lhc _ : the @xmath9 gauge boson @xmath6 can be produced at the lhc since it couples to quarks . the dominant production process is given by @xmath187 where the couplings to other quarks are suppressed by small ckm matrix elements ( see eq .
( [ eq : ckm ] ) ) .
the @xmath6 mainly decays into @xmath188 and @xmath189 pairs with their branching ratios ( br s ) as @xmath190 for small @xmath191 where we assume masses of scalar bosons with couplings which is not suppressed by @xmath175 are heavier than @xmath192 .
thus the dimuon channel provides the most clear signature of @xmath6 . to estimate the production cross section for @xmath193
, we implement the relevant interactions into calchep @xcite and use the cteq6 parton distribution functions ( pdfs ) @xcite .
[ fig : zplhc ] shows the @xmath194 at @xmath195 tev as a function of @xmath196 where we have fixed @xmath197 and applied various values of @xmath175 .
the cross section is compared with the upper bound from the atlas experiments which is indicated as red curve @xcite .
we find that @xmath198 tev is allowed for @xmath199 and the constraint is weaker for smaller value of @xmath175 .
further parameter region can be tested by searching for the dimuon signature of @xmath6 at the lhc run 2 . as a function of @xmath196 at @xmath195 tev with various values of @xmath175 .
the red curve shows the upper limit from the atlas experiment .
, width=264 ]
in this section , we perform the numerical analysis and show some predictions . first of all , we select the range of input parameters as follows : @xmath200,\ y_{n_1 } \in[10^{-5},10^{-4}],\ y_{n_2 } \in[10^{-4},10^{-3}],\ y_{n_1 } \in[10^{-3},10^{-2}],\ \delta_{cp}\in[0,2\pi],{\nonumber}\\ & m_{\nu_3}\in [ 10^{-12},10^{-9}]\ [ { \rm gev } ] , \quad m_{r}\approx m_i \in [ 3000,5000]\ [ { \rm gev } ] , \quad m_{z'}\in [ 100,1500]\ [ { \rm gev}],\end{aligned}\ ] ] where @xmath201 is dirac phase in the neutrino sector , and we fix the new @xmath9 gauge coupling to be @xmath202 . due to the type @xmath1 texture of the neutrino mass matrix , obvious predictions are as follows , which are independent of the other phenomenologies as well as the above ranges of input parameters as already discussed in ref .
@xcite . 1 . _
the third neutrino mass is almost fixed to be @xmath203 [ gev ] . _ 2 .
_ only the normal ordering of the neutrino masses is allowed .
_ two majorana cp phases @xmath204 and @xmath93 correlate each other and behave as the red line in fig . [ fig : majo - p ] , where @xmath205 with @xmath206 in ref .
@xcite . and
@xmath207 is predicted .
_ neutrinoless double beta decay is predicted to be @xmath208(0.01 ) ev , which follows from the above two predictions .
_ here we have used the global neutrino oscillation data at 3@xmath93 confidential level @xcite .
notice here that @xmath201 is allowed in all the possible range , @xmath209 $ ] .
the other properties are shown in fig . [
fig : dm - x - zp ] that satisfies the neutrino oscillation data , lfvs , lep bound , and thermal relic density of dm where the allowed region of our dm mass is at @xmath210 [ gev ] , and the mass of @xmath83 is likely to be a free parameter in the upper left figure .
the correlation between @xmath211 and @xmath212 is shown in the right upper figure , in which the lower bound comes from the assumption @xmath213 , while we took into account the constraints from lep experiment in eq .
( [ eq : lep ] ) and from neutrino trident production .
in addition , the bottom figure shows the soft correlation between @xmath175 and @xmath212 , in which the upper bound also comes from the lep experiment . and @xmath93 which are majorana phases .
, width=264 ] and @xmath83,the upper - right figure shows the scattering plot between @xmath211 and @xmath212 , and the bottom figure shows the scattering plot between @xmath175 and @xmath212 .
the red dashed line indicates the constraint from neutrino trident production for @xmath151,title="fig:",width=264 ] and @xmath83,the upper - right figure shows the scattering plot between @xmath211 and @xmath212 , and the bottom figure shows the scattering plot between @xmath175 and @xmath212 .
the red dashed line indicates the constraint from neutrino trident production for @xmath151,title="fig:",width=264 ] and @xmath83,the upper - right figure shows the scattering plot between @xmath211 and @xmath212 , and the bottom figure shows the scattering plot between @xmath175 and @xmath212 .
the red dashed line indicates the constraint from neutrino trident production for @xmath151,title="fig:",width=264 ]
in this paper , we have proposed a predictive radiative seesaw model at one - loop level with a flavor dependent gauge symmetry @xmath0 , in which we have considered the majorana fermion dark matter .
we have obtained a two zero texture with @xmath1 type that provides us the normal ordered neutrino mass spectra with @xmath214 [ gev ] .
also specific patterns of two majorana phases are obtained in fig .
[ fig : majo - p ] .
the other properties are shown in fig . [ fig : dm - x - zp ] , and we have found the allowed region of our dm mass is at @xmath210 [ gev ] , and the mass of @xmath83 is likely to be the free parameter in the upper left figure . the correlation between @xmath211 and @xmath212 is shown in the right upper figure , in which the lower bound comes from the assumption @xmath213 , and the constraints from lep experiment and neutrino trident production are taken into account .
the bottom figure shows the soft correlation between @xmath175 and @xmath212 , in which the upper bound of @xmath175 also comes from the lep experiment for each value of @xmath212 .
we also discussed phenomenology of @xmath6 which has flavor dependent couplings to sm fermions .
the flavor violating interaction in the down quark sector can induce a sizable contribution to the wilson coefficient @xmath158 which can be within @xmath173 value obtained from global fitting by lhcb data .
although magnitude of our @xmath215 is less than the best fit value it can be an explanation of anomalies in the measurements of @xmath155 . remarkably we found anomaly in lepton - universality measurement @xmath216 can be explained within our model . in addition , we estimated cross section of the process @xmath193 at the lhc 13 tev which provides a clear signature of flavor - dependent @xmath6 in our model .
in particular , @xmath6 lighter than @xmath217 tev is allowed by current data and further parameter space can be tested in the future data of lhc experiments .
h. o. is sincerely grateful for all the kias members , korean cordial persons , foods , culture , weather , and all the other things .
this work is supported in part by national research foundation of korea ( nrf ) research grant nrf-2015r1a2a1a05001869 ( pk ) , and by the nrf grant funded by the korea government ( msip ) ( no .
2009 - 0083526 ) through korea neutrino research center at seoul national university ( pk ) .
e. ma , phys .
d * 73 * , 077301 ( 2006 ) [ hep - ph/0601225 ] .
a. crivellin , g. dambrosio and j. heeck , phys .
d * 91 * , no . 7 , 075006 ( 2015 ) [ arxiv:1503.03477 [ hep - ph ] ] . c. kownacki , e. ma , n. pollard and m. zakeri , arxiv:1611.05017 [ hep - ph ] .
p. ko , y. omura and c. yu , phys .
d * 85 * , 115010 ( 2012 ) [ arxiv:1108.0350 [ hep - ph ] ] .
ko , y. omura and c. yu , jhep * 1201 * , 147 ( 2012 ) [ arxiv:1108.4005 [ hep - ph ] ] .
y. omura and c. yu , eur .
j. c * 73 * , no . 1 , 2269 ( 2013 ) [ arxiv:1205.0407 [ hep - ph ] ] .
ko , y. omura and c. yu , jhep * 1303 * , 151 ( 2013 ) [ arxiv:1212.4607 [ hep - ph ] ] .
r. aaij _ et al . _
[ lhcb collaboration ] , phys .
lett . *
113 * , 151601 ( 2014 ) [ arxiv:1406.6482 [ hep - ex ] ] .
r. aaij _ et al . _
[ lhcb collaboration ] , phys .
lett . * 111 * , 191801 ( 2013 ) [ arxiv:1308.1707 [ hep - ex ] ] .
s. descotes - genon , l. hofer , j. matias and j. virto , jhep * 1606 * , 092 ( 2016 ) [ arxiv:1510.04239 [ hep - ph ] ] . t. hurth , f. mahmoudi and s. neshatpour , nucl .
b * 909 * , 737 ( 2016 ) [ arxiv:1603.00865 [ hep - ph ] ]
. h. fritzsch , z. z. xing and s. zhou , jhep * 1109 * , 083 ( 2011 ) [ arxiv:1108.4534 [ hep - ph ] ] .
d. v. forero , m. tortola and j. w. f. valle , phys .
d * 90 * , no .
9 , 093006 ( 2014 ) [ arxiv:1405.7540 [ hep - ph ] ] .
a. m. baldini _ et al .
_ [ meg collaboration ] , eur .
j. c * 76 * , no .
8 , 434 ( 2016 ) [ arxiv:1605.05081 [ hep - ex ] ] .
j. adam _ et al . _
[ meg collaboration ] , phys .
lett . * 110 * , 201801 ( 2013 ) [ arxiv:1303.0754 [ hep - ex ] ] .
w. altmannshofer , s. gori , m. pospelov and i. yavin , phys .
* 113 * , 091801 ( 2014 ) [ arxiv:1406.2332 [ hep - ph ] ] .
s. baek , h. okada and k. yagyu , jhep * 1504 * , 049 ( 2015 ) [ arxiv:1501.01530 [ hep - ph ] ] .
p. ko , t. nomura , h. okada and y. orikasa , phys .
d * 94 * , no . 1 , 013009 ( 2016 ) [ arxiv:1602.07214 [ hep - ph ] ] .
k. griest and d. seckel , phys .
d * 43 * , 3191 ( 1991 ) .
s. schael _ et al . _
[ aleph and delphi and l3 and opal and lep electroweak collaborations ] , phys . rept . * 532 * , 119 ( 2013 ) [ arxiv:1302.3415 [ hep - ex ] ] .
g. hiller and f. kruger , phys .
d * 69 * , 074020 ( 2004 ) [ hep - ph/0310219 ] .
g. hiller and m. schmaltz , phys .
d * 90 * , 054014 ( 2014 ) [ arxiv:1408.1627 [ hep - ph ] ] .
a. belyaev , n. d. christensen and a. pukhov , comput .
commun . * 184 * , 1729 ( 2013 ) [ arxiv:1207.6082 [ hep - ph ] ] .
m. nadolsky , h. l. lai , q. h. cao , j. huston , j. pumplin , d. stump , w. k. tung and c .- p .
yuan , phys .
d * 78 * , 013004 ( 2008 ) [ arxiv:0802.0007 [ hep - ph ] ] .
m. aaboud _ et al . _ [ atlas collaboration ] , phys .
b * 761 * , 372 ( 2016 ) [ arxiv:1607.03669 [ hep - ex ] ] . | we propose a predictive radiative seesaw model at one - loop level with a flavor dependent gauge symmetry @xmath0 and majorana fermion dark matter .
for the neutrino mass matrix , we obtain an @xmath1 type texture ( with two zeros ) that provides us several predictions such as the normal ordering for the neutrino masses .
we analyze the constraints from lepton flavor violations , relic density of dark matter , and collider physics for the new @xmath0 gauge boson . within the allowed region ,
the lhcb anomalies in @xmath2 and @xmath3 with @xmath4 or @xmath5 can be resolved , and such @xmath6 could be also observed at the lhc
. kias - p17007 |
negative index materials ( nims ) were first investigated theoretically by veselago in @xcite and were innovated by nicorovici et al . in @xcite and pendry in @xcite .
the existence of such materials was confirmed by shelby , smith , and schultz in @xcite .
the study of nims has attracted a lot the attention of the scientific community thanks to their many possible applications .
one of the appealing ones is cloaking .
there are at least three ways to do cloaking using nims . the first one is based on the concept of anomalous localized resonance discovered by milton and nicorovici in @xcite .
the second one is based on plasmonic structures introduced by alu and engheta in @xcite .
the last one makes use of the concept of complementary media and was suggested by lai et al . in @xcite . in this paper , we concentrate on the last method .
the study of cloaking using complementary media faces two difficulties .
firstly , this problem is unstable since the equations describing the phenomenon have sign changing coefficients , hence the ellipticity is lost .
secondly , the localized resonance , i.e. , the field explodes in some regions and remains bounded in some others as the loss goes to 0 , might appear , see ( * ? ? ?
* figure 2 ) . in this paper
, we give a proof of cloaking using complementary media for a class of schemes inspired by the work of lai et al . in @xcite in the quasistatic regime .
to handle the localized resonance , we introduce the technique of removing localized singularity and apply a three spheres inequality .
the proof also uses the reflecting technique in @xcite . to our knowledge
, this work presents the first proof on cloaking using complementary media .
let us describe how to cloak the region @xmath0 for some @xmath1 in which the medium is characterized by a matrix @xmath2 using complementary media .
here and in what follows given @xmath3 , @xmath4 denotes the ball in @xmath5 ( @xmath6 or 3 ) centered at the origin of radius @xmath7 .
the assumption on the cloaked region by all means imposes no restriction since any bounded set is a subset of such a region provided that the radius and the origin are appropriately chosen . the idea suggested by lai et al . in @xcite ( for two dimensions ) is to construct a complementary medium of @xmath2 in @xmath8 for some @xmath9 .
inspired by their idea , we construct a cloak in two and three dimensions as follows .
our cloak consists of two parts . the first one , in @xmath8 , makes use of complementary media to cancel the effect of the cloaked region .
the second one , in @xmath10 , is to fill the space which disappears " from the cancellation by the homogeneous media .
for the first part , we slightly change the strategy in @xcite . instead of @xmath0 , we consider @xmath11 for some @xmath12 as the cloaked region in which the medium is given by the matrix @xmath13 i & \mbox { in } b_{r_3 } \setminus
b_{2 r_2}. \end{array } \right.\ ] ] the complementary medium in @xmath8 is given by @xmath14 where @xmath15 is the kelvin transform with respect to @xmath16 , i.e. , @xmath17 here and in what follows we use the standard notation @xmath18 for a diffeomorphism @xmath19 .
it follows that @xmath20 concerning the second part , the medium in @xmath10 is given by @xmath21 the reason for this choice is condition ( mentioned later in the introduction ) .
note that in the two dimensional case , the medium in @xmath10 is @xmath22 , as used in @xcite , while it is not @xmath22 in the three dimensional case .
the cloaking scheme discussed here can be extended for a large class of reflections considered in @xcite .
the cloaking setting is illustrated in figure [ fig1 ] .
consists of two parts : @xmath2 in @xmath0 ( grey and green regions ) , which is the original object , and @xmath22 in @xmath23 .
the cloaking device contains two parts .
the first part @xmath24 ( the red , orange , and pink regions ) in @xmath8 is the complementary medium of @xmath25 : the red part is complementary to the grey part , the orange part is the complementary to the green part , and the pink part is the complementary to @xmath22-layer .
the second part @xmath26 ( the blue region ) is to fill the space which disappears by the cancelation.,width=302 ] to study the problem correctly , one should allow some loss in the medium and study the limit as the loss goes to 0 . with the loss ,
the medium is characterized by @xmath27 where @xmath28 f^{-1}_*b & \mbox { in } b_{r_2 } \setminus b_{r_1 } , \\[6pt ] \big(r_3 ^ 2/r_2 ^ 2 \big)^{d-2 } i & \mbox { in } b_{r_1},\\[6pt ] i & \mbox { otherwise } , \end{array } \right.\ ] ] and @xmath29 1 & \mbox { otherwise}. \end{array } \right.\ ] ] physically , the imaginary part of @xmath30 is the loss of the medium ( more precisely the loss of the medium of negative index in @xmath8 ) . here and in what follows , we assume that , @xmath31 for some @xmath32 . here and in
what follows @xmath33 denotes the euclidean scalar product in @xmath5 .
we require in addition that @xmath34 this paper deals with the bounded setting equipped the zero dirichlet boundary condition in the quasistatic regime .
let @xmath35 be a smooth open subset of @xmath5 ( @xmath36 ) such that @xmath37 .
given @xmath38 , let @xmath39 be respectively the unique solution to @xmath40 and @xmath41 here is the main result of this paper .
[ thm1 ] let @xmath36 , @xmath38 with @xmath42 and let @xmath43 be the unique solution to and respectively .
there exists @xmath44 , depending only on @xmath45 , @xmath46 , and the lipschitz constant of @xmath25 such that if @xmath47 then @xmath48 for an observer outside @xmath49 , the medium in @xmath49 looks like the homogeneous one by ( see also ): one has cloaking .
one can verify that medium @xmath50 is of reflecting complementary property , a concept introduced in ( * ? ? ?
* definition 1 ) , by considering diffeomorphism @xmath51 which is the kelvin transform with respect to @xmath52 , i.e. , @xmath53 it is important to note that @xmath54 since @xmath55 .
this is the reason behind the choice of @xmath56 in .
the proof of theorem [ thm1 ] is based on two crucial ingredients .
the first one is a three spheres inequality .
the second one is the technique of removing localized singularity introduced in this paper . defined in in the proof of theorem [ thm1 ] . ]
the proof of theorem [ thm1 ] is inspired by the approach in @xcite where the reflecting technique is used .
the removing singularity technique is used in @xcite to study superlensing and cloaking via anomalous localized resonance .
nims have been studied extensively recently see @xcite and references therein . in @xcite and @xcite , ammari et al . and kohn et al . studied the blow up of the power ( @xmath57 ) for a general core - shell structure with piecewise constant coefficients . without the core ,
it is shown in @xcite that there is no connection between the blow up of the power and the localized resonance in general . in the works
mentioned , the localized resonance was considered only in the cases where the settings are simple enough so that the separation of variables can be applied . the paper is organized as follows . in section [ sect - pre ] , we state and prove several useful lemmas which are used in the proof of theorem [ thm1 ] .
the proof of thereom [ thm1 ] is presented in section [ sect - proof ] .
in this section , we present some lemmas which will be used in the proof of theorem [ thm1 ] .
the first one is on a three spheres inequality for lipschitz matrix - valued functions .
[ three spheres inequality ] [ lem1.1 ] let @xmath58 , @xmath59 and let @xmath60 be a lipschitz matrix - valued function defined in @xmath61 such that @xmath60 is uniformly elliptic in @xmath62 , @xmath63 , and @xmath64 for @xmath65 .
let @xmath66 be a solution to the equation @xmath67 then for all @xmath68 , there exists @xmath69 depending only on @xmath70 , @xmath71 , and the lipschitz and ellipticity constants of @xmath60 such that , for @xmath72 , @xmath73 for some positive constant @xmath74 independent of @xmath75 . * proof . * by (
* theorem 2.3 and ( 2.10 ) ) , there exists a constant @xmath76 which depends only on @xmath70 and @xmath71 such that @xmath77 here and in what follows in this proof , @xmath74 denotes a positive constant which is independent of @xmath75 and can change from one place to another .
we claim that @xmath78 where @xmath79 and @xmath80 \big/ \ln \big[r_3 / r_2 \big].\ ] ] the proof of can be proceeded as follows .
we only consider the case @xmath6 , the case @xmath81 follows similarly . since @xmath82 in @xmath83 , one can represent @xmath75 in @xmath83 as follows @xmath84 for @xmath85 ( @xmath86 ) with the convention @xmath87 . for @xmath88
, we have @xmath89 here for two nonnegative quantities @xmath90 and
@xmath91 , @xmath92 means that @xmath93 and @xmath94 .
it follows that , for @xmath95 , @xmath96 where @xmath97 .
we obtain claim . a combination of and yields @xmath98 here we use the fact that @xmath99 by taking @xmath100 large enough so that @xmath101 is close to 1 , @xmath102 is close to 0 .
the conclusion follows .
@xmath103 the second lemma of this section is standard and its proof is left to the reader .
[ lem3 ] let @xmath104 , @xmath105 , @xmath106 be two smooth open subsets of @xmath107 , and @xmath108 , @xmath109 , and @xmath110 .
assume that @xmath111 is such that @xmath112 and @xmath113 = g \mbox { on } \partial d , \quad \mbox { and } \quad [ \partial_r v ] = h \mbox { on } \partial d.\ ] ] we have @xmath114 for some positive constant @xmath74 independent of @xmath115 , @xmath116 , and @xmath117 . here and in what follows , @xmath118 , @xmath119 , and @xmath120 $ ] on @xmath121 denote the trace of a function @xmath75 from the exterior , interior of @xmath122 , and @xmath123 on @xmath121 respectively . similarly , the notations @xmath124 , @xmath125 , and @xmath126 $ ] are used on @xmath121 for an appropriate matrix @xmath127 and an appropriate function @xmath75 , where @xmath128 is the normal unit vector directed to the exterior of @xmath122 . the following lemma which is a consequence of (
* lemma 4 ) is used in the proof of theorem [ thm1 ] .
[ lem - to ] let @xmath104 , @xmath129 with @xmath130 , @xmath131^{d \times d}$ ] be a matrix valued function , and @xmath132 be the kelvin transform with respect to @xmath133 , i.e. , @xmath134 for @xmath135 , define @xmath136
. then @xmath137 if and only if @xmath138 moreover , @xmath139
let @xmath140 be a matrix valued function defined in @xmath49 such that @xmath141 in @xmath11 , @xmath142 , and @xmath140 is lipschitz .
take @xmath143 large enough such that holds for @xmath144 with @xmath145 , @xmath146 , @xmath147 , and @xmath148 .
then @xmath149 for @xmath150 which satisfies @xmath151 in @xmath49 .
multiplying by @xmath152 and integrating on @xmath35 , we obtain @xmath153 considering the imaginary part and the real part , we have @xmath154 in this proof , @xmath74 denotes a positive constant changing from one place to another but independent of @xmath155 and @xmath115 . since @xmath156 on @xmath157
, it follows that @xmath158 we derive from that @xmath159 and @xmath160 as in @xcite , let @xmath161 be the reflection of @xmath162 through @xmath163 by @xmath164 , i.e. , @xmath165 applying lemma [ lem - to ] and using the fact that @xmath166 in @xmath11 , we obtain @xmath167 define @xmath168 in @xmath169 as follows @xmath170 0 & \mbox { in } b_{r_{2}}. \end{array } \right.\ ] ] using the fact that @xmath166 in @xmath11 and applying lemma [ lem - to ] , we derive from , , and that @xmath171 , @xmath172 = - \frac{i \delta}{1 - i \delta } { \cal a } \nabla u_{\delta } \cdot \eta \big|_{+}\mbox { on } \partial b_{r_{2}}.\ ] ] it is clear from , , and that @xmath173 let @xmath174 be the unique solution to @xmath175 = - \frac{i \delta}{1 - i \delta } { \cal a } \nabla u_{\delta } \cdot \eta \big|_{+}\mbox { on } \partial b_{r_{2}}.\ ] ] then @xmath176 since @xmath177 it follows from and that @xmath178 define @xmath179 then @xmath180 is a solution to @xmath181 using , we obtain @xmath182 from and , we have @xmath183 we derive from that @xmath184 on the other hand , from , , and , we have @xmath185 a combination of , , and yields @xmath186 since @xmath187 in @xmath188 , it follows that @xmath189 a combination of , , , and implies @xmath190 where , by , @xmath191 as in @xcite , let @xmath192 be the reflection of @xmath161 through @xmath52 by @xmath193 , i.e. , @xmath194 we have @xmath195 by . @xmath6 . from , one can represent @xmath161 and @xmath192 as follows @xmath196 and @xmath197 for @xmath198 ( @xmath199 ) .
we derive from that @xmath200 and applying lemma [ lem - to ] , we have @xmath201 a combination of , , and yields @xmath202 and @xmath203 we derive from that @xmath204 from and , we obtain @xmath205 we now introduce the technique of removing localized singularity .
set @xmath206 define from @xmath207 in @xmath208 .
function @xmath209 contains high modes and creates a trouble for estimating @xmath210 on @xmath211 ( to obtain an estimate for @xmath207 ) .
however this term can be negligible for large @xmath212 since @xmath213 is small for large @xmath7 .
this is the reason for the choice of @xmath209 and this also explains the terminology : removing localized singularity " . ]
@xmath214 u_{\delta } - { \hat u}_{\delta } & \mbox { in } b_{r_3 } \setminus b_{3 r_2 } , \\[6pt ] u_{2 , \delta } & \mbox { in } b_{3 r_2}. \end{array } \right.\ ] ] it is clear from that @xmath215 we claim that @xmath216 here and in what follows we use the standard notation : @xmath217 denotes a quantity which converges to @xmath218 as @xmath219 .
indeed , from , we have @xmath220 hence follows .
we derive from that @xmath221 \|_{h^{1/2}(\partial b_{r_3 } ) } + \|[\partial_r w_{\delta } ] \|_{h^{-1/2}(\partial b_{r_3 } ) } = o(1 ) \| f\|_{l^2(\omega ) } + o(1 ) \| u_\delta\|_{l^2(\omega \setminus b_{r_3})}.\ ] ] we next consider the transmission conditions for @xmath222 on @xmath211 . since @xmath223 = u_\delta - { \hat u}_\delta - u_{2 , \delta } \quad \mbox { on } \partial b_{3 r_2},\ ] ] it follows from and that @xmath223 = ( u_\delta - u_{1 , \delta } ) + \sum_{n = 1}^\infty \sum_{\pm } \frac{i \delta}{2 ( 1 - i \delta ) } ( 3 r_2)^n e_{n , \pm } e^{i n \theta } \quad \mbox { on } \partial b_{3 r_2}.\ ] ] this implies @xmath224 \|_{h^{1/2}(\partial b_{3 r_2})}^2 \le \| u_\delta - u_{1 , \delta } \|_{h^{1/2}(\partial b_{3 r_2})}^2 \\[6pt ] + \sum_{n = 1}^\infty \sum_{\pm } \delta^{2 } n |e_{n , \pm}|^2 r_3^{2n}\frac{(3 r_2)^{2n}}{r_3^{2n}}.\ ] ] we derive from and that @xmath225 \|_{h^{1/2}(\partial b_{3 r_2})}^2 \le & \delta^{2 \beta } \|f\|_{l^2(\omega ) } \| u_\delta \|_{l^2(\omega \setminus b_r ) } + \delta \| f\|_{l^2(\omega ) } \| u_\delta \|_{l^2(\omega \setminus b_{r_3 } ) } ; \end{aligned}\ ] ] which yields @xmath226 similarly , @xmath227
\|_{h^{-1/2}(\partial b_{3 r_2})}^{2 } = o(1 ) \| f\|_{l^2(\omega)}^{2 } + o(1 ) \|
u_\delta \|_{l^2(\omega \setminus b_{r_3})}^2.\ ] ] applying lemma [ lem3 ] and using , , and , we have @xmath228 for small @xmath155 . without loss of generality , one may assume that @xmath229 as @xmath219 . from , , and
, we have @xmath230 hence @xmath231 .
since the limit @xmath232 is unique , the convergence holds for the whole family @xmath233 as @xmath219 .
the proof is complete in two dimensions .
the proof in the three dimensional case follows similarly as the one in the two dimensional case .
we just note here that , in three dimensions , @xmath161 and @xmath192 can be represented , by , as follows @xmath234 and @xmath235 the proof is complete .
* acknowledgment : * the author thanks graeme milton for pointing out the localized resonance from simulations in @xcite .
he also thanks luca rondi for enlightening discussions on the three spheres inequality .
h. ammari , g. ciraolo , h. kang , h. lee , and g. w. milton , _ spectral theory of a neumann - poincar - type operator and analysis of cloaking due to anomalous localized resonance _ , arch .
rational mech .
218 * ( 2013 ) , 667692 .
h. ammari , g. ciraolo , h. kang , h. lee , and g. w. milton , _ spectral theory of a neumann - poincar - type operator and analysis of cloaking due to anomalous localized resonance @xmath236 _ , contemporary mathematics * 615 * ( 2014 ) 114 .
g. w. milton , n. p. nicorovici , r. c. mcphedran , and v. a. podolskiy , _ a proof of superlensing in the quasistatic regime , and limitations of superlenses in this regime due to anomalous localized resonance _ , proc .
ser . a * 461 * ( 2005 ) , 39994034 .
h. m. nguyen , _
asymptotic behavior of solutions to the helmholtz equations with sign changing coefficients _ , trans .
( 2015 ) , to appear , http://dx.doi.org/10.1090/s0002-9947-2014-06305-8 . | cloaking using complementary media was suggested by lai et al . in @xcite .
the study of this problem faces two difficulties .
firstly , this problem is unstable since the equations describing the phenomenon have sign changing coefficients , hence the ellipticity is lost .
secondly , the localized resonance , i.e. , the field explodes in some regions and remains bounded in some others as the loss goes to 0 , might appear . in this paper
, we give a proof of cloaking using complementary media for a class of schemes inspired from @xcite in the quasistatic regime .
to handle the localized resonance , we introduce the technique of removing localized singularity and apply a three spheres inequality .
the proof also uses the reflecting technique in @xcite . to our knowledge
, this work presents the first proof on cloaking using complementary media .
* msc . * 35b34 , 35b35 , 35b40 , 35j05 , 78a25 , 78m35 .
* key words . * negative index materials , cloaking , sign changing coefficients , localized resonance , complementary media . |
phononic crystals ( pncs ) have emerged as novel platform in which the phonon propagation dynamics can be controlled @xcite .
pncs are usually composed from a periodically modulated elastic media with the period determined by the wavelength of the corresponding phonon waves . the long range periodic modulation in these structures results in phonon standing waves to be established that give rise to phonon bandgaps .
this concept was first put into practice in the macroscopic regime within a two - dimensional ( 2d ) periodic arrangement of stainless steel cylinders which could block sound waves at frequencies corresponding to the bandgap @xcite .
the ultimate evolution of this concept has led to the development of 2d pncs which consist of a 2d periodic arrangement of air holes typically in a silicon crystal @xcite .
the 2d pnc can not only completely block phonons at the bandgap frequencies but the introduction of point and line defects , by modulating the air holes , even enables phonons to be guided and trapped within these structures . indeed
this platform can control different flavors of phonons i.e. ultra-/hyper - sound @xcite and even heat @xcite and their high confinement efficiencies have enabled concepts such as phonon multiplexing to be established for signal processing applications @xcite .
the key to accessing such advanced functions is underpinned by the phonon dispersion relation in these structures which can be engineered by careful design of the 2d periodic air hole matrix .
+ more recently , a variation of 2d pncs has been developed in pnc waveguides ( wgs ) where phonons are confined and guided via a one - dimensional ( 1d ) array of suspended electromechanical membranes @xcite .
in contrast to 2d pnc wgs , where phonons are confined within the line defect defined wg via the phonon bandgap of the surrounding 2d pnc , the phonon confinement in the 1d pnc wg is achieved by the acoustic impedance mismatch between the suspended membrane and the local bulk environment .
however , the periodic air hole array used to suspend the membranes imparts a periodic elastic potential that is experienced by the phonon confined within the wg and it results in the emergence of phononic bands , bandgaps , slow phonons and even nonlinear dispersion which can broaden the phonon packets in the wg .
however in spite of their richness , a detailed physical understanding of these effects and how they can be harnessed , akin to 2d pncs , remains elusive .
indeed such knowledge holds the tantalizing promise of greatly enhancing the functionality of the pnc wg architecture . for instance , photonic crystals represent the apex of such engineered structures in which photon dispersion relations are designed by tuning their periodic refractive index structure @xcite .
this in turn enables a whole range of optical phenomenon to be controllably accessed including four wave mixing @xcite , supercontinuum generation @xcite and even soliton formation @xcite for a multitude of applications .
+ to that end , the purpose of this study is to elucidate the key physical parameters of the pnc wg architecture that enables the phonon dispersion relations to be designed for a specific application .
these systematic experimental investigations , underpinned by finite element method ( fem ) analysis , reveal that by adjusting both the waveguide and air hole geometry enables the phononic transmission bands , phonon bandgaps , phonon velocities and the nonlinear dispersion to be fully controlled in the pnc wg . furthermore , these studies not only show the availability of energy degenerate phonon bands , which can be resolved in temporal measurements , but they also indicate the conditions needed to selectively activate them thus adding a completely new and hitherto unconsidered functionality to the pnc architecture .
the 1d pncs reported in this study are shown in fig .
1(a ) and were synthesized from an epitaxially grown single crystal gaas / algaas heterostructure as detailed in fig .
1(b ) . from this substrate ,
3 pnc wgs with different hole pitches ( @xmath0 ) of 8 , 10 and 12 @xmath1 m were fabricated on the same chip as shown in fig .
in total 3 chips were prepared each sustaining 3 wgs of fixed widths ( @xmath2 ) of 22 , 29 and 34 @xmath1 m , which yielded 9 different pnc wgs that were investigated .
+ specifically , the pnc wgs were fabricated by first patterning a mesa structure on the heterostructure via photolithography that was subsequently wet etched using a h@xmath3po@xmath4(6%):h@xmath5o@xmath5(6%):h@xmath5o(88% ) solution .
the mesa structure was exposed to a height of approximately 500 nm in order to eliminate electrical cross - talk between the neighboring pnc wgs through the si - doped gaas layer .
next , electrodes were patterned at the edges of the wgs via photolithography and were deposited with gold using the lift - off technique .
the 1d air hole array was then defined along the [ 110 ] crystal orientation where the number of holes were chosen to fix the length of all pnc wgs to approximately 1.0 mm .
the holes with a radius of 2.5 @xmath1 m were also patterned with photolithography and were then wet etched all the way down to the al@xmath6ga@xmath7as layer with h@xmath3po@xmath4:h@xmath5o@xmath5:h@xmath5o solution . finally , each chip was immersed in hf(5@xmath8):h@xmath5o(95% ) solution to isotropically etch the al@xmath6ga@xmath7as sacrificial layer which yielded the suspended membrane array that composed the pnc wg @xcite . by varying the hf immersion time between 25 , 33 and 40 minutes ,
3 chips sustaining 3 pnc wgs with @xmath9 = 22 , 29 and 34 @xmath1 m were prepared . in order to ensure high quality mechanical performance , all the chips were also cleaned post fabrication via an ultraviolet and o@xmath3 dry stripper to remove any residual contamination on the surface of the wgs .
+ the experiments were then performed at room temperature and in high vacuum ( @xmath10 2 @xmath11 10@xmath12 pa ) where the phonons were piezoelectrically generated in the wgs with the application of alternating voltage between the gold electrode and the si - doped gaas layer from a signal generator ( nf wf1974 ) as shown in fig .
the resultant vibrations , confined to the wg , were optically detected with he - ne laser doppler interferometer ( neoark mld-230v-200 ) .
the electrical output from the interferometer s photodetector was measured either in a vector signal analyzer ( hp 89410a ) or in an oscilloscope ( agilent dso6014a ) in which case the output from the interferometer was first preamplified ( nf sa-220f5 ) .
first , in order to evaluate the bandstructure as function of the pnc geometry , the transmission of all 9 pnc wgs were spectrally measured via the protocol depicted in fig .
the spectral responses of 3 wgs with ( @xmath2 , @xmath13 ) = ( 22 , 10 ) , ( 29 , 12 ) and ( 34 , 8) in microns corresponding to devices 1 , 2 and 3 henceforth are shown in figs .
2(a)-2(c ) respectively . in addition
, the corresponding dispersion relations extracted from fem simulations ( comsol multiphysics ) are also shown in these figures . in device 1 ,
phonon propagation in the wg emerges above a cut - off frequency of 3.8 mhz which corresponds to the 1st dispersion branch from mode a detailed in fig .
however , as the wg width increases in devices 2 and 3 , the cut - off frequency reduces to 2.2 and 1.6 mhz as shown in figs .
2(b ) and 2(c ) .
physically this trend arises from the larger membranes composing the wgs in these devices sustaining lower frequency resonances @xcite .
indeed the inverse dependence of the cut - off frequency to the wg width is observed irrespective of the hole pitch and is reproduced by the fem simulations as shown in fig .
+ the transmission measurements in device 1 ( fig .
2(a ) ) also reveal a phonon bandgap at 6 mhz that is created when the half wavelength of the phonon vibrations in the wg corresponds to the hole pitch length .
this results in bragg reflection that leads to standing waves which suppress phonon propagation at the corresponding frequencies .
meanwhile , the fem analysis reveal that this bandgap corresponds to the 1st and 3rd dispersion branches that sustain mode shapes with an antinode ( a in fig .
2(d ) ) or a node ( b in fig .
2(d ) ) between the holes in the wg with the same wavelength . in other words ,
the wg can support 2 vibrations of the same wavelength but different energies corresponding to the different mode profiles . on the other hand , as the hole pitch is increased to 12 @xmath1 m in device 2 shown in fig .
2(b ) , the bandgap between the 1st and 3rd branches closes due to the greater overlap between their corresponding mode shapes as a consequence of their longer wavelengths .
consequently , this observation naturally suggests that a shorter hole pitch will yield shorter wavelengths for the phonon modes in the 1st and 3rd branches giving rise to less overlap which should result in a more distinct bandgap emerging .
however , experimentally decreasing the hole pitch to 8 @xmath1 m in device 3 shown in fig .
2(c ) reveals a much less pronounced phonon bandgap .
this unexpected observation originates from the 3rd dispersion branch in this device sustaining mode shapes labeled d ( see fig .
2(d ) ) that overlaps with the 1st bandgap at the brillouin zone boundary and thus mask it as revealed by the fem simulations in fig .
+ the dispersion relations extracted from the fem simulations , shown in figs .
2(a)-2(c ) indicate that the pnc wgs can sustain a range of phonon branches whose corresponding mode shapes are detailed in fig .
however , the phonon modes corresponding to the 2nd and 4th branches ( gray lines in figs .
2(a)-2(c ) ) arising from mode shapes labeled c and e in fig .
2(d ) make a negligible contribution in the present study due to the piezoelectric transducer s inability to activate them as a consequence of the electrodes being located at their nodal position . in contrast , the 5th branch sustains mode shapes that exhibit tension mediated intermodal interactions with the modes in the 3rd branch that results in their anti - crossing which opens up a new bandgap around 7 to 9 mhz in devices 2 and 3 as shown in figs .
2(b ) and 2(c ) @xcite .
although the spectral transmission measurements are unable to clearly resolve these higher order bandgaps as they are located far from the brillouin zone boundary , surprisingly the temporal measurements detailed below can do so .
the spectral transmission measurements in all 9 pnc wgs enable both the cut - off frequency ( described above ) and the phonon bandgap to be systematically evaluated as function of the wg s geometry and are summarized in figs .
3(a ) and 3(b ) respectively . +
this analysis unsurprisingly reveals that the spectral positions of the 1st bandgap is strongly dependent on the wg width and the hole pitch length as shown in fig .
specifically , reducing the hole separation gives rise to shorter wavelength standing waves in the wg thus yielding a bandgap at higher frequencies , whereas the wg width determines the spectral position of all the phonon branches as referenced to the cut - off frequency detailed in fig .
indeed the fem simulations confirm the former trend as shown by the solid line in fig .
3(b ) for a pnc wg with @xmath2 = 29 @xmath1 m whilst @xmath13 is continuously varied from 7 to 13 @xmath1 m .
the spectral width of the 1st bandgap also shows an inverse correlation with the hole separation .
however this trend is not captured by the widest wg with @xmath13 = 8 @xmath1 m as a consequence of the emergence of the 3rd branch which overlaps the 1st bandgap as described above .
+ somewhat surprisingly , the spectral position of the 2nd bandgap also reveals a stronger than expected correlation on the hole separation even though this bandgap arises from the interaction between the phonon modes in the 3rd and 5th branches .
however as the hole pitch is reduced for a given pnc wg width , all the phonon branches migrate to higher frequencies which consequently results in the anti - crossing between the 3rd and 5th branches also occurring at higher frequencies . again
this trend is confirmed by the fem analysis as shown by the dotted lines in fig .
+ consequently these systematic analyses reveal that the spectral positions of the bandgaps and somewhat generally their spectral widths can be tailored by simply controlling the hole pitch in the pnc wgs . in order to gain a deeper understanding of the phonon propagation dynamics in the 1d pnc wgs ,
temporal measurements are also performed . in fig .
4 , the spatially mapped phonon propagation in device 2 in response to a rectangular shaped input pulse of 2 @xmath1s duration is shown .
this measurement is executed with a pulse frequency of 5.6 mhz at which phonon modes sustained by the 3rd dispersion branch are activated as detailed in fig .
the mechanical vibrations corresponding to these phonons are excited from one edge of the pnc wg and they subsequently propagate through the wg with a constant speed of 134 m / s and a loss of 0.4 db / mm . once these phonons reach the opposite edge of the wg , they are reflected , which results in a fabry - perot resonance that can be identified by the equally spaced peaks in the spectral measurements as highlighted in fig .
+ next to evaluate the temporal characteristics of the phonon propagation in the wg with respect to the underlying bandstructure , similar time of flight measurements are performed as function of frequency in all 3 devices .
phonons in the pnc wgs are excited by rectangular input pulses from the wg s right edge with durations of 2 @xmath1s and 8 @xmath1s which are then only detected at the left edge , in contrast to the spatial mapping measurements detailed above , and are shown in figs .
5(a)-5(c ) .
as expected , phonons in all 3 devices could only be observed above the cut - off frequency irrespective of the pulse duration .
once the phonons reached the opposite edge of a wg they are reflected resulting in an enhanced signal ( highlighted in fig .
indeed multiple reflections occur as identified by the enhanced intensity fringes in all 3 devices .
although the phonons propagate with constant velocity in the wgs but when the pulse frequencies approach the edges of the bandgap , phonon slowing occurs namely the phonon group velocity ( @xmath14@xmath15 ) is reduced .
this can be clearly seen by the change in curvature of the reflection fringes as highlighted by the black arrows in figs .
5(a)-5(c ) .
the physical origin of this effect arises from a change in the curvature of a dispersion branch with respect to wavevector at frequencies around the bandgap @xcite .
indeed this group velocity dispersion ( gvd ) can also broaden the phonon pulse packet as it travels down the pnc wg as detailed below .
+ in contrast to devices 1 and 2 where the extremities of the 1st bandgap could be identified from the phonon slowing , such unambiguous identification of the bandgap in device 3 was not possible .
in fact the temporal measurements indicate the existence of an interference like response in this spectral region which is also observed in the spectral measurements depicted in the inset to fig .
these observations can be understood from the fem simulations ( also shown in fig .
2(c ) ) which indicate that the 3rd dispersion branch overlaps with the 1st bandgap .
consequently the suppression of modes a and b ( see fig .
2(d ) ) in parallel with the availability of mode d ( see fig .
2(d ) ) and the subsequent overlap of their modal profiles gives rise to the complex temporal response observed in this device for frequencies in and around the bandgap .
it should be noted that in contrast to the spectral measurements , the bandgap region in device 3 could be more easily identified in the temporal measurements due to this highly pronounced interference effect .
the time of flight measurements detailed above also enable the group velocity to be quantitatively extracted .
specifically , the first fringe corresponding to a one way trip in the pnc wg is gaussian fitted at a given frequency .
the group velocity at this frequency can then be readily determined from the ratio of the wg s length to the time obtained from the center of the gaussian fit . extending this analysis to all the frequencies displayed in figs .
5(a)-5(c ) yields the spectral dependence of the group velocity in all 3 pnc wgs as shown in figs .
6(a)-6(c ) .
+ first , the extracted group velocities show no dependence on the pulse width where clearly the narrower pulse corresponds to a spectrally broader frequency packet and vice versa .
second , in the 1st and 3rd dispersion branches ( away from the 1st bandgap and only for mode shapes a and b in fig .
2(d ) ) spanning 3 different frequency ranges of 3.5 - 10 mhz , 2 - 7 mhz and 2 - 6 mhz in devices 1 , 2 and 3 respectively , almost identical group velocities are obtained .
indeed the group velocities determined directly from the fem simulated bandstructure shown in figs .
2(a)-2(c ) , from the relation @xmath16 , where the @xmath17 and @xmath18 are frequency and wavenumber respectively , not only reproduce the experimentally observed group velocities but they also confirm that the phonon velocities are inextricably linked to these dispersion branches . finally , the extracted group velocities not only exhibit the slowing effect in the proximity of the 1st bandgap , but it can also be seen around the 2nd bandgap ( see figs .
6(b ) and 6(c ) ) even though these bandgaps can not be resolved in the spectral measurements described in figs .
2(b)-2(c ) .
these results also unsurprisingly indicate that the broader 8 @xmath1s pulse , corresponding to a spectrally pure input , can probe the rapidly changing dispersion branches in the spectral regions around the bandgaps with a higher resolution than the 2 @xmath1s pulse . + the suppression in the group velocity can also modify the temporal waveforms of the pulse fringes as depicted in figs .
6(d)-6(f ) .
this effect can be attributed to the gvd which can be spectrally identified from the second derivative of the fem simulated dispersion branches in figs .
2(a)-2(c ) , namely the gvd coefficient @xmath19 as shown in the upper panels of figs .
6(a)-6(c ) @xcite . as @xmath20 increases in the proximity of the cut - off and the 1st bandgap at 4.3 mhz in devices 1 and 2 respectively ( see figs .
6(d ) and 6(e ) ) , the reflection fringes are both broadened and distorted .
in the proximity of the band edges where @xmath20 is maximized , this results in the velocity of phonons rapidly changing which experimentally manifest itself in the broadened temporal waveforms . in contrast , this frequency corresponds to the middle of the 1st dispersion branch located away from the band edges for device 3 ( see fig .
6(c ) ) where @xmath20 is negligible and thus the corresponding group velocity is more explicitly defined . as a result
the reflection fringes at this frequency also exhibit clearly defined waveforms without any broadening as shown in fig .
similar observations are also made in devices 1 and 2 at 7.5 mhz as detailed in fig .
6(d ) and 6(e ) .
consequently , designing the phonon bandstructure by simply tuning the geometry of the pnc wg enables not only the phonon group velocity at given frequency to be selectively adjusted but the corresponding waveforms can also be modified on demand via the nonlinear dispersion . + most interestingly , the temporal measurements shown in fig .
5(c ) clearly reveal the existence of 2 different reflection fringe profiles with different curvatures above 6 mhz ( where this new fringe is highlighted in pink ) . since this effect is most clearly observed with the 2 @xmath1s input pulse ( left panel in fig .
5(c ) and lower panel in fig .
6(f ) ) , their temporal waveforms are selectively fitted with 2 different gaussians .
this analysis reveals the existence of 2 different group velocity profiles above 6 mhz in device 3 as shown in fig .
as described above , at these frequencies the 3rd dispersion branch interacts with the 5th dispersion branch and the resultant anti - crossing modifies the 3rd branch so that it can sustain 2 distinct phonon modes ( b and d shown in fig .
2(d ) ) with different wavevectors but at the same frequency .
indeed the above analysis indicates that the additional fringe structure ( see lower panel in fig .
6(f ) ) yields slower group velocities which correspond to phonons in the 3rd branch whose mode profile is described by d ( see fig .
on the other hand , the faster group velocities at these frequencies correspond to phonons whose mode shape is described by b in fig .
2(d ) from the 3rd branch as in devices 1 and 2 . + the combination of temporal measurements in the pnc wgs with the fem derived bandstructure provides a powerful approach to investigating phonon dynamics in this platform .
indeed these investigations have even opened up the path to identifying different phonon modes within a particular dispersion branch .
this new degree of freedom , where different vibration modes are available at the same frequency , is akin to engineering the orbital angular momentum of photons to increase the information throughput at the same frequency in optical fibers @xcite .
consequently , these investigations bring forth the exciting prospect of geometrically engineering pnc wgs so that their bandstructure can sustain multiple vibration modes at the same frequency which in turn can be identified from temporal measurements via for example their different group velocities .
if such functionality can be harnessed in the pnc wg architecture it will not only make the concept of phonon circuits more tantalizing but it will also open up new directions in phononics .
the 1d pnc wg architecture is systematically investigated as function of its geometry via a combination of spectral and temporal measurements in conjunction with its bandstructure extracted from fem simulations .
these investigations reveal that the cut - off frequency , the spectral position and width of a given phonon band and the bandgap in the pnc wg can be controlled by simply varying the width of the wg and the pitch of the holes used to suspend the membranes .
consequently , the speed of phonons and even their temporal profile can be controlled via the gvd at given frequency by appropriately engineering the wg geometry in accordance to the bounds extracted in this report .
furthermore , the fem simulations also indicate that phonons in only certain dispersion branches can be activated as consequence of the position of the piezoelectric transducers on the pnc wg .
indeed this analysis provides design directions in future devices where multiple piezoelectric transducers will be incorporated to selectively activate phonons in a particular dispersion branch .
most unexpectedly , the tension induced interaction between dispersion branches not only results in the creation of new bandgaps but the highly modified branches also sustained different wavevector vibration modes which coexist at the same frequency . although the spectral measurements could not resolve these degenerate states , the temporal measurements uniquely could discriminate between the coexisting phonon modes in a given dispersion branch via their differing group velocities .
+ consequently , this study vividly indicates that the combination of the fem simulated bandstructure in combination with temporal measurements are the key to ascertaining detailed knowledge of the phonon dynamics in the pnc wg architecture .
ultimately these investigations have yielded insights into the relationship of the wg s geometric parameters to its underlying bandstructure which can now be harnessed to engineer and exploit both the well - known phonon dynamics and the more exotic phononic degrees of freedom arising from the nontrivial bandstructure in pnc wgs . | the phonon propagation dynamics in a phononic crystal waveguide , realized via a suspended one - dimensional membrane array with periodic air holes , is investigated as function of its geometry .
the bandstructure of the phononic crystal can be engineered by modifying the characteristics of the phonon standing waves in the waveguide by varying the waveguide width and the pitch of the air holes .
this enables the phonon transmission bands , the bandgaps , the velocity and the nonlinear dispersion in the phononic crystal to be controlled .
indeed the engineered bandstructure can also be tuned to sustain multiple phonon modes in a given branch which whilst being spectrally degenerate can be temporally resolved via their differing group velocities .
this systematic study reveals the key geometric parameters that enable the phonon transport in phononic crystal waveguides to be fully controlled . |
the first explanation ( or rather prediction ) of the power law of primary cosmic ray spectrum made by e. fermi was purely phenomenological : it was based on the exponential growth of energy with time and exponential distribution of ages of observed particles .
both of them ignored the real space - time distribution of sources , highly inhomogeneous structure of the interstellar medium and boundedness of the galaxy .
a few years later , the cr propagation was interpreted as some kind of brownian motion and began to be calculated in terms of the ordinary ( normal ) diffusion theory @xcite .
it is easy to understand the origin of this model : fermi wrote about random walk of cr particles among magnetic clouds chaotically distributed in interstellar space , and brownian motion was the most expressive image of such a process .
however , its mathematical representation in the form of the _ local diffusion _ ( lod ) equation ( or its other local modifications ) did not involve the particle velocity in an explicit form and for this reason failed in description of front propagation and some other pecularities of the cr transport .
this is just the same task which inspired authors of the work @xcite to seek a new propagator in the frame of _ nonlocal diffusion _ ( nod ) model .
we talk here on locality because the generic sign of the diffusion approximation is the local laplace operator @xmath0 .
its application assumes uniformity and mutually independence of points of scattering like those we observe in ideal gases .
no doubt , distribution of irregularities in the interstellar magnetic field do not possess this property , they are linked with each other by means of more or less long magnetic force lines .
these lines are frozen into the material component of interstellar plasma and are involved into the common turbulent motion .
it is rather strange , why the results in turbulent diffusion theory obtained in classical works followed the kolmogorov ( 1941 ) phenomenological basis of the turbulence @xcite were completely ignored in the early development of the cr - propagation theory .
one of the generic features of the turbulent diffusion is the presence of nonlocal operators in its equation .
involvement of non - local operator @xmath1 instead of its local counterpart @xmath2 has shown that in case of a single local pulse source the knee arises without any additional assumptions except the turbulent ( called also _ fractal _ ) character of the interstellar medium . during 2001 - 2010 years
, a series of calculations were performed by lagutin et al . with the use of nod - model , but all results were obtained without account for the relativistic limitation of speed and the boundary conditions .
the speed limitation was inserted into the nod - model in @xcite and this version was later named the _ nonlocal relativistic diffusion _ ( nord ) model @xcite .
the physical interpretation of the nord model and first results obtained in its frame were described in reviews @xcite . here ,
we report some new results of numerical investigations in the frame of this model adding account of multiplicity of sources and boundedness of halo .
modeling of cr sources was performed under following assumptions . 1
. all sources are isotropic , point and instantaneous with equal powers and energy spectra @xmath3 . here
@xmath4 is an exponent of injection spectrum , and @xmath5 characterizes maximum energies of @xmath6-type nuclei accelerated at supernova remnants .
the sources are separated into two groups : relatively young supernovae ( less than @xmath7 years ) whose space - time coordinates were taken individually ( 15 objects : vela , monogem ring , geminga and others ) , and 30000 sources with spatial distribution generated according to approach proposed in @xcite and used in @xcite .
the radial distribution of discrete sources was simulated according to the density of pulsars ( see @xcite ) , @xmath8 with @xmath9 kpc and @xmath10 , the @xmath11-distribution was modeled by exponential function with @xmath12 pc . to simulate the spiral structure ,
the algorithm proposed in @xcite was used .
the example realization is presented in fig .
[ fig_sources ] .
the area of cr diffusion is the cylinder of height @xmath13 kpc with radius @xmath14 kpc surrounded by a non - reflective environment .
the galactic disk being in the middle of the cylinder parallel to bases has the thickness @xmath15 . as in ref .
@xcite , we assume that the diffusion coefficient is independent of spatial and time variables , and determined by particle rigidity @xmath16 , @xmath17 , \quad r>3gv.\ ] ] results reported here are related to values @xmath18 and @xmath19 @xmath20 . to account for the halo boundedness , solutions ( [ eq_lod ] ) should be modified : considering zero boundary conditions and neglecting by lateral surface of the cylinder , the image method can be used in local approach , @xmath23 @xmath24 here @xmath25 are @xmath26-coordinates of the image sources , @xmath27 are coordinates of an original source
. the nod - propagator in an infinite medium is expressed through fractional stable lvy - feldheim density @xcite . in case of @xmath28
it can be presented in the form @xcite
@xmath29 @xmath30 here @xmath31 is a one - sided lvy stable density ( subordinator , @xmath32 ) and @xmath33 in ( [ eq_nod ] ) be taken with the diffusion coefficient linked with @xmath34 of the nod - model by relation @xmath35^{2/\alpha}.\ ] ] if @xmath36 , then @xmath37 , where @xmath38 .
for example , for @xmath18 and @xmath39 , we have @xmath40 . formula ( [ eq_nod ] ) is suitable for monte carlo representation of the nod - propagator @xmath41 where @xmath42 is a stable random variable generated according to ( see e.g. @xcite ) @xmath43 @xmath44^{2/\alpha-1}}{[\sin ( \pi u_1)]^{2/\alpha}[\ln u_2]^{2/\alpha-1}}.\ ] ] formulae ( [ eq_diff_coeff ] ) and ( [ eq_nod_mc ] ) indicate the correspondence principle between nod- and lod - models ( when @xmath45 ) and allow us to use the standard diffusion coefficient to estimate cr characteristics in frames of the nonlocal models . in calculations of energy spectrum in frames of the nord - model ( for @xmath46 )
we have used approximate expression for propagators @xmath47 @xmath48 @xmath49 where @xmath50 is the heaviside step function .
it should be noted that zero boundary conditions for nonlocal models are fulfilled approximately , if the mean scattering length of crs much smaller than thickness of the halo . to account for restriction of the diffusion domain ( halo ) , the generalized marshak boundary conditions can be used .
the solutions have been modified by means of the image method . in this approximation , we use @xmath51 instead of @xmath33 in expressions ( [ eq_nod ] ) and ( [ eq_nod_mc ] ) using procedure ( [ eq_nord ] ) .
we estimated energy spectra accounting for spatial and temporal distribution of supernova remnants and diffusion in halo in frames of four models ( lod , lord , nod , nord ) .
note , that such calculations for the local model were performed in many papers ( see e.g. @xcite ) and for nonlocal model in refs .
all of them , however , utilize propagators ignoring the relativistic speed limit requirement .
relativistic nature of crs sufficiently modify the propagators and spectra as well .
nord - propagators are restricted to domain @xmath52 , have ballistic splash at this boundary whereas in the middle part they are spreading superdiffusively ( remind , that we restrict ourselves by values of superdiffusion parameter @xmath53 $ ] ) .
spallation effects are taken into account according to blasi and amato @xcite using formulae provided in @xcite . as is known , the standard assumptions about the origin of galactic crs : power law spectra of particles injected by snr , and diffusive propagation throughout the galaxy with a coefficient @xmath36 leads to the spectrum @xmath55 ( from the snr ensemble ) reflecting the balance between injection and escape of crs from the confinement volume . accounting for relativistic restriction ( in terms of relativistic propagators ) steepens the spectrum at high energies , and the position of the spectrum break , its sharpness and slopes of asymptotes depend on @xmath56 and @xmath57 @xcite .
the cause of this effect is universal and must be kept in mind in all calculations of propagation of high energy crs .
explanation of the spectrum steepening at high energies in the nord model is schematically represented in fig .
[ fig_steepening ] .
the grey field corresponds to the space - time area containing cr sources which do not take part in the formation of the spectrum due to the relativistic restriction .
their accounting in the ordinary diffusion model leads to a pure power law spectrum @xmath58 under the assumption about the pure power law spectrum of injected particles .
taking the relativistic restriction into account in the nord model excludes automatically these sources and this exclusion affects predominantly on high energy part of crs due to @xmath36 .
it is important to emphasize that , the nord model ( compared to the nod - version used by lagutin et al . ) operates with acceptable values of injection exponent @xmath4 ( @xmath59 ) ( in contrast with @xmath60 , e.g. @xmath61 , in recent calculations by lagutin et al .
@xcite ) .
[ proton ] , [ cno ] and [ total ] demonstrate computational results for spectrum of protons , cno - group and all nuclei obtained in frames of lod , lord , nod and nord models .
relativistic speed limit requirement taken into account in lord and nord leads to steepening of spectra as mentioned above .
the energy of the ` knee ' @xmath62 obtained in nord is very close to provided in @xcite by exponential cut - off of injection spectrum , but we observe power law nord - spectrum above @xmath62 in contrast to exponential decay obtained in @xcite .
sure , this requires greater maximum energies of particles accelerated in sources .
the exponentially falling tails for all nuclei contradict to observed power law spectrum above @xmath62 , if we consider it as produced by gcrs ( without extragalactic component ) . from this point of view nord - model
seems to be preferable .
lagutin , a. a. , nikulin , y. a. , & uchaikin , v. v. ( 2001 ) .
the ` knee ' in the primary cosmic ray spectrum as consequence of the anomalous diffusion of the particles in the fractal interstellar medium .
_ nuclear physics b proceedings supplements _ , 97(1 ) , 267 .
erlykin , a. d. , sibatov , r. t. , uchaikin , v. v. , & wolfendale , a. w. ( 2015 ) a look at the cosmic ray anisotropy with the nonlocal relativistic transport approach .
_ proceedings of science _ ( icrc2015 ) 463 .
blasi , p. , & amato , e. ( 2012 ) .
diffusive propagation of cosmic rays from supernova remnants in the galaxy .
i : spectrum and chemical composition .
_ journal of cosmology and astroparticle physics _
, 2012(01 ) , 010 .
evoli , c. , gaggero , d. , grasso , d. , & maccione , l. ( 2008 ) .
cosmic ray nuclei , antiprotons and gamma rays in the galaxy : a new diffusion model .
_ journal of cosmology and astroparticle physics _ , 2008(10 ) , 018 .
lagutin , a. a. , tyumentsev , a. g. , volkov , n. v. , & raikin , r. i. ( 2015 ) .
spectra of cosmic - ray protons and nuclei from @xmath63 to @xmath64 ev within the galactic origin scenario of cosmic rays .
_ bulletin of the russian academy of sciences : physics _ , 79(3 ) , 322 - 325 . | energy spectra of galactic cosmic rays calculated within the framework of the nord ( nonlocal relativistic diffusion ) model are presented .
the model accounts for the turbulent character of the interstellar medium and the relativistic speed limit requirement .
calculations account for spiral distribution of sources , boundedness of halo , spallation of nuclei , energy dependence of the diffusion coefficient , tempered power law injection spectrum .
we show that the knee in the background spectra from the ensemble of supernovae can arise due to relativistic speed limit requirement .
we compare our calculations with the results obtained in frames of the local models ( lod and lord ) and nonlocal nonrelativistic model ( nod ) . |
in the past few years , many more new charmmium - like states , so - called @xmath11 states , have been observed by the belle , babar and besiii collaborations @xcite .
the discovery of these states not only enriched the spectroscopy of charmmium - like states , but also provided us an opportunity to research the properties of charmmium - like states .
for example , the @xmath0 was observed in the inclusive process @xmath12 and had the decay mode @xmath13 by the belle collaboration at a mass of @xmath14 mev .
the decay width of these state was less than 52 mev at the @xmath15 c.l . which has taken into the systematics @xcite .
later belle collaboration confirmed the observation of @xmath0 with a significance of @xmath16 , they got the mass and width of @xmath0 were @xmath17 mev , @xmath18 mev . at the same time , they also observed a new charmnium - like state @xmath1 in the process @xmath19 , the mass and width of @xmath1 were @xmath20 mev , @xmath21 mev @xcite .
the observations of these @xmath11 states inspire many interests about their physical natures .
there are already many theoretical approaches which have been used to study the properties of these @xmath11 states @xcite . in this paper
, we mainly discuss the properties and decays of two @xmath22 states : @xmath0 and @xmath1 . ref .
@xcite and ref .
@xcite have assigned that the @xmath23 parity of @xmath0 and @xmath1 should be even , @xmath24 . assuming @xmath0 as @xmath25 or one of @xmath26 charmnium - like states , ref .
@xcite studied @xmath12 process by the light - cone formalism , and they considered that @xmath0 is @xmath25(@xmath3 ) . ref .
@xcite investigated the properties of @xmath0 and @xmath1 which were @xmath3 and @xmath27 , respectively .
@xcite calculated the strong decays of @xmath28 , they found that the explanation of @xmath0 as @xmath3 is possible and the assignment of @xmath1 as @xmath4 can not be excluded . in ref .
@xcite , the authors studied vector - vector interaction of @xmath1 which was basically a @xmath29 molecular state with @xmath30 . ref .
@xcite had studied the inclusive production of @xmath0 in the decay of ground bottomnium state @xmath31 by the nrqcd factorization formula , and they also considered @xmath0 as the excited @xmath3 state .
@xcite calculated the strong decay of @xmath1 which was assumed as @xmath27 , @xmath32 , @xmath33 or @xmath4 by the @xmath5 model . in ref .
@xcite , they also explored the properties of @xmath0 and @xmath1 as the @xmath3 and @xmath4 , respectively .
but their results suggested that @xmath0 as @xmath3 was established and the explanation of @xmath1 to be @xmath4 is fully excluded . using the nrqcd factorization approach , ref .
@xcite calculated the branching fractions of @xmath34 with @xmath35 or @xmath36 , they thought that the @xmath0 and @xmath1 can be explained as @xmath25 and @xmath37 charmnium - like states , respectively .
up to now , it is very difficult to confirm the constructions of @xmath0 and @xmath1 , because lack of the enough experimental data .
many more theoretical prediction and experimental data are needed for @xmath0 and @xmath1 .
the bethe - salpeter(b - s ) equation is a good way to describe the mesons .
@xcite took the b - s equation to describe the light mesons @xmath38 and @xmath39 , then they calculated the mass and decay constant of @xmath38 by the b - s amplitudes @xcite , they also studied the weak decays @xcite and the strong decays @xcite combine with the dyson - schwinger equation . in this paper , we focus on the strong decays of @xmath0 and @xmath1 as radial high excited states @xmath3 and @xmath4 by the improved b - s method . because the higher excited states have larger relativistic correction than the corresponding ground state , a relativistic model is needed in a careful study .
the improved b - s method is a relativistic model that describes bound states with definite quantum number , the corresponding relativistic form of wavefunctions are solutions of the full salpeter equations . in previous works , we have studied the properties of some @xmath11 states , such as radiative e1 decay of @xmath40 @xcite , two - body strong decay of @xmath41 which was @xmath42 state combine with the @xmath5 model @xcite , the strong decay of @xmath43 as @xmath44 state @xcite .
we also studied the productions of @xmath0 and @xmath1 in @xmath45 decays @xcite .
all the theoretical results consistent with experimental data , so the improved b - s method is good method to describe the properties and decays of the radial high excited states , we will choose this method to study @xmath0 and @xmath1 . in our method , we study the natures of heavy mesons by the coupling of @xmath46 for the quark and anti - quark in mesons . according to the @xmath46 coupling
, we show the wavefunctions of the heavy mesons in term of the quantum number @xmath47 ( or @xmath48 ) which are very good to describe the equal mass systems in heavy mesons .
the quantum numbers @xmath48 of @xmath3 and @xmath4 both are @xmath49 , the @xmath23 parities are even which agree with the results of ref .
@xcite and ref .
the corresponding okubo-zweig-iizuka ( ozi ) @xcite rule allowed two - body open charm strong decay modes are : @xmath50 and @xmath51 , while other strong decays in the final state are ruled out by the kinematic possible mass region . in order to calculate the two - body open charm strong decay ,
we adopt the @xmath5 model which assumes that a quark - antiquark pairs is created with vacuum quantum numbers , @xmath52 @xcite .
the @xmath5 model was proposed in ref .
@xcite , then ref .
@xcite and ref .
@xcite applied the @xmath5 model to study the open - flavor strong decays of the light mesons .
now , people have extended this model to study the natures of heavy - light mesons @xcite and heavy quarkonia @xcite . in ref .
@xcite and ref .
@xcite , we have calculated the ozi allowed two - body strong decays of charmnium and bottomnium in the @xmath5 model with the relativistic b - s wavefunctions . in this paper , we take the same method as ref .
@xcite and ref .
@xcite to study strong decays of @xmath0 and @xmath1 .
the paper is organized as follows . in sec .
ii , we introduce the instantaneous b - s equation ; we show the relativistic wavefunctions of initial mesons and final mesons in section .
iii ; in sec .
iv , we give the formulation of two - body open charm strong decays ; the corresponding results and conclusions are present in sec . v.
in this section , we briefly review the bethe - salpeter equation and its instantaneous one , the salpeter equation .
the bs equation is read as @xcite : @xmath53 where @xmath54 is the b - s wave function , @xmath55 is the interaction kernel between the quark and antiquark , and @xmath56 are the momentum of the quark 1 and anti - quark 2 .
we divide the relative momentum @xmath57 into two parts , @xmath58 and @xmath59 , @xmath60 @xmath61 correspondingly , we have two lorentz invariants : @xmath62 when @xmath63 , @xmath64 and @xmath65 , respectively . in instantaneous approach , the kernel @xmath55 takes the simple form @xcite : @xmath66 let us introduce the notations @xmath67 and @xmath68 for three dimensional wave function as follows : @xmath69 @xmath70 then the bs equation can be rewritten as : @xmath71 the propagators of the two constituents can be decomposed as : @xmath72 with @xmath73\ ; , \label{eq8}\ ] ] where @xmath74 for quark and anti - quark , respectively , and @xmath75 . introducing the notations @xmath76 as : @xmath77 with contour integration over @xmath78 on both sides of eq .
( [ eq6 ] ) , we obtain : @xmath79 and the full salpeter equation : @xmath80 @xmath81 @xmath82 for the different @xmath48 ( or @xmath83 ) states , we give the general form of wave functions . reducing the wave functions by the last equation of eq .
( [ eq11 ] ) , then solving the first and second equations in eq .
( [ eq11 ] ) to get the wave functions and mass spectrum .
we have discussed the solution of the salpeter equation in detail in ref .
@xcite .
the normalization condition for bs wave function is : @xmath84=2p_{0}\;. \label{eq12}\ ] ] in our model , the instantaneous interaction kernel @xmath85 is cornell potential , which is the sum of a linear scalar interaction and a vector interaction : @xmath86 where @xmath87 is the string constant and @xmath88 is the running coupling constant . in order to fit the data of heavy quarkonia
, a constant @xmath89 is often added to confine potential .
one can see that @xmath90 diverges at @xmath91 , we introduce a factor @xmath92 to avoid the divergence : @xmath93[vsvv ] it is easy to know that when @xmath94 , the potential becomes to eq .
( [ vrww ] ) . in the momentum space and the c.m.s of the bound state ,
the potential reads : @xmath95 @xmath96 where the running coupling constant @xmath97 is : @xmath98 we introduce a small parameter @xmath99 to avoid the divergence in the denominator .
the constants @xmath87 , @xmath100 , @xmath89 and @xmath101 are the parameters that characterize the potential .
@xmath102 for @xmath103 ( and @xmath104 ) system .
in this paper , we focus on the two - body open charm strong decay of @xmath0 and @xmath1 which are considered as @xmath3 @xmath4 states .
@xmath4 states have two decay modes : @xmath50 and @xmath51 .
so we only discuss the relativistic wavefunctions of @xmath83 equal to @xmath105 and @xmath106 states .
the general form for the relativistic wavefunction of pseudoscalar meson can be written as @xcite : @xmath108\gamma_5,\end{aligned}\ ] ] where @xmath109 is the mass of the pseudoscalar meson , and @xmath110 are functions of @xmath111 . due to the last two equations of eq .
( [ eq11 ] ) : @xmath112 , we have : @xmath113 where @xmath114 and @xmath115 are the masses and the energies of quark and anti - quark in mesons , @xmath116 .
the numerical values of radial wavefunctions @xmath117 , @xmath118 and eigenvalue @xmath109 can be obtained by solving the first two salpeter equations in eq .
( [ eq11 ] ) . in ref .
@xcite , we have plot the wavefunctions of @xmath0 and @xmath1 which are considered as @xmath3 and @xmath4 , respectively . according to the eq .
( [ eq10 ] ) the relativistic positive wavefunction of pseudoscalar meson in c.m.s can be written as @xcite : @xmath119{\gamma}_5,\end{aligned}\ ] ] where the @xmath120s ( @xmath121 ) are related to the original radial wavefunctions @xmath117 , @xmath118 , quark masses @xmath122 , @xmath123 , quark energy @xmath124 , @xmath125 , and meson mass @xmath109 : @xmath126 the general form for the relativistic wavefunctions of vector state @xmath128(or @xmath129 for quarkonium ) can be written as eight terms , which are constructed by @xmath130 , @xmath131 , @xmath132 and gamma matrices @xcite , @xmath133+m_{f1}{\not\!\epsilon_1}f'_5\\ \nonumber & + & { \not\!\epsilon_1}{\not\!p_{f1}}f'_6 + ( { \not\!q}_{f1\perp}{\not\!\epsilon_1}- q_{f1\perp}\cdot{\epsilon_1 } ) f'_7+\frac{1}{m_{f1}}({\not\!p_{f1}}{\not\!\epsilon_1 } { \not\!q}_{f1\perp}-{\not\!p_{f1}}q_{f1\perp}\cdot{\epsilon_1 } ) f'_8,\label{eq13}\end{aligned}\ ] ] where @xmath134 is the polarization vector of the vector meson in the final state .
the relativistic positive wavefunctions of @xmath137 state can be written as @xcite : @xmath138 where we first define the parameter @xmath139 which are the functions of @xmath140 ( @xmath137 wave functions ) : @xmath141 then we define the parameters @xmath120 which are the functions of @xmath140 and @xmath139 : @xmath142 @xmath143 @xmath144
for the two - body ozi - allowed open charm strong decays , such as @xmath145 , we adopt the @xmath5 model to calculate the strong decay amplitude .
the non - relativistic @xmath5 model describe the decay matrix elements by the @xmath146 pair - production hamiltonian : @xmath147 @xcite . according to the improved b - s method which is a relativistic model
, we can extend the non - relativistic @xmath5 model to the relativistic form : @xmath148 @xcite . here
@xmath149 is the dirac quark field , @xmath150 , @xmath151 is the quark mass of the light quark - pairs , @xmath152 is a dimensionless constant which describe the pair - production strength and can be obtained by fitting the experimental data . in this paper , we choose @xmath153 @xcite which give reasonable calculation of @xmath3 , then we use the same value to @xmath4 .
the feynman diagram of two - body open charm strong decay.,height=188 ] using the @xmath146 pair - production hamiltonian , the amplitude of two - body ozi - allowed open charm strong decays @xmath154 in fig .
[ ozistrongdecay ] , can be written as @xcite , @xmath155\nonumber\\ & = & g\int\frac{d^3\vec q}{(2\pi)^3}{\rm tr}\left[\frac{\not\!{p}}{m}\varphi^{++}_p(\vec q ) \frac{\not\!{p}}{m}\bar\varphi^{++}_{p_{f2}}(\vec
q_{f2})\bar\varphi^{++}_{p_{f1}}(\vec q_{f1})\right ] \left(1-\frac{m - w_1-w_2}{2w_{12}}\right),\end{aligned}\ ] ] where @xmath156 , @xmath157 and @xmath158 are the relativistic positive wavefunctions of initial meson @xmath159 , finial meson @xmath160 and @xmath23 , respectively .
we have given the detailed form of wavefunctions in sec .
@xmath162 , @xmath130 , @xmath163 and @xmath164 , @xmath165 , @xmath166 are the momentum and three dimension relative momentum between quark and anti - quark of initial meson @xmath159 , finial meson @xmath160 and @xmath23 , respectively .
@xmath167 , @xmath168 .
@xmath169 and @xmath170 are the three momentum of finial mesons @xmath160 and @xmath23 .
@xmath171 . using the b - s wavefunctions in sec .
iii and the formula of amplitude eq .
( [ amplitude ] ) , the two - body open charm strong decay amplitude can be defined as , @xmath172 where @xmath159 denote @xmath3 or @xmath4 , @xmath173 and @xmath174 are the polarization vector of the final mesons @xmath160 and @xmath23 .
@xmath175 and @xmath176 are the strong decay coupling constants which are related to the b - s wavefunctions . finally , using eq .
( [ amplitude1 ] ) the two - body open charm strong decay width can be written as , @xmath177 where @xmath178[m^2-(m_{f1}+m_{f2})^2]}/(2m)$ ] which is the three momentum of the final mesons .
in order to fix cornell potential in eq.([eq16 ] ) and masses of quarks , we take these parameters : @xmath179 gev@xmath180 , @xmath181 gev , @xmath182 gev , @xmath183 gev , @xmath184=0.311 gev , @xmath185=0.500 gev , @xmath186 gev , @xmath187 gev , @xmath188 @xcite , which are best to fit the mass spectra of ground states @xmath160 , @xmath189 mesons and other heavy mesons . and
we get the masses : @xmath190 gev , @xmath191 gev , @xmath192 gev , @xmath193 gev , @xmath194 gev , @xmath195 gev , @xmath196=3.942 gev , @xmath197=4.156 gev . for @xmath4 state ,
the main strong decay channels are @xmath198 and @xmath199 , @xmath200 is very small with the small phase space , and the decay @xmath201 is forbidden . in table .
[ totalwidth ] , the total two - body open charm strong decay widths of @xmath4 is @xmath7 mev .
our result is larger than the result of ref .
@xcite , but considering the uncertainties of the results , our result is close to the lower limit of @xmath1 for experimental data @xcite . in our calculation , the branching ratio @xmath202 , which is consistent with the experimental data @xmath203 @xcite .
there is another of branching ratio of @xmath4 : @xmath204 , is much larger than the upper limit of the experimental data @xmath205 which is reported by belle @xcite . in order to find out the relation of the decay width to the mass of @xmath4
, we plot the relation of different decay width and branching ratio to the mass of @xmath4 in fig .
[ x4160m - width ] and fig .
[ br ] . especially in fig .
[ br ] , the branching ratio is decreased with the increased the mass of @xmath4 it may be close the experimental data by enough large mass .
the mass and width of @xmath1 have larger uncertainties , so if we want to take @xmath4 as the candidate of @xmath1 , the precision measurement of the total decay width and the branching ratio @xmath206 is needed in future experiments . in summary , considering @xmath0 and @xmath1 as @xmath3 and @xmath4 states , we study the two - body open charm ozi - allowed strong decay of @xmath3 and @xmath4 by the improved b - s method combine with the @xmath5 model . for the strong decay of @xmath3 ,
the dominant strong decay is @xmath145 , the corresponding strong decay width is @xmath6 mev , which is close to the experimental data , therefore , @xmath3 is a good candidate of @xmath0 . for @xmath4 state ,
the main strong decay channels are @xmath198 and @xmath199 , @xmath4 can not decay to @xmath208 , which have not been observed for @xmath1 in experiment .
@xmath209 is smaller than @xmath210 , the branching ratio @xmath8 is larger than the experimental data by belle .
we also find that the branching ratio @xmath8 is dependent on the mass of @xmath4 .
finally , we calculate the strong decay width of @xmath4 : @xmath7 mev , considering the errors of the results , it s close to the lower limit of @xmath1 . so taking the @xmath4 as an assignment for @xmath1 ,
the precision measurement of the total decay width and the ratio between @xmath9 and @xmath10 need to be confirmed in future by the new experiments .
99 k. a. olive @xmath211 .
( partile data group ) , 2015 , @xmath212 c * 38 * 090001 . k. abe @xmath213 belle collaboration , 2007 , @xmath214 * 98 * 082001 .
p. pakhlov @xmath213 belle collaboration , 2008 , @xmath214 * 100 * 202001 .
x. liu , 2014 , @xmath215 * 59 * 3815 .
h. x. chen , w. chen , x. liu and s. l. zhu , 2016 , @xmath216 * 639 * 1 .
v. v. barguta , a. k. likhoded and a. v. luchinsky , 2006 , @xmath217d * 74 * 094004 .
k. t. chao , 2008 , @xmath218b * 661 * 348 .
b. q. li and k. t. chao , 2009 , @xmath219 d * 79 * 094004 .
t. h. wang and g. l. wang , 2011 , @xmath220 b * 697 * 233 .
t. h. wang , g. l. wang , y. jiang and w. l. ju , 2013 , @xmath221 g * 40 * 035003 .
t. h. wang , g. l. wang , h. f. fu and w. l. ju , 2013 , @xmath222 * 1307 * 120 .
x. w. liu , h. w. ke , x. liu and x. q. li , 2016,@xmath217d * 93 * 074013 .
h. wang , z. z. yan and j. l. ping , 2015 , @xmath223c * 75 * 196 .
r. molina and e. oset , 2009 , @xmath217d * 80 * 114013 . z. g. wang and t. huang , 2014 , @xmath217d * 89 * 054019 .
z. g. he and b. q. li , 2010 , @xmath220 b * 693 * 36 . y. c. yang , z. r. xia and j. l. ping , 2010 , @xmath217d * 81 * 094003 .
he , d. y. chen , x. liu and t. matsuki , 2014 , @xmath223c * 74 * 3208 .
r. l. zhu , 2015 , @xmath219 d * 92 * 074017 .
w. sreethaong , k. xu and y. yan , 2015 , @xmath221 g * 42 * 025001 .
p. maris and c. d. roberts , 1997 , @xmath224 * 56 * , 3369 .
p. maris , c. d. roberts and p. c. tandy , 1998 , @xmath225 * 420 * , 267 , + c. d. roberts , 1996 , nucl .
a * 605 * 475 . m. a. ivanov , j. g. k@xmath226rner , s. g. kovalenko and c. d. roberts , 2007 , phys .
d * 76 * 034018 , + m. a. ivanov , yu . l. kalinovsky , p. maris and c. d. roberts , 1998 , phys .
c * 57 * 1991 , + m. a. ivanov , yu .
l. kalinovsky , p. maris and c. d. roberts , 1998 , phys .
b * 416 * 29
. + b. el - bennich , m. a. ivanov and c. d. roberts , 2011 , @xmath224 * 83 * 025205 , + d. jarecke , p. maris and p. c. tandy , 2003 , phys .
c * 67 * 035202
. y. jiang , g. l. wang t. h. wang and w. l. ju , 2013 , @xmath227a * 28 * 1350145 . z. h. wang , y. zhang , t. h. wang , y. jiang and g. l. wang , arxiv : 1605 .
09091 , [ hep - ph ] . s. okubo , 1963 , @xmath228 . * 5 * 165 .
j. iizuka , 1966 , @xmath229 * 21 * 37
. g. zweig , 1964 , cern report no .
th 401 and th 412 .
l. micu , 1969 , @xmath230b * 10 * 521 .
a. le yaouanc , l. oliver , o. pene and j. c. raynal , 1973 , @xmath217d * 8 * 2223 .
e. s. ackleh , t. barnes and e. s. swanson , 1996 , @xmath217d * 54 * 6811 .
z. f. sun and x. liu , 2009 , @xmath217d * 80 * 074037 .
x. liu , z. g. luo and z. f. sun , 2010 , @xmath214 * 104 * 122001 .
t. barnes , s. godfrey and e. s. swanson , 2005 , @xmath217d * 72 * 054026 .
j. segovia , d. r. entem and f. fernandez , 2012 , @xmath218b * 715 * 322 . h. f. fu , x. j. chen , g. l. wang and t. h. wang , 2012 , @xmath227a * 27 * 1250027 .
salpeter and h.a .
bethe , 1951 , @xmath219 * 84 * 1232 .
salpeter , 1952 , @xmath219 * 87 * 328 . c. s. kim and g. l. wang , 2004 , @xmath218b * 584 * 285 . c. h. chang and g. l. wang , 2010 , @xmath231 g * 53 * 2005 .
g. l. wang , 2006 , @xmath218b * 633 * 492 .
g. l. wang , 2007 , @xmath218b * 650 * 15 .
z. h. wang , g. l. wang , h. f. fu and y. jiang , 2012 , @xmath220 b * 706 * 389 . | the new mesons @xmath0 and @xmath1 have been found by belle collaboration in the processes @xmath2 . considering @xmath0 and @xmath1 as @xmath3 and @xmath4 states , the two - body open charm ozi - allowed strong decay of @xmath3 and @xmath4
are studied by the improved bethe - salpeter method combine with the @xmath5 model .
the strong decay width of @xmath3 is @xmath6 mev , which is close to the result of @xmath0 measured by belle , therefore , @xmath3 is a good candidate of @xmath0 .
the strong decay width of @xmath4 is @xmath7 mev , considering the errors of the results , it s close to the lower limit of the total width of @xmath1 .
however the branching ratios @xmath8 of @xmath4 is larger than the experimental data of @xmath1 .
so taking @xmath4 as the candidate of @xmath1 , the precision measurement of the total decay width and the ratio between @xmath9 and @xmath10 is needed in future experiments .
* keywords : * @xmath0 ; @xmath1 ; strong decay ; improved bethe - salpeter method . |
in the fast growing wireless market , the frequency spectrum is one of the most scarce and valuable resources .
however , some surveys for actual measurements show that most of the allocated spectrum is largely under - utilized @xcite . cognitive radio ( cr ) is known as an efficient way to improve spectrum utilization and a promising technology to enable dynamic spectrum access by exploiting the unused spectrum in the wireless environments @xcite . in a cognitive radio network ( crn ) , there are two types of users , namely , licensed primary users ( pus ) who have the licensed spectrum access opportunities and unlicensed secondary users ( sus ) who can only utilize the spectrum which the pus do not occupy . to exploit limited spectrum efficiently , in a crn
, the sus are allowed to opportunistically to access licensed spectrum bands when pu transmission is not presented , which is able to significantly improve the spectrum utilization efficiency . during last decades
, the technology of cr has attracted plenty of interests and has been investigated extensively , among which , how to explore the access opportunities and regular the spectrum access without harming the pus are of consistent research interests @xcite .
recently , there is a particular interest to apply the queueing theory @xcite which is a natural tool to analyze the transmission in the wireless networks or the game theory @xcite which is commonly used in developing optimization algorithms to address the spectrum access problems .
intriguingly , applying queue theory with pricing strategies also brings a novel view on understanding the su s behavior and cognitive market policies , which can be traced from the original works of @xcite-@xcite that study the equilibrium behavior in queueing systems .
we call it the queueing game in the following to highlight the key issues in this area : by utilizing the information of queue and considering its own payoff , the user needs to make the decision on whether to queue or not .
recent works of @xcite-@xcite have leveraged the queueing game to the spectrum access control in the crn . in @xcite ,
the authors present an observed queue model with sus acquiring for transmission from a cognitive base station ( cbs ) .
it is also assumed that pus emerged periodically for transmission opportunities . during the pu s transmission period
, cbs stops serving sus and sus remained in the queue waiting for the re - operation of cbs .
based on this model , a su s decision strategy , i.e. , joining the queue or balk , is investigated and the pricing policy was studied .
authors of @xcite extend the work of @xcite to an unobserved case and analyzed the strategies of sus and cbs . in @xcite , the strategy of delay sensitive sus
is explicitly considered .
the authors also consider pricing and load balancing effect in the spectrum access decision - making in both monopoly and duopoly markets .
authors of @xcite also utilizes the concept of the queueing game and model users as selfish players that compete with each other by choosing the optimal transmission threshold maximizing system throughput .
it can be well observed that the queueing game is an effective tool for analyzing the sus behaviors in spectrum access .
the motivation of this work is to extend the previous works and overcome their limitations .
it can be found that the previous works focused on the case that a separate cbs is used for serving the sus .
therefore , when the pu emerges , the sus can remain in the queue of the cbs and wait for pus finishing the transmission . in such case ,
the queue of the cbs can be modeled and analyzed by a server breakdown model .
however , the installation and deployment of the cbs bring additional cost and it may not be practical in some cognitive systems .
in contrast , we consider there is only one bs in the system . the primary job of the bs is to serve the incoming pus , while it will utilize the time when the pu frequency band is not occupied to serve sus . specially , when the pu comes , the bs no long holds the information of the queue and all the sus are forced to leave the queue .
when a su decides to join the queue and wait for spectrum access opportunity , the sojourn time induces a cost and if its job has been finished , the su can receive a reward .
in addition , if pu emerges , the sus who have to leave the queue without any reward or compensate . under this model , we study the su s decision - making process , i.e. , whether to join for spectrum access or balk , and the pricing based spectrum access control . compared to previous works ,
the main contributions of this paper are as follows .
we first model the interaction among sus in a partially observed queue as a noncooperative game .
based on the queueing analysis and payoff model , we then analyze both individually equilibrium strategy and optimal social welfare strategy of the sus about whether to join or not .
furthermore , we study the bs pricing strategy for the system such that the individually equilibrium decision of sus can coincide with the socially optimal strategy that optimizes the total welfare of the whole system .
our presented analysis and algorithm are demonstrated by the simulation studies .
the rest of this paper is organized as follows .
section describes the system model including the queue model and profit of sus .
we present the individual equilibrium strategies , social welfare optimization and pricing studies in section .
our presented algorithm and analysis are demonstrated in section through simulation studies , and finally we conclude this work in section .
the system model can be found in fig . [ fig:1 ] .
we consider a crn consists of multiple sus , one bs and multiple pus .
when pus are absent , the bs can utilize the available spectrum to serve the sus .
if the pu is entering the system , due to its priority , the bs has to drop the service connection of the sus and starts to serve pu .
meanwhile , the queue consisting of the sus is no long exists and all sus should leave the queue and seek for other transmission opportunities .
we consider the data arrival rate of the su follows the poisson process at rate @xmath0 and the arrival of the pu follows poisson process at rate @xmath1 .
the service requirements of the sus are i.i.d with exponential distribution @xmath2 .
the first - come - first - served ( fcfs ) rule is applied for determining the service order of sus at the bs .
the length of the service time of the pu is also assumed to be exponentially distributed with rate @xmath3 .
the above considered queue model is a common assumption in some previous literatures , e.g. @xcite @xcite .
the state of the system at time @xmath4 is represented by a pair @xmath5 , where @xmath6 is the length of queue , i.e. , the number of sus in the system .
@xmath7 denotes the working status of the bs , with @xmath8 standing for serving sus and @xmath9 showing that the bs is serving the pus .
so based on the system model , when @xmath10 , we also have @xmath11 .
we assume when being successfully served by the bs , the su can receive a reward .
while waiting in the queue , the delay cost of the su is a function of its sojourn time .
based on the total payoff which is the difference between the reward and the cost , su can make the decision on whether to join the queue and waiting for spectrum access or not .
the payoff models are presented in the following .
we assume when being successfully served by the bs , the su receives a reward @xmath12 .
we also assume that the cost of staying in the queue is @xmath13 where @xmath14 is the sojourn time in the system representing the time that su stays in the queue .
@xmath13 should be an increasing function of @xmath14 and one simple linear example is that @xmath15 where @xmath16 is the unit cost . then the profit of the su can be given as @xmath17 accordingly , we can have the definition of the expected individual profit when there are @xmath18 sus in the queue , the expected profit of the arriving su can be defined as @xmath19 where @xmath20 is the probability that the su can be served and @xmath21 is the expected sojourn time related to @xmath18 .
moreover , we can also define the social welfare as follows .
the social welfare of the considered system is given by @xmath22 where @xmath23 is the fraction of sus that join the queue and leave after being served .
@xmath24 is the mean number of sus . in the following ,
we consider the sus are risk natural and try to maximize their profits .
we also assume the sus are identical i.e. , a symmetric game is assumed .
in addition , we also consider that when the su enters the system at time @xmath4 , all the system parameters including profit model are known except the queue length @xmath6 and su can only observe @xmath7 upon arrival , i.e. , a partially observed queue is considered .
after entering the system , the su is able to decide whether to join the queue waiting for spectrum access or not based on its own observation , i.e. the profit .
thus , there are only two pure strategies , join and balk , and a mixed strategy is specified by the joining probability @xmath25 $ ] of a su that finds the bs is not serving pu .
therefore , the effective arrival rate is @xmath26 . for the considered system
, we first obtain the stationary probabilities @xmath27 .
based on the transition rate diagram in fig .
[ fig:2 ] , we are able to obtain the stationary probabilities @xmath27 based on the following balance equations , @xmath28 @xmath29 @xmath30 @xmath31 to obtain @xmath32 and @xmath33 , we have following observation
. the stationary probability @xmath32 and @xmath33 can be given as follows , @xmath34 where @xmath35 is @xmath36 the proof is given in appendix a. in order to achieve the expected profit , from ( [ eq : ip ] ) , @xmath20 and @xmath21 should be obtained at the first place .
consider a su enters the network at state @xmath37 upon arrival and decides to join the queue .
this su may leave either due to service completion or due to a pu entering . for its service completion
, the su has to wait for a sum of @xmath38 independent exponentially distributed times with parameter @xmath2 . for the case that the pu enters
, the su has to wait for an exponentially distributed time with parameter @xmath1 .
hence , the sojourn time of the su in the system is given as @xmath39 , where @xmath40 follows a gamma distribution with parameters @xmath41 and @xmath2 .
@xmath42 is an exponentially distributed random variable with rate @xmath1 , and @xmath42 is independent of @xmath40 .
therefore , we have @xmath43 . to this end , when considering @xmath6 and @xmath7 are known to the su , we can obtain the @xmath20 and @xmath21 in ( [ eq : ip ] ) as @xmath44 where @xmath18 is the number of sus in the queue .
then , we can use ( [ eq : sp1 ] ) and ( [ eq : a1 ] ) to address @xmath45 in ( [ eq : ip ] ) and then find the expected profit of a su that enters the queue with a certain probability . with the results of the stationary probability and the expected sojourn time
, the expected profit of a su that enters the queue with probability @xmath46 can be achieved as following , when there is frequency bandwidth available for the sus , the expected profit of a su that enters the queue with probability @xmath46 given that other sus join the queue with probability @xmath47 is given by @xmath48\ ] ] the proof is given in appendix b. we can now process to find out the individually equilibrium ( ie ) strategy of a su and we have the following , in the considered model , a unique ie mixed strategy exists , with joining probability @xmath49 given by @xmath50 where @xmath51 and @xmath52 .
the proof is given in appendix c. from * theorem 1 * , we can observe that the ie is independent of the pu s service time .
this is because the su makes decision only when pu is not being served and the su can not be aware of the pu s information .
however , the socially equilibrium strategies of the sus do depend on the pricing strategies of the bs as well as the pu s arrival rate .
the ie does not imply the social welfare optimality .
thus , in the following , we investigate the social optimal strategy .
based on the definition of social welfare and previous results of the stationary probabilities , we can arrive at the following proposition , the expected social profit , given that the sus follow a mixed strategy with probability @xmath47 of joining ( i.e. arriving sus that find an bs not serving the pu enters with probability @xmath47 , while the rest choose to balk without being served ) is given by @xmath53}{(\xi + \eta)(1-x(\lambda^{'}))}.\ ] ] to address the social welfare ( [ eq : spso ] ) , we need to find @xmath24 and @xmath54 . using ( [ eq : sp1 ] ) and ( [ eq : a1 ] ) , we can obtain @xmath54 and @xmath24 as @xmath55 then , obtaining the geometric sums in ( [ eq : c1 ] ) and through ( [ eq : spso ] ) we can arrive ( [ eq : profitso ] ) . with the observation in ( [ eq : profitso ] ) , we can obtain the socially optimal strategy , which can be found in the following theorem . in the considered model
, a unique socially optimal strategy exists with probability @xmath56 of joining the queue which can be expressed as @xmath57 where @xmath58 and @xmath59 . the proof is given in appendix d. we have obtained the socially optimal strategy as well as the individual equilibrium strategy .
moreover , it can be observed that the socially optimal joining probability @xmath60 is always smaller than the individual one @xmath49 , which can also be found in fig .
[ fig:3 ] . to oblige the sus to adopt the socially optimal strategy ,
one approach is to apply a pricing mechanism to reduce the individually optimal threshold @xmath49 @xcite . in this work
, we consider that the bs will act as an agent to impose an admission fee , which is a constant given the arrival rate , service pattern , reward and cost .
the admission fee is to force the individually equilibrium probability to equal with the social optimal one .
when the admission fee is considered , the expected profit of a su is given as @xmath61 .
it can be observed that when imposing an admission fee , the social profit remains the same as it implies a transfer of income from one group to another .
thus , through [ eq : a2 ] , we can obtain @xmath62 .
we further denote that the equilibrium probability of joining by @xmath63 .
then the optimal fee @xmath64 should satisfy @xmath65 . as the monopoly market is considered here , and
a monpoly does not allow the a positive user surplus since in such a case , the price can be increased without reducing @xmath47 .
therefore , the @xmath66 can be arrived by @xmath67
the numerical parameters are @xmath68 . first , in fig .
[ fig:3 ] , we can see that the socially optimal joining probability @xmath60 is always smaller than the individual one @xmath49 , and inherently , there is a gap between the individually equilibrium arrival rate and social arrival rate as the equilibrium arrival rate and social arrival rate can be given as @xmath69 .
therefore , we can see that by imposing appropriate admission fee , the arrival rate of sus can be regulated . in fig .
[ fig:4 ] , we can see that the social benefit when @xmath60 is used is higher than the one when @xmath49 is considered
. comparing fig .
[ fig:3 ] and fig .
[ fig:4 ] , we can see that only when @xmath70 , the social benefit is higher than @xmath9 , otherwise it remains lower or equal to @xmath9 .
meanwhile , the social benefits when @xmath60 is considered is always above @xmath9 and comparable high , which indicates that appropriate admission fee can improve the social benefits of the considered system .
in this paper , we have investigated the problem of spectrum access decision - making for the sus in the crns . utilizing the queueing game approaches , the decision - making process of the sus that whether to queue or not is studied in present of arrival of the pus .
both individual equilibrium and social optimization strategies are derived .
moreover , the optimal pricing of the service provider is also investigated .
our proposed algorithms and corresponding analysis are validated through simulation studies .
from ( [ eq : sp1])-([eq : sp1.3 ] ) , we can observe @xmath71 . to obtain @xmath33 , the similar approach used in @xcite can be applied .
one can notice that ( [ eq : sp1.2 ] ) can be considered as a homogeneous linear difference equation of order @xmath72 , which is with constant coefficients and characteristic equation @xmath73 ( [ eq : o1 ] ) have too roots , @xmath74 and @xmath75 , which are @xmath76 from the standard theory of homogeneous linear difference equations ( see e.g. ( * ? ? ? * sec 2.3 ) ) , we conclude that @xmath77 where @xmath78 and @xmath79 are constants .
we can easily see that @xmath80 , thus , @xmath78 should be necessarily @xmath9 .
the constant @xmath79 can be calculated using ( [ eq : sp1.3 ] ) and * proposition 1 * can be approved .
to prove proposition 2 , first we have the expected profit when a su is able to observe @xmath18 sus and queue state in the system upon arrival ( i.e. observed queue ) and decide to enter , which can be derived from ( [ eq : a1 ] ) , @xmath81 when a su decides to join given that the bs is serving sus and others are following strategy @xmath47 , the expected profit is @xmath82 where @xmath83 is the probability that there are @xmath18 sus in the queue when a su arrives . as we have that @xmath84 , ( [ eq : profitexpected ] ) can be arrived and * proposition 2 * can be proved .
for a su , it will prefer to enter the queue if its expected profit after entering @xmath85 , and balk if it @xmath86 . considering @xmath87 , we can solve for @xmath75 and the unique solution is @xmath88 .
we notice that @xmath89 is one root of ( [ eq : o1 ] )
. then we are able to solve @xmath90 with respect to @xmath47 , which yields @xmath91}{\lambda(1-x)}= \frac{(\mu \varpi_s -c)(c + \xi \varpi_s)}{\lambda c \varpi_s } .\ ] ] we can also observe that @xmath75 is a strictly increasing function for @xmath25 $ ] as @xmath92 .
thus , @xmath93 iff @xmath94 , where @xmath52 . in other word , @xmath49 is in the interval @xmath95 iff @xmath96 .
moreover , it can be found that when @xmath97 , @xmath98 keeps positive .
thus , the su s best response is to join . in this case , `` join '' is the unique individual equilibrium . on the other hand ,
when @xmath99 , @xmath98 becomes negative . in this case , `` balk '' is the individuals equilibrium strategy .
therefore , * theorem 1 * can be proved .
it can be noticed that ( [ eq : profitso ] ) can be reformed as @xmath100 , where @xmath101}{(\xi + \eta ) ( 1-x)}$ ] .
as we can see , @xmath102 can be deduced to @xmath103 , which means that we need to solve @xmath104 , that is we can see that the discriminant of the quadratic polynomial in ( [ eq : d1 ] ) is less or equal to @xmath9 iff @xmath106 .
we can deduce that @xmath107 is increasing and so as @xmath108 .
( [ eq : d1 ] ) has two roots , which are @xmath109 * when @xmath111 , then we have @xmath112
. then s(q ) is decreasing in @xmath113 $ ] and the social optimal is @xmath114 .
we can also see that @xmath115 implies that @xmath116 .
* when @xmath117 , then we can see that the maximum @xmath108 is obtained for @xmath47 such that @xmath118 . by using ( [ eq :
o1 ] ) , substituting @xmath119 for @xmath120 and solving for @xmath47 , we obtain @xmath121 .
it can be observed that @xmath117 means @xmath122 . * similarly , when @xmath123 we can have the social optimal joining probablity is @xmath124 and @xmath125 .
shiang and m. van der schaar , `` queuing - based dynamic channel selection for heterogeneous multimedia applications over cognitive radio networks , '' _ ieee trans .
multimedia _ , vol .
896 - 909 , aug . 2008 .
l. yang , h. kim , j. zhang , m. chiang , and c. w. tan , `` pricing - based decentralized spectrum access control in cognitive radio networks , '' _ ieee / acm trans .
networking _ ,
2 , pp . 522 - 535 , apr .
2013 .
h. li and z. han , `` socially optimal queuing control in cr networks subject to service interruptions : to queue or not to queue ? '' _ ieee trans .
wireless commun .
5 , pp . 1656 - 1666 , may 2011 . c. t.
do , n. h. tran , m. v. nguyen , c. s. hong , and s. lee , `` social optimization strategy in unobserved queueing systems in cognitive radio networks , '' _ ieee commun .
1944 - 1947 , dec . 2012 .
n. h. tran , c. s. hong , s. lee , and z. han , `` optimal pricing effect on equilibrium behaviors of delay - sensitive users in cognitive radio networks , '' _ ieee j. sel .
areas commun .
_ , vol . 31 , no . 11 , pp . 2566 - 2579 , nov . | in this paper , we investigate the problem of spectrum access decision - making for the secondary users ( sus ) in the cognitive radio networks .
when the primary users ( pus ) are absent on certain frequency bandwidth , sus can formulate a queue and wait for the base station ( bs ) to serve .
the queue of the sus will be dismissed if the pu is emerging in the system . leveraging the queueing game approaches , the decision - making process of the sus that whether to queue or not is studied .
both individual equilibrium and social optimization strategies are derived analytically . moreover
, the optimal pricing strategy of the service provider is investigated as well .
our proposed algorithms and corresponding analysis are validated through simulation studies .
queueing game ; pricing ; nash equilibrium strategies ; social optimizations ; spectrum access ; cognitive radio |
in recent years , we have experienced a renaissance of deep radio surveys , usually associated with multi - waveband observational programmes mainly designed to probe the formation and evolution of galaxies and super massive black holes ( smbh ) .
these surveys , mostly carried out at 1.4 ghz , have enabled the study of sub - mjy and @xmath1jy radio source populations ( hereafter @xmath1jy population ) and in particular the determination , with robust statistics , of the radio source counts down to a few tens of @xmath1jy ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . while it is clear that the @xmath1jy population is a mixture of different classes of objects ( radio loud and radio quiet agn , starburst galaxies , spirals ) , the exact contribution of these different classes to the radio source counts is still not very well established and often dependent on the optical limit of the spectroscopic follow - up @xcite .
the role played by the cosmological evolution of the different classes of objects is even more uncertain .
recently , @xcite have developed a new classification method to separate star forming galaxies and agn based on photometric rest - frame colors for local galaxy samples .
the application of this classification method to the vla - cosmos sources @xcite shows that the radio population above @xmath6 50 @xmath1jy is a mixture of roughly 3040% of star forming galaxies and 5060% of agn galaxies , with a minor contribution ( @xmath6 10% ) of qsos .
the international cosmos ( cosmic evolution ) survey @xcite is designed to probe the correlated evolution of galaxies , star formation , agn up to high redshift .
the survey includes state of the art imaging data covering the entire wavelength range from the x - rays to the radio domain @xcite supported by extensive optical spectroscopic campaigns using mainly the vlt / vimos and the magellan / imacs instruments @xcite .
the radio observations of the cosmos field were done with the vla in a and c configuration and are described in schinnerer et al .
( 2004 , 2007 ) .
the whole 2 deg@xmath0 field was imaged in mosaic mode with a resolution of 1.5 arcsec .
the sensitivity is uniform in the inner 1 deg@xmath0 with an average rms value of 10.5 @xmath1jy / beam and it increases almost regularly to the outer regions of the 2 deg@xmath0 field . from the mosaic a catalog of radio sources , the vla - cosmos large project catalog , was extracted .
the catalog lists @xmath3 radio sources selected above the local @xmath7 threshold over an area of 2 deg@xmath0 .
a full description of the observations , data reduction and catalog extraction is given in schinnerer et al .
( 2007 , hereafter s07 ) . in this paper we present a further analysis on the vla - cosmos large project catalog aimed to quantify the impact of bandwidth smearing ( section 2 ) and the survey completeness ( section 3 ) .
finally , we derive the radio source counts and compare them with those derived from similar surveys ( section 4 ) .
a more detailed interpretation of the radio data is left to other papers which are currently in preparation .
in particular , the optical identifications of the vla - cosmos radio sources are presented by @xcite , the fraction of star - forming galaxies and agn in the sub - mjy population of the vla - cosmos field is discussed in @xcite and the dust un - biased cosmic star formation history is derived from the 1.4 ghz radio data in @xcite .
bandwidth smearing ( or chromatic aberration ) affects all synthesis observations made with a finite width of the receiver channels .
imaging sources at large distances from the phase center can yield to radial smearing , increasing the source size and reducing the peak flux density , while the integrated flux density is conserved .
the image smearing is proportional to the fractional bandwidth ( the bandwidth divided by the central frequency of observation ) and to the distance from the phase center in units of the synthesized beam ( e.g. * ? ? ?
* ) . for single pointing observation the bandwidth smearing simply increases with the radial distance from the phase centre . for a mosaic image , produced by multi - pointing observations ,
the contribution from all the pointings has to be taken into account and the resulting radial pattern will be much more complicated depending also on the adopted spacing pattern of the individual pointings . in order to quantify the effect of bandwidth smearing on the vla - cosmos observations we ran the same procedure that produced the radio catalog on each of the 23 individual pointings ( see s07 ) . for the strongest sources we compared their peak and total flux densities in the final mosaic with the corresponding peak and total flux densities in the pointing where the sources are closest to the field center .
the total flux density of each source in the mosaic and in the individual pointings are in very good agreement ( the median value of the ratio is 1.03 with an r.m.s .
dispersion of 0.04 ) confirming that the total flux density is properly recovered . on the other hand ,
as expected for observations affected by bandwidth smearing , the peak flux densities are underestimated in the final mosaic with respect to the peak fluxes in the individual pointings where the sources are closest to the field center . to quantify this effect we selected only sources which are within a radius @xmath8 from the field center in , at least , one individual pointing .
the peak flux density of a source within @xmath9 from the field center is almost unaffected ( less than 5% for the vla - cosmos observations ) by bandwidth smearing and can be compared with the peak flux in the mosaic . the image of the source in the mosaic is the combination of all the images in the individual pointings where the source is within the cut - off radius ( @xmath10 , see s07 ) . in figure [ bw_fig1 ]
we show the peak flux ratios for compact sources ( angular sizes less than 4@xmath11 ) with peak fluxes brighter than 0.2 mjy / beam and @xmath8 as a function of the radial distance from the center in the final mosaic .
the ratio is about @xmath12 for sources in the inner region of the mosaic , within 30@xmath13 from the mosaic center . in the region between 30@xmath13 and 45@xmath13
, the ratio is @xmath14 , implying a smaller correction in this area .
this is due to the fact that the number of overlapping pointings is smaller at larger radii ( see fig . [ pointings ] ) . within these two regions the effect appears to be independent from the position in the mosaic considering the dispersion of the points .
therefore , we adopted a constant bandwidth smearing correction factor of 0.8 for the sources within @xmath15 from the mosaic center and a constant value of 0.9 for those between @xmath15 and @xmath16 .
we performed this analysis in the inner 1 deg@xmath0 ( @xmath17 deg .
square region ) of the final mosaic since this region has the deepest and most uniform sensitivity and it will be used to derive the source counts in the following section . beyond @xmath16 from the mosaic center
this correction is no more valid as each source in the mosaic receives a contribution from only one pointing and therefore the bandwidth smearing correction will be a function of the distance from the center of the individual pointing where each source is located , rather than of the distance from the mosaic center .
consequently , to derive the radio source counts in the inner 1 deg@xmath0 we divided all the peak flux densities by 0.8 or 0.9 depending on the distance from the center of the mosaic and derived the new resolved / unresolved criterion following the procedure described in s07 . since the ratio between total and peak fluxes is a direct measure of the extent of a radio source , we used it to discriminate between resolved or extended sources ( i.e. larger than the beam ) and unresolved sources .
figure [ res_unres ] shows the ratio of the total integrated flux density @xmath18 and the corrected peak flux density @xmath19 as function of the signal to noise ratio . to select the resolved sources we have fitted a lower envelope in fig .
[ res_unres ] which contains @xmath20 of the sources with @xmath21 and mirrored it above the @xmath22 line .
the criterion of 95% of sources contained in the lower envelope of fig .
[ res_unres ] is slightly different from that used in s07 , where a value of 99% was adopted .
we reckon that the value adopted here is the best compromise to properly classify the sources as resolved or unresolved because it fully characterises the shape of the distribution of sources with @xmath21 and , at the same time , it excludes the most extreme values .
this envelope is described by the equation @xmath23\ ] ] it is important to note that in fig .
[ res_unres ] there is no systematic offset from the @xmath22 line at high snrs .
this is an a posteriori confirmation that we adopted an appropriate bandwith smearing correction .
considering only sources with snr@xmath24 in the inner 1 deg@xmath0 , we obtain 484 resolved sources , for which the total flux is given by the total flux of the gaussian fitting , and 1208 unresolved ones , for which the total flux is set equal to the peak flux .
it is worth noting a few points regarding the effects of the correction on the vla - cosmos large project catalog : * the correction has been determined only out to a radius of 45 arcmin from the mosaic center and should be safely adopted only in this region . *
the catalog is still selected on the old , pre - bandwidth smearing correction , peak flux densities or signal - to - noise ratio , and we applied the bandwidth smearing correction only to the sources already in the catalog . in the remaining we will refer to the signal - to - noise ratio as the original one , calculated with the uncorrected peak flux density . *
the correction affects the resolved / unresolved classification of a source and therefore also the total flux density and size for the sources , previously classified as resolved , that are found unresolved after correcting for the smearing .
this affects about 500 sources , almost one third of the total number of sources in the inner 1 deg@xmath0 with snr @xmath24 .
the vla - cosmos large project catalog contains @xmath3 radio sources extracted from a 2 deg@xmath0 area .
as shown in s07 , the noise is not uniform over the whole area and increases in the outer regions .
the central 1 deg@xmath0 area has the lowest and most uniform noise ( mean rms 10.5 @xmath1jy ) and therefore it is very well - suited to derive the source counts .
extending the analysis to the whole area and catalog would increase the number of objects at higher flux densities ( @xmath25 mjy ) , but it would not add much to the accuracy of the source counts at the lowest flux densities , which is the main focus of this paper . on the contrary , the correction factors that would need to be applied to take into account the loss of the faintest sources in the outer and noiser region would add more systematic uncertainties at the faint end of the source counts .
furthermore , in our following analysis we consider only sources with signal to noise ratio greater than or equal to 5 ( the vla - cosmos large project catalog has been selected with a signal to noise ratio threshold of 4.5 ) and without a flag for possible sidelobe residuals , in order to minimize contamination from spurious sources . given these constraints
, we have a catalog of 1692 sources with snr @xmath26 within the inner 1 deg@xmath0 region . the visibility area of the inner 1 deg@xmath0 of the vla - cosmos survey as a function of the rms noise
is shown in fig .
[ visi_fig ] .
this is the area over which a source with given peak flux density ( 5 times the noise in our case ) can be detected .
since the noise in this region is very smooth the visibility area increases very rapidly and becomes flat above an r.m.s .
noise of @xmath27 @xmath1jy / beam .
the same curve , computed over the entire 2 deg@xmath0 , becomes flat only above an r.m.s .
noise of @xmath28 @xmath1jy / beam ( see fig . 13 in s07 ) .
this plot has been derived using the original peak flux densities and not those corrected for the bandwidth smearing since the catalog is still selected on the basis of the uncorrected peak flux . besides the visibility area of each source ,
other correction factors need to be considered to estimate the completeness of the radio catalog and to derive the intrinsic source counts : the errors introduced by the fitting routines , the noise bias , and the resolution bias can all be modelled using simulated samples of radio sources .
in particular , the resolution bias can be quite important since the beam of the vla - cosmos observations is @xmath29 ( e. g. about 15 times less in area than for similar surveys carried out at the same frequency ) . since the completeness of the radio catalog
is defined in terms of the peak flux , while the source counts are derived as a function of the total flux density , corrections need to be applied to take into account resolved sources with peak fluxes below the catalog threshold and total flux densities above the nominal limit of the survey .
our strategy to estimate the correction factor has been to simulate samples of radio sources , with a realistic flux density and angular size distributions , to insert these sources in the mosaic image and to recover these mock sources using the same technique used to produce the catalog . from the comparison between
the number of sources added to the image and those effectively recovered we can estimate the correction factors for different flux density bins .
the correct choice of the flux density distribution is rather straightforward , as a broken power law derived from the observed distribution is a reasonable assumption , consistent with the extrapolation from shallower surveys .
the choice of the angular size distribution for the simulated radio sources is more uncertain .
while there is good evidence that the median of the intrinsic source size at 1 mjy is about @xmath30 and decreases with decreasing flux density at sub - mjy level , it is not known how exactly this happens and what is the best analytical relationship to model this behaviour . as we will show later
, this is an important parameter in deriving the completeness of the vla - cosmos catalog . in the past different scaling relationships between the median angular size , @xmath31 , and the total flux density , @xmath32 ,
have been used , e.g. @xmath33 with @xmath34 @xcite or @xmath35 @xcite .
more recently , new surveys have provided some results on the angular size distributions of the @xmath1jy radio sources .
@xcite obtained a complete sample of 289 radio sources with flux density @xmath36 40 @xmath1jy and found that 64% of objects are unresolved with sizes less than @xmath37 arcsec .
merlin observations at higher angular resolution ( @xmath38 ) provided consistent results for a sample of 92 radio sources , 80 of which with flux density between 40 @xmath1jy and 200 @xmath1jy @xcite .
the angular size distribution derived from the latter sample of objects , which are all resolved , shows a dominant narrow gaussian component centered at @xmath39.7 and a tail of sources ( @xmath40 of the sample ) with angular sizes between 2@xmath41 and 4@xmath41 .
we derived the median of the angular size distribution for the sources observed by muxlow et al .
( 2005 ) in the flux density range 4090 @xmath1jy .
the 55 sources in this flux density range have a median angular size of @xmath42 ( with a r.m.s .
dispersion of [email protected] ) and a median flux density of 54 @xmath1jy . for comparison
, @xcite found that the median angular size for sources in the flux density range 0.4 - 1.0 mjy is @xmath43 with a median flux density of 0.56 mjy .
the value of @xmath44 consistent with these median angular sizes at different flux densities is @xmath45 .
it is worth noting that this value for @xmath44 has to be considered as a lower limit since the angular sizes given by @xcite are the largest angular sizes while the angular sizes given in @xcite are the fwhm .
these results provide strong evidences that radio sources with flux density around 100 @xmath1jy have typically sub - arcsecond sizes even if the quality of the data ( e.g. large number of upper limits in the angular sizes derived in bondi et al .
2003 ) and the statistics ( the above result on @xmath44 is derived from less than 100 sources ) do not allow one to derive a firm estimate of @xmath44 .
given these uncertainties , we decided to use a complementary method to derive the intrinsic angular size distribution , simulating mock samples of radio sources with different angular size distributions following a general power law @xmath46 with the exponent @xmath44 ranging from 0.2 to 0.5 , and normalized to the integral angular size distribution derived from the vvds survey in @xcite .
we constructed a simulated sample of radio sources containing about 3100 radio sources down to a total flux density level of 30 @xmath1jy .
this allows us to count also sources with flux density below the limit which , because of positive noise fluctuations , might have a measured peak flux density above the threshold .
then , all the simulated sources were randomly injected in the inner 1 deg@xmath0 of the vla - cosmos mosaic and subsequently recovered using the same procedures adopted for the real sources and binned in flux density intervals .
this was repeated for 3 different samples , and for four different angular size distributions ( with @xmath44 equal to 0.2 , 0.3 , 0.4 , and 0.5 ) yielding a total of 12 simulated samples with more than 35,000 simulated radio sources . finally , from the comparison between the number of simulated sources detected in each bin and the number of sources in the simulated input sample in the same flux density bin we derived the correction factors for each value of the exponent @xmath44 .
these correction factors , listed in table [ corr_5sigma ] , account for the visibility area , noise bias , fitting errors and resolution bias . from table [ corr_5sigma ] it is evident that for the brightest sources the correction factors are @xmath47 and the differences among the correction factors for the various angular size distributions are negligible . on the other hand , below @xmath48 mjy
the correction factors become quite large with significant differences for the various angular size distributions .
therefore , it is important to infer the intrinsic angular size distribution in order to reconstruct the properly corrected source counts .
another interesting feature of the correction factors which is worth noting is that the maximum is not at the lowest flux density bin but around 0.12 mjy .
this is due to a combination of the resolution bias and the capability to classify a radio source as resolved , which depends on the signal - to - noise ratio .
a source with a given total flux density and size will be catalogued in one of the following three classes : 1 ) undetected , if the peak flux is below the sample threshold , 2 ) detected with the correct total flux density if the signal to noise ratio is high enough to allow for proper deconvolution , or , 3 ) detected with a lower total flux density if the signal to noise ratio is not high enough to recognize the source as resolved .
the sources in the last class are assigned a total flux density lower than the intrinsic one and equal to the peak flux . a significant fraction of the sources with intrinsic total flux densities @xmath49 mjy is redistributed to lower total flux densities since the signal - to - noise ratio is not sufficient to allow a proper deconvolution . to constrain the intrinsic angular size distribution we performed the following test .
we derived the median major axis of the simulated samples of radio sources for each of the tested power laws ( @xmath50 ) in the flux density range 0.25 - 0.4 mjy .
this flux density range was chosen as a compromise to minimise the number of unresolved sources ( and therefore upper limits on source size ) and allow for a statistically significant large sample of sources reaching as close as possible to the lowest fluxes .
we calculated both the median major axis of the original injected sources ( row 1 in tab .
[ median_sim ] ) and the median major axis of the simulated sources recovered by the fitting routine ( row 2 in tab .
[ median_sim ] ) .
these latter values were obtained considering also the upper limits derived for the sources fitted as unresolved ( @xmath51 15% of the total number of sources in this flux density range ) .
table [ median_sim ] shows that , in the flux density range 0.25 - 0.4 mjy , although the measured median major axis of the simulated sources is somewhat larger than that actually injected , the differences are not statistically significant for any given value of @xmath44 . on the other hand , different values of @xmath44 yield to significantly different median major axis .
therefore , the value of the median major axis in this flux density range for the real sources could be used to roughly discriminate between different intrinsic angular size distributions .
unfortunately , the measured sizes of the radio sources are affected by bandwidth smearing and they can not be used for such a test .
however , as we have seen in section 2 , the total flux densities are properly determined and the ratio between the total and the corrected peak flux densities can be used as an estimate of the area covered by the radio sources .
then we made the assumption that radio sources are extended in only one direction and we derived the major axis from the ratio between the total and the corrected peak flux densities .
this assumption can be justified since extended radio emission is generally associated to elongated features ( e.g. radio jets in agn or spiral arms in star - forming galaxies ) , and when this is not true the derived source sizes will be an upper limit of the real ones . in this way we obtained a median value of the major axis for the observed sources in the flux density range 0.25 - 0.4 mjy of @xmath52 . from the comparison with the values listed in tab .
[ median_sim ] , we conclude that the observed median major axis is more consistent with that obtained using simulated samples of radio sources following an angular size distribution with @xmath35 or @xmath45 . with both methods , the comparison between the angular size distribution obtained by @xcite and @xcite in different flux density intervals and the comparison between the estimate of the true angular sizes and those derived from the simulated samples
, we obtained consistent results which exclude values of @xmath53 and @xmath34 and suggest @xmath54 . in the following analysis
we apply the correction factors obtained for @xmath35 to the radio source counts .
we have derived the radio source counts at 1.4 ghz in the inner 1 deg@xmath0 region of the cosmos field from the vla - cosmos catalog , corrected for the effect of bandwidth smearing and with a signal - to - noise threshold of 5 .
the source counts up to @xmath55 mjy are presented in table [ counts_tab ] . for each flux density bin
we give the mean flux density , the observed number of sources , the corresponding differential source density d@xmath56/d@xmath32 ( in sr@xmath57jy@xmath57 ) , the normalised differential counts ( d@xmath56/d@xmath32)@xmath58 ( in sr@xmath57jy@xmath59 ) with their estimated poissonian errors , the incompleteness correction factors from table [ corr_5sigma ] for @xmath35 , and the cumulative number of effective sources after applying the corrections .
the correction factors above 0.2 mjy are @xmath60 .
such variations are roughly consistent with the poissonian error of the source counts and for this reason the source counts above 0.2 mjy are assigned a correction factor of 1 .
the mean flux density in each bin was calculated as the geometric mean of the flux density extrema . in fig .
[ counts1 ] and fig .
[ counts2 ] we show the raw and the corrected source counts derived from the vla - cosmos survey compared to those from other surveys at 1.4 ghz : the vla - vvds @xcite , the atca - hdf_s @xcite , the vla - hdf_n @xcite , the pdf survey @xcite , the ssa 13 field @xcite and the first survey @xcite .
the full range of source counts up to @xmath61 mjy is shown in fig .
[ counts1 ] , and a blow - up of the sub - mjy region is shown in fig.[counts2 ] . in both figures
the raw source counts from the vla - cosmos survey are shown with empty circles and those corrected for incompleteness with filled ones .
the solid line in fig .
[ counts1 ] is a linear least - squares sixth - order polynomial fit obtained using the vla - cosmos source counts above 0.06 mjy supplemented with the first source counts above 2.5 mjy and up to 1 jy .
the resulting polynomial fit is given by @xmath62=\sum_{i=0}^6 a_i[\log(s / mjy)]^i\ ] ] with @xmath63 , @xmath64 , @xmath65 , @xmath66 , @xmath67 , @xmath68 , and @xmath69 .
the residuals from the polynomial fit have an rms of about 0.06 in the logarithm of the normalized counts .
as discussed by @xcite the sixth - order polynomial is necessary to follow the different curvatures shown by the source counts . in fig .
[ counts1 ] we also show for comparison with a dashed line the sixth - order polynomial fit obtained by @xcite from the phoenix deep survey source counts .
the two fits are in reasonably good agreement with our fit slightly higher for flux density below 0.2 mjy .
it is evident that in some cases , the field - to - field differences in the radio source counts below 0.5 mjy are larger than the combined errors .
it is likely that these differences ( at least partly ) can be ascribed to instrumental defects or different recipes to account for the incompleteness . in particular ,
the assumption of the intrinsic angular size distribution of the @xmath1jy radio sources in high resolution radio surveys is certainly a factor , as we have shown in the previous section .
nevertheless , the most extreme cases ( e.g. the ssa 13 and hdf north fields ) should reflect real cosmic variance @xcite .
there is a particularly good agreement between the source counts derived from the vla - cosmos , after correction , and the hdf - s fields .
this is reassuring for the resolution bias correction we applied since the hdf - s field was observed with a resolution of about @xmath70 , and for this reason is much less affected by the precise shape of the intrinsic angular size distribution .
it is worth noting that while the vla - cosmos survey is not as deep as the ssa13 and hdf - n surveys , it is , by far , the one with the most robust statistics . at a flux density level of @xmath71 @xmath1jy
the vla - cosmos survey counts almost 400 sources while typical numbers for other surveys , in this flux density range , are between 50 and 90 sources .
a possible drop off in the radio source counts below @xmath72 @xmath1jy has been already detected and discussed by other surveys imaging various fields with different sensitivity and resolution ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
although , in every case , the drop off relies on the faintest bins it should not be immediately discounted as incompleteness in the source counts as it is accepted that the flat region of the normalised source counts below @xmath73 mjy can not continue indefinetely @xcite . from the vla - cosmos radio source counts there is some indication that the density of sources is beginning to decrease below 100 - 150 @xmath1jy producing a downturn in the observed radio source counts , even if the result is strongly dependent on the fidelity of the first flux density bin .
it is important to note that if incompleteness is responsible for the drop off observed in the first flux density bin of fig .
[ counts2 ] , it would mean that we are missing about 130 sources in the range 0.06 - 0.0735 mjy , which is extremely unlikely .
so , while incompleteness at the lowest flux density levels can still be present at some level , it can not explain by itself the drop off observed in the radio source counts .
thus , this decline can provide important limits when modeling the populations contributing to the radio source counts .
as we have explained in the previous section , the intrinsic angular size distribution adopted to derive the incompleteness correction factor has a large impact in this flux density range , but adopting distributions with @xmath74 would only enhance the drop off below @xmath75 mjy . also the uncertainties in the bandwidth smearing correction
could , in principle , significantly modify the radio source counts and the observed drop off observed at lower flux density .
we have tested this assuming two extreme cases , a constant correction over the whole area of 0.8 and 0.9 respectively .
as we can expect , the first case would produce a very small difference , well within the poissonian error boxes , since we already applied the 0.8 correction over almost 80% of the 1 deg@xmath0 . on the other hand , adopting a correction factor of 0.9 over the whole area , would significantly lower only the radio source counts in the first three bins , increasing the slope of the drop off region .
as we have shown in section 2 , the assumption of a constant correction over the whole 1 deg@xmath0 is not corroborated by the real data .
the data require a mixture of the two correction factors and with the previous test we have just verified that the drop off in the source counts below 0.15 mjy can not be produced by the applied smearing corrections .
we can conclude that given the large statistics and the fact that the incompleteness factors have been carefully investigated and applied , the decrease in the source counts below @xmath76 mjy observed in the vla - cosmos radio source counts should be considered real .
in this paper we have presented a further analysis of the vla - cosmos large project catalog with the goals of : 1 ) quantifying and correcting for the bandwidth smearing effect which affects the vla - cosmos radio mosaic and catalog ; 2 ) determining the incompleteness of the vla - cosmos radio catalog , through extensive use of monte - carlo simulations taking into account different intrinsic angular size distributions , in order to quantify the effects of noise bias and resolution bias ; 3 ) deriving the radio source counts at 1.4 ghz in the vla - cosmos field . the bandwidth smearing correction factors have been determined from the comparison of peak and total flux densities in the final mosaic and in individual pointings where the source is within @xmath9 from the center .
we have found that a two - value correction factor is a good approximation : out to a radius of @xmath15 from the center , the peak flux densities in the vla - cosmos catalog must be divided by a factor 0.8 , while in the region between @xmath15 and @xmath16 from the center by a factor 0.9 . inside these two regions
the correction factor is rather constant , within the dispersion of the data , with a sharp transition from 0.8 to 0.9 at a radial distance of @xmath15 reflecting the change in overlapping pointings .
consequently , a new classification in resolved and unresolved sources has been derived .
a number of sources , originally classified as resolved , turned out to be unresolved after the bandwidth smearing correction was applied . for these sources
also the total flux density was corrected .
we simulated several samples of radio sources with a flux density distribution compatible with the observed one and with different angular size distributions .
the angular size distribution was modelled using @xmath46 with @xmath77 .
these mock sources were added to the radio mosaic and recoverd using the same procedure that yielded to the radio catalog .
in such a way we were able to quantify the effects of the combined noise and resolution bias affecting the completeness of the radio catalog .
a comparison of the median angular size at a flux density level not significantly affected by the resolution bias for the simulated samples of sources and the observed one , strongly suggest a general distribution of radio source sizes following @xmath78 . finally , we derived the radio source counts in the vla - cosmos field ( both uncorrected and corrected for the incompleteness ) .
the source counts extend down to 60 @xmath1jy and are consistent with those derived by other surveys but with much more robust statistics .
in particular , the drop off of the source counts below @xmath72 @xmath1jy has to be considered real and not due to incompleteness .
the national radio astronomy observatory ( nrao ) is operated by associated universities , inc .
, under cooperative agreement with the national science foundation .
kj acknowledges support by the german dfg under grant schi 536/3 - 1 . | the vla - cosmos large project has imaged the 2 deg@xmath0 cosmos field with a resolution of 1.5 arcsec and a sensitivity of about 11 @xmath1jy ( 1@xmath2 ) yielding to a catalog of @xmath3 radio sources . in this paper
we present a further analysis of the vla - cosmos large project catalog of radio sources aimed to : 1 ) quantify and correct for the effect of bandwidth smearing in the catalog , 2 ) determine the incompleteness produced by the noise bias and the resolution bias in the new catalog and 3 ) derive the radio source counts at 1.4 ghz .
the effect of bandwidth smearing on the radio sources in the catalog was quantified comparing the peak and total flux densities in the final mosaic and in each of the individual pointings where the source was closest to the center of the field .
we find that the peak flux densities in the original vla - cosmos large project catalog have to be divided by a factor about 0.8 or 0.9 , depending on the distance from the mosaic center .
the completeness of the radio catalog has been tested using samples of simulated radio sources with different angular size distributions .
these simulated sources have been added to the radio image and recovered using the same techniques used to produce the radio catalog .
the fraction of missed sources as a function of the total flux density is a direct measure of the incompleteness . finally , we derived the radio source counts down to 60 @xmath1jy with unprecedented good statistics .
comparison to the findings of other surveys shows good agreement in the flux density range 0.06 - 1 mjy confirming the upturn at @xmath4 mjy and a possible decline of the source counts below @xmath5 mjy . |
hierarchical scenarios of structure formation have been very succesfull in understanding the formation histories of gravitationally bound virialized haloes .
particularly compelling has been the formulation of a formalism in which the collapse and virialization of overdense dark matter halos within the context of hierarchical clustering can be treated on a fully analytical basis .
this approach was originally proposed by press & schechter ( 1974 ) , which found a particularly useful and versatile formulation and modification in the the _ excursion set formalism _
( bond et al .
1991 ) .
it is based on the assumption that for a structure to reach a particular nonlinear evolutionary stage , such as complete gravitational collapse , the sole condition is that its _ linearly extrapolated primordial density _ should attain a certain value .
the canonic example is that of a spherical tophat overdensity collapsing once it reaches the collapse barrier @xmath0 .
the successive contributions to the local density by perturbations on a ( mass ) resolution scale @xmath1 may be represented in terms of a density perturbation random walk , the cumulative of all density fluctuations at a resolution scale smaller than @xmath1 . by identifying the largest scale at which the density passes through the barrier @xmath2 it is possible ( 1 ) to infer at any cosmic epoch the mass spectrum of collapsed halos and ( 2 ) to reconstruct the merging history of each halo ( see fig . 3 , lefthand ) . -3.0truecm -1.5truecm
-0.2truecm [ voidproc ] in this study we demonstrate that also the formation and evolution of foamlike patterns as a result of the gravitational growth of primordial density perturbations is liable to a succesfull description by the excursion set analysis . a slight extension and elaboration on the original formulation
enables us to frame an analytical theory explaining how the characteristic observed weblike megaparsec scale galaxy distribution , and the equivalent frothy spatial matter distribution seen to form in computer simulations of cosmic structure formation , are natural products of a hierarchical process of gravitational clustering .
this is accomplished by resorting to a complementary view of clustering evolution in which we focus on the evolution of the _ voids _ in the megaparsec galaxy and matter distribution , spatially _
th _ dominant component ( see e.g. muller & maulbatsch , these proceedings ) . 0.5 cm -2.0 cm -6.5 cm 0.0 cm -1.0truecm
primordial underdensities are the progenitors of voids . because underdensities are regions of suppressed gravitational attraction , representing an effective repulsive gravity , matter flows out of their interior and moves outward to the edges of these expanding _
voids_. voids _ expand _ , become increasingly _ empty _ and develop an increasingly _ spherical _ shape ( icke 1984 ) .
matter from the void s interior piles up near the edge : usually a ridge forms around the void s rim and at a characteristic moment the void s interior shells take over the outer ones . at this _ shell - crossing _ epoch
the void reaches maturity and becomes a nonlinear object expanding self - similarly , the implication being that the majority of observed voids is at or near this stage ( blumenthal et al .
1992 ) . as voids
develop from underdensities in the primordial cosmic density field , the interaction with internal substructure and external surrounding structures translates into a continuing process of hierarchical void evolution ( dubinski et al .
1993 , van de weygaert & van kampen 1993,colberg et al .
2004 ) .
-0.truecm -0.cm -0.2truecm the evolution of voids resembles that of dark halos in that large voids form from mergers of smaller voids that have matured at an earlier cosmic time ( fig . 1 , top row ) .
however , in contrast to dark halos , the fate of voids is ruled by two processes .
crucial is the realization that the evolving void hierarchy does not only involve the _ void - in - void _ process but also an additional aspect , the _ void - in - cloud _ process .
small voids may not only merge into larger encompassing underdensities , they may also disappear through collapse when embedded within a larger scale overdensity ( fig . 1 , bottom row ) .
in terms of the _ excursion set approach _
, it means that the _ one - barrier _ problem for the halo population has to be extended to a more complex _ two - barrier problem_. voids not only should ascertain themselves of having decreased their density below the _ void barrier _
@xmath3 of the _ shell - crossing _ transition . for their survival and sheer existence
it is crucial that they take into account whether they are not situated within a collapsing overdensity on a larger scale which crossed the collapse barrier @xmath2 .
they should follow a random walk path like type `` 3 '' in fig . 3 , rather than the _ void - in - cloud _ path `` 4 '' )
. the repercussions of this are far - reaching and it leads to a major modification of the void properties and distribution .
analytically , the resulting expression follows by evaluating the fraction of walks which first cross @xmath4 at @xmath5 , and which do not cross @xmath6 until after they have crossed @xmath4 ( see fig . 3 , 3rd & 4th row ) .
this distribution @xmath7 is given by ( sheth & van de weygaert 2003 ) , @xmath8 in which the relative impact of void and halo evolution on the hierarchically evolving population of voids is parameterized through @xmath9 .
a smaller value of @xmath10 corresponds to a diminished importance of the _ void - in - cloud _ process .
the above expression may be approximated by the more accessible expression , for values of @xmath11 , using the `` self - similar '' parameter @xmath12 @xmath13 the resulting distributions , for various values of @xmath10 , are shown in figure 4 .
the void size distribution is clearly _ peaked about a characteristic size _ : the _ void - in - cloud _ mechanism is responsible for the demise of a sizeable population of small voids .
the halo mass distribution diverges towards small scale masses , so that in terms of numbers the halo population is dominated by small mass objects .
the void population , on the other hand , is `` void '' of small voids and has a sharp small - scale cut - off .
four additional major observations readily follow from this analysis : ( @xmath14 ) at any one cosmic epoch we may identify a _
characteristic void size _ which also explains why in the present - day foamlike spatial galaxy distribution voids of @xmath15 are the predominant feature ; ( @xmath14 ) the _ void distribution evolves self - similarly _ and the _ characteristic void size increases with time _ : the larger voids present at late times formed from mergers of smaller voids which constituted the dominating features at earlier epochs ( fig . 4 , top panels ) ; ( @xmath14 )
volume integration shows that at any given time _
the population of voids approximately fill space _ , apparently squeezing the migrating high - density matter in between them ; ( @xmath14 ) as the size of the major share of voids will be in the order of that of the characteristic void size this observation implies that the _ cosmic matter distribution _ resembles a _ foamlike packing of spherical voids of approximately similar size and excess expansion rate_. to develop this generic picture into an encompassing view of the formation and evolution of the foamlike galaxy distribution
various additional issues need to be adressed .
one issue concerns the translation from matter into galaxy distribution , and thus the question of galaxy formation within voidlike environments . in this respect
it should be noted that the suggested scale of the peaked matter distribution still falls short of that of the observed galaxy distribution ( see e.g. vogeley 2004 ) .
another issue is the identification and location of the collapsing void population .
recent results ( fig .
2 ) have indicated that most of these events take place near and on the boundary of the large expanding voids : most small voids appear to get squashed and sheared ( see van de weygaert & babul 1996 ) there where the matter currents out of voids collide with the surrounding overdensities .
in all , from this study an enticing image of cosmic structure evolution emanates .
a continuously evolving hierarchy of voids produces a dynamical foamlike pattern whose characteristic dimension grows continuously along with the evolution of cosmic structure , a universe which at any one cosmic epoch is filled with bubbles whose size corresponds to the scale just reaching _
maturity_. 1990 _ mnras _ * 245 * , 522 1992 _ apj _ * 388 * , 234 1991 _ apj _ * 379 * , 440 2004 , these proceedings 1993 _ apj _ * 410 * , 458 1984 _ mnras _ * 206 * , 1 2004 , these proceedings 1974 , _ apj _ * 187 * , 425 2003 , _ mnras _ , in the press 1996 , _ unpublished _ 1993 , _ mnras _ * 263 * , 481 2004 , these proceedings | we present a model for the distribution of void sizes and its evolution within the context of hierarchical scenarios of gravitational structure formation . for a proper description of the hierarchical buildup of the system of voids in the matter distribution , not only the _
void - in - void _ problem should be taken into account , but also that of the _ void - in - cloud _ issue . within the context of
the excursion set formulation of an evolving void hierarchy is one involving a _ two - barrier _ excursion problem , unlike the _ one - barrier _ problem for the dark halo evolution .
this leads to voids having a peaked size distribution at any cosmic epoch , centered on a characteristic void size that evolves self - similarly in time , in distinct contrast to the distribution of virialized halo masses in not having a small - scale cut - off . |
in the hope to get a better understanding of strongly interacting systems , there has been considerable interest in ladder systems .
@xcite these ladder systems have proven to be a theoretical wonderland , both analytically @xcite and numerically .
@xcite however , much of the analytic work done on ladders has been in weak coupling ( or perturbatively in some parameter ) , namely because there are very few analytic methods at strong coupling .
exact diagonalization , monte carlo , and density matrix renormalization group methods have been the primary tools for studying these systems at strong coupling .
each of these methods has both strengths and weaknesses when considering the lattice sizes , temperatures , and couplings on can consider . with the ability to fabricate these materials ,
@xcite ladders are not just a theoretical playground .
for example , in @xmath0 , there are well separated two - leg ladders composed of @xmath1 .
@xcite also the cuprate - like material @xmath2 consists of weakly coupled @xmath3 two - leg ladders , @xcite and the material @xmath4 consists of weakly coupled @xmath3 three - leg ladders .
@xcite recently , a powerful analytic method was developed to deal with strongly coupled quasi - one dimensional systems the recurrent variational approach ( rva ) .
@xcite this method is similar in spirit to wilson s numerical renormalization group @xcite and white s density matrix renormalization group ( dmrg ) .
@xcite the key idea in all of these methods is to build up the system by adding on sites at the boundary .
however , the real power of the rva is that , though analytic , the physics of the problem is taken into account in an unbiased way and elucidated quite clearly .
for example , the importance of different configurations in the ground state wave function is determined without any outside assumptions , and the physics of these configurations is made clear . in this work
, we apply the rva to the two - leg hubbard ladder at strong coupling and small dopings .
though a considerable amount of work has been done on the hubbard ladder and many of its properties are known , we are unaware of any work which has put this information together and constructed a ground state wave function . our goal in this work is to provide a simple physical picture of what the ground state might look like and to go ahead and construct a ground state wave function .
the hamiltonian of the two - leg hubbard ladder is given by @xmath5 where @xmath6 creates a fermion at site * i * with spin @xmath7 , @xmath8 , @xmath9 is the hopping matrix element along the chain , @xmath10 is the hopping matrix element perpendicular to the chain ( i.e. , along the rung ) , and @xmath11 is the on - site coulomb repulsion .
site * i * has coordinates @xmath12 with @xmath13 and @xmath14 .
it will also be convenient to introduce the following two operators : @xmath15 which creates a singlet across sites * i * and * j * , and @xmath16 which creates a doubly occupied site .
the rest of the paper is organized as follows . in sec .
ii we consider the half - filled case . in sec .
iii we consider the hubbard ladder at small dopings . in sec .
iv we present our results for the ground state energies and compare with dmrg . finally , in sec .
v we summarize our results and present some concluding remarks .
we begin with the half - filled ladder . what are the ingredients necessary to construct a wave function which captures the physics of the half - filled hubbard ladder ?
a trail of clues has been laid by previous works .
first of all , we know that at strong coupling the half - filled = 3.25 in hubbard model is equivalent to the heisenberg model . secondly , it is known that the two - leg heisenberg ladder ( and hubbard ladder ) have a spin gap and short range spin correlations .
furthermore , it has been shown that the resonating valence bond ( rvb ) state captures the essential physics of the two - leg heisenberg ladder.@xcite hence , the rvb picture should capture the essential physics of the half - filled hubbard ladder at strong coupling .
thus we propose a generalization of the rvb state as our variational ansatz .
a key property of the rvb state is that it is the exact solution to the @xmath17 heisenberg plaquette .
therefore , we base our generalized rvb state on the exact solution to the @xmath17 hubbard plaquette . since the @xmath17 plaquette will serve as the basis for our ansatz , we discuss it in detail below .
the ground state of the @xmath17 plaquette is given by @xmath18 where @xmath19 @xmath20 is shown schematically in fig .
[ fig:2x2half ]
. we will be mainly interested in the case @xmath21 . in table
[ table:2x2half ] we list the values of the parameters for several values of @xmath11 ( with @xmath22 ) .
= 3.1 in .values of the parameters for the half - filled 2x2 plaquette ( with @xmath23 ) which give the exact groundstate . [ cols="^,^,^,^,^,^,^,^,^",options="header " , ] [ table:2x2holes ] that the ground state of the half - filled ladder transforms in the @xmath24 representation , we see that the operator that creates a hole pair out of the undoped system has @xmath25 symmetry .
@xcite these facts continue to be true for low doping in larger ladders .
also , notice that @xmath26 where @xmath27 is from the half - filled ladder .
this will play a role in writing our ansatz for the doped ladder . using the configurations of the generalized rvb state as well as the hole pair configurations , a typical configuration for the doped ladder
is shown in fig .
[ fig : typicaldoped ] .
= 2.9 in = .6 in the ground state will be a superposition of all such configurations shown in fig .
[ fig : typicaldoped ] .
fortunately , we can generate such a state _
recursively_. specifically , we build the ground state of a ladder with @xmath28 rungs and @xmath29 holes using the knowledge of the ground states of a ladder with @xmath30 rungs and @xmath31 holes .
this is achieved by recursion relations which express the ground state @xmath32 in terms of the ground states @xmath33 with @xmath34 and @xmath35.@xcite using the above ingredients , we consider the following ansatz for the doped hubbard ladder @xmath36 = 3.6 in = 2.9 in which is shown schematically in fig .
[ fig : goodstates1 ] . for completeness , the states in terms of the operators in eqs .
[ singlet ] and [ double ] are given in appendix b. a few words are in order about the configurations in our ansatz .
i ) the ansatz contains the configurations for the rvb state , as well as the configurations for holes bound in pairs .
the hole pair states are based on the exact solution to the 2x2 plaquette with 2 holes , and our ansatz reproduces this exact solution .
ii ) the states @xmath37 and @xmath38 extend over three rungs .
these states are necessary to allow the hole pairs to move smoothly through the rvb background .
iii ) note that physically , the picture of hole pairs moving through an rvb background can only be appropriate for low dopings . to compute the values of the coefficients in our ansatz ,
we treat them as variational parameters and minimize the ground - state energy with respect to these parameters
. it will be useful to define @xmath39 they are supplemented by the initial conditions @xmath40 to determine the values for the variational parameters for given ( finite ) values of @xmath41 and @xmath42 , we iterate the recursion relations and we minimize the quantity @xmath43 .
the actual recursion relations are quite unwieldly ; we have relegated them , as well as their derivation , to appendix b. what is the nature of the state we have constructed ? in order to answer this question , we plot @xmath44 , @xmath45 , and @xmath46 vs. doping .
these parameters contain most of the physics of our ansatz .
@xmath44 is the `` rvb parameter '' ; @xmath45 and @xmath46 are the weights of the hole pair configurations .
the results were obtained on a @xmath47 ladder .
first , consider fig .
[ fig : parameters](a ) .
@xmath44 begins at @xmath48 and increases ( i.e. , becomes less negative ) with doping until a critical doping , @xmath49 , where it vanishes . beyond this doping
, @xmath44 is positive .
this has also been found for the @xmath50 ladder . @xcite
upon doping , the hole pairs cause destructive interference which degrades the rvb mechanism . for @xmath51 ,
this destructive interference has driven @xmath44 positive , and it is no longer appropriate to think of our state as describing hole pairs moving through an rvb background .
@xcite similar to the t - j ladder , the difference between @xmath52 and @xmath51 can be attributed to two different internal structures of the hole pairs . for @xmath52
the hole pairs have a @xmath25 structure _ relative _ to the rvb background . for @xmath51
the hole pairs have an @xmath7-wave like symmetry relative to their background .
@xcite now consider figs .
[ fig : parameters](b ) and [ fig : parameters](c ) .
first of all , notice that @xmath53 .
this shows the importance of the diagonal frustrating bonds for all dopings .
@xcite also , notice that @xmath45 and @xmath46 both reach their maximum at @xmath54 . at @xmath54
the system is essentially a large scale reproduction of the 2x2 plaquette with 2 holes .
@xcite indeed , the values of @xmath45 and @xmath46 at @xmath54 are similar to their values for the 2x2 plaquette .
= 3.25 in = 3.25 in = 3.25 in
first , we show results for energy per site vs. @xmath11 in fig . [ fig : energyvsu ] . for comparison , dmrg results are presented for the same set of parameters .
@xcite at half - filling , as we would expect , our ansatz is most accurate for large @xmath11 and large @xmath10 .
the ground state energy per site for a @xmath47 half - filled ladder as a function of @xmath11 for various @xmath10 is shown in fig .
[ fig : energyvsu](a ) . for @xmath55 and @xmath56 ,
the energy from the rva agrees with dmrg to within @xmath57 and improves as @xmath11 or @xmath10 is increased . up to about @xmath58 , longer bonds ( extending over at least 3 rungs ) are coming into play .
these states should be included in the ansatz to further improve the overlap with the ground state . at @xmath59 and @xmath56 ,
our ansatz gives a ground state energy within @xmath60 of the dmrg result .
it should be noted that for the heisenberg ladder , the rvb state gives a ground state energy within @xmath60 of true + = 3.25 in = 3.25 in = 3.25 in ground state energy , obtained from dmrg .
@xcite a recent + = 3.0 in dmrg study of different ladder models found that the hubbard model and heisenberg model begin to agree only for rather large @xmath11 ( @xmath61 ) .
@xcite therefore , it is not surprising that the rvb picture becomes as good for the hubbard model as it is for the heisenberg model at @xmath61 .
[ fig : energyvsu](b ) shows the ground state energy of the rva ansatz as a funtion of @xmath11 for various @xmath10 for a doping of @xmath62 on a @xmath47 ladder .
again , we show energies obtained from dmrg for the same set of parameters . for @xmath59 and @xmath63 , the two energies agree to only within @xmath64 and improves slightly as @xmath10 is increased . for example , at @xmath59 and @xmath65 , the overlap of energies increases to @xmath66 .
further discrepancies occur when the doping is increased to @xmath67 ( see fig .
[ fig : energyvsu](c ) ) .
the differences in ground state energies occur due to the importance of `` pair - breaking '' configurations , like those shown in fig .
[ fig : pairbreaking ] .
the weights of these types of states increase as we move away from half - filling . consequently , to further improve the rva ansatz , such states must be included in the wave function .
note also that our ansatz would be essentially exact for the case where hole pairs are well localized on a rung . for the @xmath50 model with @xmath68 , pairs are well localized along a rung , and the ground state is essentially a product of rung singlets and rung hole pairs . however , for the hubbard model at strong coupling ( i.e. , @xmath69 ) , this is not the case .
holes would always rather occupy _ adjacent _ rungs , even for @xmath70 , since this minimizes the coulomb energy from doubly occupied sites . to see this consider a @xmath17 plaquette with 2 holes ;
let @xmath71 and @xmath72 . with 1 particle on each rung ( i.e. , one hole on each rung ) ,
the ground state energy is approximately @xmath73 ; with both particles on the same rung ( i.e. , both holes on the same rung ) , the ground state energy is approximately @xmath74 .
therefore , at large u , the particles would rather occupy adjacent rungs .
( see fig .
[ fig : adjacent ] . )
the situation we have with the doped ladder is similar to what we had for the half - filled ladder in the early stages of this work .
we found that without the states + = 2.9 in = 3.25 in @xmath75 and @xmath76 which extend over three rungs ( see fig .
[ fig : goodstates ] ) , the rva did not accurately reproduce the ground state even at extremely large @xmath11 .
however , once we included @xmath75 and @xmath76 , the results from the rva improved drastically .
based on these results , we expect the rva to greatly improve by including the states shown in fig .
[ fig : pairbreaking ] . in fig .
[ fig : xeplot1 ] we plot energy per site vs. doping for a @xmath47 ladder for @xmath55 and @xmath59 ( with @xmath63 ) in order to better understand the region of validity of our rva ansatz .
again , we see good agreement with dmrg results at half - filling . however , as soon as we dope , configurations like those shown in fig .
[ fig : pairbreaking ] are also important .
it is interesting to note that the idea of hole pairs moving through the rvb background seems to more accurately represent the ground state of the @xmath50 model than the hubbard model . using the well known relation at strong coupling , @xmath77
, the rva agrees to within @xmath78 of the true ground state energy of the @xmath50 ladder for @xmath79 @xmath80 at a doping of @xmath62 .
there are two ways to interpret this ; either the t - j model supports pairing better than the hubbard model , or we must view the hole pairs in the hubbard model as having a larger size ( i.e. larger coherence length . )
to summarize , we applied the recurrent variational approach to the two - leg hubbard ladder .
our results were in qualitative agreement with previous results on the heisenberg and t - j ladders .
for the half - filled ladder , the generalized rvb state became more accurate in the parameter regime where the hubbard and heisenberg ladders were shown to coincide .
however , comparison of the rva with dmrg for the doped ladder indicate that hole - pairs moving through an rvb background is incomplete ; it does not capture the essential physics .
`` pair - breaking '' configurations are also necessary to capture the essential physics . as we saw , the strength of the rva is the ease in which we could extract the physics .
we were able to see the importance of the configurations in our ansatz quite easily .
furthermore , the rva has a natural way in which to include longer bonds in the ansatz to more accurately represent the groundstate wavefunction .
the importance of such additional states to the physics of the ladder is not easily probed with other techniques .
generalized rvb states similar to ours have been considered previously for the half - filled hubbard ladder .
@xcite fano _ et .
al . _ were even able to produce an ansatz coming within @xmath81 of the true groundstate energy for a ( half - filled ) 2x4 ladder at @xmath59 .
( their ansatz included diagonal bonds of length @xmath82 .
) however , none of these works considered the doped case . using the approach in ref .
17 ( in terms of dimer coverings ) , it appears to be a formidable task to consider doping .
this is one of the strengths of the rva ; doping is handled rather easily .
even though our results for the doped ladder showed that hole pairs moving through an rvb background is incomplete , the rva offers a straightforward way to improve the situation , namely include `` pair - breaking '' configurations ( shown in fig .
[ fig : pairbreaking ] ) in the ansatz .
another ( and probably better ) way to improve the situation for the doped ladder is to consider a matrix product ansatz .
@xcite a matrix product ansatz can be generated by first order recursion relations .
@xcite in the rva , the size of the hole - pairs are fixed .
( in our case , the hole pairs had a size of one lattice spacing . )
however , by construction , the matrix product ansatz takes into account hole pairs of _ arbitrary _ size .
we leave this ( and other possibilities ) for future work .
we would like to thank d. j. scalapino , s. daul , s. hortikar , r. konik , c. l. martin , and m. a. martin - delgado for helpful discussions and comments .
ehk gratefully acknowledges the warm hospitality of argonne national laboratory where parts of this manuscript were written .
dd would like to thank s. daul and r. m. noack for assistance with the dmrg results presented here .
ehk was supported by nsf grant no .
dmr-9527304 , gs by dges grant pb97 - 1190 , and dd by doe grant no . de - fg03 - 85er451907 .
the states in the ansatz for the half - filled ladder of eq .
[ halffilledansatz ] are given by @xmath83 to derive the recursion relations , the following inner products are necessary : @xmath84 using these inner products , a straightforward calculation gives the following ( coupled ) recursion relations : @xmath85
for the doped ladder ( see eq .
[ dopedansatz ] ) , @xmath86 , @xmath87 , @xmath88 , @xmath89 , @xmath90 , @xmath91 , @xmath92 , and @xmath93 are the same as the half - filled case , and @xmath94 to derive the recursion relations , we use the inner products from the half - filled case as well as the the following : @xmath95 using these inner products , a straightforward calculation gives the following ( coupled ) recursion relations : @xmath96 for example see c. varma and a. zawadowski , phys .
b * 32 * , 7399 ( 1985 ) , d. g. shelton , a. a. nerseynan , and a. m. tsvelik , phys . rev .
b * 53 * , 8521 ( 1996 ) , h. h. lin , l. balents , and m. p. a. fisher , phys
b * 58 * , 1794 ( 1998 ) , and references therein . | we applied the recurrent variational approach to the two - leg hubbard ladder . at half - filling ,
our variational ansatz was a generalization of the resonating valence bond state . at finite doping
, hole pairs were allowed to move in the resonating valence bond background .
the results obtained by the recurrent variational approach were compared with results from density matrix renormalization group . 2 |
the detection and characterization of rich clusters of galaxies from low ( @xmath2 ) to high ( @xmath3 ) redshifts provides a crucial test of both cosmological models ( e.g. , carlberg et al .
1996 ; fan , bahcall , & cen 1997 ; oukbir , bartlett & blanchard 1997 ; eke et al .
1998 ; borgani et al . 1999 ; ledlow et al . 1999 ; reichart et al . 1999 ; viana & liddle 1999 ; henry 2000 , to list a recent subset of an extensive lterature ) and galaxy evolution ( see 3 for an extensive discussion ) . clusters trace structure in the universe to large scales ( e.g. , bahcall & soneira 1983 ; gramann et al . 1995 ; tadros , efstathiou , & dalton 1998 ; borgani , plionis & kolokotronis 1999 ) , and provide many examples of both strong and weak lensing ( see mellier 1999 for a recent review ) .
the obvious importance of galaxy clusters has led to numerous galaxy cluster surveys in recent years , both optically ( e.g. , gunn , hoessel & oke 1986 ; abell , corwin & olowin 1989 ; couch et al .
1991 lumsden et al .
1992 ; lidman & peterson 1996 ; postman et al . 1996 ; dalton et al . 1997 ; zaritsky et al . 1997 ; ostrander et al .
1998 ; scodeggio et al .
1999 ; gal et al .
2000 ) and with x - rays ( gioia & luppino 1994 ; scharf et al . 1997 ; rosati et al .
1998 ; vikhlinin et al . 1998
; de grandi et al . 1999 ; romer et al . 2000 ) . as pointed out by kepner et al .
( 1999 ) , each of these surveys has defined its own detection algorithm , considered by the individual authors to be appropriate to their particular data .
in essence , each of these detection algorithms makes assumptions about `` what a cluster looks like '' and then searches the relevant data to identify likely cluster locations .
[ [ section ] ] all cluster finding methods suffer from selection effects . at minimum , each will be biased against any clusters which do not fit the particular cluster definition used . even if cluster samples appear essentially congruent at different wavelengths , scatter in the individual properties of individual clusters
will still engender somewhat different samples if surveys are done at different wavelengths .
for example , results from the cnoc1 survey show that the optical richness ( important for optically selected catalogs ) and x - ray luminosity ( important for x - ray selected catalogs ) of a sample of 16 x - ray selected clusters at @xmath4 are only moderately correlated ( yee et al .
x - ray surveys , which ultimately define clusters as diffuse x - ray sources , are potentially biased against gas - poor clusters , and clusters in which the gas distribution is compact and hence unresolved .
there is some limited evidence for optically rich and massive , but x - ray weak clusters .
for example , bower et al . (
1994 ) found that a sample of 14 optically selected rich clusters at @xmath5 showed relatively weak x - ray emission , though the significance of this result is dependent on the interpretation of masses from minimal spectroscopic data .
castander et al .
( 1994 ) found similar results for a sample of optically selected clusters at @xmath6 .
optical surveys are potentially biased against optically dark clusters deep cluster - sized potential wells which , for whatever reason , have failed to form a significant galaxy population .
notably , there is little evidence for many such clusters .
the x - ray selected sample of vikhlinin et al .
( 1998 ) showed no evidence for any optically dark clusters , with a sample of @xmath1 200 x - ray selected clusters from @xmath7 to @xmath8 .
a similar result is found in the bright sharc sample ( romer et al .
2000 ) . at higher redshift , the cluster axj2019 + 112 at @xmath9 , claimed by hattori et al .
( 1997 ) to be a dark cluster , has been shown by bentez et al .
( 1999 ) to not be dark at all .
clusters which are under - luminous in galaxian light may exist , but appear to be the exception rather than the norm .
another low - mass cluster type likely missed by optical techniques are the so - called fossil groups ( ponman et al .
1994 ; vikhlinin et al .
1999 ; romer et al .
2000 ) which appear to be relatively common though not massive ( vikhlinin et al . 1999 ) .
[ [ section-1 ] ] more complicated redshift dependent biases and uncertainties also exist in most cluster catalogs , quite apart from the dominant cosmological effects of redshift .
for example , optical - ir selected samples are subject to the effects of cluster and field galaxy evolution , which are , in general , not well determined at @xmath10 .
note , however , that the evolution of the _ elliptical _ galaxy population in clusters has been shown by numerous authors ( e.g. , stanford , eisenhardt , & dickinson 1998 and references therein , hereafter sed98 ) to be remarkably simple and homogeneous .
the apparent stability and smooth evolution of this dominant cluster population implies the selection functions for clusters defined by these galaxies can be computed with some confidence , even at @xmath10 . for a given cluster finding technique to be most useful ,
the reliable computation of selection functions is as critical as the sensitivity of the algorithm ( postman et al .
1996 ) .
[ [ section-2 ] ] the general selection functions for clusters in most surveys are typically quite complicated functions of numerous parameters , including , but not limited to , the cluster redshift and the survey flux limits . however ,
in most cases , the detection rate for clusters of a given parameter set can generally be computed directly from the survey data , by inserting and attempting to recover fiducial test clusters .
so long as the test clusters are a realistic representation of the cluster populations , this procedure will produce reliable results , and such techniques have been used successfully in several cluster surveys ( e.g. , postman et al .
1996 ; ostrander et al . 1998 ; bramel , nichol & pope 2000 ) .
false positive detections are however more difficult to quantify , as a thorough understanding of such false cluster detections requires some @xmath11 @xmath12 knowledge of the contaminants such as projection in optical catalogs ( e.g. , postman et al .
1996 ) or contamination by active galactic nuclei in x - ray selected catalogs ( e.g. , vikhlinin et al . 1998 ) .
current attempts to quantify the false positive contaminations rates in samples in the absence of detailed follow - up data are typically based on some form of resampling strategy ( e.g. , bramel , nichol & pope 2000 ) .
[ [ section-3 ] ] indeed , the dominant cause of false - positive cluster candidates in optical cluster catalogs is projection . it has long been suggested that there are significant projection effects in the abell catalog ( e.g. , lucey 1983 ) , though the exact degree of contamination is a matter of some debate ( collins et al .
even more modern cluster catalogs suffer significant false cluster detections due to random projections ( postman et al .
1996 ; oke , postman , & lubin 1998 ; holden et al .
the projection - induced contamination in the abell catalog has been studied with exhaustive n - body simulations , which show this to be primarily an artifact of the large aperture used to initially define the abell catalog ( van haarlem , frenk , & white 1997 ) .
these n - body results demonstrate that the false - positive rates due to projection for simulated cluster catalogs detected by either x - rays or 2-d galaxy overdensities is in fact quite similar , provided a similar detection aperture is used ( van haarlem , frenk , & white 1997 ; also see lpez - cruz & yee 2000a , hereafter lcy00 ) .
[ [ section-4 ] ] given the extreme importance of galaxy clusters for a variety of studies , it is clear that large , well - understood samples of clusters are needed . in such a setting , it desirable to have samples which are selected in as many ways as possible , at as many wavelengths as possible , as each sample will probe a somewhat different population with somewhat different selection effects .
furthermore , the cluster finding algorithms used should be as efficient as possible in terms of the required data , since clusters are rare objects and thus only found in large numbers in observationally expensive wide - field surveys . the method proposed below addresses these requirements .
the method , which we term the cluster - red - sequence ( crs ) method , exploits the observational fact that the bulk of the early - type galaxies in all rich clusters lie along a linear color - magnitude relation .
this relation , referred to hereafter as the red sequence , has been shown to have an extremely small scatter ( e.g. , bower , lucey , & ellis 1992b ) and appears to be extremely homogeneous from cluster to cluster ( lcy00 ) . in the crs method clusters are detected as over - densities in projected angular position , color and magnitude simultaneously .
the color of the red sequence provides a precise redshift estimator for the detected clusters .
the color constraint makes the crs method extremely insensitive to projection effects , as random projections do not exhibit the necessary red sequence signature in the color - magnitude plane .
this makes the crs method fundamentally different from previous optical cluster finding algorithms .
the concept of using the red sequence as a cluster marker , or to delineate cluster superstructures , has some history in the literature ( e.g. , lopez - cruz 1997 ; gladders & yee 1998 ; kaiser et al .
1998 ; yee , gladders , & lpez - cruz 1999 ; lubin et al .
2000 ; lcy00 ) , and other cluster - finding techniques have been suggested which exploit the dominance of early - type galaxies in clusters in some way ( ostrander et al . 1998 ) .
however , the critical point of the technique suggested here is the recognition that the cluster red sequence can be effectively isolated from survey data with only two filters , so long as the filter pair samples the 4000break .
this technique precludes the need for many - band imaging , such as is required for full photometric redshift analyses ( kodama , bell & bower 1999 ) and so is efficient while still exploiting the available color information .
this paper represents the first effort to cast the crs method as a well - defined , well - motivated algorithm .
the purpose is to describe an algorithm which is amenable to stringent testing , and useful for finding clusters in a wide range of optical - ir imaging data .
[ [ section-5 ] ] we have begun a large observational program , the red - sequence cluster survey ( rcs ) , to identify and characterize clusters using the crs method .
this paper , the first of a series of two , describes the basic observational motivation for the crs method , and provides a detailed algorithm and a test of that algorithm using existing data with spectroscopic redshifts .
the second paper in the series discusses the crs method in the specific context of the rcs dataset , focussing on a detailed derivation of the rcs selection functions for clusters using extensive simulations .
[ [ section-6 ] ] this paper is arranged as follows .
we describe and motivate the crs method in the context of previous work and current observational constraints in 2 and 3 .
a detailed implementation of the method is given in 4 . in
5 we provide a simple but powerful test of the method , using real redshift data from the cnoc2 redshift survey ( yee et al .
section 6 provides a brief discussion of possible refinements to the basic method outlined here .
we summarize our findings in 7 .
a cosmology of @xmath13=70 , @xmath14=0.2 and @xmath15=0.0 is used throughout .
by definition , optical / ir cluster - finding techniques all rely upon overdensities in galaxy distributions as the signature of mass overdensities .
the basic premise shared by all such techniques is that galaxies are a reliable ( though not necessarily unbiased , kaiser 1984 ) tracer of the underlying mass distribution , which is the cosmologically interesting distribution .
so long as the mass - to - light ratio of galaxy clusters is roughly constant , or evolving in well - understood ways , this basic premise seems sound .
the most likely complication is the possible presence of dark clusters , though the current evidence indicates that such objects are rare at best ( see 1 above ) .
given this basic premise , the differences between cluster finding - techniques are based in the manner in which each attempts to identify the real three - dimensional galaxy overdensities in typically two - dimensional imaging data . in general , the more modern techniques ( the matched - filter technique , postman et al . 1996 ; kepner et al .
1999 ) are designed to be more sensitive to true clusters ( while being less sensitive to random projections ) at the expense of greater susceptability to modeling uncertainties .
the method we propose and test here makes the basic assumption that the presence of a red sequence of early - type galaxies is a ubiquitous and near - universal signature of clusters .
this relatively strong assumption is well supported by the current observational evidence , and provides a very efficient and powerful method for detecting clusters . in the following
we describe the basic advantages of this method and discuss some of the possible uncertainties , deferring a detailed description of the algorithm to 4 .
[ [ section-7 ] ] the crs method is motivated by the observation that all rich clusters have a population of early type galaxies which follow a strict color - magnitude relation .
an example of this is provided in figure 1 , which shows the measured color - magnitude diagram for the @xmath16 cluster abell 2390 , from the hst data in gladders et al .
( 1998 ) . even in this field - contaminated data ,
the cluster red sequence is easily recognized .
current evidence ( discussed in detail in 3 ) indicates that this galaxy population is remarkably homogeneous , both within individual clusters and between clusters .
moreover , it appears that the stellar population which makes up the red sequence is formed at high redshifts ( @xmath17 ) . in cold - dark - matter dominated scenarios of hierarchical structure formation
this is the expected result , as present - day clusters are correlated with the most extreme initial over - densities , which are the first to collapse .
overall , cluster elliptical galaxies appear to have the oldest and most homogeneous stellar populations , and so can be expected to be a relatively stable marker of clustering .
[ [ section-8 ] ] there are numerous observational reasons which make the red sequence an attractive target for cluster finding , apart from its apparent homogeneity .
first , elliptical galaxies generally dominate the bright end of the cluster luminosity function ( e.g. , sandage , bingelli , & tamman 1985 ; thompson 1986 ; barger et al .
1998 though see rakos , odell , & schombert [ 1997 ] for an exception to this ) , and so are the most readily seen galaxies in a flux limited survey . cluster elliptical galaxies are more luminous at higher redshifts ( e.g. , schade , barrientos , & lopez - cruz 1997 ; van dokkum et al .
1998 ) , consistent with the age - induced fading expected in a passively evolving stellar population .
in addition , the radial distribution of elliptical galaxies in regular , centrally concentrated clusters is more compact than that of other morphological types due to the morphology - density relation ( dressler et al .
1997 and references therein ) , and so presents a higher contrast against the background . even in irregular clusters , with no well defined center , the morphology density relation holds , though at lesser significance ( dressler et al .
1997 ) , and elliptical galaxies still trace the densest cluster regions . additionally , elliptical galaxies have core - dominated compact brightness profiles , and so can be morphologically selected with a high level of confidence ( abraham et al . 1994 ; gladders et al .
1998 ) [ [ section-9 ] ] an examination of figure 2 reveals the final , critical advantage of using elliptical galaxies to find clusters .
figure 2 shows model color - magnitude tracks , over a range of redshift , for galaxies of various spectral types ( from coleman , wu , & weedman 1980 ) , as well as the expected location of the cluster red sequence , at similar @xmath18 , using red sequence slopes from kodama ( 1997 ) . since cluster
elliptical galaxies likely represent the oldest stellar populations in the universe , they are as red or redder than any other galaxies at a given redshift . moreover , with properly chosen filters straddling the 4000break
, the cluster red sequence is as red or redder than other galaxies at a given redshift _ and all lower redshifts_. as can be seen in figure 2 , this means that the bulk of the contaminant galaxies within a color slice corresponding to a red sequence at a given redshift are intrinsically bluer galaxies at yet higher redshifts .
_ in other words , the crs method does not accumulate an overburden of foreground contamination when extended to higher redshifts , effectively solving the critical foreground projection problem with optical / ir cluster searches at such redshifts .
_ moreover , projection effects due to background structures at higher redshift are minimized , because such structures will generally be most significant at yet redder colors .
[ [ section-10 ] ] a secondary advantage of using the red sequence as a cluster tracer is that the red sequence color provides an extremely precise redshift indicator .
lcy00 noted that the red sequence @xmath19 ( at a fixed apparent @xmath20 ) versus redshift diagram for low @xmath18 clusters shows a remarkably small scatter of @xmath21 about the mean relation .
this had initially led lpez - cruz ( 1997 ) to suggest that the red sequence color would be a valuable redshift indicator .
smail et al .
( 1998 ) came to similar conclusions , though with a much larger uncertainty , for their cluster sample at @xmath22 .
sed98 found the red sequence color to be well correlated to redshift to @xmath23 .
notably , the redshift scatter implied by these simple photometric methods is comparable to or better than that seen in four ( or more ) band photometric redshift studies with calibration redshifts ( brunner et al .
this precision is a result of the significant reduction in photometric errors due to using many galaxies for a single color estimate , and the intrinsic similarity of red sequence galaxies .
notably , this redshift accuracy is significantly better than that achieved by the matched - filter method ( holden et al .
2000 ) , because the color - based redshift relies upon measuring the centroid of a peak in the colour distribution whereas the matched - filter redshift requires a measure of the cluster m@xmath24 .
the former can be measured with much greater precision than the latter .
[ [ section-11 ] ] in brief , the crs method defines a cluster as an overdensity on the sky which also demonstrates an overdensity in the color - magnitude plane consistent with a red sequence of early - type galaxies .
the method is remarkably insensitive to projection effects because the requisite color - magnitude relation is unlikely to be met by random projections .
the enhanced sensitivity of the crs algorithm to real clusters and its robust insensitivity to projections is a result of making the assumption that all significant real clusters have a red sequence of early - type galaxies .
though this assumption must break down at some point in the early formation of a cluster s galaxies , it is observed to hold to at least @xmath25 in a variety of cluster environments and a range of cluster masses , as detailed below .
[ [ section-12 ] ] observations of the red sequence in clusters generally indicate that the bulk of the stars in the cluster early - type galaxies formed at a high redshift .
two major lines of evidence in support of this idea have been pursued by numerous authors .
one is based on various analyses of the color evolution of the early - type galaxies .
the other is based on analyses of the luminosity evolution of the early - type galaxies .
each general class of analysis is discussed separately below .
furthermore , since the observational data at @xmath10 are much less complete , the apparent properties of clusters at these redshifts are discussed separately .
all current data indicate that the red sequence is a universal signature of galaxy clusters .
[ [ section-13 ] ] bower , lucey , & ellis ( 1992a , b ) completed the first comprehensive photometric analysis of the red sequence properties in the coma and virgo clusters , and concluded that the early - type galaxies in each were indistinguishable . considering that the coma and virgo clusters are prototypical clusters of completely different types coma is rich , massive , centrally concentrated , elliptical rich , and x - ray luminous ; and virgo is relatively poor , much less massive , irregular and spiral rich
this is a striking result .
this basic result has been extended to a much larger sample in the exhaustive low - redshift cluster survey of lcy00 .
this multi - color imaging survey of 45 x - ray - selected abell clusters provides a sample which encompasses a wide range of optical richnesses ( abell richness class 0 to @xmath26 ) , bautz - morgan classes ( i - iii ) , masses and rood - sastry classes ( see lpez - cruz 1997 for details of the sample ) . despite this optical heterogeneity ,
every cluster in the sample has a red sequence ; and the k - corrected slopes , scatters and colors of these red sequences are indistinguishable . at least at @xmath27 ,
the red sequence is a universal and homogeneous feature of galaxy clusters ( lcy00 ) .
[ [ section-14 ] ] other surveys extend the conclusions of lcy00 to higher redshifts .
smail et al .
( 1998 ) studied 10 optically - selected luminous x - ray clusters at @xmath28 , and barrientos ( 1999 ) studied 8 optically - selected clusters at @xmath29 . in each sample , the clusters have red sequences with colors and scatters which are remarkably homogeneous , and imply a high formation redshift .
notably , barrientos ( 1999 ) suggests based on the @xmath5 data alone , that although cluster red sequence formation as low as @xmath30 is possible with appropriate arrangement of the cosmology , red sequence metallicity and star formation synchronicity , the most reasonable formation redshift from these data is @xmath312.0 .
a similar result comes from ellis et al .
( 1997 ) based on three clusters at @xmath32 .
[ [ section-15 ] ] the studies of barrientos ( 1999 ) , lcy00 and smail et al . (
1998 ) are the most comprehensive photometric studies of heterogeneous clusters samples over restricted redshift ranges .
most other surveys analyze clusters over a larger redshift range , up to redshifts of @xmath30 .
analyses of the evolution of the color ( aragn - salamanca et al .
1993 ; rakos & schombert 1995 ; sed98 ; kodama et al .
1998 ) all point to a high redshift of formation for the red sequence stellar populations .
complementary analyses of the scatter ( sed98 ) and evolution of the slope ( gladders et al .
1998 ; sed98 ) over large redshift baselines indicate similar conclusions .
the dominant uncertainty in much of this analysis is the implicit identification of the higher redshift clusters with lower redshift clusters of similar richness .
as pointed out by kauffmann ( 1995 ) , in a hierarchical universe this identification is likely not correct , as the high redshift systems correspond to yet more massive clusters in the present day , and have a more vigorous merger history immediately prior to the epoch of observation when compared to present day clusters of similar mass .
however , for our purposes , it is sufficient to recognize that the red sequences of the highest redshift clusters in these @xmath33 samples all satisfy the photometric properties of a high formation redshift .
[ [ section-16 ] ] a final line of evidence for the universality of the red sequence in clusters comes from the spectroscopic and photometric analysis of nine @xmath34 optically selected cluster candidates by oke , postman , & lubin ( 1998 ) .
of the nine candidates , spectroscopic data indicate that six are real clusters . of the six apparently real clusters ,
detailed information is available for three .
the first , cl0023 + 0423 , is in fact the projection of two sub- abell richness class 0 groups ( lubin , postman , & oke 1998 ) .
cl0023 + 0423 shows a galaxy morphology mix similar to the field ( lubin et al . 1998 ) , though 6 of the 24 spectroscopically confirmed members of both components of cl0023 + 0423 have the colors of cluster ellipticals at that redshift ( postman , lubin , & oke 1998 ) .
the second , cl1604 + 4304 , is a massive , centrally concentrated cluster , and shows a strong , apparently old red sequence ( sed98 ) .
the third , cl1324 + 3011 , also shows a prominent red sequence ( gladders et al .
note that the well - formed clusters both show a red sequence , and even the projected groups show some evidence of red galaxies .
in addition , another candidate , cl0231 + 0048 , was initially included in the sample precisely because it showed an anomalous lack of a red sequence ( j.b .
oke , private comm . ) .
despite being a significant overdensity on the sky ; spectroscopic analysis shows that it is not a real cluster ( oke et al .
1998 ) . furthermore
, the fact that the oke et al . sample includes cluster candidates which have a broad range of galaxy mixtures again argues against suggestions that optical cluster candidate catalogs such as these are biased to only clusters with old , well - developed galaxy populations .
rather , the optical samples available to date appear to sample a wide range of cluster types , and all true clusters in the samples appear to have a red sequence .
[ [ section-17 ] ] the measured luminosity evolution of cluster early - type galaxies provides further evidence of the high formation redshift for the stars in these galaxies .
numerous fundamental - plane studies of clusters at redshifts up to 0.83 have demonstrated a change in the mass - to - light ratios of cluster galaxies consistent with that expected from passive evolution of stellar populations formed at high redshift ( e.g. , van dokkum & franx 1996 ; kelson et al .
1997 ; bender et al .
1998 ; van dokkum et al . 1998 ; jorgensen et al . 1999 ) .
barrientos , schade , & lpez - cruz ( 1996 ) , schade et al .
( 1996 ) and schade , barrientos , & lpez - cruz ( 1997 ) demonstrated similar evolution in the @xmath35 relation ( a projection of the fundamental plane ) ; as did pahre , djorgovski , & de carvalho ( 1996 ) , barger et al .
( 1998 ) , and ziegler et al .
( 1999 ) using the evolution of the kormendy relation ( kormendy 1977 ) .
further evidence for luminosity evolution comes from the evolution of the mg@xmath36-@xmath37 relation ( e.g. , bender , ziegler , & bruzual 1996 ; ziegler & bender 1997 ) .
taken together , these studies provide convincing evidence of the essentially passive evolution of the stellar populations in cluster ellipticals
. ziegler et al . ( 1999 ) do note that there are a number of possible uncertainties in such studies however , the general conclusion that the bulk of the stellar population in these galaxies must have formed at @xmath38 still holds . as in the case of photometric color - based studies , it is difficult to provide more stringent limits , as this requires observations of clusters at higher redshifts than those explored so far .
[ [ section-18 ] ] the universality of the red sequence is not yet well tested at @xmath10 , and analysis in the context of galaxy formation models is quite limited ( e.g. , sed98 ) .
this is simply due to a lack of candidate clusters at these redshifts , and the observational cost required to observe them .
despite this , optical and ir imaging of a portion of the limited sample of clusters does point to the universal presence of the red sequence .
the highest redshift cluster discovered to date in a blank field imaging survey is cig j0848 + 4453 at @xmath39 ( stanford et al .
this cluster was first detected as a spatially compact overdensity of red galaxies in an optical - ir imaging survey essentially as a detection of the red sequence .
a nearby , possibly associated cluster , rx j0848.9 + 4452 , is described by rosati et al .
( 1999 ) and also shows a strong red sequence .
the radio galaxy 3c324 at @xmath40 is also embedded in a large cluster , which again shows a strong red sequence ( dickinson 1995 ) .
a further interesting case is the cluster axj2019 + 112 at @xmath9 .
as discussed in 1 , this cluster was first thought to be an optically dark cluster ( hattori et al .
1997 ) , but was later revealed to be a normal cluster at @xmath9 with a strong red sequence ( bentez et al .
da costa et al .
( 1999 ) present color - magnitude diagrams for two @xmath30 cluster candidates detected in the eso imaging survey ( olsen et al .
1999 ) ; both show a red sequence of early - type galaxies at the color appropriate to the estimated redshift , and are thus both likely real clusters .
these candidates were selected on the basis of galaxy overdensities using the matched - filter technique , which makes no assumptions about galaxy colors , and so the presence of a red sequence is not expected _ a priori_.
notably , neither of these clusters is particularly rich ( m. franx , private comm . ) , which is expected given the total area from which the candidates were selected ( 1.1 deg@xmath41 , olsen et al .
1999 ) .
[ [ section-19 ] ] it is striking that all the high - redshift clusters studied so far have red sequences . except for cig j0848 + 4453 , the selection techniques for these clusters do not in general require such a signature indeed , the cluster axj2019 + 112 was selected by a combination of x - ray emission and the lensing of a background quasar ( hattori et al .
1997 ) without any reference to cluster galaxies whatsoever .
moreover , the richness or mass range encompassed by these clusters is quite large ( smail & dickinson 1995 ; hattori et al .
1997 ; stanford et al .
1997 ) , with implied richnesses ranging from abell richness classes 0 to 2 , and most of the clusters being relatively poor .
the implication is that the evolution of the cluster galaxy population towards a red sequence occurs at relatively small mass scales and thus prior to the assembly of large clusters .
a recent weak - lensing analysis of ms105403 ( a massive cluster at @xmath42 ) demonstrates this as well .
ms105403 is clearly not relaxed , and appears to be composed of several sub - clumps with velocity dispersions ( presuming an isothermal distribution ) of @xmath43 km s@xmath44 ( hoekstra , franx , & kuijken 2000 ) . despite this
, it already has a strong red sequence of early - type galaxies .
moreover , hst imaging and keck spectroscopy of ms105403 reveals the presence of numerous galaxy - galaxy mergers , many of which are occurring between galaxies which are already red ( franx et al .
again , the implication is that the entrenchment of a red sequence in the cluster ( i.e. , the formation of the bulk of the stellar population in the eventual red sequence galaxies irrespective of the dynamical state of the host galaxies of said stars ) occurs before the formation of the cluster as a massive relaxed system . in support of this general picture
is the recent result of ponman , cannon , & navarro ( 1999 ) . on the basis of an analysis of the entropy of the x - ray gas in clusters and groups , ponman et al .
conclude that some form of pre - heating is require to explain the observed trends in x - ray temperature with cluster mass .
ponman et al .
argue that this requires that the galaxies in clusters must form the bulk of their stellar populations prior to the assembly of the cluster , as is generally expected from hierarchical structure formation models .
[ [ section-20 ] ] present observations thus indicate a paucity of mass overdensities ( clusters , in the cosmological sense ) which do not have a red sequence of early - type galaxies , even at @xmath10 .
while it still remains possible that such clusters do exist , perhaps at some very early , low - mass stage of cluster assembly , it seems likely that they account for only a very small fraction of all clusters . clearly , any search for clusters using the crs method will tend to be biased against such clusters .
paper ii discusses this in further detail , by examining the effect of variations of the blue fraction on detection efficiency . in general
, it seems eminently reasonable to search for clusters using the red sequence .
the implementation of the crs method considers four main parameters for each detected galaxy in a survey : positions @xmath45 and @xmath46 , magnitude @xmath47 , and color @xmath48 .
the color has an associated color uncertainty , @xmath49 , and the positions and magnitude are considered to be error - free .
photometric errors on the magnitude are ignored because the red sequence is nearly horizontal in a color - magnitude diagram ( cmd ) . from the four input parameters for a set of observed galaxies , we wish to locate those positions which likely
correspond to a galaxy cluster , and estimate a redshift , @xmath18 . in summary , this is done as follows : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \1 ) define a series of overlapping color slices in the @xmath48-@xmath47 plane , guided by a fiducial red sequence model .
\2 ) select a subset of all galaxies as belonging to each slice , based on on the probability ( evaluated from @xmath47 , @xmath48 , @xmath49 and the model ) that each galaxy belongs to that slice .
\3 ) for a given subset , compute weights for all selected galaxies .
the weight is based both on the galaxy magnitude , and the probability that each selected galaxy belongs to that slice .
\4 ) compute the weighted surface density of galaxy positions .
all the slices taken together define a form of volume density in @xmath45 , @xmath46 and redshift .
\5 ) identify the peaks in this volume and select clusters candidates using some significance cut .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the detailed implementation of this method is described in the following subsections .
[ [ section-21 ] ] first , the expected photometric behavior of the red sequence as a function of redshift is determined from a default model of the red sequence , which provides the color and slope of the red sequence as a function of redshift . in the context of typical galaxy spectral synthesis codes , defining the model essentially consists of selecting a cluster formation redshift , @xmath50 , and deducing a luminosity - metallicity calibration so that the observed slope of the red sequence at low redshift is reproduced ( see 5 for an example ) .
the details of the model are relatively unimportant , as any application to a real survey should include a verification of the model colors by comparison to observations of known clusters over the redshift range of interest .
even if the model is not particularily well calibrated , errors in the model colors will tend to manifest as biases in secondary derived parameters such as the redshift , and will have a minimal effect on the @xmath11 @xmath12 chance of detecting a given cluster . moreover , minor errors in the model slopes are likely to have a minimal effect , since the slopes are small to begin with .
[ [ section-22 ] ] on the basis of this model , the color - magnitude diagram of the entire sample is divided into a number of regions bordered by the expected red sequence at various values of @xmath18 .
the choice of these bounding @xmath18 values is driven by the data . over the redshift range set by the filter combination , bounding red sequences are placed at color separations , @xmath51 , corresponding to : @xmath52 where @xmath53 is the measured color error of galaxies at a given color and magnitude .
this fiducial color error is evaluated at values of @xmath48 and @xmath47 which correspond to a fixed absolute magnitude ( @xmath54 is used here ) at various redshifts .
clearly , the color slices can be thought of as redshift slices , with a width in redshift , @xmath55 , which is a function of @xmath51 and the details of how color is mapped onto redshift .
the second term , @xmath56 , is the intrinsic scatter in color seen in cluster red sequences at moderate redshifts .
a value of @xmath57 is used here ( barrientos 1999 ; sed98 ) .
each `` color slice '' defined in this manner thus has a width sized relative to both the expected photometric scatter and the intrinsic scatter of the red sequence at the chosen redshift . in practice ,
the color slices are actually overlapped to ensure that no clusters are poorly sampled by being on the border between two slices .
figure 3 shows an example of the construction of color slices using data from the cnoc2 redshift survey patch cnoc0223 + 00 ( yee et al .
the red - sequence model used in figure 3 is described in detail in 5 .
[ [ section-23 ] ] for each color slice , defined by an upper and lower bounding red sequence , the probability that each galaxy in the sample belongs to that slice is then computed .
this is done by assuming that the true colors are distributed normally about the measured colors , with gaussian widths given by the measured color errors .
the integration of each presumed error distribution in the desired color range then gives the probability that any given galaxy s true color places it in the slice .
a subset of galaxies based on a probability cut is then selected for each color slice .
a probability cut of 10% has been used here .
this value represents a compromise between sampling deeply in to the available data ( as all faint galaxies have low probabilities of belonging to a given color slice because of large color errors ) and computational efficiency ( which improves with the smaller number of galaxies selected by a higher probability cut ) .
an example of this probability assessment , using the same data and models as in figure 3 , is shown in figure 4 .
galaxies which pass the probability cut trace out an envelope which is essentially as wide as the color slice at bright magnitudes , but widens to fainter magnitudes as the size of the color errors becomes comparable to the width of the slice .
apparent outliers at brighter magnitudes are galaxies which have larger than typical color errors , often due to object crowding effects .
[ [ section-24 ] ] it is well known that field galaxy counts are in general steeper than the faint end of the typical cluster luminosity function ( e.g. , lin et al .
1999 ; lpez - cruz & yee 2000b ) .
thus , at fainter magnitudes in a given color slice the field counts will eventually overwhelm the counts from any clusters .
without color selection , a similar effect happens toward brighter magnitudes , as the cluster counts are overwhelmed by counts from lower redshift galaxies .
however , in any given color slice , because lower redshift galaxies are excluded by the color cut , any cluster elliptical galaxies likely dominate the bright end of a particular color slice , and so this effect is not expected at the bright end . this recognition that cluster galaxies of different magnitudes have a differing contrast against the background mandates the use of magnitude - based weighting of the galaxies , in addition to the weights taken directly from the color - based probability assessment discussed in 4.2 .
[ [ section-25 ] ] the relative weight which should be assigned to a galaxy of a given magnitude can be deduced from the data , without reference to models of the cluster and field luminosity functions .
this is done by computing the surface density of all objects in the color slice ( with probability weights based only on the colors [ 4.2 ] , and using the density estimator described in 4.4 below ) .
galaxies located at the most significant peaks in the density are then designated as cluster galaxies ( with field contamination ) and the rest are designated as field galaxies . a comparison of these two samples of galaxies , after correcting for the field contamination in the cluster sample , then allows one to evaluate the probability that a galaxy of a given magnitude in the uncorrected cluster sample is in fact a cluster member .
put simply , if @xmath58 is the field - corrected , area - normalized counts for the cluster galaxies and @xmath59 is similarly for the field galaxies , then the probability that a galaxy of a given absolute magnitude @xmath60 in the cluster sample is in fact a cluster member is : @xmath61 note that the early - type galaxy model is used to compute the evolved value of m@xmath24 , with the measured m@xmath24 of lcy00 taken as the local value .
[ [ section-26 ] ] the construction of @xmath62 will depend on the available data .
if each color slice contains enough data that a reliable @xmath62 can be computed , it is best to find @xmath62 on a slice - by - slice basis .
this then allows for variations in the relationship between clusters and the field with color ( i.e. , redshift ) .
however , in the data discussed in 5 ( and illustrated in figure 5 ) the survey area is not large enough , and so @xmath62 has been computed cumulatively over the entire color range considered .
furthermore , it is clear that for a particular cluster , @xmath62 is not absolute , as it will scale with the richness of that cluster .
however , generally the variation in the ratio between values of @xmath63 at different @xmath60 is small , over a factor of a few in cluster richness counts .
furthermore , since the originally identified cluster galaxy sample is to first order representative of true clusters in the sample , this probability is adequate .
an illustration of the computation of @xmath62 , using initial peaks with a range of significances ( and hence richnesses ) , is shown in figure 5 .
it shows the run of @xmath62 with magnitude for the data discussed in 5 , using two different significance cuts to identify initial peaks , and hence the `` cluster '' subsample .
the size of the two different `` cluster '' subsamples differ by an order of magnitude , in that they represent 2% and 20% of all galaxies in the sample , and yet the difference between the relative values of @xmath62 , expressed by the slope of linear fits to the @xmath62 vs. @xmath60 relation , is only a factor of 1.5 . based on this
, we use a cut of 10% in the computations presented in 5 , as this adequately spans the extreme values illustrated in figure 5 .
moreover , note that in any sample @xmath62 will be dominated by the contribution of galaxies in the poorest systems selected by the chosen cut , as these systems contain a much larger fraction of all galaxies .
therefore , regardless of the survey size , or the chosen cut , @xmath62 is naturally weighted towards the poorer systems , for which proper weighting is more important .
[ [ section-27 ] ] an additional modification is made to @xmath62 , in which @xmath62 for all galaxies brighter than the expected cluster @xmath54 is set equal to @xmath64 .
this correction is based on the observation that without such a modification , an unacceptable level of stochastic noise is introduced from the few bright galaxies in a given slice . with a final @xmath62 computed , a weight , @xmath65 , is then assigned to each galaxy in a given subset .
the assigned weight is simply the product of the probability of being included in the color - slice subset ( subject to some lower threshold as detailed in 4.2 ) and @xmath62 .
[ [ section-28 ] ] with a list of positions , @xmath45 and @xmath46 , and associated weights , @xmath65 , now in hand , the next step is to evaluate the surface density of objects . we have considered a number of estimators for this , and have settled on simple fixed - kernel smoothing . we considered in detail the more complicated adaptive kernel smoothing used by gal et al . ( 2000 ) , but have decided that this estimator is actually inappropriate for this particular application .
fundamentally , the tendency of the adaptive kernel to break up regions of high density ( i.e. , clusters in the data ) into smaller features may not be desirable .
estimators such as the adaptive kernel are motivated on mathematical grounds ( vio et al .
1994 ) and may not be appropriate to every single physical situation , as is the case here .
though appealing in general , we have found that the benefit of the adaptive smoothing is outweighed by the additional computational cost in this case .
[ [ section-29 ] ] the fixed - kernel estimator is used to produce a surface density map of the galaxies in the color slice .
this map is simply the weighted convolution of the kernel with the data points at positions @xmath45 , @xmath46 , evaluated over a grid spanning the range of the data ( see vio et al . for details ) .
hereafter , the value of the kernel - smoothed density at a particular location @xmath66 in the map is referred to as @xmath67 .
two different forms were considered for the smoothing kernel : @xmath68 and @xmath69 . when appropriately scaled , both give profiles akin to several standard forms used to describe cluster radial density profiles .
however , experiments on data from the cnoc2 project ( see 5 ) showed the former kernel to be better suited , as the cuspy core of the latter kernel tends to produce density maps with spurious strong peaks associated with individual bright ( and hence large weight ) galaxies which lie at the kernel cusp .
the final form of the kernel used is : @xmath70 where @xmath71 is the proper distance from the kernel center . in this normalization
, the radius @xmath71 is expressed in units of the scale radius of the more familiar navarro , frenk , & white ( 1997 ) ( nfw ) profile .
the exponential constant of 1.965 is derived by requiring that the adopted profile fit the nfw profile at intermediate radii , and is constant provided that @xmath71 is expressed in units of the nfw scale radius .
the relation between the two profiles considered and several other standard forms for cluster density profiles is given in figure 6 .
the various data for the standard profiles are taken from adami et al .
( 1998 ) . also , in equation 3 , @xmath72 is a normalizing factor chosen so that : @xmath73 a scale radius of 0.33 @xmath74 mpc has been used for the tests described in 5 below , a value suggested by measurements of the cnoc1 cluster sample ( carlberg et al .
1996 , 1997b ) . the maximum radius , @xmath75 , is taken as four times the scale radius .
it should be noted that the inability of other authors to strongly distinguish between possible profile shapes ( e.g. , king , hubble or nfw [ adami et al .
1998 ] ; hernquist or nfw [ carlberg , yee , & ellingson 1997 ] ) even with large cluster samples indicates that the exact shape of the kernel is not of overwhelming significance .
lubin & postman ( 1996 ) similarly found that the precise shape of the radial filter in the pdcs had little effect on the detectability of clusters in that survey .
[ [ section-30 ] ] prior to the merging of the individual 2d density maps into a data cube , the @xmath67s must be transformed to some standard measure ( such as @xmath37 of detection ) .
this is necessary because the meaning of a given density value changes between the slices ( both because of the apparent change in the size of the fixed - metric kernel , and because of the change in the mean density of objects relative to the kernel size ) . as in gal et al .
( 2000 ) , the bootstrap technique is used to evaluate the significance of peaks relative to the background distribution .
however , it is worth noting that a simple application of the bootstrap technique is likely not correct , as one of the fundamental underpinnings of the bootstrap process is the assumption that the data points to be analyzed are independently and identically distributed ( press et al . 1992 ) . in data presumed to contain clusters
this is clearly not true .
however , as the clusters occupy only a small fraction of the total surface area of any map , it is possible to exclude some fraction of the highest valued regions ( and lowest , to preserve symmetry ) of the map .
the same regions in the bootstrap realization maps of the data ( acquired by sampling the data with replacement and re - running the kernel smoothing ) are excluded in turn .
the probability distribution function ( pdf ) for density values in the remaining areas of the bootstrap maps then closely matches the background distribution of the initial map .
experimentation shows that an exclusion cut of only the 10% highest and lowest density values is enough to solve this problem .
figure 7 illustrates the result of this process , using the density map from a single color slice of the data discussed in 5 .
several derived background distributions of @xmath67 are shown .
the first , and best fitting distribution ( solid line ) is based on the bootstrap method . for comparison , another distribution ( dashed line ) , constructed by replacing the correlated galaxy positions with random positions ( but keeping the same weights as the real data ) is also shown . obviously , the restricted bootstrap dataset provides a good representation of the background distribution found in the input data , while the dataset based on random positions does not .
[ [ section-31 ] ] the probability of a given value of @xmath67 occurring at random can then be evaluated from the restricted bootstrap dataset .
this is best done by a direct analysis of the bootstrap pdf as it is often non - normal .
each value of @xmath67 in the input map is transformed into a probability , @xmath76 , by evaluating the likelihood that a value of @xmath67 or greater occurs in the the restricted bootstrap dataset . in other words : @xmath77 the raw probabilities can then be transformed to more familiar gaussian sigmas if desired , though such a change is essentially cosmetic .
[ [ section-32 ] ] once probability maps have been constructed for each slice , the slices can be stacked together to form a data - cube .
the generation of a cluster catalog from this data - cube is then simply a matter of selecting high probability regions , subject to some predetermined probability cut .
the details of how to do this optimally will not be described here , as the process is data dependent , though not complicated .
a further exploration of this topic , in the context of a particular survey and detailed simulated data , can be found in paper ii ( gladders & yee 2000b ) , or in 5 below .
it should be noted that other authors have used relatively simple algorithms in this final step with good success ( postman et al . 1996 ) ; similar simple algorithms should work as well .
[ [ section-33 ] ] with a cluster candidate identified , the final step is to estimate a redshift for each candidate .
this is done directly from the above data - cube , as the redshift of a given cluster candidate is encoded by the location of the peak probability along the color axis of the stacked density maps at the @xmath45-@xmath46 position of the candidate .
[ [ section-34 ] ] the cnoc2 dataset ( yee et al .
2000b ) provides an useful testbed for cluster - finding algorithms , and has kindly been provided in advance of publication by the rest of the cnoc2 group .
this intermediate redshift field galaxy survey consists of 4 patches of data .
each patch is @xmath78 deg@xmath79 in size .
the surveys consists of @xmath80 imaging and spectroscopy for a large sample of galaxies ( @xmath140,000 in the photometric catalog to the 100% completeness limit of @xmath20=23 , with @xmath16000 spectroscopic redshifts for galaxies typically as faint as @xmath20=21.5 ) .
however , the layout of each patch is complex , as the patches were designed to sample the correlation function , rather than to minimize patch edges , and so are not optimal for cluster detections .
additionally , the cnoc2 redshift catalog is sparsely sampled , and is somewhat biased against detecting compact structures ( due to the unavoidable exclusion of an above average number of galaxies in dense regions when using a multi - object spectrograph
see yee et al .
2000b for details ) .
nevertheless , the availability of a large spectroscopic sample outweighs any such considerations , and makes the cnoc2 survey an indispensable aid in assessing the suitability of the crs method .
note that in considering the output of the photometric cluster - finding process , it is necessary to make some analysis of the cnoc2 redshift sample in order to identify real bound structures in the redshift data .
this process is complex , particularly because of the sparseness of the redshift sampling , and the complex weighting functions incurred by the use of a multi - object spectrograph .
a global and rigorous analysis of bound structures in the cnoc2 redshift survey data will be presented by the cnoc2 group ( carlberg et al .
this paper uses a simplified process , described below , to provide a first - pass comparison with the crs method .
[ [ section-35 ] ] to test the crs method , a dataset was constructed over the entire cnoc2 survey area using a subset of the available photometry .
only fields in which @xmath81 magnitudes , @xmath82 colors , @xmath83 color errors , and @xmath45 and @xmath46 positions are available were used , and these provide the input to the crs algorithm . the total area covered by
this subset is 1.48 square degrees . over the cnoc2
spectroscopically unbiased redshift range ( @xmath84 ) , this area corresponds to a volume of @xmath85 @xmath86 mpc@xmath87 .
note that rich clusters occur with a density of about 1 every @xmath88 @xmath86 mpc@xmath87 ( e.g. , bramel , nichol & pope 2000 ) , so we expect to find only poor clusters and groups in this volume .
the crs algorithm was then run across all four patches , with the redshift range restricted to @xmath89 ( somewhat larger than the redshift range over which spectroscopic confirmation of the clusters can be expected ) .
the red - sequence model used is based on a 0.5 gyr burst ( drawn from the gissel library , bruzual & charlot 1993 ) completed at z=3 , evolving to lower redshifts with a metallicity - luminosity relationship which reproduces the red - sequence slope in the coma cluster at @xmath2 ( gladders et al .
a portion of the results of the crs cluster - finding is shown in figure 8 for the cnoc0223 + 00 patch .
a number of significant sources are obviously detected . in these data values of @xmath67
have been re - cast in gaussian sigmas , referred to hereafter as @xmath90 .
[ [ section-36 ] ] individual sources were then selected from the data - cubes using the clump finding algorithm of williams , de geus , & blitz ( 1994 ) .
this algorithm was originally developed for locating structure in position - position - velocity radio data .
williams et al .
discuss the accuracy of their clump finding technique in detail ; in summary , it works by contouring the data in 3-d at multiple levels , and following clumps through the various levels from an initial peak value , identify sub - clumps as appropriate . because the algorithm assumes no clump profile , it is readily applied to the crs position - position - redshift data - cubes . via extensive simulations , williams et al .
recommend that the data be contoured at levels which are separated by @xmath91=@xmath92 . in the original application
, @xmath93 is the root - mean - square error in the temperature of the position - position - velocity data .
a similar statistic can be estimated from the crs data - cubes by examining random realizations of the position and weight data used to construct the data - cubes .
these random realizations , which are distinct from the bootstrap realizations discussed in 4.4 , have the same weights as the original input data but randomized positions , and are used within the algorithm primarily to account for irregular patch edges . in the particular cnoc2 data discussed here ,
the rms of the random realizations is about @xmath94=0.55 .
the rms is not directly indicated by the distribution of @xmath90s in the real data - cube as a whole ( i.e. the bootstrap normalized data - cube , cast in @xmath90s , does @xmath95 have a noise of @xmath94=1 ) , as this represents a scaling relative to the distribution of the data as a whole , including all real structures . hence , for the data discussed here , the williams et al .
clump - finder was run using @xmath96=1.1 , down to a contour level of @xmath97 .
this means that the lowest significance peaks detected have a maximum significance of 3.5@xmath37 ( 1 contour above the lowest contour level ) and all detected peaks , regardless of maximum significance , are traced to 2.4@xmath37 .
[ [ section-37 ] ] for each selected clump in the data - cube , a redshift was estimated from the position of the clump maximum projected along the @xmath18-axis of the data - cube .
all cnoc2 galaxies with spectroscopic redshifts within a projected distance of 0.5 @xmath74 mpc of the projected clump in @xmath45 and @xmath46 were then selected . a simple friends - of - friends analysis of these galaxies ( compared to the entire redshift sample )
was then performed .
a fixed linking length of 0.5 @xmath74 mpc in angular space , and 400 km s@xmath44 in redshift space was used throughout .
this simple analysis serves to identify likely associated objects in the redshift sample .
a visual examination of the redshift distribution of the selected subset of galaxies , taking into account both the distance from the clump center and the number of friends each galaxy has is then used to select ( if possible ) a structure corresponding to the identified clump . in cases where more than one structure appears in the redshift subsample , the structure closest to the @xmath45-@xmath46 position
is used . in such cases , there are often other identified clumps which appear associated with the other structures , but untangling the correspondence between various clumps and redshift structures is straightforward .
figure 9 shows three examples of this process using data from a portion of the central block of the patch cnoc0223 + 00 ( the right hand portion of the @xmath90 maps shown in figure 8) .
there are three significant peaks in this region , all in the redshift interval @xmath98 .
each is readily associated with a different redshift structure , despite the close proximity of the peaks both in redshift and angle .
the first two of these example redshift structures are particularly close separated by only 1700 kms@xmath44 in redshift , and 0.9 @xmath74 mpc projected on the sky yet each is clearly identified as a separate redshift structure , and as a separate crs peak .
[ [ section-38 ] ] figure 10 presents the comparison of the crs algorithm output to the cnoc2 redshift sample .
crs cluster candidates both with and without a redshift counterpart are shown .
crs cluster candidates with peak significance in either the first or last color slice , indicating that the predicted redshift may be in error , are flagged .
this possible error is small at the lower redshift end ( where the color errors are small and hence color slices are narrow in redshift ) but could be quite significant at the upper redshift end .
the photometric redshifts have been adjusted by a quadratic fit to an initial comparison between the spectroscopic and raw photometric redshifts in order to account for any uncertainties in the early - type galaxy spectral energy distribution model . as mentioned
4.1 , any real cluster survey is likely to include the training data necessary for this calibration process .
several points of note are apparent from figure 10 .
[ [ section-39 ] ] first , considering the subsample at @xmath99 , the false - positive rate is very low , with only 1 of 23 candidates not detected in the spectroscopic sample , and even that one is at a relatively high @xmath18 of @xmath100 . at @xmath8 , the apparent false - positive
rate is much higher , with 4 of 11 candidates not showing any spectroscopic counterparts .
this is likely a result of incompleteness in the redshift sample . as the cnoc2 redshift data is sparsely sampled , very poor cluster / group systems ( with only a few bright galaxies in total ) , or systems in the higher redshift range ( with only a few bright galaxies in the spectroscopic sample ) may not be well represented .
[ [ section-40 ] ] second , the redshift accuracy of the method is extremely good , bearing out suggestions made in previous work ( lcy00 ; smail et al .
the standard deviation of the difference between the color - estimated redshifts and the spectroscopic redshifts at @xmath99 is only @xmath101 .
the apparent redshift scatter at @xmath8 is much higher , with one nominally confirmed candidate with a redshift error of 0.18 .
this particular candidate is found only in the highest redshift slice ( as are most unidentified candidates ) , and this large redshift error may thus be due to a misidentification of a system at a real redshift beyond the spectroscopic reach of cnoc2 which has been falsely identified with a lower redshift system which is along a similar line of sight and is too poor to have been found by the crs method as applied here .
[ [ section-41 ] ] regardless , the results at @xmath99 clearly indicate that the crs method is very successful in identifying real overdensities , and that the redshift estimates are extremely accurate .
there are three reasons for the ability of the crs method to obtain such high accuracy using only two filters : 1 ) the use of a known and empirically calibrated spectral enery distribution , 2 ) the fact that e / s0 galaxies have the strongest spectral breaks and hence provide a strong color signature with redshift , and 3 ) the estimation of the redshift for any given cluster by using colors for a number of individual galaxies simultaneously .
[ [ section-42 ] ] finally , it should be recalled that the volume probed by the survey considered here is quite small ( in cluster terms ) and that it is expected that most of the systems identified above should be quite poor .
this expectation is well borne out by a weak - lensing examination of several cnoc2 fields ( hoekstra , franx & kuijken 1999b ) and a dynamical analysis of the groups in the cnoc2 survey ( carlberg et al .
these independent analyses indicate that most of the systems have 1-d velocity dispersions of only 200 - 400 km s@xmath44 , well below the lower end of abell richness class 0 ( @xmath102 kms@xmath44 , yee et al .
2000a ) , and generally much poorer than systems identified using previous cluster finding algorithms .
though it is almost certain that the crs results shown here do not represent a 100% complete sample at such masses , the success in finding @xmath103 such low - mass systems at these intermediate redshifts using only 2-color photometry is a strong validation of the general method .
in addition , the extremely low ( @xmath104% ) false positive rate compares very favorably to the generally higher contamination in optical cluster catalogs constructed by other means ( e.g. , oke et al . 1998 ) .
furthermore , the upper limit of redshift explored here was selected to reflect the spectroscopic limits of the cnoc2 survey ; there is no reason why the crs method can not be applied at yet higher redshifts . limiting the required wide - field data to that from ccds ,
a reasonable upper limit is about @xmath105 corresponding to having the 4000break at the red end of typical ccd sensitivities .
[ [ section-43 ] ] a number of refinements or modifications of the general method described above are possible . for example , given enough information to construct appropriate field and cluster models , one could replace the slicing , smoothing and stacking approach used above by a maximum - likelihood based approach ( postman et al .
1996 ; kepner et al .
such models can likely be deduced internally from a large enough cluster - search dataset . additionally , entirely different density estimators could be used .
for example , the voronoi tessellation approach of kim et al .
( 1999 ) is likely applicable with only minor modification of the general algorithm described above .
many of the approaches used in x - ray astronomy to detect diffuse collections of points embedded in a background ( e.g. , campana et al .
1999 , and references therein ) are applicable here , with the galaxies treated analogously to single photons .
the purpose of this paper is not to lay out the definitive algorithm for using the red sequence to locate clusters .
rather , the intention has been to describe a particular , and well - motivated algorithm which can be used to demonstrate the effectiveness of the basic premise of the crs method applied to real data .
[ [ section-44 ] ] one major modification of the basic method outlined here would be the addition of morphological information .
it is possible , even with ground based imaging , to deduce the likely morphological classification of a galaxy using simple and computationally tractable estimators ( e.g. , abraham et al .
specifically , one can assign a probability of being an early - type galaxy to all galaxies in a survey .
including such analysis into the algorithm described above would be relatively simple , as this probability can be incorporated into the weights assigned to each galaxy .
however , constructing the morphological probability assessment of a given galaxy in real imaging data is a significant task , as this assessment is a function of numerous factors ( i.e. , the seeing , the sampling , the depth of the imaging relative to the galaxy magnitude , the galaxy redshift and intrinsic size etc . ) .
furthermore , in the cnoc2 data discussed above , the imaging is also affected by serious camera distortions .
we have thus not attempted to incorporate morphological data into the above analysis .
however , given the basic premise of the crs method , it is obvious that morphological filtering will be of benefit .
numerous large studies to @xmath30 have now demonstrated that the red sequence of early - type galaxies is a universal and homogeneous feature of galaxy clusters . based on this strong observational fact
, we have developed a new technique , the crs method , for locating clusters in 2-filter optical or infrared imaging data . using appropriate filters
, the crs method circumvents problems with foreground , and to some extent background , projection effects which have been so problematic in previous optical cluster surveys .
a specific algorithm has been constructed to implement this general method .
this algorithm constructs 2-d smoothed density maps of color slices corresponding to redshifts bin , which are then normalized and reassembled into a 3-d density data - cube .
peaks in this data - cube are then defined as cluster candidates . using data from the cnoc2 survey
we have tested the crs method over a significant redshift range . from this test , and the general data on the evolution of cluster early - type galaxies available in the literature , we conclude the following : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \1 ) the apparent universality of the red sequence in available data , even to redshifts greater than one , indicates that the crs method should be applicable to the entire redshift range accessible to ccds ( @xmath106 ) .
\2 ) the crs method is a powerful new approach to finding galaxy clusters in imaging data at optical and infrared wavelengths .
the performance of the method shows that it is able to reliably detect rather poor bound structures ( @xmath107 km s@xmath44 ) with a very low contamination rate of less than 5% , using data of only moderate depth .
\3 ) the redshift accuracy of the method is @xmath108 over the redshift range @xmath109 , or equivalently 10% in redshift .
this accuracy is comparable to that achieved for single galaxies using four ( or more ) filter photometric redshift methods . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we have begun a large survey , the rcs , designed to exploit the power of the crs method ( e.g. , gladders & yee 2000a ) .
the primary survey data are 100 square degrees of imaging in two filters ( @xmath110 and @xmath111 ) , to the depth necessary to find clusters to @xmath105 .
paper ii of this series ( gladders & yee 2000b ) discusses the application of the crs method to the rcs data specifically , and further explores the applicability and completeness of the method to redshifts as high as @xmath105 . unlike this paper ,
the main results of paper ii are based on a broad and comprehensive set of tests performed on simulated data , created using detailed and realistic galaxy and cluster models .
thanks the natural sciences and engineering research council ( nserc ) of canada for support via pgsa and pgsb graduate scholarships .
thanks nserc for support via an operating grant .
thanks ray carlberg for numerous useful discussion on statistics and the bootstrap method .
we are particularly indebted to the cnoc2 group for providing access to the cnoc2 dataset in advance of publication .
we also thank the referee , bob nichol , for his detailed comments . | numerous methods for finding clusters at moderate to high redshifts have been proposed in recent years , at wavelengths ranging from radio to x - rays . in this paper
we describe a new method for detecting clusters in two - band optical / near - ir imaging data .
the method relies upon the observation that all rich clusters , at all redshifts observed so far , appear to have a red sequence of early - type galaxies .
the emerging picture is that all rich clusters contain a core population of passively evolving elliptical galaxies which are coeval and formed at high redshifts .
the proposed search method exploits this strong empirical fact by using the red sequence as a direct indicator of overdensity .
the fundamental advantage of this approach is that with appropriate filters , cluster elliptical galaxies at a given redshift are redder than all normal galaxies at lower redshifts .
a simple color cut thus virtually eliminates all foreground contamination , even at significant redshifts . in this paper ,
one of a series of two , we describe the underlying assumptions and basic techniques of the method in detail , and contrast the method with those used by other authors .
we provide a brief demonstration of the effectiveness of the technique using a real photometric sample with redshift data , and from this conclude that the method offers a powerful yet simple way of identify galaxy clusters .
we find that the method can reliably detect structures to masses as small as groups with velocity dispersions of only @xmath0 , with redshifts for all detected structures estimated to an accuray of @xmath110% . |
multicore processors have become the norm in many execution platforms in various fields .
such architectures come with a cache memory hierarchy made of several levels , shared or not between cores .
figure [ cache - archi ] represents the typical cache hierarchy that can be found in a multicore architecture with two levels of cache , noted @xmath1 and @xmath2 where @xmath3 is the number of the core ( in the figure @xmath3 ranges from 1 to 4 ) . in such architectures ,
the @xmath2 cache is larger but provides slower access time than the @xmath0 cache .
generally , the @xmath4 cache is unified and shared among all cores , while at the @xmath5 level data and instruction caches are separated and private to each core .
however , in this work we focus only on the data caches .
using such multicore architectures for developing hard real - time systems is an important research area .
multicore are generally tuned for optimizing performance for the average case , while hard real - time systems must consider worst - case scenarios due to certification constraints .
a major problem lies in the management of cache for mastering the impact of conflicts on the worst - case execution time ( wcet ) of each task .
designing cache - aware scheduling policies is becoming a popular research area . in this work ,
we show _ how a static scheduling that minimizes @xmath0 data cache misses between hard real - time tasks on a multicore architecture can be computed_.
@xcite focuses on the memory - to-@xmath2 traffic in the cache hierarchy of soft real - time systems .
they propose a two steps method to discourage the co - scheduling of the tasks generating such traffic .
first , the tasks that may induce significant memory - to-@xmath2 traffic are gathered into groups . then at runtime , they use a scheduling policy that reduces concurrency within groups .
@xcite also proposes several global multi - core scheduling strategies for soft real - time systems to minimize the @xmath2 cache trashing .
co - scheduling of the tasks of a same group is used to optimize the efficient use of the @xmath2 shared cache .
task promotion is another example of a studied scheduling policy .
when considering hard real - time systems , to the best of our knowledge we are only aware of @xcite .
cache - partitioning techniques are used to avoid interferences , at the @xmath2 cache level , between the tasks that are running simultaneously .
in addition to regular temporal constraints used within a schedulability test , cache constraints due to cache - partitionning are added and steer the computation of the scheduling .
they propose a linear programming formulation to solve this problem and an approximation of this formulation for larger task sets .
the closest work to ours is @xcite .
while proposed cache - aware scheduling strategies are evaluated using a soft - real time kernel , the results can also be used for hard real - time systems .
they propose a bin packing approach to evenly distribute the working set size ( wss ) of the tasks on all cores in order to reduce conflicts .
the resolution algorithm is based on the next fit decreasing heuristic applied on the tasks ordered by their decreasing wss .
besides , they rely on a notion of distance between caches of non - uniform memory architectures to further optimize the solution . this is only used for the tasks that share some common memory area and are gathered into groups .
however , it is unclear how the common memory area defines a group as well as how groups are reduced when the heuristic fails to allocate a group . to summarize ,
most of the existing cache - aware scheduling proposal have focused on the efficient use of the @xmath2 cache .
let @xmath6 be a set of @xmath7 independent , synchronous , preemptible and periodic tasks .
@xmath8 is handled using the implicit deadline periodic task model .
each task @xmath9 has the following temporal parameters @xmath10 .
@xmath11 is the period of the task and @xmath12 is the wcet .
a job @xmath13 represents an instance of a task with @xmath12 its wcet .
let @xmath14 be the hyper - period of task set .
it equals to the least common multiple of all periods of tasks in @xmath8 . as in @xcite
, the hyper - period @xmath14 is divided in intervals , an interval being delimited by two task releases .
a job can be present on several intervals , and we note @xmath15 the weight of job @xmath16 on interval @xmath17 .
we denote by @xmath18 the set of intervals and @xmath19 the duration of the @xmath20 interval , @xmath21 .
the weight of each job is the amount of processor necessary to execute job @xmath13 on interval @xmath19 only ( it is not an execution time but a fraction of it ) .
@xmath22 is the job set of all jobs of @xmath23 scheduled during the hyper - period @xmath14 .
then , temporal schedulability constraints are expressed using a linear program described in @xcite to compute the optimal job weights on each interval for all tasks @xmath24 .
first , the sum of all job weights on an interval does not exceed the processor maximum capacity : @xmath25 then each job weight does not exceed each processor maximum capacity : @xmath26 finally , jobs must be completely executed : @xmath27
the problem we address in this work is to reduce @xmath5 data cache misses when scheduling hard real - time tasks on a multicore architecture . to this end , we maximize the co - scheduling on a same core of tasks that exchange data while still ensuring temporal schedulability constraints . we assume a static knowledge of data exchange between the tasks of an application .
therefore , we extend the classical periodic task model with the wss parameter for each task and model this problem using a variant of the knapsack problem .
we also assume that the system is schedulable and hence we only seek to optimize the allocation of the tasks on the @xmath5 caches .
in addition , we suppose that the size of a @xmath5 cache enables to host several tasks simultaneously , a valid hypothesis in the case studies we consider .
we leave as future work the management of cache conflicts , using techniques such as in @xcite .
the multicore platform is made of @xmath28 cores and we note @xmath29 the capacity of each data cache @xmath0 ( we assume the @xmath0 caches to have an equal size ) .
finally , from the data of the application we can calculate @xmath30 which is the wss of job @xmath31 .
this is computed by adding the size of each data section from the binary of an application .
we note @xmath32 the affinity between the job @xmath13 and the job @xmath33 .
the affinity between two tasks is defined as the number of communication flows between them .
higher is the number of communication flow , higher is the affinity between two tasks .
communication flows between tasks are extracted using the software architecture of the considered hard real - time application .
then , integrating into the previously described schedulability constraints require to introduce a decision variable representing the allocation of the tasks on cores for each intervals .
let @xmath34 be this decision variable that is equal to @xmath35 if the job @xmath13 is assigned to cache @xmath16 during time interval @xmath36 and @xmath37 otherwise .
the sum of the wss of the jobs allocated to a cache should not exceed its capacity : @xmath38 in addition , each job must be assigned to a single cache : @xmath39 besides , to link the temporal schedulability constraints with the aforementioned cache constraints , the following relationship can be define : if @xmath40 then @xmath41 and if @xmath42 then @xmath43
. finally , since our aim is to maximise affinity @xmath0 , we obtain the following objective function : @xmath44
in this work , we show how we can extend classical ( temporal ) schedulability constraints to minimize @xmath0 data cache miss between communicating hard real - time tasks on a multicore architecture . in future work , we plan to generalise the formulation to address other level of a cache memory hierarchy . as our formulation of the problem uses a _
quadratic knapsack _ , known to be _ np - hard _
@xcite , another next step is therefore the linearization of the objective function .
then , we plan to implement it using the cplex solver to generate the static scheduling of hard real - time tasks .
finally , we plan to test our method on several hard real - time industrial applications .
m. lemerre , v. david , c. aussagus , and g. vidal - naquet , _ equivalence between schedule representations : theory and applications _ , proc . of the ieee real - time and embedded technology and applications symp . , 2008 . | the growing need for continuous processing capabilities has led to the development of multicore systems with a complex cache hierarchy .
such multicore systems are generally designed for improving the performance in average case , while hard real - time systems must consider worst - case scenarios .
an open challenge is therefore to efficiently schedule hard real - time tasks on a multicore architecture . in this work ,
we propose a mathematical formulation for computing a static scheduling that minimize @xmath0 data cache misses between hard real - time tasks on a multicore architecture using communication affinities . |
this talk is based on the work in ref . .
i am grateful to the organizers of the sixth alexander friedmann international seminar on gravitation and cosmology , for inviting me to this interesting conference .
t. d. li and c. n. yang , _ phys .
rev . _ * 104 * , 254 ( 1956 ) ; y. kobzarev , l. okun and i. pomeranchuk , _ yad . fiz . _ * 3 * , 1154 ( 1966 ) ; s. blinnikov and m. khlopov , _ astron .
* 60 * , 632 ( 1983 ) ; b. holdom , _ phys .
_ * b166 * , 196 ( 1985 ) ; s. l. glashow , _ phys .
* b167 * , 35 ( 1986 ) ; m. khlopov _
_ , _ astron .
* 68 * , 42 ( 1991 ) ; r. foot , h. lew and r. volkas , _ phys .
lett . _ * b272 * , 67 ( 1991 ) ; r. foot , _ mod .
* a9 * , 169 ( 1994 ) ; r. foot and r. volkas , _ phys .
rev . _ * d52 * , 6595 ( 1995 ) ; z. silagadze , _ phys . at .
nucl . _ * 60 * , 272 ( 1997 ) ; s. blinnikov , astro - ph/9801015 , astro - ph/9902305 ; z. berezhiani , d. comelli and f. l. villante , _ phys .
lett . _ * b503 * , 362 ( 2001 ) , and references therein . | the cosmological ( upper ) limit on the peccei - quinn constant , related to the primordial oscillations of the axion field , can be relaxed for a mirror axion model .
the simple reason is that the mirror world is colder and so the behavior of the axion temperature - dependent mass is dominated by the contribution from the mirror sector .
so the coherent oscillations start earlier and correspondingly the axion mass density @xmath0 is reduced .
the cp violating term @xmath1 , in the qcd lagrangian , leads to observable effects ( e.g. non zero neutron electric dipole moment ) , experimentally not observed .
this fact is referred to as the strong cp problem ( for a general reference see ref . ) . in the peccei - quinn ( pq )
mechanism,@xcite which is the most appealing solution of this problem , the parameter @xmath2 becomes a dynamical field , the axion @xmath3 , whose potential is minimized for the cp even configuration @xmath4 .
the axion emerges as the pseudo - goldstone mode of a spontaneously broken global axial symmetry @xmath5 . here
@xmath6 is a constant , with dimension of energy , called the pq constant , and @xmath7 stands for the color anomaly of @xmath5 current.@xcite in the following , we will usually refer to the constant @xmath8 , which characterizes the axion phenomenology .
astrophysical considerations exclude all the value of @xmath9 up to @xmath10 gev.@xcite on the other hand , the cosmological limit , related to the primordial oscillations of the axion field , demands the upper bound @xmath11
gev.@xcite let us briefly review the origin of this last limit . in most axion models ,
the pq symmetry breaking occurs when a complex scalar field , @xmath12 , which carries pq charge , acquires a vacuum expectation value ( vev ) @xmath13 .
this occurs as the temperature of the universe cools down below the value of the pq temperature .
as said , the axion is identified with the angular degree of freedom @xmath14 . today
@xmath2 is settled in the cp - conserving minimum @xmath15 . however , after the pq symmetry breaking , as the temperature is of the order of the pq scale , the axion is massless and the initial value of the phase @xmath2 is chosen stochastically .
we indicate this initial value with @xmath16 . when the universe is cold enough for the qcd phase transition , instanton effects curve the potential so that the axion acquires a non zero mass @xmath17 , where @xmath18 is the zero temperature limit:@xcite @xmath19 the function @xmath20 for temperature @xmath21 ( @xmath22 mev is the qcd scale ) was calculated in ref . and can be approximated as:@xcite @xmath23 where @xmath24 and @xmath25 .
this power law is valid when @xmath26 while for @xmath27 we have @xmath28 and the mass becomes independent of the temperature .
we use this relation until @xmath29 , that is until @xmath30 mev , and assume that @xmath31 for @xmath32 .
then @xmath2 starts to roll down , slowed by the hubble expansion , following the equation : @xmath33 where @xmath34 is the hubble expansion rate .
when the curvature term dominates on the ( hubble ) friction term , @xmath2 begins to oscillate with the frequency @xmath35 .
this happens at the temperature value @xmath36 defined by the equation @xmath37 , where @xmath38 gev is the plank mass and @xmath39 is the effective number of the particle degrees of freedom at @xmath36 . for @xmath40 gev , 86.25 for 100 gev @xmath41 5 gev , 75.75 for 5 gev @xmath41 2 gev , 61.75 for 2 gev @xmath42 , 17.25 for @xmath43 100 mev , 10.75 for 100 mev @xmath41 1 mev .
] for temperatures @xmath44 , when the coherent oscillations commence , eq .
( [ theta ] ) can be considerably simplified , and leads to the quite simple result @xmath45 const , where @xmath46 is the axion number density .
in other words , the number of axions in a comoving volume remains in fact constant during the universe expansion .
we further assume that no entropy production takes place after the moment when the axion field begins to oscillate .
then the axion number density to the entropy density ratio , @xmath47 , remains constant at @xmath48 , and thus the present energy density of axions is @xmath49 , where @xmath50 is the present entropy density , @xmath51 and @xmath52 k is the cmb temperature . at the initial moment @xmath53
, the axion number density is @xmath54 .
comparing to the critical density @xmath55 gev@xmath56 , we obtain : @xmath57 where @xmath58 and @xmath59 gev . substituting ( [ f ] ) in the expression @xmath37 , which defines the temperature @xmath60
, we obtain : @xmath61 where @xmath62 and the uncertainties of about 20 percent are related to uncertainties in parameters @xmath63 and @xmath64 .
this estimation very weakly depends on @xmath65 and thus for the peccei - quinn scales of the cosmological relevance , starting from the astrophysical bound @xmath66 gev to @xmath67 order gut scale @xmath68 gev , @xmath60 changes within interval 2 gev 200 mev and hence @xmath69 .
this estimation is valid , however , until @xmath70 mev .
instead we have @xmath71 , @xmath72 and hence @xmath73 gev .
this condition holds for @xmath74 gev .
] therefore , we obtain : @xmath75 demanding @xmath76 , the natural choice @xmath77 , leads to @xmath78 .
one implies the _ rms _ average from a uniform distribution of initial values from @xmath79 to @xmath80 , @xmath81 .
thus , in this case we obtain an upper limit on @xmath82 .
however , if the pq symmetry were broken before or during inflation , then the value of @xmath16 is the same in the whole universe within the single inflationary patch , and in fact it is an arbitrary parameter selected by some random choice . ]
it is worth underlining that the only upper limit on the pq scale comes from cosmology .
this limit is a few order of magnitude less than the typical gut scale .
thus , due to cosmological considerations , it seems quite improbable to insert the axion in any gut model .
the obvious question is then : how universal is relation ( [ omegah2-st ] ) ? we will show that for a mirror axion model this can change and ,
consequently , the cosmological bound can be relaxed .
the idea of mirror world was introduced many years ago and has been successfully applied in different sectors of astrophysics and cosmology.@xcite the mirror world consists of another , parallel , sector of `` mirror '' particles and interactions with the lagrangian completely similar to that of the ordinary particles . in other words , it has the same gauge group and coupling constants as the ordinary world , so that the lagrangian of the whole theory is invariant with respect to the mirror parity ( m - parity ) which interchanges the two sectors .
the possibility to implement the pq mechanism in the mirror world scenario is discussed in ref . .
the general feature is the following : the total lagrangian must be of the form @xmath83 , where @xmath84 represents the ordinary lagrangian , @xmath85 the mirror one and @xmath86 is an interaction term . for @xmath87 ,
the total lagrangian contains two identical @xmath88 symmetries , while the @xmath86 term breaks these in just the usual @xmath5 .
thus only one axion field results , which is defined in both sectors .
as long as the m - parity is an exact symmetry , the particle physics is exactly the same in the two worlds , and so the strong cp problem is simultaneously solved in both sectors . in particular , the axion couples to both sectors in the same way and their non - perturbative qcd dynamics produce the same contribution to the axion effective potential .
this situation does not bring drastic changes of the axion properties ; just the axion zero temperature mass is increased by a factor of @xmath89 with respect to the standard expression ( [ a - m ] ) . on the other hand , on a cosmological ground ,
the physics is less trivial .
cosmologically the mirror sector can not be identical to ours .
if so , the energy density due to the extra ( mirror ) degrees of freedom would be much bigger than what is allowed in the standard cosmology scenario .
the mirror sector must then be colder than ours by a factor @xmath90 .
extra neutrinos , in evident contradiction with the observation of the helium abundance predicted by big bang nucleosynthesis ( bbn ) .
the value @xmath91 corresponds to @xmath92 . for more details
see ref . .
for a complete discussion on the limit on @xmath93 allowed by bbn , e.g. see ref . and references therein .
] as a consequence , the energy density of the whole universe is essentially the ordinary one , and the hubble expansion is completely driven by our sector : @xmath94 , where @xmath95 counts only the ordinary degrees of freedom .
let us consider , now , the primordial oscillations of the axion field for a generic mirror axion model .
for this case , we have @xmath96 in the zero temperature limit .
however , at finite temperatures the behavior of the axion mass is dominated by the contribution from the mirror sector , which has a lower temperature than the ordinary one , @xmath97 , and thus we have @xmath98 .
therefore , the axion mass grows faster with the temperature , and the temperature @xmath60 at which the axion field starts to oscillate scales as @xmath99 .
correspondingly , the axion mass density @xmath0 roughly scales as @xmath100 .
e.g. for @xmath101 , we obtain @xmath0 about 3 times smaller than in the estimate ( [ omegah2-st ] ) .
however , this situation is valid until the condition @xmath102 , where @xmath103 is defined as the temperature at which @xmath71 .
this condition roughly translates in @xmath104 . in this case
we obtain : @xmath105 and hence : @xmath106 which , for @xmath77 , leads to the upper limit @xmath107 .
in conclusion , we have considered here a very simple model for mirror axion .
in fact , in this , the axion properties are essentially the same as for the standard invisible axion , and only the mass is increased by a factor @xmath108 . on the other hand , we have shown that its cosmology is non trivial , because of the different temperature of the mirror sector .
this leads to a weaker cosmological bound on the pq scale .
in particular it can be closer to the gut scale . |
nonequilibrium dynamics has been a topic of intensive research over the last few decades .
the fact that models displaying identical equilibrium behaviour can be differentiated on the basis of critical relaxation , coarsening and persistence behaviour , has enhanced the interest in this field . traditionally critical and off critical
dynamical behaviour were studied for magnetic systems with different dynamical rules and constraints @xcite .
more recently , physicists have been able to construct dynamical models in problems which are interdisciplinary in nature , where one can investigate how such models can be classified into different dynamical classes .
one very popular interdisciplinary topic is opinion dynamics , where a number of models @xcite have been proposed and studied by physicists , many of which can also be regarded as spin models .
the dynamical rules in all the well studied opinion dynamics models with binary opinions ( @xmath0 ) in one dimension lead to the consensus state .
the question whether the intrinsic dynamics for ordering are equivalent has been asked in context of the generalised @xmath1 state voter model @xcite , generalised glauber models @xcite and sznajd models @xcite .
one of the quantities which is computed to resolve this important question is the exit probability ( ep ) which denotes the probability @xmath2 that one ends up in a configuration with all opinions equal to 1 starting from @xmath3 fraction of opinions equal to 1 .
ep has been shown to be identical @xcite in the generalised glauber model and the sznajd model ; the latter was originally claimed to have a different dynamical scenario . in the voter model ( or ising glauber model ) in one dimension , @xmath2 is simply equal to @xmath3 , corresponding to the conservation in the dynamics @xcite . in the nonlinear voter model , long ranged sznajd model and long ranged glauber model ,
it is a nonlinear continuous function of @xmath3 @xcite . in all these cases ,
the results are also independent of finite system sizes indicating there is no scaling behaviour .
the claim that the exit probability for sznajd model is a continuous function has been however questioned in @xcite .
but analytical and numerical study of @xmath1 state nonlinear voter model ( sznajd model corresponds to nonlinear voter model with @xmath4 ) show that ep is a continuous function of @xmath3 @xcite .
interestingly , the mean field result was shown to be exact in the nonlinear @xmath1 voter model which includes the sznajd model and independent of the range @xcite .
on the other hand , in two dimensions or on networks , the exit probability shows a step function behaviour in many models which has been interpreted as a phase transition .
@xmath2 also shows finite size dependence in that case @xcite .
however , strictly speaking , one should interpret this as the existence of a separatrix between two different regions of basin of attraction - where the attractors are the states with all opinions equal to @xmath5 or @xmath6 . in this paper
, we have proposed a model ( the weighted influence model , wi model hereafter ) which shows completely different behaviour as the ep has a step function behaviour even in one dimension .
the result also shows clear deviation from mean field theory , although the latter provides a reasonable first order estimate .
the model includes one parameter which allows one to obtain a relevant `` phase diagram '' and also show the presence of universal behaviour .
apart from studying the exit probability , we also investigate the dynamical behaviour of the wi model by studying the density of domain walls @xmath7 and persistence probability @xmath8 as functions of time @xmath9 starting from an initial disordered state .
the latter is the probability that a spin has not flipped till time @xmath9 @xcite .
it is known that in conventional coarsening processes , @xmath10 and @xmath8 shows a scaling behaviour @xmath11 in many systems @xcite where @xmath12 and @xmath13 are the dynamic ( growth ) and persistence exponents respectively . by calculating @xmath12 and @xmath13 ,
the dynamical class of the model can be identified and compared to models for which these exponents are known , for example , for the zero temperature ising model in one dimension , @xmath14 and @xmath15 are exactly known results .
the rest of the paper is arranged as follows : sec .
[ sec2 ] describes the model .
[ sec3 ] discusses the mean field theory in the context of the present work .
numerical results obtained from extensive simulations are presented in sec .
dynamical properties of the model are given in sec .
[ sec5 ] and finally in sec .
[ sec6 ] , concluding remarks are made .
the wi model is a stochastic model with opinions taking values @xmath16 , and there is a bias towards one type of opinion controlled by a relative weight factor .
it can not be true that an initial majority always wins ( in that case the same candidate / party will go on winning elections every time ) @xcite and a huge majority of people may give up to an initial minority view @xcite .
the relative weight factor included in the model takes care of this idea .
this weight factor represents the relative strength that may arise due to monetary factors , local factors , muscle power , larger accountability , traditional , religious or cultural influence , recent incidents which has a great impact on the population etc .
the model is described in the following way .
let there be two groups of people with opposing opinion in the immediate spatial neighbourhood of an individual . under the influence of these two groups ,
she / he will be under pressure to follow one group .
the pressure is proportional to the size of the group . denoting the two neighbouring opposing group sizes
as @xmath17 and @xmath18 for opinion = @xmath0 respectively , an individual takes up opinion @xmath5 with probability @xmath19 where @xmath20 is the relative influencing ability of the two groups and can vary from zero to @xmath21 .
probability to take opinion value @xmath6 is @xmath22 the model is considered in one dimension . in case
an individual with opinion @xmath5 ( @xmath6 ) has both nearest neighbours with opinion @xmath6 ( @xmath23 ) , her opinion will change deterministically .
the unweighted model corresponds to @xmath24 @xcite . in this model ,
the dynamics is completely stochastic .
a quasi - deterministic model in which the neighbouring domain sizes determine the state of a spin had been proposed earlier ( bs model hereafter ) @xcite .
the bs model takes into consideration the sizes of neighbouring domains in the dynamics and the larger neighbouring domain always dictates the opinion , irrespective of its size . only in the case when the neighbouring domains are of equal size ( which occurs rarely ) is the dynamical rule stochastic . in the bs model , @xmath25 and @xmath26 (
both the exponents are different from the ising model ) .
we find some interesting effects of the nature of the stochasticity in the wi model , especially regarding the persistence behaviour , which is revealed when compared to the bs model and ising model dynamical results . regarding the binary opinion values as ising spin states ( up / down ) , the absorbing states are the all up / all down states in the wi model .
probability of attaining these consensus states however depends on the value of @xmath20 instead of being simply 1/2 ( as in the ising or voter model ) even when one starts from a completely random initial configuration ( @xmath27 ) . in the limit @xmath28
, all spins will be up as @xmath29 for any initial value of @xmath3 while for @xmath30 , the final state will be all down for any value of @xmath31 .
thus the threshold values @xmath32 for which the final state will be all up is zero for @xmath28 and 1 for @xmath30 .
the exit probability is trivially a step function in these extreme limits .
the question is what happens for other values of @xmath20 , including @xmath24 .
the dynamics can be studied in terms of the motion of domain walls as only the spins adjacent to domain walls can flip . in a glauber like process
, one considers the flipping of a random spin in time @xmath33 with the time unit being such that @xmath34 , where @xmath35 is the total number of spins .
initially , there will be many domains of size one , but they will quickly vanish as it is a deterministic process . assuming no domain of size one remains in the system and using a mean field approximation , one can write down a microscopic equation for the ( average ) fraction of up spins at time @xmath36 given that there was a fraction @xmath3 at @xmath9 . it may be noted that in this approximation , the fluctuations in the flipping probabilities @xmath37 and @xmath38 can be ignored and they can be taken to be site - independent .
the equation for @xmath39 is then given by @xmath40 \nonumber \\ & & + r(t ) [ ( x(t)+1/l)p_1 + x(t)p_{-1 } ] \nonumber \\ & & + ( 1 - 2r(t ) ) x(t).\end{aligned}\ ] ] here @xmath41 is the density of domain walls , @xmath42 when domains have length at least 2 .
@xmath43 are also in general time dependent .
the first two terms on the right hand side correspond to cases where the up and down spin at the boundary are chosen for flipping respectively while the last term is for the case when a spin within a domain is selected ( @xmath3 remains same in the last case obviously ) .
thus one gets @xmath44 . \label{fixed}\ ] ] this equation can not be solved without knowing the dynamical equation for @xmath41 which is again expected to involve @xmath45 in a complicated manner .
however , the fixed points of the equation , in which we are actually interested , are easily obtained ; a trivial fixed point @xmath46 and the other one is @xmath47 = 0 $ ] . for the ising or voter model
, @xmath48 is equal to @xmath38 which corresponds to the result that @xmath49 independent of @xmath3 .
this leads to the known result that the exit probability is simply equal to @xmath3 .
all points are fixed points here . in case one
gets a single fixed point @xmath32 from eq .
( [ fixed ] ) , it will indicate the existence of the step function like behaviour associated with the exit probability .
the mean field approximation of course neglects all correlations and fluctuations . in the wi model , @xmath38 and @xmath48 in mean field approximation
can be estimated by taking @xmath50 and @xmath18 proportional to @xmath3 and @xmath51 respectively ( at the fixed point ) in eq .
( [ prob1 ] ) .
there is no reason to take the constant of proportionality to be different ( i.e. there is no bias to either type of domain ) such that @xmath52 $ ] and we get @xmath53 although the mean field result involves many assumptions it is tempting to accept this result as it coincides with the limiting results that @xmath54 for @xmath55 , @xmath56 for @xmath30 .
the mean field result also predicts @xmath57 for @xmath24 .
@xmath58 corresponds to the model with unweighted influence , and here if one starts with @xmath27 the system will go to @xmath23 state with @xmath59 probability ( by the argument of symmetry ) .
if there is any initial bias ( @xmath60 ) in the system then it will win at the end .
ep will be zero for @xmath61 and equal to 1 for @xmath62 .
hence one expects that at @xmath58 , @xmath63 as given by ( [ eq - mf ] ) .
having obtained the evidence of a single value of @xmath32 from mean field approximation , our next job is to find out numerically whether there exists a separatrix and and whether finite size effects exist .
also , the deviation from mean field theory , if any , will be investigated in the following section .
we calculate the exit probability @xmath64 for system sizes ranging from @xmath65 to @xmath66 and repeat the simulations for over at least @xmath67 configurations for each system size .
@xmath64 against initial concentration @xmath3 is shown in fig .
[ epvsups ] for two values of @xmath20 .
it indeed shows a sharp rise close to a value of @xmath68 , henceforth called the separatrix point .
the shape of the exit probability @xmath64 plotted against @xmath3 immediately shows that it is nonlinear , moreover , @xmath2 curves show strong system size dependence and intersect at a single point @xmath32 for different values of @xmath35 .
the behaviour of ep indicates that it shows a step function behaviour in the thermodynamic limit .
finite size scaling analysis can be made using the scaling form ( when @xmath20 is constant ) : @xmath69 where @xmath70 for @xmath71 and equal to 1 for @xmath72 ( i.e. , a step function in the thermodynamic limit ) .
@xmath73 gives the the value of ep at the separatrix point .
the data collapse , shown in fig .
[ epvsups ] , takes place with @xmath74 for all values of @xmath20 . for fixed @xmath3
, the finite size scaling form for exit probability can be written as : @xmath75 where @xmath76 for @xmath72 , equal to 1 for @xmath71 .
both the scaling forms ( eqs ( [ epscaling ] ) and ( [ delscaling ] ) ) give @xmath77 independent of the exact location of the separatrix point .
@xmath32 as a function of @xmath20 denotes the trajectory of the separatrix point as @xmath20 is varied and for convenience we call it the `` phase boundary '' .
so one can conclude that universal behaviour exists along the entire phase boundary .
estimating @xmath32 for different values of @xmath20 , we plot the phase boundary in fig .
[ delvsep ] .
the phase boundary is not exactly given by the mean field estimate ( [ eq - mf ] ) but shows systematic deviation from this equation ( except at @xmath78 and @xmath30 ) as shown in fig .
[ meanfield ] .
there is a systematic difference from the mean field results away from @xmath79 which vanishes at @xmath80 and also as @xmath30 .
however , the difference is mostly less than ten percent for other values of @xmath20 and therefore the mean field result can be taken as a first order estimate . in principle
, one may consider an expression of @xmath32 given by a polynomial in @xmath81 as obviously it deviates from a simple linear form .
however , introducing only a second order term is not sufficient and we therefore attempt to fit the numerically obtained values of @xmath32 accurately by a single correction term , the form of which is conjectured by the known values of @xmath32 at @xmath82 and @xmath30 .
we assume the following form for @xmath32 : @xmath83 here @xmath84 in the correction term ( second term on right hand side of eq .
( [ eqcorec ] ) ) takes care that the term vanishes for @xmath85 .
if one compares the numerical data with the mean field result ( fig .
[ meanfield ] ) , the former gives larger values of @xmath32 for @xmath86 ( and lower values of @xmath32 for @xmath87 ) .
so @xmath88 will be there in the correction term instead of @xmath89 .
one also needs a factor proportional to a power of @xmath20 in the denominator of the correction term which should be nonzero for @xmath28 and make the term vanish in the limit @xmath30 .
such a term is chosen as @xmath90 where @xmath91 should be greater than @xmath92 and @xmath93 .
we indeed find that eq .
( [ eqcorec ] ) fits the curve quite nicely with @xmath94 and @xmath95 .
we also investigate the behaviour of @xmath73 @xmath96 @xmath97 as a function of @xmath20 ; although a monotonic increase is found , no obvious functional form appears to fit the data .
next we consider the dynamical behaviour by studying the density of domain walls @xmath7 and persistence probability @xmath8 as functions of time @xmath9 . in this context ,
comparison with the zero temperature dynamics in the ising model and the bs model @xcite will be interesting .
coarsening study for @xmath98 : we start with a completely disordered state where @xmath24 is the estimated separatrix point .
the scaling behaviour of @xmath7 is compatible with a value of @xmath99 at @xmath58 .
as @xmath20 deviates from @xmath5 , the coarsening process becomes very fast : obviously , in the extreme limit @xmath28 or @xmath21 , the system goes to the all up / down configuration almost instantaneously .
in fact for any value of @xmath100 , the power law behaviour for @xmath7 is no longer valid .
this is not surprising , it is known that power law scalings are valid only on the transition point ( e.g. , in the ising model , the order parameter shows exponential decay to its equilibrium value away from the critical temperature ) .
the growth exponent @xmath12 in coarsening phenomena in spin systems can be found out by studying several quantities apart from the domain density @xmath7 which varies as @xmath101 .
these include the variations of the absolute magnetisation with time , total time to reach equilibrium as a function of system size and the fraction of spin flips again as a function of time @xmath9 .
the last quantity , @xmath102 , is expected to follow the same behaviour as @xmath7 since spins at the domain boundary can flip only ( it will be less than @xmath7 in magnitude though ) .
this is true for all models . for @xmath103 and @xmath24 , in the wi model @xmath102
thus varies with time as @xmath104 with @xmath25 .
we have shown in fig .
[ flip ] the scaling behaviour of the flipping probability as this quantity is useful in understanding the persistence behaviour .
the following dynamical scaling form for the persistence probability is used @xcite to obtain both @xmath12 and @xmath13 . @xmath105 the persistence probability saturates at a value @xmath106 at large times in finite systems where
@xmath107 is related to the spatial correlation of the persistent spins at @xmath108 .
so the scaling function @xmath109 with @xmath110 for @xmath111 and @xmath112 constant at large @xmath113 .
we have estimated the exponents @xmath13 and @xmath12 from the above scaling relation for @xmath79 , giving @xmath114 and @xmath115 .
the raw data as well as the scaled data are shown in fig .
[ colper ] . these exponents are universal in the sense that if one starts at any point on the phase boundary ( i.e. , with the initial fraction of up spins equal to @xmath116 ) , one gets the same values . in the wi model , @xmath25 as in the bs model but the persistence exponent is different ( by more than ten percent numerically ) .
in fact , @xmath117 obtained here is less than the bs model value ( @xmath118 ) , which is a bit counter - intuitive as the wi model is completely stochastic while the bs model is not . in the bs model
, there is a ballistic motion of the surviving domain walls , which in the course of their motion will flip all the spins that appear on one of their boundaries . with high probability
, these spins will flip once only so that if any flipping occurs it is more likely to affect the persistence probability . in the wi model
, there will be motion of the domain walls in both directions ( though less in comparison to the ising model where pure random walk is executed before the domain walls are annihilated ) , such that the same spin may flip more than once with higher probability and obviously persistence will not be affected when a spin flips more than once . to check this ,
we have computed the distribution @xmath119 of the number ( @xmath120 ) of times a spin flips in both the models .
the results , shown in fig .
[ probdistri ] , exhibit indeed that in the bs model , @xmath121 is much higher compared to the wi model while probability that a spin flips a large number of times ( @xmath122 ) is much less .
( roughly , for @xmath123 , @xmath124 for the bs model and @xmath125 for the wi model with @xmath79 . )
we have already seen that the probability of flipping @xmath102 shows the same scaling behaviour in the two models and this shows why the persistence decays in a slower manner in the present model . in comparison , in the ising model
, the persistence probability decays fastest although the domain walls perform pure random walk . here
, @xmath126 implies that domains survive much longer and as a result a larger number of spins are flipping at every step .
so although @xmath119 has a much slower decay the persistence probability decays much faster showing that the effect of the slower decay in the number of domain walls is more important than the unbiased random walk motion of each .
stochasticity in the bs model has been introduced in several ways earlier by incorporating a parameter in the model @xcite .
in fact in @xcite , the domain sizes had been considered in the dynamics in a different manner . however
, even such stochasticity in the bs model could not lead to any new dynamical behaviour ( @xmath12 and @xmath13 were found to be equal to @xmath127 and @xmath118 respectively ) . on the other hand ,
the stochastic model considered here shows a different result for @xmath13 even for the unweighted model ( @xmath24 ) .
we have presented a stochastic binary opinion dynamics model with one parameter where the exit probability has a step function like behaviour even in one dimension , in contrast to other familiar models .
one obtains a separatrix which is similar to that appearing in magnetic systems at zero field , separating regions of positive and negative magnetisation although in the latter one considers strictly the equilibrium behaviour .
the results show finite size scaling where the scaling argument is @xmath128 indicating that the width of the region where @xmath64 is not equal to unity or zero decreases as @xmath129 .
the unique behaviour of the exit probability may be present due to the effective long range interactions in the wi model .
however it has been shown previously for the generalised voter and sznajd models that the exit probability does not change its nature even if one makes the range of interaction infinite @xcite .
this indicates that the dynamical rule of wi model which handles the range of interaction in a subtly different manner could be responsible for the behaviour of the exit probability .
a thorough study of similar models with domain size dependent dynamics is in progress to check this @xcite .
one may also attempt to check the dependence of @xmath130 when the problem is considered in higher dimensions .
it is also observed that there is a deviation from the mean field result unlike other models in one dimension .
this deviation is attributed to the fact that the fluctuations which have been ignored ( e.g. , by taking @xmath43 independent of location and replacing all domain sizes by an average value ) are indeed relevant . however , the deviation from mean field estimates is still small such that the mean field result can be considered as a first order calculation .
the other important and interesting result is that the persistence exponent of wi model is not only different but lowest among the well known models , including those where domain size dependent dynamics have been used .
thus the wi model is claimed to belong to a unique dynamical class in opinion dynamics models .
we are grateful to deepak dhar for a critical reading of an earlier version of the manuscript and for some useful discussions .
ss acknowledges support from the ugc dr .
d. s. kothari postdoctoral fellowship under grant no .
f.4 - 2/2006(bsr)/13 - 416/2011(bsr ) .
sb thanks the department of theoretical physics , tifr , for the use of its computational resources .
ps thanks upe ( ugc ) project for computational support .
p. c. hohenberg and b. i. halperin , rev .
phys . * 49 * , 435 ( 1977 ) . a. j. bray , adv .
phys . * 43 * , 357 ( 1994 ) .
d. stauffer , in _ encyclopedia of complexity and systems science _ edited by r. a. meyers ( springer , new york , 2009 ) .
c.castellano , m.a.munoz , r.pastor-satorras , phys .
e * 80 * , 041129 ( 2009 ) . c. castellano and r. p. satorras phys .
e * 83 * , 016113 ( 2011 ) .
k. sznajd - weron and j. sznajd , int .
c * 11 * , 1157 ( 2000 ) .
f. slanina , k. sznajd - weron and p. przybyla , europhys .
* 82 * 18006 ( 2008 ) . c. castellano , s. fortunato and v. loreto , rev .
phys * 81 * , 591 ( 2009 ) .
r. lambiotte and s. redner , europhys . lett . * 82 * , 18007 ( 2008 ) . s. galam and a. c. r. martins , europhys .
48005 * 95 * ( 2011 ) _ and the references therein_. p. przybyla , k. sznajd - weron , and m. tabiszewski , phys .
e * 84 * , 031117 ( 2011 ) d. stauffer , a. o. sousa and s. m. de oliveira , int . j. mod .
c , * 11 * , 1239 ( 2000 ) .
n. crokidakis and p. m. c. de oliveira , j. stat .
( 2011 ) p11004 and the references therein .
b. derrida , a. j. bray and c. godreche , j. phys . a * 27 * , l357 ( 1994 ) .
available election data do not provide the dynamical evolution of a voters decision but mainly the distribution of of votes obtained by the candidates ( see e.g. a. chatterjee , m. mitrovic and s. fortunato , scientific reports * 3 * , 1049 ( 2013 ) ) .
thus comparing real data directly with our model at present is not possible .
however , the fact that a change in government / winner happens quite often in elections implies that the loser in one election can emerge as a winner in the next , and the dynamics must be responsible for this effect .
s. moscovici , _
silent majorities and loud minorities _ , communication yearbook , 14 , j.a .
anderson ( sage publications inc . ,
1990 ) ; s. galam , eur . phys . j. b , * 25 * , 403 ( 2002 ) .
the @xmath24 model was very briefly discussed in @xcite , only the dynamic results were mentioned which _ are not _ fully correct .
s. biswas and p. sen , phys .
e * 80 * , 027101 ( 2009 ) .
g. manoj and p. ray , phys .
e * 62 * , 7755 ( 2000 ) ; j. phys .
a * 33 * , 5489 ( 2000 ) .
s. biswas , a. k. chandra , and p. sen , phys .
e * 78*,041119 ( 2008 ) .
p. sen , phys .
e * 81 * , 032103 ( 2010 ) .
s. biswas , p. sen and p. ray , journal of physics : conf .
series * 297 * , 012003 ( 2011 ) .
p. roy , s. biswas and p. sen ( to be published ) . | we introduce a stochastic model of binary opinion dynamics in which the opinions are determined by the size of the neighbouring domains .
the exit probability here shows a step function behaviour indicating the existence of a separatrix distinguishing two different regions of basin of attraction .
this behaviour , in one dimension , is in contrast to other well known opinion dynamics models where no such behaviour has been observed so far .
the coarsening study of the model also yields novel exponent values .
a lower value of persistence exponent is obtained in the present model , which involves stochastic dynamics , when compared to that in a similar type of model with deterministic dynamics .
this apparently counter - intuitive result is justified using further analysis . based on these results
it is concluded that the proposed model belongs to a unique dynamical class . |
a new era of x - ray astronomy has begun with the launch of the @xmath0 @xmath21-@xmath22 @xmath23 on 1999 july 23 , thanks to its very high sensitivity , broad energy range and high angular resolution ( weisskopf et al . 2002 ) .
the two deepest x - ray surveys ever conducted , 2 ms @xmath0 deep field north ( brandt et al .
2003 , alexander et al . 2003 ) and 1 ms @xmath0 deep field south ( giacconi et al . 2002 ; rosati et al .
2001 ) , were obtained using the advanced ccd imaging spectrometer detector ( acis ; garmire et al .
2003 ) on @xmath0 @xmath21-@xmath22 @xmath23 .
these two surveys are @xmath14 50 times more sensitive than the deepest pre-@xmath0 observations in the soft x - ray band ( 0.5 2.0 kev , e.g. , hasinger et al .
1998 ) and greater than 100 times more sensitive than those deepest pre-@xmath0 observations in the hard x - ray band ( 2.0 10.0 kev , e.g. , ueda et al . 1999
; fiore et al .
1999 ) . with these surveys , the `` diffuse '' x - ray background discovered four decades ago ( giacconi et al . 1962 )
has been almost entirely resolved into discrete sources ( i.e. @xmath20 90% in the soft band , and @xmath14 80% in the hard band ) .
many other deep x - ray blank - sky surveys from @xmath0 ( stern et al . 2002a ; yang et al .
2003 ; mushotzky et al .
2000 ; manners et al .
2003 ) and xmm ( hasinger et al . 2001 ) were also performed in the past few years . along with multi - band observations of the detected x - ray sources ,
these deep surveys have brought us many interesting results and more important science issues to be addressed .
these include the large scale structures from 2d ( yang et al .
2003 ) and 3d ( gilli et al . 2003 ; barger et al .
2002 ) analyses , type 2 qsos ( norman et al . 2002 ; stern et al .
2002b ) , very high redshift x - ray selected active galactic nuclei ( agn , e.g. , barger et al .
2003 ) , and much more . here
we present a new deep ( 172 ks ) @xmath0 acis exposure , obtained originally for the follow - up of ly@xmath24 sources from the large area lyman alpha ( lala ) survey s botes field .
the large area lyman alpha ( lala ) survey ( rhoads et al . 2000 , 2003 ; rhoads & malhotra 2001 , malhotra & rhoads 2002 ) was designed to search for lyman @xmath24 emitters at high redshifts through narrowband imaging .
the survey comprises two primary fields , 36@xmath136 in size each , located in botes ( at 14:25:57 + 35:32 j2000.0 ) and in cetus ( at 02:05:20 -04:55 j2000.0 ) .
both fields were chosen to be inside the large areas of the noao deep wide - field survey ( ndwfs ) , which is a deep optical and ir ( @xmath25 ) imaging survey of 18 @xmath26 of the sky with the primary goal of studying the evolution of large - scale structure from @xmath27 ( jannuzi and dey 1999 ; brown et al .
2003 ; jannuzi et al .
2004 , in preparation ) .
five broadband optical images ( @xmath9 , @xmath7 , @xmath11 from ndwfs , and @xmath10 , @xmath12 as part of lala ) and eight narrowband images are used to search for lyman-@xmath24 emitters at z @xmath14 4.5 , 5.7 , and 6.5 respectively . the x - ray image presented here
was originally obtained to investigate the x - ray properties of the detected high redshift ly@xmath24 emitters .
this study was presented by malhotra et al .
( 2003 ) , where we placed stringent upper limits on the typical x - ray flux of ly@xmath24 sources and conclude that agn ( obscured or otherwise ) can not constitute a dominant portion of the ly@xmath24 source population .
this @xmath0 exposure ( 172 ks ) is among the deepest yet obtained by @xmath0 of the extragalactic sky ; only the cdf - n and cdf - s are substantially deeper . in this paper
we present a full catalog of the detected x - ray sources , along with an analysis of the x - ray source counts , and the @xmath7-band magnitudes ( or 3@xmath8 upper limits ) for their optical counterparts . to study the nature of these sources , spectroscopic follow - up observations for these sources are under way .
the present paper is structured as follows : we present the x - ray observations and data reduction in section 2 , source detection and catalog in section 3 , logn - logs in section 4 , and optical identifications in section 5 .
our conclusions and summary are presented in section 6 .
an 178 kilo - second exposure , composed of two individual observations , was obtained using the advanced ccd imaging spectrometer ( acis ) on the _ chandra x - ray observatory _ in the very faint ( vfaint ) mode .
the first observation , with 120 ks exposure , was taken on 2002 april 16 - 17 ( _ chandra _ obs i d 3130 ) . the second observation , with 58 ks exposure , was taken on 2002 june 9 ( obs i d 3482 ) .
all four acis - i chips and acis - s2 , acis - s3 chips were used , with the telescope aimpoint centered on the acis - i3 chip for each exposure . the aimpoint of obs i d 3130 is 14:25:37.791 + 35:36:00.20 ( j2000.0 ) , and the aimpoint of obs i d 3482 is 14:25:37.564 + 35:35:44.32 , 16@xmath28 away from that of obs i d 3130 .
the galactic column density n@xmath29 towards our field is 1.15 @xmath1 10@xmath30 @xmath4 ( hartmann et al .
due to their large off - axis angle during the observations , the acis - s chips have worse spatial resolution and lower effective area relative to the acis - i chips . in this paper , data from any acis - s ccd were therefore ignored .
the data were reduced and analyzed using the @xmath0 interactive analysis of observations ( ciao ) software ( version 2.2.1 , see http://asc.harvard.edu/ciao ) .
the data were reprocessed to clean the acis particle background for very faint mode observations , and then filtered to include only the standard event grades 0,2,3,4,6sec : grades ] .
all bad pixels and columns were also removed , not only from the photon events files , but also when calculating the effective exposure maps .
the high background time intervals were manually removed by checking the total event rates .
the total net exposure time is 172 ks ( 120 ks from obs i d 3130 , and 52 ks from obs i d 3482 ) .
the offset between the astrometry of the two observations was obtained by registering the x - ray sources showing up in both exposures .
the two event files were merged after correcting the small offset ( 0.3 ) , and the combined data have the same coordinate system as obs i d 3130 .
we present the color composite x - ray image of the combined exposure in fig .
[ color ] .
three images were extracted from the combined event file for source detection : a soft image ( 0.5 2.0 kev ) , a hard image ( 2.0 7.0 kev ) and a total image ( 0.5 7.0 kev ) .
the hard and total bands were cut at 7 kev since the effective area of @xmath0 decreases above this energy , and the instrumental background rises , giving a very inefficient detection of sky and source photons . the wavdetect program ( freeman et al .
2002 ) , which is included with the ciao software package , was run on the extracted images . a probability threshold of 1 @xmath1 10@xmath31 ( corresponding to 0.5 false sources expected per image ) , and wavelet scales of 1,2,4,8,16 pixels ( 1 pixel = 0.492 ) were used .
a total of 168 x - ray sources were detected : 160 in the total band ( 0.5 7.0 kev ) , 132 in the soft band ( 0.5 2.0 kev ) , and 111 in the hard band ( 2.0 7.0 kev ) .
we present the catalog of the detected sources in table 1 .
note that the number of total detected x - ray sources will be increased to 196 if we use a probability threshold of wavdetect ( 1 @xmath1 10@xmath32 ) , however , considering the increased number of possible false detections ( @xmath14 5 expected per image ) , we prefer to publish a conservative catalog with fewer false detections .
we also tried to search for x - ray sources with scales much larger than the psf by running wavdetect with wavelet scales of 32,64,128 pixels on the images .
however , we did nt detect any new source beyond those in the catalog .
below we give explanations to the columns in table 1 .
column(1)-(4 ) : the source i d , iau name , right ascension and declination respectively .
the iau name for the sources is cxolala1 jhhmmss.s+ddmmss .
the positions were determined by wavdetect .
whenever possible , we use positions derived in the soft band , which has the best spatial resolution among the three bands . for sources which are not detected in the soft band , we use total band positions instead
if available , and hard band positions were quoted for those sources detected only in the hard band .
column(5 ) : 3 @xmath8 uncertainties of the centroid positions directly given by wavdetect .
column(6)-(8 ) : the net counts in the soft , hard and total bands .
the counts were calculated using circular aperture photometry . for each source
, we defined a source region which is a circle centered at the position given in column ( 3 ) and ( 4 ) , with radius r@xmath33 set to the 95% encircled - energy radius of @xmath0 acis psf at the source position .
r@xmath33 varies in the range of 2@xmath28 to 15@xmath28 from the center to the edge of the field .
source photons were then extracted from the regions , and the local background was extracted from an annulus with outer radius of 2.4 @xmath1 r@xmath33 and inner radius of 1.2 @xmath1 r@xmath33 , after masking out nearby sources . the aperture correction ( @xmath1 1/0.95 )
was applied to the source counts .
the derived net counts and 1 @xmath8 poisson uncertainties ( e.g. , gehrels 1986 ) are given for each source in each band .
column(9 ) : indication of source detection .
we mark the source with `` t '' , `` s '' and `` h '' for source detected in total , soft and hard band respectively .
multiple letters are used for sources detected in more than one band .
for example , `` ts '' means detections in both the total band and the soft band , but non - detected in the hard band .
column(10 ) : hardness ratio defined as @xmath34 = ( h - s)/(h+s ) , where h and s are the vignetting - corrected net counts in the hard and soft band respectively .
the hardness ratios vs 0.5 10.0 kev band x - ray fluxes for the detected x - ray sources are plotted in fig .
[ hr ] . assuming a power - law spectrum with the galactic hi column density ( 1.15 @xmath1 10@xmath30 @xmath4 )
, the observed hardness ratio can be converted to the photon index @xmath35 of the spectrum , which is also presented in the figure . as presented in earlier surveys ,
harder sources are seen at fainter fluxes , most of which are believed to be obscured agns .
column(11)-(13 ) : x - ray fluxes ( galactic absorption corrected ) of three bands in the unit of 10@xmath3 ergs @xmath4 s@xmath5 .
a power - law spectrum with the galactic column density was assumed to calculate the conversion factors from net counts to x - ray fluxes .
the photon index of the power - law was chosen at @xmath35 = 1.4 , which was also used in giacconi et al .
( 2002 ) and stern et al .
( 2002a ) .
three band net count rates were calculated by dividing the net counts in column 6 - 8 by the effective exposure time at each source position in each band , and then converted into x - ray fluxes of 0.5 10.0 kev , 0.5 2.0 kev , and 2.0 10.0 kev respectively .
these settings make our results directly comparable with those from other surveys .
the conversion factors used were 1.25 @xmath1 10@xmath36 ergs @xmath4 count@xmath5 from the 0.5 7.0 kev band observed count rates to the galactic absorption corrected x - ray fluxes in the band 0.5 10.0 kev , 4.67 @xmath1 10@xmath37 ergs @xmath4 count@xmath5 from the 0.5 2.0 kev band count rates to the 0.5 2.0 kev band fluxes , and 2.96 @xmath1 10@xmath36 ergs @xmath4 count@xmath5 from the 2.0 7.0 kev band count rates to the 2.0 10.0 kev band fluxes .
note the total band ( 0.5 10.0 kev ) flux is not equal to the sum of the soft and hard band fluxes if the actual photon index differs from 1.4 .
to calculate the fluxes assuming a power - law spectrum with different photon index @xmath35 , the conversion factors f@xmath38/f@xmath39 for different bands are needed ( fig .
we can see from the figure that the soft band ( 0.5 2.0 kev ) flux is not sensitive to the photon index @xmath35 , while the hard ( 2.0 10.0 kev ) and total ( 0.5 10.0 kev ) band fluxes correlate with it strongly .
one thus has to be careful while using these fluxes .
column(14)-(16 ) : the offsets of the detected optical counterparts from the x - ray source positions ( @xmath40 = ra@xmath41 - ra@xmath42 , @xmath43 = dec@xmath41 - dec@xmath42 ) , and the @xmath7-band auto magnitudes ( kron - like elliptical aperture magnitudes , bertin & arnouts 1996 ) with 1 @xmath8 uncertainties - band zeropoint . ] . for sources which are not detected in @xmath7-band , 3@xmath8 upper limits of the magnitudes
are given when available .
see section 5 for details .
column(17 ) : the fwhm ( full - width half - maximum ) of the optical counterparts in ndwfs @xmath7-band image ( see section 5 ) .
in order to calculate the cumulative source counts , n(@xmath20s ) , one need count the number of sources with fluxes @xmath20s , and also compute the summed sky in the field where these sources can be detected ( the sky coverage ) . in fig .
[ cutoff ] , we present the soft and hard band net counts vs off - axis angle for faint sources ( with net counts @xmath18 50 )
. detected and non - detected sources ( detected in other bands but having net counts from x - ray photometry ) are displayed with different symbols .
it s clear from the figure that the x - ray sources with larger off - axis angles need more net counts to be detected because of the larger psf size .
this indicates that the detection limits of net counts vary with the psf sizes .
two dashed lines ( net counts = a + b @xmath1 psf@xmath44 respectively ) are added to fig .
[ cutoff ] , with the parameters a and b chosen for both bands by visual inspection to exclude all non - detected sources and include maximum number of detected sources .
we can see that although some detected sources are located below the threshold dashed lines , they are mixed up with these non - detected sources , i.e. , the detection is incomplete below the lines , and the sky coverage for these sources is thus not available . in this paper
, we use only these sources with net counts c @xmath20 a + b @xmath1 psf@xmath44 to calculate logn - logs . using the above cutoffs ,
the sky coverages are derived and presented in fig .
the derived logn - logs for both soft and hard band are presented in fig .
[ lognlogs ] .
a maximum - likelihood power - law was used to fit the slope of the logn - logs in each band .
for the 0.5 2.0 kev band we find @xmath45 and for the 2.0 10.0 kev band , we find : @xmath46 the above procedures of calculating logn - logs were also run on the 1 ms cdf - n data ( brandt at al .
2001 ) , 1 ms cdf - s data ( rosati et al .
2002 ) , and 184.7 ks lynx data ( stern et al .
the independently derived logn - logs from the above three fields match the published ones to within 1@xmath8 , and are plotted in fig .
[ lognlogs ] for comparison .
the source densities ( n@xmath20s ) and 1@xmath8 uncertainties at the faint end ( 2.0 @xmath1 10@xmath2 ergs @xmath4 s@xmath5 in 0.5 2.0 kev band , and 2.0 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 in the 2.0 10.0 kev band ) are plotted in the inserts . in the soft band , there is no significant difference among the source counts from the four deep surveys . in the hard band ,
obvious fluctuations of the source counts are seen at the faint end : lala botes field has the highest source density at 2.0 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 which is 33% higher than that of cdf - s , while cdf - n is 23% higher than cdf - s and lynx field is 14% lower .
note that similar field - to - field variances of the hard band source counts have been reported previously ( see tozzi 2001a , cowie et al .
this is believed to be due to the clustering of the x - ray sources .
the non - detection of field - to - field variance in the soft band is also consistent with the results from previous studies ( yang et al .
2003 ; tozzi et al .
2001a ) , indicating that the soft x - ray sources are less correlated than the hard x - ray sources .
what fraction of the hard x - ray background is resolved by our deep 172 ks @xmath0 imaging ? in the range of 1.7 - 100 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 ,
the integrated hard x - ray flux density in the 2.0 10.0 kev band is 1.2 @xmath1 10@xmath36 ergs @xmath4 s@xmath5 deg@xmath47 .
note for sources bright than 10@xmath48 ergs @xmath4 s@xmath5 , the integrated hard x - ray flux density from our survey is 0.4 @xmath1 10@xmath36 ergs @xmath4 s@xmath5 deg@xmath47 , consistent with the value derived by della ceca et al .
( 1999 ) from a wider area asca survey .
the total integrated hard x - ray flux density we obtained is 1.6 @xmath1 10@xmath36 ergs @xmath4 s@xmath5 deg@xmath47 , down to 1.7 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 in 2.0 10.0 kev band .
this is equal to the high energy astronomy observatory 1 ( heao 1 ) value ( marshall et al .
1980 ) , but 10%-30% lower than the more recent determinations from asca and @xmath49 ( e.g. , ueda et al . 1999 ; vecchi et al .
we conclude that our 172 @xmath0 deep imaging resolves @xmath6 70% of the 2.0 10.0 kev x - ray background , and the main uncertainty comes from the value of total x - ray background itself .
our deep chandra pointing is within the ndwfs botes subfield ndwfs j1426 + 3531 ( which is 35@xmath135 in size ) centered at ra 14:26:00.8 , dec + 35:31:32.0 ( j2000 ) .
five broadband optical images ( @xmath9 , @xmath10 , @xmath7 , @xmath11 , and @xmath12 ) are available to search for the optical counterparts of the x - ray sources .
three of these ( @xmath9 , @xmath7 , and @xmath11 ) and @xmath11 band images are available from the noao science archive ( http://www.archive.noao.edu/ndwfs/ )
. the @xmath7-band image , which is deeper than the currently released version , will be available from the noao science archive within 12 months . ]
are from the noao deep wide field survey ( ndwfs ; jannuzi & dey 1999 ) , while the remaining two ( @xmath10 and @xmath12 ) were obtained as part of lala .
the limiting kron - cousins system magnitudes ( 3@xmath8 , 4 diameter aperture ) of the ndwfs images are @xmath50 , @xmath51 , and @xmath52 .
the corresponding limits in the lala broadband data are @xmath53 and @xmath54 ( where the @xmath12 limit is on the ab system ) .
our primary optical image for identifications in this paper is the @xmath7-band image from ndwfs , which is substantially deeper than the others .
all the optical images were obtained using the kitt peak national observatory mayall 4 m telescope and the mosaic-1 camera ( muller et al .
1998 ; jannuzi et al .
2004 , in preparation ) .
these images were produced following the procedures described in version 7.01 of `` the noao deep wide - field survey mosaic data reductions '' .
a general description of the software used is provided by valdes ( 2002 ) and the complete details of the ndwfs data reduction will be provided by jannuzi et al .
( 2004 ) .
sextractor package ( bertin & arnouts 1996 ) v2.1.0 was run on the ndwfs deep @xmath7-band image to generate the @xmath7-band source catalog
. the registration of the x - ray to optical coordinates was done by cross - correlating the x - ray and the above @xmath7-band catalogs ( e.g. , see giacconi et al . 2002 ,
stern et al .
average shifts in ( ra@xmath41 - ra@xmath42 , dec@xmath41 - dec@xmath42 ) of ( 0.1 , -0.2 ) were found from the ndwfs @xmath7-band to x - ray imaging , but no obvious rotation or plate - scale effects were discovered . after correcting the average shifts , we matched the optical and x - ray source catalogs using a 1.5 radius aperture for host identifications , which was generally used by other @xmath0 surveys ( giacconi et al .
2002 , stern et al .
the 3@xmath8 x - ray positional uncertainties in table 1 are used if larger than 1.5 .
optical counterparts for 144 of the 168 x - ray sources were detected in the @xmath7-band image down to 3@xmath8 .
one of them ( no . 122 ) has multiple possible optical counterparts .
calculations show that the possibility of having one optical source down to 3@xmath8 in a 1.5 circle by chance is 10% .
this means that at least 90% of the optical counterparts we found should correspond to the x - ray sources , and at worst we might have @xmath14 14 false matches .
we present the offsets from the x - ray source positions ( @xmath40 = ra@xmath41 - ra@xmath42 , @xmath43 = dec@xmath41 - dec@xmath42 ) , the derived @xmath7-band auto magnitudes ( kron - like elliptical aperture magnitudes , bertin & arnouts 1996 ) for all optical counterparts in table 1 , and the optical cutouts in fig .
[ cutout ] .
iraf package `` radprof '' was used to measure the radial profiles of these counterparts in ndwfs @xmath7 image , and we present the measured fwhm ( full - width half - maximum ) in table 1 .
these sources which have fwhm larger than the average seeing of the ndwfs @xmath7 image ( fwhm = 1.16 ) were examined visually .
we conclude that it is safe to consider all the bright sources ( @xmath7 @xmath18 23 ) with fwhm @xmath55 1.3 as optically resolved .
the only exception is source no
. 144 , which is too close to another bright source , thus we can not tell if it is resolved or not .
the total number of optically resolved sources is 15 .
the search for optical counterparts was extended to @xmath9 , @xmath11 , @xmath10 and @xmath12 band for the 24 x - ray sources without @xmath7-band optical counterparts , which brought us optical counterparts for 8 more x - ray sources .
the positions of these counterparts are also given in table 1 ( by listing ra - ra@xmath42 , dec - dec@xmath42 ) .
the limiting magnitude of the @xmath7-band image was given as upper limits for 22 of the x - ray sources .
note source no .
37 and 160 are overlapped by bleeding trails in the ndwfs @xmath7 image , and the upper limits of their @xmath7-band magnitudes are not available . the first survey ( faint images of the radio sky at twenty - cm , becker , white , & helfand 1995 ) detection limit in our field is 0.96 mjy / beam .
radio counterparts for 4 of our x - ray sources were found , which are source no . 1 , 66 , 91 and 141 , with integrated flux density ( mjy ) of 6.02 , 1.24 , 2.17 and 8.30 respectively . the first 3 of the 4 radio sources are resolved r@xmath1821 red galaxies , which are some of the brightest optical counterparts to x - ray sources in our catalog .
the remaining source is an r - band non - detection more than 10 " from any r@xmath1821 galaxies and is presumably at z@xmath201 . in fig .
[ xr ] we plot the x - ray - to - optical flux ratio of our x - ray sources ( in soft and hard band respectively ) , and in fig .
[ xrhr ] we plot the log(@xmath15/@xmath16 ) in the hard band vs the hardness ratio of x - ray emission . for the 22 x - ray sources without @xmath7-band counterparts detected ,
the lower limits of log(@xmath15/@xmath16 ) were plotted , which shows a consistent distribution with that of the optically faint ( @xmath7 @xmath20 24 ) sources . from the figures we can see that the majority ( @xmath14 76% ) of the x - ray sources fall within log(@xmath56/@xmath16 )
= 0.0@xmath171 , which is typical of agns . there is a significant population ( @xmath14 10% ) of x - ray faint / optically bright sources ( log(@xmath56/@xmath16 ) @xmath18 -1.0 ) , most of which are extended in the @xmath7 band optical image and x - ray soft ( see fig . [ xrhr ] ) .
these sources should be nearby , bright normal galaxies ( see tozzi et al . 2001b ; barger et al . 2001 ; hornschemeier at al . 2001 ) .
other optically extended sources falling within log(@xmath56/@xmath16 ) = 0.0@xmath171 or x - ray hard are most likely to be low redshift seyfert galaxies , or low luminosity agns .
in addition , we also have a population of sources which are x - ray @xmath19 for their optical magnitudes ( log(@xmath56/@xmath16 ) @xmath20 1.0 ) , especially in the hard band , including most of the 22 @xmath7-band nondetected x - ray sources .
we plot in fig .
[ hist ] the histogram distribution of the hardness ratios of the optically faint x - ray bright sources ( log(@xmath57/@xmath16 ) @xmath20 1.0 ) , comparing with these sources with lower x - ray - to - optical flux ratio ( 1.0 @xmath20 log(@xmath57/@xmath16 ) @xmath20 0.0 ) .
the optically faint x - ray bright sources are obviously harder in x - ray , suggesting that most of them are obscured agns , for which high column density absorption shields both the optical and the soft x - ray emission .
there are 22 x - ray sources which are not detected in the @xmath7-band down to @xmath7 = 25.7 ; 5 of them are detected in redder bands ( @xmath11,@xmath12 ) ; and 3 of them in bluer bands ( @xmath9 , @xmath10 ) ; 14 x - ray sources have no optical counterparts found in any band , which corresponds to a sky surface density of 145 deg@xmath47 . the integrated hard x - ray ( 2.0 10.0 kev ) flux density from these 22 sources is 0.14 @xmath1 10@xmath36 ergs @xmath4 s@xmath5 deg@xmath47 , down to a flux limit of 1.7 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 .
this contributes 6@xmath149% of the total 2.0 10.0 kev band x - ray background .
the 19 @xmath7-band nondetected sources , which are not detected in bluer bands ( @xmath9 , @xmath10 ) , are possible candidates for @xmath58 @xmath20 5 quasars , because of the absorption from the lyman transitions of hydrogen along our line of sight . however ,
our calculations indicate that even for a compton thick agn ( with intrinsic photon power - law index @xmath35 = 2.0 and absorption column density n@xmath29 = 10@xmath59 @xmath4 ) , the hardness ratio hr is expected to be @xmath18 0 at @xmath58 @xmath20 5 .
thus 8 of them with harder x - ray spectra ( hr @xmath20 0 ) can not be at @xmath58 @xmath20 5 . in fig .
[ xrhr ] , we also marked the 5 x - ray sources without @xmath7-band counterparts , but showing up in redder bands ( @xmath11,@xmath12 ) .
preliminary results from photometric redshift analyses show that the two softest ones of the 5 are at photo - z @xmath14 4.3 , and the other 3 which have harder x - ray spectra at photo - z @xmath14 1 - 2 .
this agrees well with the above argument that high redshift objects should be soft in x - ray .
excluding these sources with hr @xmath20 0 , there are 11 @xmath7-band nondetections left , which can be possibly at @xmath58 @xmath20 5 . assuming a constant agn density up to @xmath58 = 10 , gilli , salvati & hasinger ( 2001 ) predicted a @xmath58 @xmath20 5 agn density of 500 deg@xmath47 for a limiting flux of 2.3 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 in the 0.5 2.0 kev band .
this model predicts @xmath14 50 @xmath13 5 sources in our field , which is clearly too high . assuming the agn density above @xmath58 = 2.7 decreases by a factor of 2.7 per unit redshift as found for optical quasars ( schmidt , schneider , & gunn 1995 ) , gilli et al . predicted a much lower source density of @xmath58 @xmath20 5 agns ( 25 deg@xmath47 ) for the same limiting flux .
it corresponds to @xmath142 @xmath58 @xmath20 5 objects in our field , which is very plausible with the current data , considering that those 11 sources are unlikely to be all at @xmath58 @xmath20 5 .
actually , several optically faint @xmath0 sources from other surveys have already been confirmed to be at @xmath58 @xmath18 5 by spectroscopic observations ( e.g. , see alexander et al .
we present a deep , 172 ks @xmath0 acis exposure of the lala botes field .
this paper describes the details of the observations , data reduction , source detection , logn - logs analysis , and presents the x - ray source catalog along with @xmath7-band magnitudes of their optical counterparts .
a total of 168 x - ray sources were detected , 160 in the total band ( 0.5 7.0 kev ) , 132 in the soft band ( 0.5 2.0 kev ) , and 111 in the hard band ( 2.0 7.0 kev ) .
near the aim point , the detection is down to a flux limit of 1.5 @xmath1 10@xmath2 ergs @xmath4 s@xmath5 in soft ( 0.5 2.0 kev ) band , and 1.0 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 in hard ( 2.0 10.0 kev ) band .
logn - logs was compared with those from other deep surveys , and we find obvious field - to - field fluctuations of the hard band source counts .
these fluctuations are believed to be due to the large scale clustering of the detected x - ray sources .
calculations show that our deep imaging resolves @xmath6 70% of the x - ray background in 2.0 10.0 kev band .
optical counterparts for 90% of the x - ray sources were found from deep optical images . among the @xmath7-band
non - detected sources , 11 of them can be possibly at @xmath13 5 , based on the hardness ratios of their x - ray emission and the fact that they are not detected in bluer bands ( @xmath9,@xmath10 ) . spectroscopic follow - up observations of these counterparts are being undertaken and will be present in future paper .
this work was supported by the cxc grant go2 - 3152x and the national optical astronomy observatory which is operated by the association of universities for research in astronomy ( aura ) , inc . under a cooperative agreement with the national science foundation .
we would like to thank the ndwfs team , and in particular melissa miller ( noao ) , who assisted in the reduction of the @xmath7-band image used extensively in this paper .
we would like to thank n. brandt for providing the cdf - n logn - logs .
we also thank the referee for a prompt and helpful report .
alexander , d. m. , brandt , w. n. , hornschemeier , a. e. , garmire , g. p. , schneider , d. p. , bauer , f. e. , & griffiths , r. e 2001 , aj , 122 , 2156 alexander , d. m. , et al .
2003 , aj , in press , astro - ph/0304392 barger , a. j. , cowie , l. l. , mushotzky , r. f. , & richards , e. a. 2001 , aj , 121 , 662 barger , a. j. , cowie , l. l. , brandt , w. n. , capak , p. , garmire , g. p. , hornschemeier , a. e. , steffen , a. t. , & wehner , e. h. 2002 , aj , 124 , 1839 barger , a. j. , cowie , l. l. , capak , p. , alexander , d. m. , bauer , f. e. , brandt , w. n. , garmire , g. p. , & hornschemeier , a. e. 2003 , apj , 584 , 61 becker , r. h. , white , r. l. , & helfand , d. j. 1995 , apj , 450 , 559 bertin , e. and arnouts , s. 1996 , 117,393 brandt , w. n. , et al .
2001 , 122 , 2810 brandt , w. n. , et al .
2003 , head , 35 , 3601b brown , m , dey , a , jannuzi , b. t. , lauer , t. r. , tiede , g. p. , & mikles , v. j. 2003 , apj , in press , astro - ph0306128 cowie , l. l. , garmire , g. p. , bautz , m. w. , barger , a. j. , brandt , w. n. , & hornschemeier , a. e. 2002 , apj , 566 , l5 della ceca , r. , castelli , g. , braito , v. , cagnoni , i. , & maccacaro , t. 1999 , apj , 524 , 674 fiore , f. , la franca , f. , giommi , p. , elvis , m. , matt , g. , comastri , a. , molendi , s. , & gioia , i. 1999 , mnras , 306 , l55 freeman , p.e .
, kashyap , v. , rosner , r. , & lamb , d. q. 2002 , apjs , 138 , 185 garmire , g. p. , bautz , m. w. , ford , p. g. , nousek , j. a. , & ricker , g. r. 2003 , proc .
spie , 4851 , 28 gehrels , n. 1986 , apj , 303 , 336 giacconi , r. , gurksy , h. , paolini , f.r . , & rossi , b.b .
1962 , phys .
, 9 , 439 giacconi , r. , et al .
2002 , apjs , 139 , 369 gilli , r. , salvati , m. , & hasinger , g. 2001 , a&a , 366 , 407 gilli , r. et al .
2003 , apj letter accepted hornschemeier , a. e. et al .
2001 , apj , 554 , 742 hartmann , d , kalberla , p. m. w. , burton , w. b. , & mebold , u. 1999 , a&as , 119 , 115 hasinger , g. , burg , r. , giacconi , r. , schmidt , m. , trumper , j. , & zamorani , g. 1998 , a&a , 329 , 482 hasinger , g. et al .
2001 , a&a , 365 , l45 jannuzi , b. t. , & dey , a. 1999 , in asp conf .
191 , photometric redshifts and high - redshift galaxies , ed .
r. j. weymann , l. j. storrie - lombardi , m. sawicki , & r. j. brunner ( san francisco : asp ) , 111 malhotra , s. & rhoads , j. e. 2002 , apj , 565 , l71 malhotra , s. , wang , j. x. , rhoads , j. e. , heckman , t. m. , & norman , c. a. 2003 , apj , 585 , l25 manners , j. c. , et al .
2003 , mnras , 343 , 293 muller , g. p. , reed , r. , armandroff , t. , boroson , t. a. , & jacoby , g. h. 1998 , proc .
spie , 3355 , 577 mushotzky , r. f. , cowie , l. l. , barger , a. j. , & arnaud , k. a. 2000 , @xmath60 , 404 , 459 norman , c , et al .
2002 , apj , 571 , 218 rhoads , j. e. , malhotra , s. , dey , a. , stern , d. , spinrad , h. , & jannuzi , b. t. 2000 , , 545 , l85 rhoads , j. e. , & malhotra , s. 2001 , apj , 563 , l5 rhoads , j. e. , et al .
2003 , aj , 125 , 1006 rosati , p. , et al .
2002 , apj , 566 , 667 schmidt , m. , schneider , d. p. , & gunn , j. e. 1995 , aj , 110 , 68 stern , d. , et al .
2002a , apjs , 123 , 2223 stern , d. , et al . 2002b , apj , 568 , 71 tozzi , p. 2001a
, in proc .
marseille 2001 conf . where s the matter
l. tresse & m. treyer , astro - ph/0111036 tozzi , p. et al .
2001b , apj , 562 , 42 ueda , y. , et al .
1999 , apj , 518 , 656 valdes , f. g. 2002 , in automated data analysis in astronomy , eds .
ranjan gupta , harinder p. singh , and coryn a. l. bailer - jones , ( new delhi : narosa publishing house ) , 309 .
vecchi , a. , molendi , s. , guainazzi , m. , fiore , f. , & parmar , a. n. 1999 , a&a , 349 , l73 weisskopf , m. c. , brinkman , b. , canizares , c. , garmire , g. , murray , s. , & van speybroeck , l. p. 2002 , pasp , 114 , 1 yang , y. , mushotzky , r. f. , barger , a. j. , cowie , l. l. , sanders , d. b. , & steffen , a. t. 2003 , apj , 585 , l85 1 & j142620.3 + 35370814:26:20.3635:37:08.7 & 0.2 & @xmath61 & @xmath62 & @xmath63 & tsh & -0.52@xmath64 & @xmath65 & @xmath66 & @xmath67 & 0.3 & -0.1 & @xmath68 & 1.3 + 2 & j142555.5 + 35352814:25:55.5035:35:28.5 & 0.5 & @xmath69 & @xmath70 & @xmath71 & tsh & 0.07@xmath72 & @xmath73 & @xmath74 & @xmath75 & 0.6 & -0.0 & @xmath76 & 1.1 + 3 & j142547.2 + 35362714:25:47.2535:36:27.6 & 0.4 & @xmath77 & @xmath78 & @xmath79 & tsh & -0.27@xmath80 & @xmath81 & @xmath82 & @xmath83 & 0.4 & -0.1 & @xmath84 & 1.2 + 4 & j142547.1 + 35395414:25:47.1035:39:54.9 & 0.2 & @xmath85 & @xmath86 & @xmath87 & tsh & -0.81@xmath88 & @xmath89 & @xmath90 & @xmath91 & 0.2 & -0.1 & @xmath92 & 1.1 + 5 & j142546.4 + 35352214:25:46.4835:35:22.3 & 0.4 & @xmath93 & @xmath94 & @xmath95 & tsh & -0.08@xmath96 & @xmath97 & @xmath98 & @xmath99 & 0.3 & -0.1 & @xmath100 & 1.4 + 6 & j142545.3 + 35344914:25:45.3335:34:49.7 & 0.3 & @xmath101 & @xmath102 & @xmath103 & tsh & -0.19@xmath104 & @xmath105 & @xmath106 & @xmath107 & 0.3 & -0.2 & @xmath108 & 1.0 + 7 & j142543.3 + 35354814:25:43.3135:35:48.1 & 0.4 & @xmath109 & @xmath110 & @xmath111 & tsh & -0.15@xmath112 & @xmath113 & @xmath114 & @xmath113 & -0.1 & -0.4 & @xmath115 & 0.8 + 8 & j142542.5 + 35335814:25:42.5335:33:58.2 & 0.4 & @xmath116 & @xmath117 & @xmath118 & tsh & -0.37@xmath80 & @xmath119 & @xmath120 & @xmath121 & 0.4 & -0.3 & @xmath122 & 0.9 + 9 & j142539.5 + 35335714:25:39.5535:33:57.9 & 0.4 & @xmath123 & @xmath124 & @xmath125 & tsh & -0.43@xmath126 & @xmath127 & @xmath74 & @xmath128 & 0.2 & 0.2 & @xmath2025.7 & ... + 10 & j142538.7 + 35361814:25:38.7235:36:18.0 & 0.4 & @xmath129 & @xmath130 & @xmath131 & ts & -0.32@xmath132 & @xmath133 & @xmath134 & @xmath135 & -0.0 & -0.0 & @xmath136 & 1.1 + 11 & j142537.9 + 35361214:25:37.9435:36:12.0 & 0.3 & @xmath137 & @xmath138 & @xmath139 & ts & -0.47@xmath140 & @xmath141 & @xmath142 & @xmath143 & 0.4 & -0.3 & @xmath144 & 1.3 + 12 & j142536.3 + 35363414:25:36.3235:36:34.2 & 0.3 & @xmath145 & @xmath146 & @xmath147 & tsh & 0.06@xmath148 & @xmath149 & @xmath150 & @xmath151 & -0.6 & -0.0 & @xmath152 & 1.2 + 13 & j142534.7 + 35340714:25:34.7635:34:07.8 & 0.3 & @xmath153 & @xmath154 & @xmath155 & tsh & 0.26@xmath156 & @xmath157 & @xmath158 & @xmath159 & 0.0 & -0.1 & @xmath160 & 1.0 + 14 & j142533.5 + 35384514:25:33.5135:38:45.6 & 0.3 & @xmath161 & @xmath162 & @xmath163 & tsh & -0.32@xmath164 & @xmath165 & @xmath166 & @xmath167 & -0.1 & -0.5 & @xmath168 & 1.1 + 15 & j142530.7 + 35391114:25:30.7135:39:11.3 & 0.3 & @xmath169 & @xmath170 & @xmath171 & tsh & 0.12@xmath164 & @xmath172 & @xmath173 & @xmath174 & ... & ... & @xmath2025.7 & ... + 16 & j142529.2 + 35341214:25:29.2435:34:12.2 & 0.4 & @xmath175 & @xmath176 & @xmath177 & ts & -0.66@xmath178 & @xmath179 & @xmath180 & @xmath181 & -0.1 & -0.0 & @xmath182 & 1.1 + 17 & j142525.4 + 35362214:25:25.4035:36:22.8 & 0.5 & @xmath183 & @xmath184 & @xmath185 & tsh & 0.11@xmath186 & @xmath187 & @xmath188 & @xmath189 & 0.2 & -0.1 & @xmath190 & 1.1 + 18 & j142520.7 + 35331114:25:20.7535:33:11.7 & 0.3 & @xmath191 & @xmath192 & @xmath193 & tsh & -0.46@xmath194 & @xmath195 & @xmath196 & @xmath197 & 0.4 & -0.3 & @xmath198 & 1.0 + 19 & j142614.2 + 35363114:26:14.2935:36:31.5 & 1.0 & @xmath199 & @xmath200 & @xmath201 & tsh & -0.55@xmath202 & @xmath203 & @xmath204 & @xmath187 & -0.5 & -0.1 & @xmath205 & 1.1 + 20 & j142614.2 + 35383314:26:14.2135:38:33.1 & 0.6 & @xmath206 & @xmath207 & @xmath208 & tsh & -0.47@xmath209 & @xmath210 & @xmath211 & @xmath212 & 0.5 & -0.0 & @xmath213 & 1.0 + 21 & j142609.5 + 35321314:26:09.5635:32:13.8 & 0.7 & @xmath214 & @xmath215 & @xmath216 & tsh & -0.39@xmath217 & @xmath218 & @xmath219 & @xmath220 & 0.0 & -0.1 & @xmath152 & 1.0 + 22 & j142607.6 + 35335114:26:07.6635:33:51.4 & 0.6 & @xmath221 & @xmath222 & @xmath223 & tsh & 0.82@xmath209 & @xmath224 & @xmath225 & @xmath226 & 0.6 & 0.1 & @xmath227 & ... + 23 & j142605.8 + 35350814:26:05.8335:35:08.7 & 0.4 & @xmath228 & @xmath229 & @xmath230 & tsh & -0.25@xmath217 & @xmath231 & @xmath232 & @xmath233 & 0.1 & -0.1 & @xmath234 & 1.0 + 24 & j142602.4 + 35360514:26:02.4235:36:05.6 & 0.6 & @xmath235 & @xmath236 & @xmath237 & tsh & -0.50@xmath238 & @xmath239 & @xmath240 & @xmath241 & -0.1 & -0.0 & @xmath242 & 1.1 + 25 & j142601.1 + 35415114:26:01.1835:41:51.1 & 0.7 & @xmath243 & @xmath244 & @xmath245 & tsh & -0.26@xmath246 & @xmath247 & @xmath248 & @xmath249 & 0.5 & -0.7 & @xmath250 & 1.1 + 26 & j142557.6 + 35344514:25:57.6535:34:45.9 & 0.4 & @xmath251 & @xmath252 & @xmath103 & tsh & -0.23@xmath253 & @xmath75 & @xmath143 & @xmath254 & 0.0 & 0.1 & @xmath255 & 1.0 + 27 & j142557.7 + 35351214:25:57.7035:35:12.4 & 0.9 & @xmath256 & @xmath257 & @xmath258 & th & 0.60@xmath259 & @xmath260 & @xmath261 & @xmath127 & 0.6 & -0.5 & @xmath262 & 1.1 + 28 & j142556.9 + 35384514:25:56.9435:38:45.1 & 0.5 & @xmath263 & @xmath264 & @xmath265 & tsh & -0.33@xmath238 & @xmath266 & @xmath267 & @xmath268 & 0.2 & -0.3 & @xmath269 & 1.0 + 29 & j142556.3 + 35401814:25:56.3435:40:18.4 & 0.5 & @xmath270 & @xmath271 & @xmath272 & tsh & -0.51@xmath273 & @xmath274 & @xmath275 & @xmath276 & -0.0 & -0.2 & @xmath277 & 1.0 + 30 & j142555.9 + 35324014:25:55.9535:32:40.6 & 0.5 & @xmath278 & @xmath279 & @xmath280 & tsh & -0.23@xmath238 & @xmath281 & @xmath282 & @xmath283 & 0.1 & 0.2 & @xmath284 & 1.0 + 31 & j142555.4 + 35365014:25:55.4035:36:50.4 & 0.7 & @xmath285 & @xmath286 & @xmath287 & ts & -0.56@xmath288 & @xmath289 & @xmath290 & @xmath291 & 0.4 & -0.4 & @xmath2025.7 & ... + 32 & j142554.1 + 35323714:25:54.1135:32:37.1 & 0.5 & @xmath292 & @xmath293 & @xmath294 & tsh & -0.08@xmath164 & @xmath295 & @xmath296 & @xmath297 & -0.1 & -0.4 & @xmath298 & 0.9 + 33 & j142553.8 + 35382614:25:53.8035:38:26.0 & 0.6 & @xmath299 & @xmath300 & @xmath301 & tsh & -0.25@xmath302 & @xmath303 & @xmath304 & @xmath305 & -0.5 & -0.5 & @xmath306 & 0.8 + 34 & j142553.6 + 35331414:25:53.6835:33:14.9 & 0.3 & @xmath307 & @xmath308 & @xmath309 & tsh & -0.60@xmath310 & @xmath311 & @xmath312 & @xmath313 & -0.0 & -0.2 & @xmath314 & 1.0 + 35 & j142552.7 + 35344814:25:52.7235:34:48.2 & 0.7 & @xmath315 & @xmath316 & @xmath317 & ts & -0.54@xmath318 & @xmath319 & @xmath134 & @xmath320 & -0.3 & 0.2 & @xmath321 & 1.1 + 36 & j142552.4 + 35373014:25:52.4535:37:30.3 & 0.5 & @xmath322 & @xmath258 & @xmath323 & ts & -0.39@xmath324 & @xmath181 & @xmath325 & @xmath326 & -1.1 & -0.6 & @xmath327 & 1.3 + 37 & j142552.2 + 35382314:25:52.2735:38:23.7 & 0.4 & @xmath328 & @xmath329 & @xmath330 & th & 0.72@xmath331 & @xmath332 & @xmath134 & @xmath333 & ... & ... & ... @xmath334 & ... + 38 & j142551.3 + 35340414:25:51.3835:34:04.4 & 0.4 & @xmath335 & @xmath336 & @xmath337 & ts & -0.70@xmath178 & @xmath338 & @xmath180 & @xmath339 & 0.1 & -0.1 & @xmath340 & 1.3 + 39 & j142551.0 + 35330714:25:51.0035:33:07.2 & 0.8 & @xmath341 & @xmath342 & @xmath343 & ts & -0.39@xmath302 & @xmath344 & @xmath345 & @xmath99 & 0.0 & -0.6 & @xmath346 & 1.0 + 40 & j142550.8 + 35303314:25:50.8135:30:33.6 & 1.0 & @xmath347 & @xmath348 & @xmath349 & tsh & -0.66@xmath350 & @xmath351 & @xmath352 & @xmath353 & 0.2 & -0.2 & @xmath354 & 1.0 + 41 & j142550.6 + 35374314:25:50.6735:37:43.1 & 0.5 & @xmath355 & @xmath356 & @xmath323 & ts & -0.50@xmath357 & @xmath358 & @xmath359 & @xmath360 & 0.1 & -0.0 & @xmath361 & 1.0 + 42 & j142550.4 + 35324714:25:50.4935:32:47.9 & 0.8 & @xmath362 & @xmath363 & @xmath364 & ts & -0.53@xmath365 & @xmath121 & @xmath366 & @xmath367 & 0.0 & -0.1 & @xmath368 & 1.0 + 43 & j142549.9 + 35320314:25:49.9035:32:03.8 & 0.5 & @xmath369 & @xmath370 & @xmath371 & tsh & -0.41@xmath372 & @xmath373 & @xmath374 & @xmath375 & 0.1 & -0.1 & @xmath376 & 1.1 + 44 & j142549.8 + 35361914:25:49.8135:36:19.0 & 0.6 & @xmath377 & @xmath378 & @xmath110 & th & 0.93@xmath379 & @xmath74 & @xmath380 & @xmath381 & 0.6 & -0.4 & @xmath382 & 1.1 + 45 & j142549.2 + 35361514:25:49.2635:36:15.3 & 0.6 & @xmath383 & @xmath384 & @xmath385 & ts & -0.44@xmath386 & @xmath387 & @xmath388 & @xmath389 & 0.3 & 0.1 & @xmath2025.7 & ... + 46 & j142548.5 + 35350714:25:48.5835:35:07.8 & 0.4 & @xmath390 & @xmath391 & @xmath300 & tsh & 0.02@xmath392 & @xmath393 & @xmath389 & @xmath394 & -0.1 & -0.4 & @xmath395 & 1.4 + 47 & j142548.2 + 35304114:25:48.2135:30:41.1 & 0.4 & @xmath396 & @xmath397 & @xmath398 & tsh & -0.27@xmath273 & @xmath399 & @xmath400 & @xmath401 & 0.2 & 0.1 & @xmath402 & 0.7 + 48 & j142547.6 + 35304314:25:47.6735:30:43.0 & 0.7 & @xmath403 & @xmath404 & @xmath405 & th & 0.43@xmath406 & @xmath393 & @xmath407 & @xmath408 & 0.3 & -0.4 & @xmath409 & 1.1 + 49 & j142547.2 + 35372814:25:47.2535:37:28.0 & 0.9 & @xmath410 & @xmath131 & @xmath411 & t & -0.04@xmath412 & @xmath188 & @xmath413 & @xmath414 & ... & ... & @xmath2025.7 & ... + 50 & j142546.9 + 35324014:25:46.9535:32:40.3 & 0.5 & @xmath415 & @xmath416 & @xmath417 & tsh & -0.59@xmath418 & @xmath419 & @xmath99 & @xmath420 & -0.0 & -0.4 & @xmath421 & 2.0 + 51 & j142546.3 + 35334914:25:46.3335:33:49.4 & 0.8 & @xmath422 & @xmath423 & @xmath424 & t & 0.56@xmath425 & @xmath426 & @xmath427 & @xmath428 & ... & ... & @xmath2025.7 & ... + 52 & j142546.1 + 35335414:25:46.1335:33:54.9 & 0.4 & @xmath429 & @xmath430 & @xmath316 & tsh & -0.12@xmath431 & @xmath432 & @xmath433 & @xmath434 & 0.1 & -0.6 & @xmath2025.7 & ... + 53 & j142546.0 + 35382614:25:46.0235:38:26.3 & 0.5 & @xmath435 & @xmath430 & @xmath436 & tsh & 0.20@xmath437 & @xmath438 & @xmath325 & @xmath439 & 0.3 & -0.8 & @xmath440 & 0.9 + 54 & j142545.8 + 35384914:25:45.8535:38:49.2 & 0.4 & @xmath441 & @xmath138 & @xmath442 & tsh & -0.30@xmath443 & @xmath444 & @xmath445 & @xmath446 & 0.6 & -0.3 & @xmath447 & 1.4 + 55 & j142545.7 + 35322814:25:45.7835:32:28.4 & 0.7 & @xmath448 & @xmath449 & @xmath450 & ts & -1.00@xmath451 & @xmath452 & @xmath291 & @xmath453 & 0.1 & -0.2 & @xmath454 & 1.0 + 56 & j142545.6 + 35315214:25:45.6935:31:52.9 & 0.4 & @xmath455 & @xmath456 & @xmath457 & tsh & -0.38@xmath238 & @xmath149 & @xmath458 & @xmath459 & -0.1 & -0.0 & @xmath460 & 1.0 + 57 & j142545.1 + 35361614:25:45.1735:36:16.3 & 0.8 & @xmath130 & @xmath461 & @xmath462 & t & 0.34@xmath463 & @xmath387 & @xmath464 & @xmath465 & 0.2 & -0.5 & @xmath466 & 1.2 + 58 & j142544.7 + 35395414:25:44.7035:39:54.1 & 0.8 & @xmath467 & @xmath468 & @xmath469 & tsh & -0.19@xmath470 & @xmath471 & @xmath359 & @xmath472 & 0.0 & -0.2 & @xmath473 & 1.0 + 59 & j142541.6 + 35410714:25:41.6635:41:07.2 & 0.7 & @xmath474 & @xmath475 & @xmath476 & ts & -0.38@xmath477 & @xmath119 & @xmath478 & @xmath420 & 0.4 & 0.5 & @xmath479 & 1.1 + 60 & j142538.7 + 35334214:25:38.7335:33:42.1 & 0.4 & @xmath480 & @xmath481 & @xmath482 & tsh & -0.01@xmath483 & @xmath484 & @xmath134 & @xmath485 & -0.0 & -0.3 & @xmath486 & 1.1 + 61 & j142538.5 + 35342914:25:38.5135:34:29.9 & 0.4 & @xmath487 & @xmath488 & @xmath489 & th & 0.57@xmath490 & @xmath491 & @xmath492 & @xmath187 & 0.2 & -0.4 & @xmath493 & 1.0 + 62 & j142538.2 + 35343814:25:38.2935:34:38.0 & 0.4 & @xmath494 & @xmath495 & @xmath496 & tsh & -0.36@xmath497 & @xmath472 & @xmath304 & @xmath498 & 0.2 & -0.2 & @xmath499 & 1.1 + 63 & j142537.1 + 35430914:25:37.1535:43:09.8 & 1.3 & @xmath500 & @xmath501 & @xmath502 & ts & -0.56@xmath503 & @xmath504 & @xmath505 & @xmath240 & 0.5 & -0.4 & @xmath506 & 1.1 + 64 & j142536.9 + 35320914:25:36.9935:32:09.3 & 0.4 & @xmath507 & @xmath170 & @xmath508 & tsh & -0.41@xmath202 & @xmath509 & @xmath498 & @xmath510 & 0.4 & -0.3 & @xmath511 & 1.1 + 65 & j142535.9 + 35410714:25:35.9935:41:07.1 & 0.3 & @xmath512 & @xmath513 & @xmath514 & tsh & -0.40@xmath194 & @xmath515 & @xmath516 & @xmath517 & 0.2 & -0.2 & @xmath100 & 1.0 + 66 & j142534.5 + 35355114:25:34.5235:35:51.3 & 0.4 & @xmath518 & @xmath130 & @xmath519 & ts & -0.54@xmath520 & @xmath326 & @xmath521 & @xmath304 & 0.1 & 0.1 & @xmath522 & 1.2 + 67 & j142533.8 + 35413814:25:33.8035:41:38.1 & 0.7 & @xmath523 & @xmath524 & @xmath525 & tsh & -0.46@xmath372 & @xmath333 & @xmath526 & @xmath527 & -0.1 & -0.3 & @xmath528 & 1.0 + 68 & j142533.0 + 35364814:25:33.0935:36:48.5 & 0.5 & @xmath529 & @xmath530 & @xmath531 & tsh & -0.45@xmath532 & @xmath533 & @xmath366 & @xmath534 & 0.2 & -0.0 & @xmath535 & 0.8 + 69 & j142531.1 + 35392114:25:31.1735:39:21.6 & 0.6 & @xmath536 & @xmath537 & @xmath538 & th & 0.73@xmath539 & @xmath540 & @xmath541 & @xmath542 & ... & ... & @xmath2025.7 & ... + 70 & j142530.6 + 35342014:25:30.6335:34:20.3 & 0.3 & @xmath543 & @xmath544 & @xmath356 & tsh & -0.42@xmath545 & @xmath157 & @xmath546 & @xmath547 & ... & ... & @xmath2025.7 & ... + 71 &
j142529.9 + 35364014:25:29.9135:36:40.0 & 0.4 & @xmath548 & @xmath549 & @xmath550 & ts & -0.53@xmath551 & @xmath150 & @xmath552 & @xmath553 & 0.1 & -0.5 & @xmath554 & 0.8 + 72 & j142529.2 + 35324814:25:29.2235:32:48.4 & 0.4 & @xmath555 & @xmath556 & @xmath258 & tsh & -0.47@xmath557 & @xmath558 & @xmath559 & @xmath560 & ... & ... &
@xmath2025.7 & ... + 73 & j142526.6 + 35314014:25:26.6835:31:40.8 & 0.8 & @xmath561 & @xmath562 & @xmath563 & tsh & -0.13@xmath564 & @xmath565 & @xmath553 & @xmath566 & ... & ... & @xmath2025.7 & ... + 74 & j142525.7 + 35384114:25:25.7235:38:41.5 & 0.4 & @xmath87 & @xmath110 & @xmath567 & tsh & 0.72@xmath568 & @xmath569 & @xmath114 & @xmath570 & 0.6 & -0.5 & @xmath571 & 1.9 + 75 & j142524.7 + 35303714:25:24.7035:30:37.1 & 0.5 & @xmath572 & @xmath573 & @xmath574 & tsh & -0.54@xmath575 & @xmath576 & @xmath408 & @xmath577 & 0.7 & 0.1 & @xmath578 & 1.1 + 76 & j142523.6 + 35282414:25:23.6935:28:24.5 & 0.7 & @xmath579 & @xmath580 & @xmath581 & tsh & 0.62@xmath194 & @xmath582 & @xmath583 & @xmath584 & 0.2 & -0.1 & @xmath585 & 1.1 + 77 & j142521.8 + 35340714:25:21.8735:34:07.8 & 0.5 & @xmath586 & @xmath363 & @xmath322 & tsh & -0.14@xmath72 & @xmath587 & @xmath588 & @xmath587 & 0.1 & -0.4 & @xmath589 & 1.0 + 78 & j142521.3 + 35302914:25:21.3835:30:29.5 & 0.9 & @xmath590 & @xmath591 & @xmath592 & tsh & 0.42@xmath593 & @xmath594 & @xmath345 & @xmath595 & -0.4 & -0.2 & @xmath596 & 1.1 + 79 & j142517.7 + 35382214:25:17.7035:38:22.2 & 0.7 & @xmath597 & @xmath598 & @xmath109 & tsh & -0.19@xmath564 & @xmath599 & @xmath600 & @xmath601 & 0.5 & -0.3 & @xmath602 & 1.8 + 80 & j142514.3
+ 35391814:25:14.3735:39:18.2 & 0.6 & @xmath603 & @xmath604 & @xmath605 & tsh & -0.21@xmath606 & @xmath607 & @xmath608 & @xmath609 & 0.5 & -0.9 & @xmath610 & 1.2 + 81 &
j142514.3 + 35361814:25:14.3435:36:18.8 & 0.4 & @xmath611 & @xmath612 & @xmath613 & tsh & -0.59@xmath310 & @xmath311 & @xmath614 & @xmath615 & 0.1 & -0.2 & @xmath255 & 1.2 + 82 & j142513.1 + 35332314:25:13.1835:33:23.4 & 0.6 & @xmath616 & @xmath617 & @xmath618 & tsh & -0.50@xmath372 & @xmath619 & @xmath620 & @xmath621 & 0.4 & -0.3 & @xmath585 & 1.4 + 83 & j142512.0 + 35312414:25:12.0035:31:24.3 & 0.7 & @xmath622 & @xmath623 & @xmath624 & tsh & -0.53@xmath625 & @xmath626 & @xmath627 & @xmath628 & 0.6 & 0.1 & @xmath629 & 1.0 + 84 & j142511.3 + 35385714:25:11.3235:38:57.8 & 0.6 & @xmath630 & @xmath631 & @xmath632 & tsh & -0.47@xmath310 & @xmath633 & @xmath634 & @xmath635 & -0.2 & 0.2 & @xmath636 & 1.5 + 85 & j142509.3 + 35435614:25:09.3935:43:56.1 & 1.1 & @xmath637 & @xmath638 & @xmath639 & tsh & -0.52@xmath194 & @xmath640 & @xmath641 & @xmath642 & 0.2 & -0.6 & @xmath643 & 1.2 + 86 & j142504.5 + 35410714:25:04.5835:41:07.9 & 0.6 & @xmath644 & @xmath645 & @xmath646 & tsh & -0.42@xmath647 & @xmath648 & @xmath649 & @xmath650 & 0.6 & 0.3 & @xmath651 & 1.1 + 87 & j142457.1 + 35351814:24:57.1635:35:18.8 & 0.9 & @xmath652 & @xmath653 & @xmath654 & tsh & -0.53@xmath246 & @xmath655 & @xmath165 & @xmath656 & 0.5 & -0.7 & @xmath657 & 1.0 + 88 & j142618.8 + 35421814:26:18.8435:42:18.4 & 1.5 & @xmath658 & @xmath659 & @xmath660 & tsh & -0.37@xmath217 & @xmath661 & @xmath662 & @xmath663 & 0.9 & 0.1 & @xmath664 & 1.1 + 89 & j142618.6 + 35375414:26:18.6035:37:54.2 & 2.2 & @xmath665 & @xmath666 & @xmath667 & th & 0.55@xmath668 & @xmath669 & @xmath670 & @xmath671 & -0.3 & -0.7 & @xmath672 & 1.2 + 90 & j142611.9 + 35423014:26:11.9235:42:30.5 & 1.7 & @xmath673 & @xmath674 & @xmath675 & ts & -0.56@xmath253 & @xmath676 & @xmath677 & @xmath678 & 0.6 & -0.0 & @xmath242 & 1.0 + 91 & j142610.8 + 35420214:26:10.8935:42:02.5 & 1.5 & @xmath679 & @xmath680 & @xmath681 & tsh & 0.41@xmath682 & @xmath683 & @xmath684 & @xmath685 & 1.1 & -0.2 & @xmath686 & 1.3 + 92 & j142606.7 + 35315114:26:06.7235:31:51.3 & 1.5 & @xmath687 & @xmath688 & @xmath689 & ts & -0.44@xmath690 & @xmath691 & @xmath484 & @xmath634 & -0.3 & 0.3 & @xmath692 & 1.1 + 93 & j142604.7 + 35301514:26:04.7835:30:15.3 & 1.7 & @xmath693 & @xmath694 & @xmath695 & tsh & 0.10@xmath104 & @xmath696 & @xmath697 & @xmath698 & 1.1 & -0.2 & @xmath699 & 1.2 + 94 & j142603.7 + 35324614:26:03.7335:32:46.2 & 1.6 & @xmath700 & @xmath364 & @xmath701 & t & 0.21@xmath702 & @xmath703 & @xmath704 & @xmath303 & -0.4 & 0.1 & @xmath705 & 1.3 + 95 & j142600.2 + 35344214:26:00.2435:34:42.4 & 0.8 & @xmath706 & @xmath707 & @xmath708 & th & 1.00@xmath709 & @xmath358 & @xmath710 & @xmath711 & 0.3 & -0.6 & @xmath712 & 1.1 + 96 & j142558.1 + 35321614:25:58.1335:32:16.1 & 1.5 & @xmath713 & @xmath714 & @xmath715 & ts & -0.33@xmath716 & @xmath717 & @xmath359 & @xmath697 & ... & ... & @xmath2025.7 & ... + 97 & j142552.4 + 35272414:25:52.4735:27:24.1 & 1.8 & @xmath718 & @xmath719 & @xmath720 & tsh & -0.14@xmath721 & @xmath722 & @xmath526 & @xmath723 & -0.6 & -0.5 & @xmath724 & 1.0 + 98 & j142551.4 + 35343314:25:51.4335:34:33.1 & 0.7 & @xmath118 & @xmath707 & @xmath125 & th & 1.00@xmath725 & @xmath225 & @xmath726 & @xmath727 & 0.1 & -0.1 & @xmath728 & 1.3 + 99 & j142551.1 + 35301514:25:51.1635:30:15.7 & 1.5 & @xmath729 & @xmath730 & @xmath731 & ts & -0.56@xmath732 & @xmath733 & @xmath734 & @xmath133 & 0.1 & 0.2 & @xmath735 & 0.9 + 100 & j142547.4 + 35272014:25:47.4335:27:20.2 & 1.4 & @xmath736 & @xmath737 & @xmath738 & tsh & -0.41@xmath739 & @xmath740 & @xmath741 & @xmath742 & 0.4 & -0.3 & @xmath743 & 1.0 + 101 & j142545.5 + 35300314:25:45.5335:30:03.6 & 1.3 & @xmath744 & @xmath745 & @xmath746 & ts & -0.03@xmath747 & @xmath141 & @xmath748 & @xmath749 & 0.7 & 0.6 & @xmath750 & 1.0 + 102 & j142544.7 + 35404614:25:44.7735:40:46.6 & 1.1 & @xmath448 & @xmath111 & @xmath751 & th & 0.38@xmath752 & @xmath472 & @xmath704 & @xmath187 & 0.1 & -0.7 & @xmath753 & 1.2 + 103 & j142544.2 + 35401814:25:44.2035:40:18.4 & 0.9 & @xmath754 & @xmath755 & @xmath756 & th & 0.86@xmath757 & @xmath727 & @xmath758 & @xmath241 & ... & ... & @xmath2025.7 & ... + 104 & j142541.8 + 35424914:25:41.8035:42:49.7 & 1.3 & @xmath759 & @xmath562 & @xmath760 & tsh & -0.13@xmath564 & @xmath761 & @xmath326 & @xmath615 & -0.9 & 0.6 & @xmath762 & 1.2 + 105 & j142539.6 + 35394214:25:39.6535:39:42.6 & 0.8 & @xmath708 & @xmath763 & @xmath764 & ts & -0.24@xmath765 & @xmath260 & @xmath766 & @xmath498 & 0.4 & -0.5 & @xmath767 & 1.1 + 106 & j142536.8 + 35314714:25:36.8135:31:47.2 & 1.0 & @xmath768 & @xmath769 & @xmath770 & th & 0.63@xmath771 & @xmath772 & @xmath773 & @xmath774 & 0.2 & -0.2 & @xmath775 & 2.0 + 107 & j142533.3 + 35313014:25:33.3335:31:30.6 & 0.8 & @xmath776 & @xmath777 & @xmath778 & th & 0.49@xmath779 & @xmath780 & @xmath781 & @xmath587 & 0.5 & -0.2 & @xmath782 & 1.0 + 108 & j142532.6 + 35283814:25:32.6635:28:38.8 & 1.4 & @xmath783 & @xmath784 & @xmath785 & ts & -0.66@xmath786 & @xmath787 & @xmath788 & @xmath485 & 0.9 & 0.9 & @xmath789 & 1.4 + 109 & j142531.4 + 35445214:25:31.4135:44:52.5 & 2.1 & @xmath790 & @xmath791 & @xmath792 & tsh & 0.35@xmath793 & @xmath794 & @xmath135 & @xmath795 & -0.0 & 1.4 & @xmath796 & 1.2 + 110 & j142530.8 + 35320814:25:30.8235:32:08.6 & 1.1 & @xmath797 & @xmath531 & @xmath798 & t & 0.03@xmath799 & @xmath800 & @xmath114 & @xmath282 & -0.1 & -0.4 & @xmath801 & 1.2 + 111 & j142528.2 + 35395814:25:28.2435:39:58.4 & 0.9 & @xmath802 & @xmath803 & @xmath804 & th & 0.64@xmath805 & @xmath282 & @xmath806 & @xmath807 & 0.5 & -0.3 & @xmath168 & 1.0 + 112 & j142527.5 + 35401214:25:27.5935:40:12.1 & 0.8 & @xmath808 & @xmath809 & @xmath810 & tsh & 0.10@xmath811 & @xmath812 & @xmath766 & @xmath813 & ... & ... & @xmath2025.7 & ... + 113 & j142527.4 + 35310114:25:27.4935:31:01.0 & 1.2 & @xmath814 & @xmath815 & @xmath816 & ts & -0.34@xmath817 & @xmath505 & @xmath818 & @xmath819 & 0.4 & -0.1 & @xmath820 & 1.3 + 114 & j142526.6 + 35332714:25:26.6635:33:27.2 & 0.7 & @xmath821 & @xmath822 & @xmath823 & ts & -0.53@xmath824 & @xmath559 & @xmath114 & @xmath825 & -1.0 & -0.1 & @xmath826 & 1.2 + 115 & j142524.2 + 35292814:25:24.2235:29:28.6 & 1.4 & @xmath827 & @xmath828 & @xmath829 & tsh & -0.14@xmath830 & @xmath691 & @xmath831 & @xmath832 & ... & ... & @xmath2025.7 & ... + 116 & j142523.8 + 35412214:25:23.8035:41:22.4 & 0.8 & @xmath833 & @xmath834 & @xmath835 & ts & -0.59@xmath836 & @xmath594 & @xmath837 & @xmath838 & 0.0 & 0.2 & @xmath839 & 1.0 + 117 & j142523.6 + 35321614:25:23.6935:32:16.7 & 1.1 & @xmath840 & @xmath841 & @xmath436 & tsh & 0.14@xmath842 & @xmath121 & @xmath290 & @xmath843 & 0.6 & -0.5 & @xmath844 & 0.8 + 118 & j142523.0 + 35431314:25:23.0035:43:13.4 & 1.1 & @xmath845 & @xmath846 & @xmath847 & tsh & -0.28@xmath238 & @xmath848 & @xmath849 & @xmath850 & 0.3 & 0.8 & @xmath851 & 1.0 + 119 & j142522.8 + 35311614:25:22.8235:31:16.8 & 1.0 & @xmath852 & @xmath323 & @xmath853 & th & 0.68@xmath854 & @xmath855 & @xmath781 & @xmath856 & 0.7 & -0.2 & @xmath857 & 1.1 + 120 & j142522.4 + 35351714:25:22.4035:35:17.2 & 0.6 & @xmath858 & @xmath859 & @xmath329 & t & 0.40@xmath860 & @xmath819 & @xmath413 & @xmath861 & ... & ... & @xmath2025.7 & ... + 121 & j142521.1 + 35334014:25:21.1035:33:40.0 & 0.7 & @xmath862 & @xmath863 & @xmath317 & ts & -0.73@xmath864 & @xmath780 & @xmath142 & @xmath158 & 0.0 & -0.6 & @xmath865 & 1.1 + 122 & j142517.6 + 35345314:25:17.6135:34:53.0 & 0.5 & @xmath866 & @xmath867 & @xmath868 & ts & -0.20@xmath765 & @xmath434 & @xmath290 & @xmath358 & 0.7 & 0.3 & @xmath869 & 1.0 + & & & & & & & & & & & -0.2 & -1.3 & @xmath870 & 1.3 + 123 & j142516.6 + 35412914:25:16.6735:41:29.6 & 1.3 & @xmath871 & @xmath872 & @xmath873 & ts & -0.44@xmath690 & @xmath504 & @xmath874 & @xmath614 & 0.2 & 0.1 & @xmath875 & 1.2 + 124 & j142516.4 + 35294014:25:16.4335:29:40.7 & 1.2 & @xmath876 & @xmath877 & @xmath878 & tsh & -0.33@xmath879 & @xmath880 & @xmath452 & @xmath881 & 0.2 & 0.8 & @xmath882 & 1.2 + 125 & j142516.0 + 35314214:25:16.0035:31:42.3 & 1.1 & @xmath883 & @xmath884 & @xmath885 & t & -0.22@xmath886 & @xmath465 & @xmath521 & @xmath887 & -0.1 & -0.8 & @xmath888 & 1.0 + 126 & j142514.9 + 35400414:25:14.9035:40:04.1 & 0.9 & @xmath889 & @xmath890 & @xmath891 & tsh & -0.30@xmath690 & @xmath241 & @xmath892 & @xmath400 & 1.0 & 0.2 & @xmath893 & 1.2 + 127 & j142510.9 + 35360214:25:10.9435:36:02.8 & 1.0 & @xmath894 & @xmath895 & @xmath341 & tsh & 0.18@xmath896 & @xmath897 & @xmath734 & @xmath898 & -0.9 & 0.3 & @xmath899 & 1.0 + 128 & j142510.1 + 35341114:25:10.1935:34:11.4 & 0.9 & @xmath900 & @xmath901 & @xmath902 & tsh & -0.23@xmath903 & @xmath73 & @xmath772 & @xmath904 & -0.6 & 0.1 & @xmath905 & 1.2 + 129 & j142509.5 + 35424414:25:09.5135:42:44.9 & 2.1 & @xmath906 & @xmath907 & @xmath908 & th & 0.58@xmath909 & @xmath910 & @xmath366 & @xmath911 & -0.8 & 2.0 & @xmath912 & 1.4 + 130 & j142509.5 + 35374714:25:09.5035:37:47.5 & 0.6 & @xmath913 & @xmath914 & @xmath915 & tsh & 0.42@xmath916 & @xmath917 & @xmath918 & @xmath919 & 0.5 & -0.1 & @xmath920 & 1.2 + 131 & j142509.2 + 35352714:25:09.2335:35:27.5 & 0.8 & @xmath921 & @xmath922 & @xmath923 & tsh & 0.74@xmath924 & @xmath925 & @xmath818 & @xmath926 & -0.3 & -0.0 & @xmath927 & 1.0 + 132 & j142505.1 + 35405414:25:05.1635:40:54.9 & 0.8 & @xmath928 & @xmath929 & @xmath930 & tsh & -0.27@xmath575 & @xmath931 & @xmath419 & @xmath932 & -0.5 & 0.6 & @xmath933 & 1.0 + 133 & j142504.4 + 35351414:25:04.4735:35:14.6 & 1.5 & @xmath934 & @xmath935 & @xmath201 & ts & -0.33@xmath80 & @xmath936 & @xmath937 & @xmath938 & 0.3 & -0.7 & @xmath939 & 1.0 + 134 & j142503.9 + 35334514:25:03.9035:33:45.7 & 1.5 & @xmath940 & @xmath810 & @xmath941 & tsh & 0.37@xmath942 & @xmath943 & @xmath944 & @xmath945 & -0.9 & -0.2 & @xmath946 & 1.0 + 135 & j142503.1 + 35342114:25:03.1635:34:21.2 & 1.2 & @xmath947 & @xmath948 & @xmath949 & tsh & 0.56@xmath950 & @xmath951 & @xmath304 & @xmath952 & 1.1 & 0.5 & @xmath2025.7 & ... + 136 & j142502.9 + 35351214:25:02.9435:35:12.8 & 1.0 & @xmath953 & @xmath954 & @xmath955 & tsh & -0.24@xmath164 & @xmath956 & @xmath957 & @xmath958 & -0.6 & -0.2 & @xmath959 & 1.1 + 137 & j142501.2 + 35362214:25:01.2035:36:22.0 & 1.2 & @xmath960 & @xmath961 & @xmath962 & tsh & -0.42@xmath238 & @xmath373 & @xmath963 & @xmath964 & 0.5 & 0.0 & @xmath965 & 1.0 + 138 & j142457.3 + 35362714:24:57.3335:36:27.5 & 1.2 & @xmath966 & @xmath967 & @xmath968 & tsh & 0.04@xmath969 & @xmath970 & @xmath971 & @xmath972 & 0.5 & 0.2 & @xmath2025.7 & ... + 139 & j142454.8 + 35343114:24:54.8035:34:31.9 & 0.9 & @xmath973 & @xmath974 & @xmath975 & tsh & -0.25@xmath217 & @xmath976 & @xmath977 & @xmath978 & 0.2 & -0.7 & @xmath979 & 1.1 + 140 & j142616.9 + 35372914:26:16.9435:37:29.4 & 1.9 & @xmath980 & @xmath981 & @xmath982 & tsh & 0.35@xmath950 & @xmath983 & @xmath478 & @xmath984 & 0.9 & 0.7 & @xmath985 & 1.5 + 141 & j142613.9 + 35344514:26:13.9735:34:45.5 & 1.6 & @xmath986 & @xmath987 & @xmath988 & ts & -0.50@xmath989 & @xmath990 & @xmath360 & @xmath991 & ... & ... & @xmath2025.7 & ... + 142 & j142601.7 + 35403714:26:01.7235:40:37.5 & 1.3 & @xmath992 & @xmath993 & @xmath994 & ts & -0.28@xmath995 & @xmath141 & @xmath188 & @xmath996 & 0.2 & -0.8 & @xmath997 & 1.0 + 143 & j142601.7 + 35280214:26:01.7135:28:02.8 & 2.7 & @xmath998 & @xmath999 & @xmath1000 & t & 0.17@xmath1001 & @xmath393 & @xmath290 & @xmath1002 & 0.8 & 0.3 & @xmath985 & 1.8 + 144 & j142530.9 + 35275614:25:30.9535:27:56.8 & 2.4 & @xmath1003 & @xmath1004 & @xmath1005 & ts & -0.64@xmath1006 & @xmath1007 & @xmath282 & @xmath1008 & 0.6 & -0.9 & @xmath1009 & 1.3 + 145 & j142527.5 + 35265614:25:27.5335:26:56.2 & 3.1 & @xmath1010 & @xmath1011 & @xmath1012 & ts & -0.67@xmath551 & @xmath1013 & @xmath458 & @xmath1014 & -0.3 & 0.9 & @xmath1015 & 1.1 + 146 & j142524.4 + 35254214:25:24.4035:25:42.1 & 2.5 & @xmath1016 & @xmath1017 & @xmath1018 & ts & -0.27@xmath879 & @xmath1019 & @xmath1020 & @xmath1021 & -0.7 & -0.4 & @xmath1022 & 1.0 + 147 & j142519.7 + 35443214:25:19.7935:44:32.2 & 2.1 & @xmath660 & @xmath1023 & @xmath1024 & ts & -0.43@xmath879 & @xmath1025 & @xmath420 & @xmath1026 & -0.5 & 0.4 & @xmath1027 & 1.0 + 148 & j142517.7 + 35375414:25:17.7135:37:54.3 & 1.1 & @xmath841 & @xmath1028 & @xmath1029 & t & 0.79@xmath1030 & @xmath534 & @xmath758 & @xmath1031 & 0.7 & -0.6 & @xmath1032 & 1.3 + 149 & j142516.0 + 35432514:25:16.0235:43:25.8 & 2.3 & @xmath1033 & @xmath1034 & @xmath1035 & ts & -1.00@xmath1036 & @xmath1037 & @xmath1038 & @xmath1039 & ... & ... & @xmath2025.7 & ... + 150 & j142505.2 + 35372914:25:05.2535:37:29.5 & 1.6 & @xmath1040 & @xmath1041 & @xmath1042 & ts & -0.78@xmath1043 & @xmath352 & @xmath366 & @xmath1044 & 0.2 & -0.2 & @xmath2025.7 & ... + 151 & j142503.5 + 35385914:25:03.5735:38:59.2 & 1.7 & @xmath1045 & @xmath1046 & @xmath700 & ts & -0.20@xmath1047 & @xmath1048 & @xmath553 & @xmath1049 & -0.6 & -0.2 & @xmath1050 & 1.2 + 152 & j142501.5 + 35400014:25:01.5835:40:00.3 & 1.9 & @xmath1051 & @xmath1052 & @xmath1053 & ts & -0.34@xmath690 & @xmath594 & @xmath1054 & @xmath1055 & 1.0 & -1.6 & @xmath1056 & 1.0 + 153 & j142501.2 + 35305714:25:01.2835:30:57.4 & 2.2 & @xmath1057 & @xmath1058 & @xmath1059 & tsh & 0.18@xmath437 & @xmath1060 & @xmath1061 & @xmath1062 & -1.7 & 0.1 & @xmath1063 & 1.0 + 154 & j142459.2 + 35375114:24:59.2435:37:51.4 & 1.8 & @xmath1064 & @xmath1065 & @xmath1066 & ts & -0.21@xmath1067 & @xmath1068 & @xmath135 & @xmath1069 & -0.6 & -0.4 & @xmath1070 & 1.2 + 155 & j142611.8 + 35330114:26:11.8535:33:01.9 & 1.7 & @xmath1071 & @xmath1072 & @xmath1073 & t & 0.83@xmath1074 & @xmath991 & @xmath758 & @xmath904 & -0.4 & 0.6 & @xmath1075 & 1.1 + 156 & j142540.2 + 35462314:25:40.2135:46:23.0 & 3.8 & @xmath1076 & @xmath1077 & @xmath1078 & t & 0.57@xmath1079 & @xmath1080 & @xmath1061 & @xmath1081 & -0.7 & 1.1 & @xmath1082 & 1.6 + 157 & j142539.8 + 35443814:25:39.8735:44:38.6 & 2.4 & @xmath1083 & @xmath1084 & @xmath1085 & t & 0.23@xmath1086 & @xmath394 & @xmath552 & @xmath1087 & -1.3 & 0.4 & @xmath1088 & 1.2 + 158 & j142531.6 + 35265914:25:31.6035:26:59.0 & 3.5 & @xmath1089 & @xmath1090 & @xmath1091 & th & 0.69@xmath1092 & @xmath1093 & @xmath944 & @xmath1094 & 2.1 & 2.0 & @xmath1095 & 1.0 + 159 & j142448.4 + 35322714:24:48.4835:32:27.4 & 2.8 & @xmath1096 & @xmath1097 & @xmath1098 & t & -0.29@xmath1099 & @xmath1100 & @xmath432 & @xmath1101 & -0.6 & 1.0 & @xmath1102 & 1.1 + 160 & j142447.8 + 35313814:24:47.8035:31:38.8 & 3.3 & @xmath1103 & @xmath1104 & @xmath1105 & t & -0.22@xmath1106 & @xmath1107 & @xmath225 & @xmath990 & ... & ... & ... @xmath334 & ... + 161 & j142528.8 + 35334214:25:28.8035:33:42.0 & 0.6 & @xmath1108 & @xmath1109 & @xmath1110 & s & -0.50@xmath1111 & @xmath82 & @xmath1112 & @xmath290 & 0.6 & -0.1 & @xmath1113 & 1.3 + 162 & j142558.2 + 35314414:25:58.2935:31:44.4 & 1.6 & @xmath1114 & @xmath364 & @xmath1115 & s & 0.04@xmath1116 & @xmath465 & @xmath1117 & @xmath1118 & 0.5 & -0.2 & @xmath1119 & 1.3 + 163 & j142458.1 + 35391914:24:58.1435:39:19.4 & 2.5 & @xmath1120 & @xmath1121 & @xmath1122 & s & -0.32@xmath1123 & @xmath127 & @xmath1038 & @xmath121 & 0.8 & 0.1 & @xmath321 & 0.9 + 164 & j142535.9 + 35310214:25:35.9635:31:02.8 & 1.1 & @xmath1124 & @xmath1125 & @xmath1126 & h & 1.00@xmath1127 & @xmath181 & @xmath1128 & @xmath587 & -0.2 & -0.4 & @xmath1129 & 1.5 + 165 & j142527.4 + 35325714:25:27.4635:32:57.9 & 0.6 & @xmath1130 & @xmath707 & @xmath125 & h & 1.00@xmath1131 & @xmath135 & @xmath710 & @xmath1132 & 0.8 & -0.6 & @xmath1133 & 1.3 + 166 & j142545.4 + 35271114:25:45.4435:27:11.8 & 2.5 & @xmath1134 & @xmath1135 & @xmath1136 & h & 0.89@xmath1137 & @xmath1138 & @xmath427 & @xmath1139 & -1.9 & -0.0 & @xmath1140 & 1.2 + 167 & j142523.4 + 35351214:25:23.4935:35:12.4 & 0.7 & @xmath715 & @xmath1125 & @xmath1141 & h & 1.00@xmath1142 & @xmath158 & @xmath1128 & @xmath1118 & 0.1 & -0.4 & @xmath1143 & 1.3 + 168 & j142621.6 + 35393114:26:21.6335:39:31.1 & 3.2 & @xmath1144 & @xmath1145 & @xmath1146 & h & 0.87@xmath1147 & @xmath439 & @xmath427 & @xmath1148 & 1.4 & -2.1 & @xmath2025.7 & ... + | we present an analysis of a deep , 172 ks @xmath0 observation of the large area lyman alpha survey ( lala ) botes field , obtained with the advanced ccd imaging spectrometer ( acis - i ) on the @xmath0 x - ray observatory .
this is one of the deepest @xmath0 images of the extragalactic sky ; only the 2 ms cdf - n and 1 ms cdf - s are substantially deeper . in this paper
we present the x - ray source catalog obtained from this image , along with an analysis of source counts , and optical identifications .
the x - ray image is composed of two individual observations obtained in 2002 , and reaches 0.5 2.0 and 2.0 10.0 kev flux limits of 1.5 @xmath1 10@xmath2 and 1.0 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 , respectively , for point sources near the aim point .
a total of 168 x - ray sources were detected : 160 in the 0.5 7.0 kev band , 132 in the 0.5 2.0 kev band , and 111 in the 2.0 7.0 kev band . the x - ray source counts were derived and compared with those from other @xmath0 deep surveys ; the hard x - ray source density of the lala botes field is 33% higher than that of cdf - s at the flux level of 2.0 @xmath1 10@xmath3 ergs @xmath4 s@xmath5 , confirming the field - to - field variances of the hard band source counts reported by previous studies .
the deep exposure resolves @xmath6 72% of the 2.0 10.0 kev x - ray background .
our primary optical data are @xmath7-band imaging from noao deep wide - field survey ( ndwfs ) , with limiting magnitude of @xmath7 = 25.7 ( vega , 3@xmath8 , 4 diameter aperture ) .
we have found optical counterparts for 152 of the 168 @xmath0 sources ( 90% ) ; 144 of these are detected on the @xmath7-band image , and 8 have optical counterparts in other bands ( either @xmath9,@xmath10,@xmath11,@xmath12 ) . among the @xmath7-band non - detected sources ,
not more than 11 of them can possibly be at @xmath13 5 , based on the hardness ratios of their x - ray emission and nondetections in bluer bands ( @xmath9,@xmath10 ) .
the majority ( @xmath14 76% ) of the x - ray sources are found to have log(@xmath15/@xmath16 ) within 0.0@xmath171 , which are believed to be agns .
most of the x - ray faint / optically bright sources ( log(@xmath15/@xmath16 ) @xmath18 -1.0 ) are optically extended , which are low - z normal galaxies or low luminosity agns .
there is also a population of sources which are x - ray @xmath19 for their optical magnitudes ( log(@xmath15/@xmath16 ) @xmath20 1.0 ) , which are harder in x - ray and are probably obscured agns . |
among the many important physics processes which can be studied at a future high - energy photon linear collider ( plc ) @xcite , the production of higgs bosons , @xmath6 , stands out .
in particular , identification of an intermediate - mass standard model higgs boson via its ( dominant ) @xmath7 decay channels seems feasible and would provide an important measurement of the @xmath0 and @xmath1 partial decay widths . the dominant ( continuum )
background to @xmath8 comes from the standard model processes @xmath9 .
various techniques can be used to suppress these .
the most effective appears to be to _ polarize _ the initial
state photons : higgs bosons are only produced from an initial state with @xmath10 , whereas the leading - order @xmath11 ( @xmath12 ) backgrounds are dominantly produced from the @xmath13 initial state .
more precisely , production from the @xmath10 state is suppressed by @xmath14 @xcite .
hence in the region of the higgs resonance @xmath15 and the @xmath16 and @xmath17 backgrounds are heavily suppressed . unfortunately , the @xmath18 suppression of the @xmath2 background does not survive beyond leading order in qcd perturbation theory .
in particular , there is no corresponding mass suppression of the @xmath19 @xmath20 @xmath10 amplitude squared .
although this gives rise notionally to three - jet configurations , in certain circumstances ( for example , collinear @xmath21 splitting followed by @xmath22 compton scattering ) these can mimic the two - jet higgs @xmath23 signal .
of course a combination of @xmath24-tagging of both jets and three - jet veto can _ in principle _ eliminate the bulk of this background , and one is left considering the perturbative corrections from virtual and soft real gluon emission .
these have the same @xmath25 suppression as the leading - order contribution helicity projection .
] however the perturbation series now contains large double logarithmic corrections @xmath26 at each order @xmath27 , which potentially can give sizable corrections .
apart from their phenomenological importance , these logarithms are of theoretical interest .
they do not have an obvious ` sudakov ' exponentiating behavior , arising from the fact that they originate in non - trivial regions of soft gluon phase space .
furthermore , for any particular virtual gluon diagram they are mixed up with the genuine infra - red logarithms @xmath28 , where @xmath29 is a fictitious gluon mass introduced to control the soft divergences . in ref .
@xcite the double logarithmic contributions were studied in considerable detail through two loops , i.e. up to and including @xmath30 contributions .
inspection of the one - and two - loop coefficients led to the claim that , for @xmath31 and @xmath32 , the perturbation series was reasonably well approximated by the lowest two orders .
however from both a theoretical and practical pont of view it is clearly desirable to have more information about the behavior of the double - logarithmic ( dl ) terms to all orders , in particular to see if resummation is possible .
this is the goal of the present paper .
as we shall see , a careful study of all the relevant two- and three - loop diagrams _ does _ reveal the pattern of the dl terms , and allows the all - orders result to be derived and resummed . in this paper
we shall restrict ourselves to the theoretical study of the origin and calculation of the dl terms . in a future paper
we will investigate their numerical effect on realistic plc cross sections .
we begin , in section 2 , by deriving the leading order result for the @xmath10 cross section , identifying the origin of the @xmath18 suppression . in section 3
we discuss the structure of the one - loop corrections , show how the dl contributions arise , and compare these with the known one loop result @xcite .
the main part of the paper is section 4 , where we calculate all three - loop dl contributions and deduce the all - orders result .
section 5 contains a summary of our results .
we begin with a rederivation of the underlying born process . recall that for an initial state with @xmath2 the cross section vanishes in the massless limit . in other words ,
the background to the standard higgs production process via a fermion loop and subsequent decay into a @xmath33 or @xmath34 final state is suppressed by a factor of @xmath35 .
higher order radiative ( bremsstrahlung ) corrections remove this suppression in principle @xcite .
one therefore needs to keep at least terms of @xmath36 in the calculation of the born amplitudes shown in fig .
[ fig : ppqqb ] .
we are also only interested in the region of large scattering angle @xmath37 , which is the dominant event topology of the higgs signal @xcite . keeping only terms of @xmath38
, we find it convenient to separate out the mass dependence of the spinors according to @xcite @xmath39 where the spinors on the r.h.s .
are massless and @xmath40 is an arbitrary reference four - momentum with @xmath41 .
the separation of the mass terms then allows for the convenient usage of the polarization vector @xcite @xmath42 where here we denote the arbitrary massless reference momentum by @xmath43 . in the following
we will also use simply @xmath44 for an arbitrary fermion mass , as the results derived do not depend on any heavy quark approximation .
it is understood that for any application , the relevant fermion masses for the background to higgs production below the @xmath45 threshold are the @xmath24- and @xmath46-quark masses . the cross section of the latter is enhanced by a relative factor of @xmath47 from the different charges of the quarks . using the above notation we find for the born amplitude @xmath48 with an initial @xmath2 state : @xmath49 with @xmath50 , @xmath51 the electromagnetic coupling with magnitude @xmath52 and @xmath37 the scattering angle .
the kronecker-@xmath53 s allow for a non - zero contribution only in case all four helicities are the same .
the result agrees with refs .
@xcite up to @xmath54 and displays the somewhat surprising feature of a bosonic angular dependence ( as in bhabha scattering for instance ) , with no @xmath37-terms in the numerator .
this feature leads one to suspect that by keeping only the mass term of the virtual fermion propagator , one should obtain eq .
[ eq : ppqqb ] very easily . a detailed analysis of this contribution , however , yields : @xmath55 \label{eq : ppqqbm}\end{aligned}\ ] ] while it does not agree with the exact result in eq .
[ eq : ppqqb ] , it is also evident that for small angles and high energies both solutions do agree .
it was found in ref .
@xcite that this feature holds also through one and two loops if one is interested only in the leading logarithmic accuracy at large angles .
the reason for this seemingly contradictory behavior is that while the angle between the outgoing fermion ( antifermion ) and the beamline is large , the angle between the incoming photon and the emitted soft parton is small .
therefore , while it is wrong to keep only the @xmath44-term in the propagator of the virtual fermion for the born amplitude , it actually _ does _ give the correct leading logarithmic structure of higher order contributions with a soft fermion line . for higher order corrections
it is also convenient to use the following expression for the @xmath2 @xmath56 amplitude : @xmath57 in the following we describe how to obtain the double logarithms in the one loop approximation and introduce the techniques applied in the later stages of the paper .
at the one loop level , the only diagram that can contribute large double logarithms is the box diagram with a virtual gluon connecting the outgoing @xmath58 pair .
while the general method of employing sudakov parametrizations and integrating over the perpendicular components of soft loop momenta is well known @xcite and has been used extensively @xcite , we still find it useful to briefly review some basic properties . in section [ sec : res ]
we will also remark on some subtleties one encounters with higher order virtual corrections .
large logarithms occur only when there are small and large scales connected in a feynman diagram . while the particular feynman diagram is gauge dependent , the leading logarithmic contribution is not . in our case , the box diagram is connected with vertex and self energy corrections to form a gauge invariant cross section .
however , in any gauge only the four - point diagram can yield the desired pair of large logarithms .
it is therefore independent of the particular choice of gauge .
a familiar and analogous example of the same mechanism is the gauge invariance of the infra red @xmath59 term in the context of dimensional regularization for a massless box diagram . in fig .
[ fig:1l ] we list the four distinct topologies that allow a doubly logarithmic contribution from the box diagram .
the `` blob '' always denotes a hard momentum flowing through the omitted propagator and in each case there is a corresponding soft momentum flowing through the propagator facing the hard blob .
the feature which makes this process interesting and very different from the standard sudakov form factor is the fact that we have soft _ fermion _ contributions as well .
`` soft '' means in this context soft compared to the hard blob .
the sudakov decomposition of loop momenta should be chosen such that the external four momenta joining the soft parton line are used as the base vectors . and can lead to technical difficulties . ] for the graphs of fig .
[ fig:1l ] , however , we take : @xmath60 at large angles and in the doubly logarithmic ( dl ) approximation , we have @xmath61 the phase space for the large logarithms is given by @xmath62 , with @xmath63 the integrations over the transverse momenta of the soft particles is performed by taking half of the residues in the corresponding propagators : @xmath64 the dl - contribution of a particular feynman diagram is thus given by @xmath65 where the @xmath66 are given by integrals over the remaining sudakov parameters at the @xmath27-loop level : @xmath67 the factors @xmath68 are specific to the given feynman diagram and contain its color factors and the restrictions on the occurring sudakov variables necessary to ensure that the matrix element has logarithmic behavior in each of them .
the range of integration is restricted by demanding that there is at least one pole term on the opposite side of the contour of integration from the other propagator terms , since cauchy s theorem would otherwise give a vanishing contribution .
as a simple example , we now apply the above techniques to the one loop diagrams of fig .
[ fig:1l ] . for the standard infrared form factor of topology @xmath69
the sudakov phase space is schematically given in fig .
[ fig : spsp ] and leads to the following result : @xmath70 \nonumber \\ & = & - c_f \frac{\alpha_s}{2 \pi } \left ( \frac{1}{2 } \log^2 \frac{m^2}{s } + \log \frac{m^2}{s } \log \frac{\lambda^2}{m^2 } \right ) \label{eq : sff}\end{aligned}\ ] ] the result is identical to the usual dl - sudakov form factor @xcite . for the hard form factors of topologies @xmath71 and @xmath72 we have analogously : @xmath73 taken together , and noting that topology @xmath72 has a multiplicity factor @xmath74 , we thus have rederived the one loop structure of refs . @xcite and @xcite in the dl - approximation .
the fact that the born amplitude factors out is trivial for topology @xmath69 . for topologies
@xmath71 and @xmath72 the exact numerator factorization was derived in ref . @xcite . in the former ,
it is connected with the aforementioned purely bosonic angular dependence of @xmath56 of eq .
[ eq : ppqqb ] , as it contains a hard subprocess @xmath75 . in the latter
, one needs to exploit the dirac equation and gauge invariance in order to factor out the hard subprocess @xmath76 and the born amplitude .
details are given in @xcite .
with these results , we can now proceed to investigate the higher order structure of the leading logarithmic corrections from insertions of multiple gluons in the basic contributions shown in fig .
[ fig:1l ] .
the next section gives explicit results through three loops which enable us to arrive at an all orders resummation of both hard as well as soft logarithmic corrections .
in this section we derive the expressions for the leading logarithmic corrections to all orders for the @xmath2 initial state @xmath77 process .
while it is possible to rigorously derive these corrections to all orders for topologies @xmath69 and @xmath71 , for topology @xmath72 only the new hard `` non - sudakov '' logarithms can be treated analytically . for soft insertions
we derive explicit three loop results and show that essentially the same factorization structure emerges as in @xmath71 . this behavior
is then extrapolated to all orders . at this point
we note that we have recalculated the virtual two loop corrections presented in ref .
@xcite and agree with their results . in the following ,
we therefore restrict ourselves to a discussion of the three loop level .
all resummed formulae were checked explicitly through one- , two- and three - loops .
we begin with the sudakov form factor of topology @xmath69 . at three loops ,
the relevant feynman diagrams are given in fig .
[ fig : a3ls ] .
the summation is analogous to the soft vertex corrections where the exponentiation to all orders for a color singlet final state was proved in ref .
@xcite to leading order by employing an asymptotic hamiltonian as well as the coherent state formalism .
explicit three loop calculations were performed in the leading logarithmic approximation in ref .
@xcite and related work can also be found in refs .
@xcite . for all orders corrections to topology @xmath69
we thus find the familiar result : @xmath78 the non - abelian color factors of diagrams with an `` abelian '' topology are canceled by the contributions of the purely non - abelian topologies . in this context
it is worth noting that for the three gluon vertex coupling , there is an effective factor of @xmath79 stemming from two contributions in numerator terms that are needed to obtain the required number of logarithms .
this is only the case if one has a denominator of the type @xmath80 . for one gluon very soft ,
there is a term @xmath81 and one @xmath82 , @xmath83 being the other ( relatively ) soft gluon momentum and @xmath84 an external hard fermion momentum .
the lorentz structure of the three gluon vertex is such that only these two terms give dl - contributions ( the third connects the same fermion line ) .
the effective soft contribution is therefore @xmath85 , compared with a factor @xmath86 from the emission of a soft gluon off a fermion line . the interesting feature of the higher order corrections to the one - loop diagrams of fig . [ fig:1l ]
is that there are novel purely hard `` non - sudakov '' logarithms due to the presence of the `` soft '' fermion line in topologies @xmath71 and @xmath72 . in the case where a gluon insertion does not end on an external ( hard ) fermion line ,
the soft fermion line serves essentially as a cutoff for gluons that are soft compared to the hard blob .
the diagrams are listed in figs .
[ fig : b3l ] and [ fig : c3l ] and denoted by @xmath87 , as they contain no soft sudakov contributions .
besides a different color factor , both contributions are identical and lead to the same structure as in the soft case , except for the last two integrations .
they have the additional effect of eliminating @xmath88 terms , which are very important in the soft case ( see fig . [
fig : spsp ] ) .
the reason is simply that they are automatically fulfilled since the soft fermion line induces a @xmath89 cutoff for the soft gluon sudakov variables .
the integrations over the gluonic variables can thus be performed easily and summing over @xmath90 gluon insertions we have : @xmath91 using partial integration @xmath90-times ( differentiating @xmath92 , integrating @xmath93 ) , we find : @xmath94 ^ 1_\frac{m^2}{s } \nonumber \\ & = & - \sum^\infty_{i=0 } \left ( \frac{\alpha_s}{2\pi } \right)^{i+1 } \frac{(-1)^i \ ; \gamma ( i+1)}{\gamma ( 2i+3 ) } \log^{2i+2 } \frac{m^2}{s } \nonumber \\ & = & - \frac{\alpha_s}{4\pi } \log^2 \frac{m^2}{s } \ ; _ 2f_2 ( 1,1;2,\frac{3}{2 } ; -\frac{\alpha_s}{8\pi } \log^2 \frac{m^2}{s } ) \label{eq : hff}\end{aligned}\ ] ] in order to compare this result with the known one- and two - loop leading logarithmic results we list them below explicitly and find agreement with refs .
@xcite : @xmath95 the asymptotic behavior of the hypergeometric function @xmath96 for @xmath97 can be obtained by using an integral representation @xcite and taking the leading pole term . in our case
we find @xmath98 the last line follows from the residue of the integral representation at @xmath99 , closing the contour in the negative @xmath100-plane .
all higher negative pole contributions are suppressed by at least a factor of @xmath101 . with the @xmath102 suppression contained in the born amplitude ( [ eq : ppqqb ] ) , the high energy limit is well behaved .
it is also possible to reconstruct the asymptotic series based on the above integral representation .
the higher order poles are simple ones and can thus be evaluated easily .
we find @xmath103 although it possesses the ominous factorial growth of the coefficients in the series for the higher order residue contributions , the equivalent original expansion in terms of the confluent hypergeometric function in eq .
[ eq : hff ] means that it is borel summable and that the high energy limit in eq .
[ eq : ahff ] does indeed give the correct behavior for @xmath104 .
[ fig : hff ] contains a comparison of the exact hard form factor of eq .
[ eq : hff ] and its asymptotic logarithmic divergence given by the first term in eq .
[ eq : ihff ] .
they can be seen to agree in the high energy limit . in addition
the figure demonstrates that at a fixed order in perturbation theory , the result alternates between @xmath105 for high energies .
the numerical effect of the higher order terms is shown in fig .
[ fig : rel ] where the all orders hard form factor is compared to the one- and two - loop corrections at realistic future collider energies .
as explained in the figure , only if the experimental accuracy reaches the percentile level are higher than second order contributions important .
the relative size of the all orders correction is larger for @xmath106 than for the b - quark mass . a more detailed comparison of the behavior of the higher order residues is given in fig .
[ fig : asres ] .
the first four pole contributions are plotted against the exact hard form factor and agree with eq .
[ eq : hff ] in the asymptotic regime . at lower energies the @xmath107 behavior of the first three simple residues of eq .
[ eq : ihff ] is clearly visible . in order to check this derivation explicitly at the three loop level
, we have calculated the corresponding diagrams denoted by @xmath43 in figs .
[ fig : b3l ] and [ fig : c3l ] and find the following results ( modulo color factors and chosing for the sudakov decomposition @xmath108 for topology @xmath71 and @xmath109 for topology @xmath72 ) : @xmath110 the remaining terms @xmath111 and @xmath112 in each figure cancel the non - abelian color factor stemming from the crossed box insertion @xmath113 , i.e. @xmath114 .
it can easily be seen from the explicit form in eqs .
[ eq:3lh1 ] and [ eq:3lh2 ] that the contributions of the gluon insertions are cut off effectively by the fermion line sudakov variables for which @xmath89 is the minimum value attainable .
[ fig:3lh ] contains a comparison of the explicit three loop results given by @xmath115 and the third term in the series [ eq : hff ] .
the agreement is below the percentile level . for completeness
, we give the form factors explicitly for the two topologies @xmath71 and @xmath72 below , including the coupling and their differing color structure : @xmath116 where @xmath117 is the one - loop hard form factor given in eq .
[ eq:1lhff ]
. while the novel hard logarithms of the previous section are the most interesting new feature of the @xmath2 @xmath118 process , technically the mixed hard and soft contributions are a considerable challenge .
in fact for topology @xmath72 we are only able to show numerically through three loops that the same structure emerges as for topology @xmath71 .
we therefore start with the diagrams given in fig .
[ fig : b3l ] for topology @xmath71 which contain a soft ( s ) contribution . the diagram @xmath119 is evidently given by the product of the two - loop purely hard and the one - loop sudakov form factors .
also for the sum of diagrams @xmath120 , it can easily be seen that we again have a factorization , this time the purely hard one - loop times the two - loop sudakov form factor .
the nice feature of this topology is that corrections connecting the original one - loop hard form factor with gluon insertions between the outgoing fermions vanish in pairs .
the reason is simply a relative spinor line minus sign when a gluon starting from a fermion line of the original one - loop topology is connected with one of the final state fermions .
one always has in this case an analogous correction which ends on the anti - fermion line , thus yielding a relative minus sign .
[ fig : b3lc ] indicates the cancellation in pairs . in this particular case ,
for instance , we find ( chosing the sudakov parametrization @xmath121 for @xmath122 and @xmath123 for @xmath124 with respect to the gluon going across the hard blob ) :
@xmath125 the @xmath124 contribution is minus that of @xmath122 after the replacement @xmath126 has been made .
the two diagrams therefore cancel each other .
this feature holds to all orders and thus leads to a product of the hard form factor in eq .
[ eq : bhff ] and the sudakov form factor in eq .
[ eq : softff ] . in other words
we have @xmath127 \label{eq : bhsff}\ ] ] where we have again used the notation introduced in section [ sec:1l ] for the one - loop form factors .
@xmath128 refers to the mixed hard ( h ) and soft ( s ) nature of the form factor .
for the diagrams given in figs .
[ fig : c3l ] and [ fig : c3lsna ] there is no obvious way of generalizing a cancellation between diagrams to all orders .
we therefore chose to calculate these corrections explicitly through three loops and compare with the `` expected '' analogous result found for topology @xmath71 in eq .
[ eq : bhsff ] .
the results are given in appendix [ sec : c3l ] .
[ fig : c3lhhs ] shows a comparison of the contribution from the second term of the hard form factor from eq .
[ eq : chff ] multiplied by the one - loop soft form factor of eq .
[ eq : sff ] with the three loop corrections given by @xmath129 in fig .
[ fig : c3l ] . only the diagrams @xmath130 actually contribute , the others are either zero ( @xmath131 and @xmath132 ) or cancel in the sum ( @xmath133 ) in the dl - approximation .
the leading logarithmic behavior is shown to agree for both @xmath134 as well as for @xmath135 , keeping @xmath136 fixed .
the agreement is better than @xmath137 and improves for large values of the logarithms . for the other mixed hard and soft contribution at the three loop level , namely the one - loop hard form factor in eq .
[ eq:1lhff ] multiplied by the second term in the sudakov exponential of eq . [ eq : softff ] , many more diagrams contribute .
the remaining soft diagrams of figs . [ fig : c3l ] and [ fig : c3lsna ] are not only more numerous but also contain technical subtleties worth mentioning . in particular , diagrams containing a soft gluon radiating off the `` soft '' fermion line facing the hard blob
need careful consideration . in those cases
one has two ( for diagrams @xmath138 and @xmath139 of fig .
[ fig : c3l ] even three ) regions in the phase space contributing to the dl - corrections , corresponding to either side of the `` soft '' fermion line going on shell . for instance
the two loop diagram in fig .
[ fig:2l ] possesses this ambiguity .
it actually does not give a contribution in the dl - approximation due to an asymmetry of the diagram with respect to the external momenta @xmath140 , while the large variables ( see eq .
[ eq : dlapprox ] ) do not differ in the dl - approximation if we make the replacement @xmath141 .
the three loop diagrams @xmath131 and @xmath132 in fig .
[ fig : c3l ] vanishes for the same reason .
the easiest way to see this on a technical level is therefore to parametrize @xmath142 choosing @xmath140 for the sudakov decomposition . expanding around the left side of the `` soft '' fermion line in fig .
[ fig:2l ] we find : @xmath143\end{aligned}\ ] ] shifting the fermion loop momentum , effectively @xmath144 , and expanding around the right side of the gluon insertion gives analogously : @xmath145\end{aligned}\ ] ] where we have omitted the common color factor .
the second contribution is evidently ( @xmath146 ) the first after the replacements @xmath147 and @xmath126 . for the soft contributions of the three loop diagrams @xmath119
, @xmath148 , @xmath149 , @xmath150 , @xmath138 and @xmath139 where the antisymmetry is removed by the additional gluon , we find it technically more convenient to choose @xmath151 for the sudakov decomposition for the genuine dl - corrections listed in appendix [ sec : c3l ] . a careful treatment should in any case give the same result , however .
[ fig : c3lhss ] shows that the three loop contribution of the sum of diagrams @xmath152 is indeed equal to the product of the one - loop hard form factor in eq .
[ eq:1lhff ] multiplied by the second term in the sudakov exponential of eq .
[ eq : softff ] .
the leading logarithmic behavior is shown to agree for both @xmath134 as well as for @xmath135 , keeping @xmath136 fixed .
the relative difference actually decreases with increasing values of the large logarithms .
it is therefore more than suggestive that the structure which prevails for the one- , two- and three - loop dl - contributions exponentiates to all orders analogously to that of topology @xmath71 in eq .
[ eq : bhsff ] , namely that @xmath153 \label{eq : chsff}\ ] ]
in this paper we have derived a novel series of hard double logarithms in polarized @xmath154 collisions to all orders .
the resulting series is given by a confluent hypergeometric function @xmath96 which possesses a @xmath155 high energy limit .
taking into account the @xmath102 suppression contained in the born amplitude , the new corrections are well behaved as @xmath156 .
the series expansion agrees with the known one and two loop results as well as our explicit three loop calculation .
it was furthermore shown that at three loops all additional doubly logarithmic corrections are described by the exponentiation of sudakov logarithms at each order in the expansion of the new hard form factor .
this behavior is already present at one and two loops and can thus safely be extrapolated to all orders .
in summary we list the complete virtual dl - form factor contribution to the process @xmath157 : @xmath158 \nonumber \\ & & \left . \;\;\;\;\;\;\;\;\;\;\;\;\;\ ; + 2 \ ; { \cal f } \;\ ;
_ 2f_2 ( 1,1;2,\frac{3}{2 } ; \frac{c_a}{4 c_f } { \cal f } ) \ ; \left [ \ ; \exp \left ( { \cal f}_{\cal a } \right ) -1 \right ] \right\ } \\ & = & { \cal m}_{born } \left\ { 1 + { \cal f } \;\ ; _ 2f_2 ( 1,1;2,\frac{3}{2 } ; \frac{1}{2 } { \cal f } ) + 2 \ ; { \cal f } \;\ ; _ 2f_2 ( 1,1;2,\frac{3}{2 } ; \frac{c_a}{4 c_f } { \cal f } ) \right\ } \exp \left ( { \cal f}_{\cal a } \right ) \nonumber \label{eq : dl}\end{aligned}\ ] ] where @xmath159 and @xmath117 denote the soft and hard one - loop form factors of eqs .
[ eq : sff ] and [ eq:1lhff ] respectively .
it should be pointed out that the contributions containing the soft form factor @xmath159 cancel in a physical cross section with the infrared divergent contributions from the emission of real soft gluons .
for realistic future collider applications fig .
[ fig : rel ] demonstrated that the higher than second order contributions are important if an accuracy in the percentile regime could be reached .
a more thorough study of the phenomenological implications of this work including realistic experimental cuts is in progress @xcite .
* acknowledgements * + the authors would like to thank m. wsthoff for valuable discussions and m. heyssler for helpful advice on using `` paw '' .
this work was supported in part by the eu fourth framework programme ` training and mobility of researchers ' , network ` quantum chromodynamics and the deep structure of elementary particles ' , contract fmrx - ct98 - 0194 ( dg 12 - miht ) .
in this appendix we give explicit three loop results for the @xmath68 terms of eq .
[ eq : nsudbc ] of topology @xmath72 .
the particular contributions are listed in fig .
[ fig : c3l ] with the corresponding notation .
for the three loop contribution of topology @xmath71 in fig .
[ fig : b3l ] the left and right side of the hard blob factorize for the graphs @xmath160 . on a technical level
they are therefore effectively two loop corrections .
@xcite for explicit results . in the following we distinguish between corrections with one and those with two purely hard form factors .
we also omit color factors at this point , as the exponentiation of the soft gluons ensures that all `` higher '' order @xmath161 contributions are canceled by purely non - abelian corrections @xcite .
the correct color factors of the sum of the graphs given in figs .
[ fig : c3l ] , [ fig : c3lsna ] and [ fig : c3lz ] are contained in the results of section [ sec : res ] .
the following results for the @xmath68 correspond to the respective graphs of fig .
[ fig : c3l ] with two potentially soft gluons .
the sudakov parametrization for the three soft momenta ( as indicated in the figure for @xmath162 ) was chosen to be @xmath163 for the soft fermion line @xmath164 in all graphs , @xmath165 for @xmath142 in @xmath166 and for @xmath167 in @xmath168 and @xmath169 for @xmath142 in @xmath170 and for @xmath167 in @xmath171 .
below we give the results for the @xmath68 corresponding to the respective graphs of fig .
[ fig : c3l ] with one potentially soft gluon .
the sudakov parametrization for the three soft momenta ( as indicated in the figure for @xmath162 ) was @xmath163 for the soft fermion line @xmath164 and for @xmath142 in all graphs and for @xmath167 in @xmath173 , @xmath165 for @xmath167 in @xmath174 and @xmath169 for @xmath167 in @xmath175 .
strictly speaking , the above cancellation takes place only for the @xmath179 terms of amplitude @xmath180 . the @xmath181 term , absent in graphs @xmath182 and @xmath183 of fig . [ fig : c3l ] ,
is canceled by the analogous contribution of graphs @xmath184 and @xmath185 of fig .
[ fig : c3lsna ] . in the same way the above mentioned cancellation of `` higher order '' @xmath161 terms takes place with all non - abelian graphs of fig .
[ fig : c3lsna ] . the remaining diagrams of topology @xmath72 in fig .
[ fig : c3lz ] all cancel among themselves in groups of two or three . | the measurements of the @xmath0 and @xmath1 partial decay widths of an intermediate - mass standard model higgs boson are among the most important goals of a future photon linear collider . while in an initially polarized @xmath2 state
the background process @xmath3 is suppressed by @xmath4 , radiative corrections remove this suppression and are known to be very sizable at the one - loop level . in particular a new type of purely hard double logarithmic ( dl ) correction can even make the cross section negative at this order in perturbation theory .
the second order term of the series is also known and enters with a positive sign at the two loop level . from a theoretical as well as
a practical point of view it is clearly desirable to resum this series and to know its high energy behavior . in this paper , we derive the series of these novel `` non - sudakov '' logarithms to all orders and calculate the high energy limit analytically .
we also give explicit three loop corrections in the dl - approximation as a check of our results and show that the three loop structure also reveals the higher order behavior of mixed hard and soft double logarithms .
we are thus able to resum the virtual dl - corrections to @xmath5 to all orders .
hep - ph/9807332 + dtp/98/42 + july 1998 + * all - orders resummation of leading logarithmic + contributions to heavy quark production + in polarized @xmath0 collisions * + + |
the concept of geometric phase ( gp ) in quantum system was originally introduced by berry @xcite when he studied the dynamics of a closed quantum system which undergoes an adiabatic cyclic evolution .
he found that besides the usual dynamical phase , the system also acquires an additional phase which only depends on the geometry of the path traversed by the system during its adiabatic evolution .
since then this important notion has attracted much attention @xcite .
it has been generalized in various aspects , e.g. the gp for non - adiabatic evolution @xcite and for noncyclic evolution @xcite .
the gp has been observed experimentally in optical @xcite , nmr @xcite , and superconducting electronic circuit experiments @xcite .
recently , the renewed interest in the investigation of gp comes from the application of gp to implement the logic gates in quantum computation @xcite .
the purely geometric nature of the phase makes such computation intrinsically fault - tolerant and robust against certain types of classical fluctuation noise @xcite .
however , any realistic quantum system is inevitably coupled to its surrounding environment , which would result in the loss of quantum coherence ( i.e. the decoherence ) of the quantum system itself and hence limit the implementation of the geometric quantum computation .
therefore , the study of the gp in open quantum systems becomes an important issue .
for the gp of mixed states in open systems , uhlmann was the first to make the attempt to define the mixed - state gp via state purification @xcite .
proposed an alternative definition for the nondegenerate mixed - state density matrix under unitary evolution based on the interferometry @xcite .
this definition was further generalized to degenerate mixed state by singh _
_ @xcite and to the nonunitary evolution by tong _
_ @xcite using the kinematic approach .
the gp associated with such a nonunitary evolution is defined as @xcite @xmath0 where @xmath1 and @xmath2 , respectively , are the eigenvalues and the eigenstates of the reduced density matrix of the quantum system , and @xmath3 is the time after the system completes a cyclic evolution when it is isolated from the environment .
taking the environment into account , the system no longer undergoes a cyclic evolution . here a quasicyclic process with @xmath4 , where @xmath5 is the frequency of the system , is considered in eq .
( [ gpt ] ) .
defined a mixed - state gp in the context of pantcharatnam formula @xcite via mapping the density matrix to a nonunit vector ray in complex projective hilbert space @xcite .
the mixed - state gp has been observed in nmr system @xcite . because the environment induced decoherence would affect the performance of the quantum computation using gp , the study of environment effects on the gp is highly desired .
many works along this line have been performed within markovian approximation @xcite , which is valid only when the interaction between the system and the environment is very weak and the environmental correlation time is very small . however , in many quantum information experiments , these conditions are not completely satisfied . for example , in cavity qed experiment the imperfection of the cavity mirrors makes the cavity field having a lorentzian spectrum expansion , which acts as an environment , would exert a strong non - markovian effect on the atom in it @xcite .
there are also some works on the non - markovian effect on the gp in dephasing environments @xcite , where there is no energy / information exchange between the system and its environment . in this work ,
we extend the study of the gp of open two - level system to the situation where there has an energy / information exchange between the system and its environment .
the environment is at zero - temperature and has a lorentzian spectral density , which corresponds to the radiation field as an environment being confined in a leaky cavity .
we mainly concentrate on how the non - markvoain effect affects the gp in different parameter regimes of the spectral density , and how the gp manifests its robustness against the decoherence .
this paper is organized as follows . in sec .
[ mode ] , we introduce the model of a qubit interacting with a vacuum environment and analyze its decoherence behavior . in sec .
[ geometric ] we evaluate the gp of the qubit via performing analytical and numerical calculations .
the correction effect exerted by the environment on the gp in different parameter regimes of the lorentzian spectral density is analyzed .
finally , a brief discussion and summary are given in sec .
let us consider a system consisted of a two - level atom ( qubit ) coupled to a radiation field at zero temperature as an environment .
the hamiltonian of the system is scully97 @xmath6 where @xmath7 and @xmath8 are the inversion operators and transition frequency of the qubit , @xmath9 and @xmath10 are the creation and annihilation operators of the @xmath11-th mode with frequency @xmath12 of the radiation field , and @xmath13 is the coupling strength between the qubit and reservoir . throughout this paper
we assume @xmath14 .
the system models the decoherence process of the atom via the amplitude decaying under the born - markovian approximation , which results in the spontaneous emission of the two - level atom @xcite in quantum optics .
the model is exactly solvable .
the decoherence dynamics of the qubit is governed by the master equation @xcite @xmath15+\gamma ( t)[2\sigma _ { -}\rho ( t)\sigma _ { + } \notag \\ & & -\sigma _ { + } \sigma _ { -}\rho ( t)-\rho ( t)\sigma _ { + } \sigma _ { - } ] , \label{mstt}\end{aligned}\]]where the time - dependent parameters are given by @xmath16,~\gamma ( t)=-\text{re}[\frac{% \dot{c}(t)}{c(t ) } ] .
\label{coe}\]]it is shown that @xmath17 satisfies @xmath18 where @xmath19 is the environmental correlation function with the spectral density defined as @xmath20 and the initial condition @xmath21 .
the time - dependent parameters @xmath22 and @xmath23 play the roles of lamb shifted frequency and decay rate of the qubit , respectively .
the integro - differential equation ( [ c0 ] ) contains the memory effect of the reservoir registered in the time - nonlocal kernel function and thus the dynamics of qubit displays non - markovian effect .
if the time - nonlocal kernel function is replaced by a time - local one , then eq . ( [ mstt ] ) recovers the conventional master equation under markovian approximation .
it is obvious that the memory effect registered in the kernel function @xmath24 is essentially determined by the spectral density @xmath25 . in this work
we explicitly consider that the spectral density has a lorentzian form @xcite @xmath26 where @xmath27 is the coupling constant between the qubit and the environment , and @xmath28 defines the spectral width of the coupling at the resonance point @xmath8 .
the lorentzian spectral density describes that the vacuum radiation field as the environment is confined in a leaky cavity . due to the leakage of the cavity field induced by the imperfection of the cavity mirrors
, the spectrum of the cavity field displays a broadening at the resonance point regarding the atomic transition frequency @xmath5 . in this case
one can verify that the correlation function decays exponentially @xmath29 @xcite , which means that the parameter @xmath28 characterizes the correlation time of the environment as @xmath30 .
if @xmath31 is comparable with the typical time scale of the system , i.e. @xmath32 , then the memory effect of the environment should not be neglected and the decoherence dynamics in this situation is non - markovian . while @xmath33 , the memory effect of the environment is negligible and the decoherence dynamics is markovian . in the ideal cavity limit @xmath34
, we have @xmath35 , which corresponds to a constant kernel @xmath36 .
then the system reduces to the jaynes - cummings model with a vacuum rabi frequency @xmath37 . going back to the general case
, one can obtain the analytical form of @xmath17 by substituting the exponentially decaying @xmath38 into eq .
( [ c0 ] ) as @xmath39,\]]where @xmath40 .
so the parameters in the master equation can be calculated readily @xmath41 one can see from eqs .
( [ gf ] ) when @xmath28 is much larger than other frequency scale , the decay rate tends to a constant value @xmath42 , which just characterizes the decoherence behavior of the qubit under the markovian dynamics .
in the following we compute explicitly the gp of the qubit .
we assume that the initial state of the qubit is chosen as @xmath43 where @xmath44 and @xmath45 are the excited and ground states of the qubit , respectively .
this state corresponds to a vector in bloch sphere with polar angle @xmath46 .
the time - dependent reduced density matrix of the qubit under the initial condition ( [ istate ] ) can be obtained straightforwardly from the master equation ( [ mstt ] ) @xmath47 it is noted that besides the off - diagonal elements , the diagonal elements of the reduced density matrix @xmath48 also change with time in our model .
it is just this time - dependence of the diagonal elements of @xmath48 characterizing the energy exchange between the qubit and its environment that makes our system shows dramatic difference to the dephasing model @xcite . to calculate the gp of the qubit
, we must firstly get the eigensolution of the reduced density matrix ( [ rdm ] ) .
the eigenvalues of the above reduced density are readily calculated , @xmath49 , \label{eigenv}\ ] ] it is obvious that the eigenvalue @xmath50 , which , from eq . ( gpt )
, means that the component of the eigenstate corresponding to the eigenvalue @xmath51 gives no contribution to the gp .
thus we only need to consider the eigenstate corresponding to the eigenvalue @xmath52
@xmath53where@xmath54 below we calculate the gp .
( [ gf ] ) shows that the frequency shift of the qubit is zero for the lorentzian spectral density @xcite
. so the period of the environment disturbed atom is the same as the one for a bare atom .
then the gp of the qubit acquired after a period @xmath55 can be calculated as @xmath56 here we point out that the expression of the gp ( [ gp2 ] ) has been obtained by using the kinematic approach @xcite .
however , it can be shown that the expression is exactly the same as the one obtained by the definition of pantcharatnam s phase of a nonunit vector ray @xcite , which exhibits some essential features of the gp in such systems . from eq .
( [ gp2 ] ) , one can notes that the effect of the environment enters into the gp solely via the time - dependent factor @xmath57 of the excited - state population .
when the environment is absent , then @xmath58 and the gp reduces to @xmath59 , which is just the gp acquired by a two - level atom in the unitary dynamics . depending on the decoherence dynamics being markovian or non - markovian , the short - time dynamics of @xmath57 shows remarkably different behaviors . in the markovian dynamics
, @xmath57 decays monotonically and finally approaches zero . in this case , the larger the decay rate @xmath60 is , the larger the correction of the gp should be . in
the non - markovian dynamics , contributed from the memory effect of the environment @xmath57 shows transient oscillation with time , which naturally induces a correction of the gp different to the one in markovian dynamics , as shown in the following . up to the first order of coupling strength @xmath61 , i.e. the weak coupling limit , we have@xmath62 where @xmath63 $ ] . besides the leading term @xmath64 corresponding to the well - known gp acquired under the unitary dynamics , the second term , which is quadratic in @xmath65 with a @xmath66-dependent coefficient @xmath67 , is the lowest - order correction to the gp induced by non - unitary dynamics due to the interaction with the environment .
it is easy to check that @xmath67 is a monotonically decreasing function with the increase of @xmath66 , so we can expect that the gp shows a larger deviation to @xmath64 for a small @xmath66 than a large one . in particular , in the ideal cavity limit @xmath34 , the function @xmath67 arrives at its maximum @xmath68 , where the gp has a largest correction in this weak coupling ( or small @xmath27 ) regime . on the other hand ,
when @xmath69 , one can verify @xmath70 .
consequently , the gp in this markovian limit is @xmath71 which shows very similar form to the result of @xcite obtained from the born - markovian master equation . in the following
we give the exact gp via numerically evaluating the integration in eq .
( [ gp2 ] ) . in fig .
[ sle ] we plot the exact gp as a function of the coupling constant and the initial polar angle . when @xmath72 , the correlation time of the environment @xmath73 is much less than the typical time scale of the system @xmath74 .
in this situation , the memory effect of the environment is negligible and the dynamics of the system is markovian .
consequently , the energy and the information flow single - directionally from the system to the environment , which results in the dissipation to the system . the dissipation time scale in this markovian dynamics is characterized by @xmath75 .
we can see in fig .
[ sle](a ) that the gp in this situation decreases monotonically with the increase of the coupling constant .
this is understandable that a larger coupling constant would induce a stronger dissipation and a short decoherence time scale @xmath76 to the qubit system .
with the increase of @xmath27 , @xmath76 approaches @xmath3 , which means that the dissipation becomes more and more notable within the time scale @xmath3 . on the other hand , when @xmath27 is small the gp shows small deviation to the unitary one @xmath64 .
in this situation , @xmath76 is larger than @xmath3 , making the dissipation negligible within the time scale @xmath3 during which the gp is accumulated .
this interesting phenomenon just reflects the resilient ability of the gp to the environment s disturbance , especially in weak coupling limit .
a similar phenomenon is also observed in the dephasing model @xcite .
when @xmath77 , @xmath78 and the non - markovian effect induced by the memory effect of the environment to the system would show its distinct impact on the the dynamics of the system .
besides the dissipation , the environment also exerts a dynamical backaction on the system @xcite . the backaction effect is reflected in that the energy and the information flow back and forth between the system and the environment .
in this situation we can see in fig .
[ sle](b ) that the gp oscillates with the increase of @xmath27 and then tends to a definite value , which is qualitatively different to the above markovian situation .
the oscillation is just the manifestation of the dynamical backaction .
it is very clear that the non - zero character of the gp in the case of large @xmath27 is due to the counteracted competition between the backaction and dissipation exerted by the non - markovian environment partially , which weakens the decoherence of the qubit system . a more clear comparison of the gp with different @xmath66 as a function of @xmath27 when @xmath79 is shown in fig .
[ td ] . as comparisons ,
the perturbed results for @xmath80 and @xmath81 are also presented in this figure and they agree well with the corresponding exact one only in the weak coupling limit .
we can see obviously that in the weak coupling regime the environment with a large @xmath66 induces a smaller correction to the gp than the one with a small @xmath66 .
that is , the non - markovian effect has a strong correction to the gp in this weak coupling regime .
this is consistent with the result obtained in ref .
@xcite , where a phenomenological analysis of the non - markovian effect on the gp is performed .
while in the strong coupling regime the situation is opposite , because the system with a large-@xmath66 shows stronger decoherence than the one with a small @xmath66 where the non - markovian effect has a strong correction to the gp . even till
today it is still difficult to access experimentally the strong coupling regime in the cavity qed platform .
for example , in a recent microtoroidal resonator experiment @xcite , the achieved coupling strength between the atom and the cavity field is 90mhz , while the damping rate of the cavity field is about 180mhz .
such a large damping rate would induce a noticeable spectrum expansion of the cavity mode .
it is of course interesting to examine the gp in this bad - cavity and weak coupling regime . in fig .
[ ws ] we plot the gp as a function of the initial polar angle in the weak coupling regime for different spectrum width @xmath66 .
one can see that the gp approach more and more closely the unitary gp @xmath64 with the increase of @xmath66 .
this accounts for once again that the gp shows more strong resilient ability to the noise of the markovian environment with a large @xmath66 .
since this resilient character is obtained in weak coupling and bad - cavity condition , we argue that it is accessible by modern cavity qed experiment @xcite .
we have studied exactly the gp for a qubit in an amplitude decaying environment using the kinematic approach .
we have evaluated the non - markovian effect on the gp .
it is demonstrated that the non - markovian environment has a more significant correction to the gp in the weak coupling regime due to the strong short - time correlation between the qubit and the environment .
interestingly , our result also indicates the insensitivity of the gp to the lorentzian environment in the markovian regime when the coupling is weak .
it just manifests the resilient ability of the gp to the environmental noise .
this results elucidate that the gp in this cavity qed system is fault - tolerant not only against the classical noise induced by the parameter fluctuation @xcite , but also against the quantum noise .
this has significant meaning in geometric quantum information processing .
our model , as a basic model of quantum optics , is particularly relevant to the cavity qed experiments . in this respect
some remarkable experiments , such as the efficient coupling of the trapped atoms with cavity field @xcite , have been performed successfully . practically , any phase variation is observed only via some kind of interferometry between the involved state and certain selected reference state .
for example , the gp for mixed state has been observed via designing a quantum network in nmr system to realize the interferometry @xcite .
this provides a clue to observe gp in cavity qed system .
although the gp has not been observed in the cavity qed system , a quantum network using the recent developed microtoroidal resonator @xcite has been proposed @xcite based on the input - output process of photons @xcite .
if an effective interferometry could be realized in this quantum network , then the gp would be expected to be observable in cavity qed system .
our work on the assessment of the environmental effect on the gp , especially in non - markovian regime , is of great importance in using the gp in cavity qed system to implement the quantum gates .
this work is supported by nsf of china under grant no .
10604025 , gansu provincial nsf of china under grant no .
0803rjza095 , the program for ncet , and nus research grant no .
r-144 - 000 - 189 - 305 . | we study the geometric phase of a two - level atom coupled to an environment with lorentzian spectral density . the non - markovian effect on the geometric phase
is explored analytically and numerically . in the weak coupling limit the lowest - order correction to the geometric phase
is derived analytically and the general case is calculated numerically .
it is found that the correction to the geometric phase is significantly large if the spectral width is small and in this case the non - markovian dynamics has a significant impact to the geometric phase .
when the spectral width increases , the correction to the geometric phase becomes negligible , which shows the robustness of the geometric phase to the environmental white noises .
the result is significant to the quantum information processing based on the geometric phase . |
nanomagnets offer the ultimate thermodynamic limits of computation , as expressed by landauer s principle @xcite .
therefore , spintronics technology that uses electron spin for information processing and communication presents a favorable option to implement future low - power devices .
most compelling proposals of spintronic devices rely on either the spin - transfer torque ( stt ) or the dipolar coupling effects between nanomagnets to accomplish switching of the logic device @xcite .
the analysis presented in this paper considers nanomagnet reversal through stt effect @xcite .
a major challenge for spintronics logic , in general , is the inherent stochasticity in nanomagnet dynamics resulting from ( i ) the presence of thermal noise and ( ii ) the existence of two basins of attraction that make the dynamic evolution of magnetization extremely sensitive to its initial state @xcite .
there is a two - fold complexity associated with spintronic circuit design due to the stochastic nanomagnet behavior as discussed in our prior work @xcite .
first , the circuit delay exhibits large variability that can increase the effective - delay of the circuit depending on the desired error tolerance .
second , for interconnected logic networks , the delay distributions of individual nanomagnets combine non - linearly increasing the circuit delay complexity significantly @xcite . due to the complex nature of the equation governing the behavior of the nanomagnet body ,
complete analytical descriptions for the delay distributions of a nanomagnet are unavailable .
previous work has suggested that the effect of thermal noise may be approximated by knowing the initial angle of the nanomagnet and neglecting the thermal noise during reversal @xcite . while these models do provide significant insight into the nature of nanomagnet reversal , numerical simulations presented in this paper show that these analytical distributions are only accurate for rapid reversal times ( @xmath0 200 ps ) .
therefore , for the case when the nanomagnet is under influence of spin currents comparable in magnitude to the critical spin current for reversal , new types of distribution functions to describe the magnetization dynamics must be sought . in this paper
, we demonstrate the applicability of the frchet distribution that contains double exponentials to more accurately capture the evolution of magnetization over a very broad range of time scales .
the frchet distribution is also compatible with the results obtained in @xcite by analytically solving the landau - lifshitz - gilbert ( llg ) equation with specific boundary conditions .
once the nanomagnet delay is characterized using the frchet distribution function , we evaluate the circuit - level performance of series- and parallel - connected network of nanomagnets .
we show that the delay variation is larger when nanomagnets are evaluated in parallel , while a series - connected network yields a much higher average circuit delay .
the increase in average delay for series - connected nanomagnets is far greater than any increase in delay one may have to tolerate due to increased delay variation .
therefore , for delay - critical paths , we propose to design spintronic circuits with a larger fan - out and lower logic depth .
we use the pdf of the nanomagnet delay to analyze the performance of an asl - and gate . in an asl device ,
an electric current flowing through a transmitting nanomagnet is used to generate a spin current in a non - magnetic channel . upon reaching the receiving nanomagnet
, the spin current imparts a torque to the nanomagnet .
the operation of the asl device is pictorially represented in fig .
[ fig : asl_op ] . majority boolean logic gates , such as and/or ,
are implemented by connecting multiple input nanomagnets to a receiving nanomagnet , where a sum of the input spin currents determines the functionality of the logic gate @xcite . due to the summing nature of asl , the magnitude of the spin current delivered to the output nanomagnet will vary depending on the input values and will alter the delay distributions of the output nanomagnets .
the delay distributions presented in this work incorporate the impact of input signal pattern on the overall circuit delay .
the asl device consists of two main components : nanomagnets to store data and transmit logical values and spin channels to allow the transmission of spin current . in this section ,
we discuss the physical model and the simulation parameters of the nanomagnet .
the model for the channel in the asl device and considerations of dipolar coupling field are presented in the appendices . the phenomenological equation describing the physics of a nanomagnet under the effects of stt is the llg equation @xcite , which is stated as @xmath1 where @xmath2 is the unit - vector along the direction of the magnetization , @xmath3 is the gyromagnetic ratio , @xmath4 is the free space permeability , @xmath5 is the gilbert damping coefficient , @xmath6 is the electron charge , @xmath7 is the saturation magnetization , and @xmath8 is the spin current perpendicular to the magnetization .
it is important to note that ordinarily in magnetic - tunnel - junction ( mtj ) structures , a field - like ( f - l ) torque must be considered @xcite .
however , as mentioned in appendix a , only metallic channels are considered in this work ; therefore , the f - l torque ids negligible and is not considered in this work . for alternate asl structures where this f - l torque is significant
, the delay distribution must be reevaluated . however , the idealized situation considered in this manuscript is a useful introduction in obtaining the delay distributions of complex spintronic circuits .
an @xmath9 is chosen for the simulations in this paper .
vectors denoted with a ( @xmath10 ) are unit vectors and have a constant magnitude of unity , while vectors denoted with an ( @xmath11 ) have variable magnitudes .
@xmath12 is the number of spins in the magnet and is given as @xmath13 where @xmath14 is the volume of the magnet . in ( [ eq : llg ] ) ,
@xmath15 is the effective field on the nanomagnet and is given as @xmath16 the uniaxial anisotropy field , @xmath17 is given from the stoner - wahlforth model as @xcite @xmath18 where @xmath19 is the uniaxial anisotropy energy density .
the focus of this work is on thin - film pma nanomagnets ; hence , the anistropy field is assumed to be in the @xmath20 direction .
@xmath21 is the field due to the shape anisotropy ( demagnetization ) and is given as @xmath22 where @xmath23 , @xmath24 , and @xmath25 are the demagnetization factors determined by the shape of the nanomagnet @xcite .
for a thin - film nanomagnet , such as the one assumed in this paper , the demagnetization field is largely in the @xmath20 direction ; therefore , we assume @xmath26 , and @xmath27 .
the thermal noise manifests itself as fluctuations in the internal anisotropy field and is added to the internal field of the magnet through the term @xmath28 in ( [ eq : field_sum ] ) @xcite .
according to the theoretical formulation presented in @xcite , the thermal field can be defined as an isotropic vector process @xcite .
the thermal field can be modeled as a three - dimensional wiener process @xcite : @xmath29 this result is equivalent to the expressions derived using the fokker - planck analysis @xcite .
the llg equations are numerically simulated using the heun numerical method implemented in custom cuda simulations @xcite .
the accuracy of the simulator was verified by comparing results of the simulations against known analytical solutions for the sllg shown in section iii .
we define the nanomagnet reversal delay as the difference in time it takes for the nanomagnet to cross the @xmath30 plane from the time the spin current was initially applied to the nanomagnet . for all simulations in this paper ,
nanomagnet dimensions are chosen as 100 nm @xmath31 100 nm @xmath31 4 nm , @xmath32 = @xmath33 @xmath34 @xcite , @xmath35 = @xmath36 @xmath37 @xcite , @xmath5 = 0.01 , @xmath38 = 1.32 ma , and input spin current @xmath39 = 1.5 ma .
simulations are carried at a lattice temperature of 300 k. this work considers idealized nanomagnet and channel structures .
fabrication imperfections such as edge roughness or pinning sites may alter the results .
however , analysis under idealized conditions is a useful first step in evaluating complex nanomagnet systems and allows us to benchmark the upper limit of device performance .
in addition , a single - domain approximation is used to model the reversal mechanics of the nanomagnet . since a magnetic body typically consists of multiple domains that interact through an exchange field
, the magnetization evolution can be incoherent @xcite .
while the effect of incoherency is omitted from the delay distribution analysis , the models presented in this work have been fitted to experimental data and can still be applied to real - world setups assuming some fitting parameters @xcite .
the physics of pma nanomagnet reversal has been exhaustively studied in @xcite . assuming a perfectly square nanomagnet , the energy landscape of the nanomagnet is greatly simplified allowing analytical solutions of the llg equation .
the implicit analytical equation for the dynamics of the nanomagnet is given as @xcite @xmath40 where @xmath41 is the time it takes for the polar angle of the magnetization to transition from @xmath42 to @xmath43 .
@xmath44 is the time scale of the magnetization dynamics and is given as @xmath45 in ( [ eq : pma_delay ] ) , @xmath46 is the ratio of the spin current entering the nanomagnet and the critical spin current of the nanomagnet .
@xmath47 is mathematically given as @xcite @xmath48 where @xmath49 is the elementary charge .
( [ eq : pma_delay ] ) reduces to the expression derived in the seminal work of j.z .
sun @xcite under the condition that the input spin current of the nanomagnet vastly exceeds its critical current . assuming a large energy barrier between the two stable states of the nanomagnet
, the probability distribution of the initial angle of the magnetization of the nanomagnet is given as @xcite @xmath50 @xmath51 during fast reversals , it is expected that the thermal noise has little effect on the nanomagnet during its transition .
instead , the thermal noise only sets the initial angle , which affects the reversal delay according to ( [ eq : pma_delay ] ) . using this assumption , three different probability distribution models for nanomagnet switching
were derived in @xcite .
the analytical cumulative distribution functions ( cdfs ) are given below for completeness : @xmath52 \right \ } \label{eq : pdf_delay_1}\ ] ] @xmath53 \right \ } \label{eq : pdf_delay_2}\ ] ] @xmath54 \right \ } \label{eq : pdf_delay_3}\ ] ] [ eq : pdf_delay ] eq .
( [ eq : pdf_delay_1 ] ) corresponds to the derivation of the cdf through ( [ eq : pma_delay ] ) assuming a large energy barrier .
the pdfs of the delay can be found by taking the derivative of the cdfs .
note that the cdfs are denoted by `` @xmath55 '' while pdfs are denoted by `` @xmath56 '' .
assuming @xmath57 , ( [ eq : pdf_delay_1 ] ) can be further simplified to ( [ eq : pdf_delay_2 ] ) . finally , ( [ eq : pdf_delay_3 ] ) assumes both the initial and final magnetization angles are small .
figure [ fig : lui_pdf_comp ] compares the pdfs of ( [ eq : pdf_delay_1]-[eq : pdf_delay_3 ] ) to numerical results . in @xcite
, it is shown that an @xmath58 is sufficient to accurately describe the delay variation of the nanomagnet . however , in this work we note that these pdfs only become accurate at much larger values of @xmath59 .
this effectively means that the analytical pdfs are accurate only when the reversal time of the nanomagnet is @xmath0 100 ps . under such conditions , the electrical current required at the transmitting nanomagnet would vastly exceed the maximum threshold for electromigration of both the nanomagnet and the non - magnetic metallic channel ; therefore , there will be reliability concerns that will reduce the mean - time - to - failure of the asl device .
hence , other pdfs are needed to capture the magnetization dynamics accurately for reversal times on the order of several hundreds of picoseconds or nanoseconds .
the analytical solution of the llg becomes formidable for conditions where the input spin current is comparable to the critical spin current of the nanomagnet . to obtain a pdf that best represents the nature of the delay variation of the nanomagnet
, we consider the following situation .
it is shown in the previous sub - section that as the spin current is increased , the pdf of the delay tends toward a double exponential function .
in addition , it is known that if the spin current drops below the critical current , nanomagnet reversal occurs only when the magnetization angle of the nanomagnet becomes large enough such that the sub - critical current can overcome the reduced energy barrier of the nanomagnet@xcite .
this process of nanomagnet reversal , primarily through thermal activation , is known to be a single exponential function .
therefore , we seek a pdf solution to describe the nanomagnet reversal that can be seamlessly adjusted from a single- to a double - exponential function depending on the value of @xmath59 . in this work , we consider the frchet distribution that meets the above criteria .
the frchet distribution is mathematically given as @xcite @xmath60 where @xmath61 is the scale parameter , @xmath5 is the shape parameter , and @xmath41 is the delay of the nanomagnet . for the purposes of this paper , @xmath5 and @xmath61 are treated as fitting parameters @xcite .
to prove the suitability of the frchet distribution , fig .
[ fig : lui_pdf_comp ] shows the best fits of the frchet distribution to the delay curves of a nanomagnet driven by various spin currents @xcite .
the figure clearly shows improved accuracy of the frchet distribution to capture the numerical simulation data and the applicability of the frchet distribution to nanomagnet delay under various reversal regions .
[ fig : lui_pdf_comp ] demonstrates that the @xmath5 parameter increases greatly if @xmath57 .
since a smaller @xmath5 suggests a larger left lean , this suggests that reversal distributions for nanomagnets under large spin currents have less left lean .
for the remainder of this paper , the delay of a nanomagnet under the influence of a specific critical field will be represented by a frchet distribution that has been fitted to numerical data . because the delay has been shown to be a random variable , the probability that the nanomagnet delay will exceed some time @xmath62 will always be nonzero . when designing circuits , this probability can be referred to as the nanomagnets error rate ( @xmath63 ) . using the frchet distribution ( [ eq : pdf_frechet ] )
it is possible to derive @xmath62 as a function of @xmath63 .
mathematically , this relationship is given as @xmath64{-\ln\left [ 1-er \right ] } } , \ ] ] where @xmath5 and @xmath61 are defined previously in ( [ eq : pdf_frechet ] ) .
in any complex circuit , devices can be arranged in two ways .
they can be operated in parallel , where their outputs arrive at the same time .
alternatively , devices can be cascaded in series , where the output of one is fed into the input of of another . to analyze the total delay of a circuit , both
these cases need to be studied carefully . for the case of a circuit with multiple devices in parallel ,
the output delay ( denoted by random variable @xmath65 ) of the circuit is the maximum of the output delay of the devices given as @xmath66 where @xmath67 is a random variable representing the delay of a single nanomagnet .
hence , the cdf of the parallel device circuit becomes @xmath68 where @xmath55 corresponds to the cdf of the particular device . knowing this , the pdf of a circuit with multiple devices in parallel can be found using ( [ eq : par_pdf_2 ] ) and ( [ eq : pdf_frechet ] ) and is plotted in fig .
[ fig : multi_par ] . assuming a frchet distribution , the pdf of multiple nanomagnets in parallel
is given analytically as @xmath69 where @xmath70 is the number of devices in parallel .
100 nm @xmath31 4 nm , pma - type nanomagnet with @xmath32 = @xmath33 @xmath71 and @xmath35 = @xmath36 @xmath37 and is under the influence of a @xmath72 anti - parallel spin current .
the pdf of a single nanomagnet is estimated by fitting a frchet distribution to numerical data .
multi - nanomagnet results are calculated numerically.,width=336 ] for the case of devices connected in series , the output delay ( y ) of the circuit is the addition of the device delays connected in series and is given as @xmath73 the pdf of y is given as @xmath74 figure [ fig : multi_ser ] shows the pdfs of multiple nanomagnet devices arranged in series .
this situation corresponds to a repeater - chain circuit .
as the number of devices increases , the distribution of the delay gets more symmetrical and normal , following the central limit theorem @xcite .
100 nm @xmath31 4 nm , pma - type nanomagnet with @xmath32 = @xmath33 @xmath71 and @xmath35 = @xmath36 @xmath37 and is under the influence of a @xmath72 anti - parallel spin current .
the pdf of a single nanomagnet is estimated by fitting a frchet distribution to numerical data .
multi - nanomagnet results are calculated numerically.,width=336 ] often when developing spin - based circuits , it is possible to achieve similar functionalities using many devices driven in parallel , or many devices cascaded off eachother .
a prime example of this is the many variations of vlsi adder designs . since devices arranged in parallel increase the nanomagnet variation , it can be argued that highly parallel circuits may have longer delays than serialized circuits given a particular @xmath63
. however , fig .
[ fig : multi_comp ] demonstartes that this not the case .
while the number of devices in parallel does increase the circuit delay , this increase is still dwarfed by the increase in average delay caused by arranging the nanomagnet in series .
the previous sections introduced the key concepts required to analyze the delay variation of a complex logic circuit . for this paper , a four - input asl - and circuit is designed and analyzed .
an example of a four - input and ( and4 ) gate is shown in the inset of fig .
[ fig : pdf_and4_par ] .
as mentioned earlier , the summing nature of the inputs in asl logic naturally create a majority logic . to create an and gate from this majority logic ,
the input of the logic nanomagnet must be weighted such that the input to the nanomagnet is only positive when all the inputs are positive . in the case of the asl - and circuit shown
, a fixed magnet generating a constant @xmath75 is added such that the output nanomagnet will only receive a @xmath76-orientated spin current if all the nanomagnet are oriented along @xmath76 .
the polarity of this bias can be changed to switch between and and or logic .
it is assumed that the input nanomagnets have been reversed at time @xmath77 and the voltage supplies at these nanomagnets are turned on at this time @xcite .
an important aspect of current - based computation is the fact that the input current to the out nanomagnet will not be constant .
in fact , depending on the logical combination of the input devices , the current being fed into the out nanomagnet will vary .
assuming the inputs of the nanomagnet are equi - probable , the likelihood that the out nanomagnet is reversed by a spin current of a particular magnitude is given by table [ tab : in_prob_cur_2 ] .
the probabilities of each of these magnitudes follows an ordering , which can be described by pascal s triangle .
at worst , the nanomagnet will be driven by a current magnitude @xmath78 .
this corresponds to the case where all the inputs , except one , are oriented along the @xmath76 direction , and only one of the input nanomagnets is oriented along the @xmath79 direction .
this worst - case scenario is shown in fig .
[ fig : pdf_and4_par ] .
however , as mentioned previously , depending on the input pattern , the driving spin current to out is likely larger than @xmath80 . for larger spin current magnitudes ,
the nanomagnet is expected to reverse over much shorter timescales . by calculating the pdfs of nanomagnet reversal at each of the different spin current magnitudes ( by fitting the frchet distribution to numerical results )
, the pdfs can be combined to find the input - aware pdf of the circuit delay as shown in fig .
[ fig : pdf_and4_par ] .
this new input - aware pdf has several peaks corresponding to each of the possible spin current magnitudes during the operation of the circuit .
as the number of inputs increases , the number of peaks will also increase , but the area under each of the peaks will decrease since the area of the entire pdf must remain equal to unity . assuming a very small error - rate , one
is mainly concerned with the right - most peak that is associated with nanomagnet reversal under the minimum operating current .
this suggests that for a given delay , the circuit reliability is improved as the number of inputs increases . in other words , as the circuit becomes more complex , it tends to operate more reliably given a certain delay . .
input - probability aware distribution analyzes the out nanomagnet reversal considering the variable input spin current magnitudes considering different input combination .
inset : schematic of parallelized asl and4 gate .
blue square represent thin - film nanomagnets while purple bars represent channels .
orange arrows represent spin orientation.,width=336 ] .assuming all logical input combinations are equally possible , table demonstrates the probability that a nanomagnet is reversed with a particular spin current magnitude .
spin current magnitude is normalized against minimum reversal current . [
cols="<,<",options="header " , ]
in this paper , analytical and numerical methods have been developed to analyze the delay variation of a nanomagnet - based circuit by using the probability distribution functions that describe the stochastic dynamics of the nanomagnet .
firstly , the reversal delay of a single pma nanomagnet under the effects of thermally - induced stochasticity is extensively analyzed .
it is shown that the previously published models are mainly accurate in the limiting case of large input spin current or short reversal timescales of the nanomagnet . here , we study the reversal of nanomagnets that are operating near critical conditions that allow reversal over several hundreds of picoseconds or nanosecond timescales .
we use the frchet distribution function to model the nanomagnet reversal over a broad range of input spin currents and reversal times .
the frchet distribution is fitted to numerical simulation data obtained from solving the stochastic llg equation .
note , the delay distributions analyzed here are for a single - domain nanomagnet . while @xcite found that the probability density functions of ( [ eq : pdf_delay ] ) can be fitted to experimental reversal data , further work is needed to understand how magnetization incoherency will affect the delay distribution curves . knowing the delay of a single nanomagnet
, it is possible to analyze nanomagnet circuits where nanomagnets are evaluated in either parallel or series configurations .
it is shown that a circuit with nanomagnets evaluated in parallel has a larger variation but a smaller average delay than a circuit with nanomagnets in series .
finally , these concepts are applied to the evaluation of a asl - and gate .
it is shown that having more inputs to an and circuit is beneficial since the fraction of the time the nanomagnet is operating under the minimum input spin current is reduced .
the work in this paper is an initial step towards building and evaluating much larger nanomagnetic circuits .
further work is need to evaluate the delay variation of circuits implemented in alternate magnetic technologies and evaluating the performance metrics of complex spintronic logic circuits .
the amount of electric current , @xmath81 , pumped into the transmitter that reaches the receiver is quantified through spin injection and transport efficiency ( site ) . here
, we use the mathematical models for site derived in @xcite to obtain the amount of @xmath81 required to achieve a specific amount of spin current at the receiver .
the models take into account size effects in ultra - scaled metallic channels . as shown in fig .
[ fig : current_combined_nm ] , for a channel length of 500 nm and in the absence of grain - boundary ( @xmath82 ) and sidewall scatterings ( @xmath83 ) , the amount of electrical current to obtain 1.5 ma of spin current at the receiver nanomagnet is 5.3 ma and 4.5 ma for copper and aluminum channels , respectively .
the required electrical current increases in the presence of realistic size effects .
the inset plot of fig .
[ fig : current_combined_nm ] shows the electrical current density through the nanomagnet as a function of channel length for different values of channel width . while increasing the width of the channel reduces the electrical current density through the nanomagnet and improves reliability of the asl device , it also increases the overall device footprint and will limit the device scalability . the delay associated with spin diffusion through the channel
is given as @xmath84 where @xmath85 is the diffusion coefficient of electrons in the channel .
using @xmath85 = 126 @xmath86 and 80 @xmath86 for cu and al , respectively , the diffusion delay through a 500-nm long spin channel is only about 20 ps @xcite .
this delay is more than an order of magnitude lower than the nanomagnet switching delay and will not be considered in this work .
it has been previously argued that in asl circuits , nanomagnet coupling becomes an issue for only very short interconnects @xcite . at channel lengths of several hundreds of nanometer considered in this work ,
the dipolar field generated by one nanomagnet to the center of the other would be @xmath87 @xcite . since
the magnitude of the dipolar field is three orders of magnitude smaller than the @xmath20-anistropy of the nanomagnet , the effect of dipolar coupling is omitted in this work .
this project was supported by the nanoelectronics research corporation ( nerc ) , a wholly - owned subsidiary of the semiconductor research corporation ( src ) , through the institute for nanoelectronics discovery and exploration ( index ) .
the authors would also like to thank prof .
andrew kent and his group at new york university for their help and insight .
we would also like to thank dr .
daniele pinna from cnrs for useful discussions .
m. hosomi _ et al .
_ , `` a novel nonvolatile memory with spin torque transfer magnetization switching : spin - ram , '' in _ electron devices meeting , 2005 .
iedm technical digest .
ieee international_.1em plus 0.5em minus 0.4emieee , 2005 , pp .
459462 .
n. kani , s. dutta , and a. naeemi , `` analysis of coupling strength in multi - domain magneto - systems , '' in _ device research conference ( drc ) , 2015 73rd annual_.1em plus 0.5em minus 0.4emieee , 2015 , pp .
111112 .
g. bertotti , i. d. mayergoyz , and c. serpico , `` critical fields and pulse durations for precessional switching of thin magnetic films , '' _ magnetics , ieee transactions on _ , vol . 39 , no . 5 , pp . 25042506 , 2003 .
n. kani and a. naeemi , `` pipeline design in spintronic circuits , '' in _ proceedings of the 2014 ieee / acm international symposium on nanoscale architectures_.1em plus 0.5em minus 0.4emacm , 2014 , pp . 110115 .
s. manipatruni , d. e. nikonov , and i. a. young , `` modeling and design of spintronic integrated circuits , '' _ circuits and systems i : regular papers , ieee transactions on _ , vol .
59 , no . 12 , pp .
28012814 , 2012 .
k. roy , s. bandyopadhyay , and j. atulasimha , `` metastable state in a shape - anisotropic single - domain nanomagnet subjected to spin - transfer - torque , '' _ applied physics letters _
101 , no .
16 , p. 162405
, 2012 .
d. pinna , a. kent , a. , and d. stein , `` thermally - assisted spin - transfer torque magnetization reversal of uniaxial nanomagnets in energy space , '' _ magnetics , ieee transactions on _ , vol .
49 , no . 7 , pp . 31443146 , 2013 .
v. calayir _ et al .
_ , `` static and clocked spintronic circuit design and simulation with performance analysis relative to cmos , '' _ circuits and systems i : regular papers , ieee transactions on _ , vol . 61 , no . 2 , pp .
393406 , 2014 .
s. rakheja , s .- c .
chang , and a. naeemi , `` impact of dimensional scaling and size effects on spin transport in copper and aluminum interconnects , '' _ electron devices , ieee transactions on _ , vol . 60 , no . 11 , pp .
39133919 , 2013 .
s. rakheja and a. naeemi , `` roles of doping , temperature , and electric field on spin transport through semiconducting channels in spin valves , '' _ nanotechnology , ieee transactions on _ , vol .
12 , no . 5 , pp .
796805 , 2013 . | in this paper we systematically evaluate the variation in the reversal delay of a nanomagnet driven by a longitudinal spin current while under the influence of thermal noise .
we then use the results to evaluate the performance of an all - spin - logic ( asl ) circuit .
first , we review and expand on the physics of previously - published analytical models on stochastic nanomagnet switching .
the limits of previously established models are defined and it is shown that these models are valid for nanomagnet reversal times @xmath0 200 ps .
second , the insight obtained from previous models allows us to represent the probability density function ( pdf ) of the nanomagnet switching delay using the double exponential function of the frchet distribution .
the pdf of a single nanomagnet is extended to more complex nanomagnet circuit configurations .
it is shown that the delay - variation penalty incurred by nanomagnets arranged in parallel configuration is dwarfed by the average delay increase for nanomagnets arranged in a series configuration .
finally , we demonstrate the impact of device - level performance variation on the circuit behavior using asl logic gates .
while the analysis presented in this paper uses an asl - and gate as the prototype switching circuit in the spin domain , the physical concepts are generic and can be extended to any complex spin - based circuit . |
the large - scale distribution of galaxies contains the signature of acoustic waves that propagated through the universe prior to the epoch of recombination .
this signal , referred to as baryon acoustic oscillations ( bao ) , appears as a modulation in the amplitude of the galaxy power spectrum , @xmath12 , and a broad peak in the large - scale two - point correlation function , @xmath0 @xcite .
the wavelength of the oscillations in @xmath12 and the location of the peak in @xmath0 can be associated with the maximum distance that these acoustic waves can travel before the decoupling of matter and radiation , that is , the sound horizon at the drag redshift , @xmath13 . as this scale
can be constrained with high accuracy from observations of the cosmic microwave background ( cmb ) , the acoustic scale inferred from the clustering of galaxy samples at different redshifts can be used as a standard ruler to measure the distance - redshift relation , providing a powerful and robust probe of the expansion history of the universe @xcite .
the bao signal was first detected in the clustering of the two - degree field galaxy redshift survey ( 2dfgrs , * ? ? ?
* ; * ? ? ?
* ) by @xcite and the luminous red galaxy ( lrg , * ? ? ?
* ) sample of the sloan digital sky survey ( sdss , * ? ? ?
* ) by @xcite .
since then , subsequent analyses on various galaxy samples have provided bao measurements with increasing precision @xcite . using these results
it is now possible to construct a hubble diagram based entirely on bao distance measurements .
it has become standard practice to use this information , in combination with additional data sets , when deriving constraints on cosmological parameters .
separate measurements of the acoustic scale in the directions parallel and perpendicular to the line of sight can be used to obtain constraints on the hubble parameter , @xmath14 , and the angular diameter distance , @xmath15 , through the alcock paczynski test @xcite .
however , the bao signal on angle - averaged clustering measurements such as @xmath12 or @xmath0 provide estimates of the average distance @xmath16 .
although most analyses have focused on angle - averaged quantities , the large volumes probed by present - day galaxy samples make it possible to extend these analyses to anisotropic clustering measurements @xcite using the full power of the bao test .
the clustering of galaxies encodes additional information beyond that contained in the bao signal that can significantly improve the cosmological constraints derived from large - scale structure ( lss ) data sets .
this extra information is particularly important for anisotropic clustering measurements , where the signature of the so - called redshift - space distortions ( rsd ) can be used to constrain the growth rate of cosmic structures @xcite . in this way
, anisotropic clustering measurements can provide information of the expansion history of the universe and the growth of density fluctuations , which can be used to distinguish between the dark energy and modified gravity scenarios for the origin of cosmic acceleration .
the most accurate bao measurements to date have been obtained from the baryon oscillation spectroscopic survey ( boss , * ? ? ?
* ) , which is one of the four component surveys of sdss - iii @xcite .
after applying a modified version of the reconstruction technique of @xcite , the bao signal in the galaxy clustering of boss sdss data release 9 ( dr9 , * ? ? ?
* ) provided a 1.7 per cent accuracy measurement of the average distance @xmath17 at @xmath18 @xcite , as well as separate constraints on @xmath15 and @xmath14 at the same redshift with 3 and 8 per cent accuracy , respectively @xcite .
these measurements have been complemented by analyses of the full shape of isotropic and anisotropic clustering measurements @xcite . besides galaxy clustering analyses ,
a sample of high - redshift quasars from boss has been used to detect for the first time the signature of the bao in the fluctuations of the lyman-@xmath19 forest at @xmath20 @xcite . in this paper
we use information from the full - shape of the two - point correlation function and the clustering wedges statistic @xcite measured from boss data to derive constraints on cosmological parameters .
we extend the analyses of @xcite based on a high - redshift galaxy sample from boss dr9 to the data corresponding to dr10 @xcite and dr11 ( internal data - release ) , including results from the low - redshift boss galaxy sample .
as the statistical uncertainties characterizing different cosmological observations become smaller , it is important to explore potential systematics that can be introduced by the analysis techniques and models applied to the data .
the comparison of the results obtained by applying multiple methods to the same data can be used to identify the presence of systematics errors .
our analysis is part of a series of papers examining the clustering properties of the boss dr10 and dr11 galaxy samples with different methodologies .
tojeiro et al .
( in preparation ) and @xcite analyse the isotropic and anisotropic bao signal in these samples and explore their cosmological implications .
@xcite study the sensitivity of these bao measurements to the properties of the galaxy population being analysed .
@xcite investigate the potential systematic errors affecting anisotropic bao measurements .
@xcite perform a detailed analysis of the effect of the uncertainties in the covariance matrices determined from mock catalogues on the obtained constraints .
these analyses are complemented by those of @xcite , @xcite and @xcite , who analyse the full shape of the monopole - quadrupole pair in configuration and fourier space .
these studies attempt to condense the information of the clustering measurements into a few numbers reflecting the geometric constraints and the measurements of the growth of structures , that are then compared with the predictions from different cosmological models .
we follow an alternative approach in which we perform the comparison with cosmological models at the level of the galaxy clustering measurements themselves .
the consistency of the results presented here and those of our companion papers is a reassuring indication of the robustness of our results .
the outline of this paper is as follows . in section [ sec : data ]
we describe our galaxy sample , the procedure followed to obtain our clustering measurements and their respective covariance matrices , as well as the additional data sets included in our analysis . in section [
sec : method ] we review our model of the full shape of the correlation function and the clustering wedges and our methodology to obtain cosmological constraints . in section [ sec : results ] we present the constraints on cosmological parameters obtained from different combinations of data sets and parameter spaces .
finally , section [ sec : conclusions ] contains our main conclusions .
we use the lowz and cmass samples of boss corresponding to sdss dr10 @xcite and dr11 , which will become publicly available with the final data release of the survey .
these galaxy samples were selected on the basis of the sdss multicolour photometric observations @xcite to cover the redshift range @xmath21 with a roughly uniform comoving number density @xmath22 ( * ? ? ?
* ; * ? ? ?
* padmanabhan et al . in preparation ) .
up to @xmath23 and 2 per cent of lowz and cmass targets , respectively , were observed during the sdss i / ii surveys @xcite and thus already have a redshift .
the remaining redshifts were measured from the spectra obtained with the double - armed boss spectrographs @xcite by applying the minimum-@xmath24 template - fitting procedure described in @xcite and @xcite .
the lowz sample consists primarily of red galaxies that lie in massive haloes , with a satellite fraction of 12 per cent @xcite .
the cmass sample is approximately complete down to a limiting stellar mass of @xmath25 @xcite , and has a @xmath2610 per cent satellite fraction @xcite .
although this sample is dominated by early type galaxies , it contains a significant fraction of massive spirals ( @xmath2626 per cent , * ? ? ?
@xcite describes the construction of catalogues for lss analyses based on these samples .
we use these samples separately , restricting our analysis to the redshift ranges @xmath27 for the lowz sample , and @xmath28 for the cmass galaxies .
we study the clustering properties of these galaxy samples by means of the angle - averaged correlation function , @xmath0 , and the clustering wedges statistic @xcite , @xmath29 , which corresponds to the average of the full two - dimensional correlation function @xmath30 over the interval @xmath31 , that is @xmath32 we use two wide clustering wedges , @xmath1 and @xmath2 , defined for the intervals @xmath33 and @xmath34 , respectively .
the basic procedure implemented to obtain these measurements from the lowz and cmass samples is analogous to that of @xcite and @xcite .
here we summarize the most important points and refer the reader to these studies for more details .
we convert the observed redshifts into distances assuming a flat @xmath3cdm fiducial cosmology characterized by a matter density parameter of @xmath35 .
we use the estimator of @xcite to compute the full correlation function @xmath30 of the lowz and cmass samples , with random samples following the same selection function as the original catalogues but containing 50 times more objects .
the value of @xmath36 of a given pair is defined as the cosine of the angle between the separation vector , @xmath37 , and the line - of - sight direction at the midpoint of @xmath37 .
we infer the correlation function @xmath0 and the clustering wedges @xmath1 and @xmath2 by averaging the full @xmath30 over the corresponding @xmath36 intervals . as discussed in @xcite , this procedure correctly accounts for the @xmath36 dependence of the random - random counts , which is ignored when the estimator of @xcite is applied to the averaged counts directly , leading to a bias in the recovered clustering measurements .
when computing the pair counts , we assign a series of weights to each object in our catalogue . first , we apply a radial weight designed to minimize the variance of our measurements @xcite given by @xmath38 where @xmath39 is the expected number density of the catalogue at the given redshift and @xmath40 is a scale - independent parameter , which we set to @xmath41 .
we also include angular weights to account for redshift failures and fibre collisions .
for the cmass sample we apply additional weights to correct for the systematic effect introduced by the local stellar density and the seeing of the observations , as described in detail in @xcite .
the left panels of figs .
[ fig : measurdr10 ] and [ fig : measurdr11 ] show the resulting angle - averaged correlation function @xmath0 of the sdss - dr10 and dr11 lowz ( upper ) and cmass ( bottom ) samples , respectively , while the left panels show the corresponding clustering wedges @xmath1 ( circles ) and @xmath2 ( squares ) .
the anisotropic clustering pattern generated by redshift - space distortions leads to significant differences in the amplitude and shape of the two clustering wedges , with @xmath2 showing a lower amplitude and a stronger damping of the bao peak than @xmath1 .
the dashed lines correspond to the best - fitting @xmath3cdm model obtained from the combination of the lowz and cmass dr11 clustering wedges with cmb observations from the planck satellite @xcite and the cmb polarization measurements from wmap @xcite as described in section [ sec : lcdm ] , which provide an excellent description of all our measurements . when comparing our boss clustering measurements with theoretical predictions we assume a gaussian likelihood function of the form @xmath42 .
the calculation of the @xmath24 value of a given model requires the knowledge of the inverse covariance matrix of our measurements , which we estimate using mock catalogues matching the selection functions of the lowz and cmass samples .
these mocks were constructed from two sets of pthalos realizations @xcite , corresponding to our fiducial cosmology , as described in @xcite and manera et al .
( in preparation ) .
our cmass mocks are based on 600 independent simulations with a box size of @xmath43 , while those of the lowz sample were constructed from a separate set of 500 boxes with the same volume . in the construction of these mocks , the northern galactic cap ( ngc ) and southern galactic cap ( sgc ) components of the survey
were considered as being independent , and sampled separately from the same pthalos realizations .
the volume of the lowz sample allowed us to obtain two separate ngc and sgc mocks per pthalos realization , leading to 1000 independent combined ngc+sgc lowz mock catalogues .
the larger volume of the cmass sample makes it more difficult to construct mocks of the ngc and sgc components from the boxes without overlap .
this means that the ngc and sgc cmass mocks drawn from the same box are not independent . for dr10
the overlap between the ngc and sgc mocks is approximately 75 per cent of the area covered by the sgc , while for dr11 the whole of the southern component is also covered by the ngc . to account for this overlap in our covariance matrix estimations
we construct two sets of 300 independent ngc+sgc cmass mocks , drawing the matched components from different boxes . | we explore the cosmological implications of the angle - averaged correlation function , @xmath0 , and the clustering wedges , @xmath1 and @xmath2 , of the lowz and cmass galaxy samples from data release 10 and 11 of the sdss - iii baryon oscillation spectroscopic survey .
our results show no significant evidence for a deviation from the standard @xmath3cdm model .
the combination of the information from our clustering measurements with recent data from the cosmic microwave background is sufficient to constrain the curvature of the universe to @xmath4 , the total neutrino mass to @xmath5 ( 95% confidence level ) , the effective number of relativistic species to @xmath6 , and the dark energy equation of state to @xmath7 .
these limits are further improved by adding information from type ia supernovae and baryon acoustic oscillations from other samples . in particular , this data set combination is completely consistent with a time - independent dark energy equation of state , in which case we find @xmath8 .
we explore the constraints on the growth - rate of cosmic structures assuming @xmath9 and obtain @xmath10 , in agreement with the predictions from general relativity of @xmath11 .
cosmological parameters , large scale structure of the universe |
our knowledge of uv radiation fields and energy injection into the igm at @xmath3 is fairly tenuous .
there are two main constraints : observations of the integrated background light ( madau & pozzetti 2000 , bernstein et al 1999 ) , and the fact that no gunn - peterson trough is observed in the spectra of the highest - redshift quasar to date ( fan et al 2000 ) , implying that the universe must be reionized by @xmath4
. the upcoming next generation space telescope ( ngst ) will be able to image high - redshift star clusters or agns in rest frame uv continuum emission ( haiman & loeb 1997,1998 ) , and their redshifts may be obtained via h@xmath2 observations ( oh 1999 ) .
nonetheless , the redshift - binned number counts will be fairly sparse , and one is unlikely to probe sufficiently far down the luminosity function to get a good measure of the comoving emissivity as a function of redshift .
observations of gamma - ray blazars ( `` grazars '' ) probe extragalactic ir and uv radiation fields , by observing the pair production opacity to @xmath5 rays at the high energy end ( gould & schreder 1967 , stecker , de jager & salamon 1992 , madau & phinney 1996 , primack et al 1999 ) .
all theoretical models have confined their predictions to low redshift grazars , with the exception of salamon & stecker ( 1998 ) , who computed the @xmath5-ray opacity up to z=3 .
they concluded that because the stellar emissivity peaks between z=1 and z=2 , the @xmath5-ray opacity shows little increase at high redshift , and thus is not dependent on the initial epoch of galaxy formation . to date
, egret has detected 66 gamma - ray loud blazars ( hartman et al 1999 ) , out to redshifts @xmath6 . the next generation of gamma - ray telescopes ( glast , celeste , stacee , magic , hess , veritas , and milagro ) should greatly enlarge this sample .
if the low redshift correlation between black hole mass and bulge mass ( magorrian et al 1998 ) continues to high redshift , then it is possible that high - redshift halos could host mini - quasars ( haiman & loeb 1998 , haehnelt , natarajan & rees 1998 ) , which should be detectable in rest frame uv emission by ngst and x - ray emission by chandra ( haiman & loeb 1998 , 1999 ) in the redshift range @xmath7 .
this raises the exciting possibility that grazars could be detected at similarly high redshifts .
it is worth noting that egret has detected @xmath8 sources at high galactic latitudes @xmath9 ( mukherjee , grenier & thompson 1997 ) , with no known counterparts at other wavelengths .
their spatial distribution and log n- log s plot can be well fit by a galactic component plus an isotropic , extragalactic contribution
. some of these may well be unidentified high - redshift blazars . in this paper ,
i point out that if grazars are detected at high redshifts @xmath10 , the pair production opacity to gamma ray photons can be used to probe the comoving emissivity longward of the lyman break at these extremely high redshifts , independent of the star formation rate at lower redshifts . due to the small escape fraction of ionizing photons @xmath11 from host galaxies , as well as the high photoelectric opacity of the igm at these wavelengths , the comoving number density of uv photons exhibits a sharp drop at the lyman edge at all redshifts .
thus , there is only a limited pathlength over which a gamma - ray photon can pair produce against uv photons , before it redshifts to energies which require uv photons above the lyman edge for pair production to take place . for @xmath12 ,
the universe becomes optically thin to the gamma - ray photon .
thus , the detection of an absorption edge in a high - redshift grazar places an immediate constraint on the mean radiation field over the redshifts @xmath13 .
furthermore , measurement of the different absorption at a given observed energy between blazars at redshifts @xmath14 places an immediate constraint on the radiation field in the redshift range @xmath15 .
detection of grazars at a number of redshifts would then enable one to probe the uv emissivity history of the universe .
in all numerical estimates , we assume a background cosmology given by the concordance values of ostriker & steinhardt ( 1995 ) : @xmath16 .
a detailed study of the detectability of high - redshift blazars is beyond the scope of this paper . in this section ,
i merely show that it is plausible that glast will be able to detect high redshift blazars . with a point source sensitivity of @xmath17photons @xmath18 ,
egret has detected @xmath19 high - redshift blazars out to @xmath20 ( hartman et al 1999 ) .
the associated gamma - ray luminosities correspond to @xmath21 , and typically dominate the bolometric luminosity of the agn , with @xmath22 .
the upcoming gamma - ray telescope glast ( see http://glast.gsfc.nasa.gov ) will be 2 orders of magnitude more sensitive , with a detection threshold of @xmath23photons @xmath24 for a 5 @xmath25 detection with a 50 hour integration , and @xmath26 photons @xmath18 ( these thresholds correspond to the same detection limit for a @xmath27 source spectrum where @xmath28 ) .
goals for glast include a broad energy coverage from @xmath29 , with a spectral resolution of @xmath30 in the @xmath31 10 gev range , a field of view of @xmath31 3 sr , and a source location determination accuracy of 30 arcsec- 5 arcmin . during its lifetime
, it will perform an all - sky survey similar to that conducted by egret .
sources of the same or somewhat fainter luminosity as those detected by egret may be seen by glast out to extremely high redshifts , @xmath32 .
will such luminous sources will be present at high redshift ?
if the agn is assumed to emit all its energy at gamma - ray wavelengths at the eddington luminosity , the inferred black hole mass is extremely high , @xmath33 . however , there are two reasons why gamma ray sources of high apparent luminosity do not require such massive black holes : ( i ) even if all photons are radiated isotropically , if most of the radiation emerges at high energies ( as appears to be the case in gamma - ray blazars ) , then klein - nishina effects must be taken into account ( dermer & gehrels 1995 ) .
the inferred black hole mass , given by @xmath34 $ ] ( where @xmath35 is the observed flux between the lower and upper bandpass limits @xmath36 ) typically drops by 3 orders of magnitude , so a @xmath37 source only requires a @xmath38 black hole .
( ii ) there is strong evidence for relativistic beaming in blazars ( e.g. , through the observation of superluminal jets ( von montigny et al 1995 ) ) . in fact , if blazars were not beamed , we would not be able to see any gamma - rays from them due to the high pair production opacity at the source ; beaming reduces the luminosity / radius ratio by a factor @xmath39 , allowing photons to escape ( maraschi et al 1992 ) . here ,
if @xmath2 is the spectral index of the source , then @xmath40 for a moving sphere in the synchrotron self - compton ( ssc ) model , while @xmath41 in the external radiation compton ( erc ) model ; @xmath42^{-1}$ ] is the relativistic doppler factor , and @xmath5 is the lorentz factor . if @xmath43 is the initial intrinsic luminosity of the jet in gamma - ray emission , beaming boosts the observed luminosity of the jet to @xmath44 . for @xmath45 , then @xmath46 , and the observed luminosity
is amplified by a factor of thousands .
the strong relativistic beaming reduces the fraction of sources which are visible , since they can only be seen when viewed along the jet axis . for instance , for @xmath47 , and @xmath28 , in the ssc model the observed luminosity is reduced by an order of magnitude from its maximum if the jet is pointing @xmath48 from our line of sight , and two orders of magnitude if the jet is pointing @xmath49 from our line of sight .
note that the black hole masses derived for a number of low redshift blazars from variability timescale and transparency arguments lie in the range @xmath50 ( cheng et al 1999 , hartman et al 1996 , becker & kafatos 1995 , romero et al 2000 ) .
the luminosity of these blazars is so high they could be seen at high redshift with glast , and black hole masses of @xmath51 are reasonably abundant at high redshift in certain models of agn formation ( haiman & loeb 1998 ) .
note that for a set of blazars of constant intrinsic luminosity @xmath52 and comoving number density , the redshift distribution of detected sources in a flux - limited survey flattens considerably and extends to higher redshifts as the lorentz factor increases ( dermer & gehrels 1995 ) . at present
, the modelling of even the low - redshift population of gamma - ray blazars is a matter of considerable debate .
models which attempt to account for the unresolved gamma ray background with faint blazars either extrapolate the observed @xmath5-ray luminosity function obtained with egret ( chiang & mukherjee 1998 ) or use an assumed conversion between the observed radio loud agn luminosity function and the blazar luminosity function ( stecker & salamon 1996 ) . in this paper ,
i use a highly simplified model to estimate the detectability of high - redshift blazars .
i assume that the intrinsic luminosity @xmath52 of the jet in gamma - ray emission ( prior to beaming ) scales with the luminosity of the accretion disk @xmath53 , where the optical b - band luminosity is taken to be an accurate reflection of @xmath54 ( in particular , assuming the median quasar spectrum of elvis et al ( 1994 ) , a 1 @xmath55 black hole shining at the eddington luminosity has a b - band luminosity of @xmath56 ) . such a jet - disk correlation is observed in the ratio of observed radio ( i.e. , after beaming ) to optical luminosities ( falcke , malkan & biermann 1995 ) .
i also assume relativistic beaming with @xmath57 , where @xmath58 ( and @xmath28 , the average spectral index observed in the egret blazars ) .
the change in the observed luminosity function due to the effects of relativistic beaming is given by ( urry & padovani 1995 ) : @xmath59 where the probability of observing luminosity @xmath60 given the intrinsic luminosity @xmath52 is given by : @xmath61 i use the fit to the observed b - band luminosity function @xmath62 from pei ( 1995 ) .
i adjust the relation @xmath63 and the lorentz factor @xmath5 ( note that since the distribution of blazar lorentz factors is unknown , i assume they all have the same lorentz factor ) to fit the number of blazars detected by egret and their redshift distribution .
i ignore the effects of blazar flaring , which increases the luminosity by some factor @xmath64 ( where typically @xmath65 ) some fraction @xmath66 of the time ( where typically @xmath67 ) , which results in a second term @xmath68 , since these two degrees of freedom , the normalization ( chosen by selecting @xmath5 and thus the beaming angle ) and luminosity boost ( choosen by the combination @xmath69 ) , are already present in our model .
i find that @xmath70 ( where @xmath71 and @xmath47 provides a good fit ( see fig 1 , top panel ) .
note that since @xmath72 ( corresponding to @xmath73 ) , these relations result in a sed which is in reasonable agreement with the observed seds of gamma - ray blazars ( see fig 1 in ghisellini et al 1998 ) , where @xmath74 typically , although it can range from @xmath75 .
furthermore , the lorentz factor @xmath76 is in reasonable agreement with the relativistic doppler factor @xmath77 from models which take into account the sed , time variability , and gamma - ray transparency of blazars ; the derived @xmath78 ( ghisellini et al 1998 ) .
the somewhat lower doppler factors i have adopted conservatively underestimate the number of high - redshift blazars .
i then extrapolate this model to high redshift by applying it to the press - schechter based model of haiman & loeb ( 1998 ) for quasars .
in this model , which is calibrated to the observed luminosity function of pei ( 1995 ) at lower redshifts , each halo with @xmath79k hosts a black hole with mass @xmath80 which shines at the eddington luminosity for @xmath81 years .
the result is shown in the bottom panel of fig [ blazar_fig ] , which shows that gamma - ray blazars may be detected out to @xmath82 . by the time glast
is launched , a large database of quasars with known redshifts will be available ( e.g. from the sloan digital sky survey , sdss ( york et al 2000 ) ) , and many high - redshift blazars can be selected simply by identifying their optical counterparts .
i emphasize once again that this highly simplified model is only intended to serve as a plausibility argument .
the main point is that while the press - schechter formalism predicts that massive halos @xmath83 expected to host supermassive black holes of the requisite luminosity become exponentially rare at high redshift , processes which increase the luminosity of a lower luminosity population ( beaming @xmath57 , flaring @xmath84 ) create a power - law tail of bright sources .
i have neglected a tail to the distribution of lorentz factors , or flaring , which could further flatten the redshift distribution of detectable sources , increasing the maximum redshift out to which sources can be seen .
finally , gravitational lensing could bring otherwise undetectable sources into view , although the low optical depths for strong lensing ( e.g. @xmath85 for @xmath86 , porciani & madau 2000 ) imply that this should only have a small impact on number counts .
the pair - production optical depth for a photon observed at energy @xmath87 and emitted from a source at redshift @xmath88 is given by ( e.g. , madau & phinney 1996 ) : @xmath89 where @xmath90 , and @xmath91 is given by the criterion that pair production can take place , which requires that @xmath92 .
for a given energy e , the pair production cross - section @xmath93 rises sharply from the threshold energy @xmath91 , reaches a peak of @xmath94 at @xmath95 , and finally falls off as @xmath96 for @xmath97 . to calculate the optical depth of the universe to gamma - ray photons emitted at high redshift
, we need to know the number density of photons as a function of energy and redshift . before doing so in detail
, i make some simple estimates .
one can perform a simple order of magnitude estimate to show that the minimal comoving emissivity required to reionize the universe implies a high pair - production opacity for @xmath98 , but a low opacity for @xmath99 , where we define the threshold energy in the rest frame of the source for pair production against uv photons at the lyman edge as @xmath100gev .
this is because the number density of uv photons plummets at wavelengths shortward of the lyman limit ; the pair production opacity is thus dominated by soft photons @xmath101ev . for
the universe to be reionized , a minimum of 1 ionizing photon per baryon must be emitted .
thus , the comoving number density of ionizing photons is @xmath102 , where @xmath103 is the average number of times each baryon recombines during the reionization epoch .
the pair production cross section peaks when @xmath104 , with a value of @xmath105 .
thus , across a hubble volume , the optical depth for photons with @xmath99 is @xmath106 , which is undetectably small .
what is the pair production opacity for photons with @xmath98 ?
we can scale the number density of photons longward of the lyman limit with respect to the number of ionizing photons produced .
there are 3 factors involved : ( i ) @xmath107 .
observations of our own galaxy ( dove , shull , & ferrara 2000 , bland - hawthorn & maloney 1999 ) and local starbursts ( leitherer et al 1995 ) find the escape fraction of ionizing photons from galaxies into the igm is of order @xmath108 , and radiative transfer calculations find that the escape fraction should decrease strongly with redshift ( woods & loeb 1999 , ricotti & shull 2000 ) . the actual comoving number density of ionizing photons produced in starbursts is @xmath109 , where @xmath110 is the escape fraction of ionizing photons from the halo .
( ii ) @xmath111 .
photoelectric absorption in stellar atmospheres produces a sharp drop in the flux at the lyman edge . for a salpeter imf with metallicity @xmath112
, there are @xmath113 times as many photons emitted longward of the lyman limit as there are shortward of it , integrated over the history of the starburst .
( iii ) @xmath114 , due to intergalactic absorption . the energy density @xmath115 where @xmath116 where @xmath117 is the mean free path of an photon of frequency @xmath118 .
while the universe is optically thin to photons longward of the lyman limit , it is optically thick to ionizing photons , which have a much shorter mean free path . the energy density of ionizing photons is lower by a factor @xmath119 due to their high absorption rate .
the mean free path of ionizing photons decreases strongly with redshift and for @xmath120 the radiation field is largely local , due to the large increase in the number of adsorbers ( e.g. , madau , haardt & rees 1999 ) .
thus , we have @xmath121 .
this implies that the pair production opacity longward of @xmath0 over a hubble length is much greater , @xmath122 .
a similar estimate may be obtained by normalizing to the observed metallicity of the igm at @xmath123 , @xmath124 ; this can be shown to correspond to @xmath125 ionizing photons per igm baryon ( miralda - escude & rees 1997 ) , or a comoving number density of @xmath126 for photons longward of the lyman break , compared to the previous estimate of @xmath127 .
the expectation of a sharp drop in the intensity of the ambient intergalactic radiation field at the lyman edge is fairly robust .
a recent composite spectrum of 29 lyman break galaxies ( lbgs ) exhibit an observed flux ratio l(1500)/l(900)=@xmath128 , which is consistent with no internal photoelectric absorption , i.e. @xmath129 ( steidel , pettini & adelberger 2000 ) .
note that the authors themselves stress this result should be treated as preliminary ; the result could be due to a large number of uncertainties or selection effects , among them the fact that these galaxies were selected from the bluest quartile of lbgs .
even so , if this is typical of all high redshift galaxies , the observed lyman break ( likely due to absorption by stellar atmospheres ) and expected processing by intergalactic absorption at high redshift imply a lyman edge @xmath130 . similar considerations of internal and intergalactic photoelectric absorption apply if quasars dominate the ionizing radiation field . as a photon redshifts
, it has to interact with higher energy photons to pair produce .
a photon is able to pair produce until it redshifts below @xmath131 , i.e. , the universe is optically thin to a photon once it redshifts below @xmath132 , where @xmath133 is the original energy of the photon at source redshift @xmath88 .
thus , there is a fairly well - defined pathlength @xmath134 over which a photon may pair produce .
this has the fortunate consequence that low redshift photons do not affect photons emitted near @xmath131 ; thus , near the threshold energies one is always probing the average radiation field at redshifts comparable to the source redshift .
the flux decrement at a given frequency measures the average number density of photons redward of the lyman break over the associated redshift interval , @xmath135 .
gamma ray absorption measurements thus provide a fairly clean measurement of radiation fields at high redshift which are uncomplicated by radiative transfer effects since ( apart from unimportant @xmath136 opacity effects ) the universe is optically thin to photons redward of the lyman limit . in particular ,
photons longward of the lyman limit establish a homogeneous , isotropic radiation field early in the history of the universe .
as the pathlengths for pair production opacity to become significant are typically of order a hubble volume , opacity fluctuations due to poisson fluctuations or source clustering are insignificant . the measured background radiation field may be compared directly against the expected background from measurements of the source luminosity function by direct imaging of high redshift sources in mid - ir with ngst , where fast photometric redshifts may be obtained using the gunn - peterson break .
a comparison of the two should in principle allow one to check the completeness of a survey at ngst flux limits .
let us now use a specific model of high - redshift star formation to calculation the abundance of uv photons at high redshift .
the solution of the cosmological radiative transfer equation yields the mean specific intensity of the radiation background at the observed frequency @xmath137 , as seen by an observer at redshift @xmath138 as ( peebles 1993 ) : @xmath139 where @xmath140 , and @xmath141 is the effective optical depth of the igm to radiation emitted at @xmath142 and observed at @xmath138 at frequency @xmath137 .
redward of the lyman edge , radiative transfer is particularly simple as the universe is optically thin , and only the redshifting of photons is important ( there is one caveat to this statement : the optical depth of the igm in the lyman resonance lines prior to reionization is very large ; thus whenever a photon redshifts into a ly@xmath143 or higher order lyman resonance , it is reprocessed into a ly@xmath2 and balmer or lower order line photon ( see haiman , rees & loeb 1997 ) ) .
however , this merely causes a modulation in the spectrum in the 11.213.6 ev range , redistributing photons to the ly@xmath2 and balmer wavelengths .
the large typical energy intervals of target photons ( see fig 3 , top panel ) implies that this redistribution causes the pair - production opacity to remain the same or increase .
i therefore ignore this complication ) .
the number density of photons in an energy interval is then given by @xmath144 .
i model the star formation history @xmath145 of the high redshift universe with the semi - analytic models of haiman & loeb ( 1997 ) , in which a fixed fraction @xmath146 of the gas ( normalised to reproduce the observed igm metallicity at @xmath147 of @xmath148 ) in halos with @xmath79k ( which are able to undergo atomic cooling ) fragment to form a starburst lasting for @xmath149 years , and the halo collapse rate is given by the press - schechter formalism ( sasaki 1994 ) : @xmath150 where @xmath151 is the growth factor , and @xmath152 is the threshold above which mass fluctuations collapse . in this case
the star formation rate is given by : @xmath153 an approximate fit to the comoving star formation rate in the interval @xmath154 can be given by @xmath155 , where @xmath156 .
the comoving emissivity is : @xmath157 where @xmath158 is the stellar population spectrum , defined as the power radiated per unit frequency per unit initial mass by a generation of stars with age @xmath159 .
i obtain this spectrum from the bruzual & charlot ( 1993 ) code for a @xmath160 population ( i.e. , extremely low metallicity ) , assuming a salpeter imf with lower and upper mass cutoffs at 0.1 and 100 @xmath55 , and only computing the emissivity redward of the lyman break .
i ignore the effects of dust extinction , which should be negligible at these high redshifts when the metallicities are very low .
is the number density of recombination line photons such as ly@xmath2 sufficiently high to cause significant pair production opacity ?
it has been emphasized ( e.g. , loeb & rybicki 1999 ) that apart from possible dust attenuation , ly@xmath2 photons are not absorbed by the igm but resonantly scattered until they redshift out of resonance .
indeed , the ly@xmath2 radiation intensity has been predicted to be particularly strong prior to the epoch of reionization ( haiman , rees & loeb 1997 , baltz , gnedin & silk 1998 ) .
this is because the optical depth in the lyman series is very high , and all the lyman series lines except ly@xmath2 are absorbed immediately and redistributed .
thus , the ly@xmath2 and balmer lines are considerably brighter because they receive the energy of the lyman series .
this has spawned suggestions of detecting the epoch of reionization by a sharp drop in intensity at the rest - frame ly@xmath2 wavelength ( baltz , gnedin & silk 1998 , shaver et al 1999 ) .
in particular , baltz , gnedin & silk ( 1998 ) find that ly@xmath2 is about 3 times brighter and h@xmath2 is about 30 times brighter immediately prior to the reionization redshift .
if this is indeed the case , ly@xmath2 photons might contribute significantly to the pair production opacity .
let us define @xmath161 , the jump in the number density of photons longward of the ly@xmath2 wavelength .
this is given by the number of photons @xmath162 injected at the ly@xmath2 wavelength , over the total number of photons @xmath163 , include those redshifting into resonance .
most studies calculate the number of ly@xmath2 recombination photons by summing over the igm recombination rate in numerical simulations , by direct estimation of gas clumping in the simulations .
in fact , this underestimates the number of ly@xmath2 photons produced : if the escape fraction of ionizing photons is small , most recombinations occur in the hii regions of dense halos where star formation takes place , and the primary source of ly@xmath2 photons are the ionising sources themselves .
if each ionizing photon is converted into a ly@xmath2 photon ( plus lower energy photons ) at the source , then @xmath164 , where @xmath107 is the average escape fraction of ionizing photons from a souce .
thus , inserting @xmath165 into equation ( [ rad_transfer ] ) , i obtain for @xmath166 , the solution @xmath167 where @xmath168 is the production rate of ionizing photons in proper coordinates , _ prior _ to attenuation by the ionizing photon escape fraction .
this is a lower bound on @xmath169 since it does not include the conversion of photons trapped in higher order lyman resonances to ly@xmath2 photons .
i find that for the adopted stellar spectra , in which the lyman edge is typically @xmath113 , then typically @xmath170 , which implies that ly@xmath2 photons are comparable to uv continuum photons as a source of opacity . the jump factor @xmath171 may easily be understood as the ratio of the total number of lyman alpha photons produced ( or , the total number of ionizing photons produced ) against the total number of photons in the @xmath172ev range .
if the break at the lyman edge is reduced , then @xmath171 is increased and ly@xmath2 photons are more important in contributing to the pair - production opacity . in particular , for a power law spectrum with no discontinuity at the lyman edge ( e.g. as for quasars ) ,
ly@xmath2 photons are the dominant source of opacity .
ly@xmath2 photons are also the dominant source of opacity in scenarios where the universe is reionized by zero metallicity stars .
the higher effective temperature of these pop iii stars imply that their spectrum is much harder , and significantly fewer photons are emitted longward of the lyman break ( tumlinson & shull 2000 ) .
ly@xmath2 photons could thus an important source of pair production opacity provided the escape fraction is small , i.e. @xmath173 , and dust attenuation while the ly@xmath2 photons resonantly scatter in the igm is unimportant . if ly@xmath2 photons are the dominant source of pair production opacity , this would be extremely interesting , as it would provide a indirect census of the ionizing photon emissivity , prior to attenuation within the host sources .
this could prove to be a good measure of the comoving star formation rate .
one way to check if this is the case would be to directly image sources in ngst in rest frame uv longward of ly@xmath2 ( sufficiently far away from the ly@xmath2 damping wing ) , as well as in rest frame balmer line emission ( which should also be directly proportional to the production rate of ionizing photons , @xmath174 )
. this will give a sense as to whether ly@xmath2 or uv continuum photons are a greater source of opacity , provided the trend observed in bright sources extrapolates down to lower luminosities .
fig [ tau_fig ] shows the predicted attenuation factor as a function of observed photon energy , for grazars at redshifts @xmath175 , for both high ( @xmath176 ) and low ( @xmath177 ) star formation efficiencies .
there are several important features to note .
firstly , the shape of the attenuation curve for ly@xmath2 is similar to that for uv continuum photons .
it is not possible to immediately distinguish between scenarios where ly@xmath2 photons and uv continuum photons are the dominant source of opacity .
secondly , for lower star formation efficiencies , attenuation both sets in at only at higher energies and the attenuation curve is significantly shallower ( i.e. , it reaches full attenuation after a much longer energy interval ) .
this is easy to understand .
gamma - ray photons of higher energy have both a longer path - length to travel before they redshift to @xmath99 , and a higher number density @xmath178 of photons to pair produce against , since @xmath179 is lower .
thus , if the overall number density of uv photons is lower , one must go to higher energies to achieve the same attenuation . in figure [ dtaudx_fig
] , i show the relative contribution of different rest - frame target photon energies and redshift intervals for gamma - ray photons with @xmath180 , for a variety of source redshifts .
the shape of these curves is largely dependent on the overall number density of target uv photons .
for a lower sfr , the energy @xmath87 at which @xmath180 increases , and the target photon interval and redshift interval contributing to the resultant opacity broadens . note that most of the opacity comes from redshifts comparable to that of the source .
also , the opacity arises from a fairly narrow target photon energy interval @xmath181ev ; this varies weakly with source redshift .
thus the results do not depend strongly on the assumed spectral slope of the uv sources .
i also obtained by direct computation the contribution to the opacity due to ionizing photons shortward of the lyman edge . depending on one s assumptions for @xmath182 , it is smaller by 24 orders of magnitude , and is completely negligible . an important technique for isolating the contribution of high - redshift radiation fields would be to use measurements of _ differential _ absorption between blazars at different redshift .
consider two blazars with at redshifts @xmath183 and @xmath184 , with @xmath185 . at
a given observed energy @xmath87 , photons from each blazar will encounter identical optical depths for @xmath186 .
thus , any additional opacity seem in the spectrum of the blazar at @xmath184 must arise from the redshift interval @xmath15 alone , @xmath187 .
note that this identification is independent of any uncertainties in the spectral slope or redshift evolution of the uv background . as argued above
, radiation field intensity fluctuations are unimportant since the pair - production opacity arises on scales of order a hubble length .
thus , in figure ( [ tau_fig ] ) , any difference between the @xmath188 curves is solely due to radiation fields between @xmath189 and @xmath190 ; if star formation ceases for @xmath191 , the curves would lie on top of one another .
if there is little difference between two absorption curves from sources at @xmath183 and @xmath184 , one can get an upper bound on the uv emissivity in the range @xmath192 ; likewise , if two curves are so widely separated that meaningful measurements of @xmath193 can not be obtained ( in particular , if absorption saturates in one of the curves ) , one can get a lower bound on the uv emissivity in the range @xmath192 . in figure ( [ tau_evolve ] ) , i show how the attenuation at a fixed observed photon energy @xmath87 is expected to rise with redshift , due to the opacity provided by high redshift uv fields . for lower star formation efficiency ,
the attenuation rises more slowly ; if there is no star formation at high redshift the curve would be flat . in figure ( [ tauone_redshift ] ) i show how the energy @xmath87 for which @xmath180 is expected to fall with increasing redshift ; due to the increased opacity provided by high redshift photons , attenuation sets in at lower energies .
again , if there were no star formation at high redshift , this curve would show no evolution .
in fact , differential absorption measurements provide a powerful cross - check on the technique : if @xmath194 , then the assumed unabsorbed blazar spectrum must be evolving with redshift , due to a change in internal absorption , or underlying spectral index . note that if star formation rates are high , the attenuation occurs rapidly over a small energy interval , and uncertainties due to the extrapolation from the unabsorption portion of the spectrum is small ; for lower star formation rates the attenuation occurs over a larger energy interval , and uncertainties due to extrapolation are greater .
in this paper , i have suggested that if blazars can be detected at high redshift , detection of gamma - ray absorption due to pair production against high - redshift uv photons will provide a valuable probe of high - redshift uv radiation fields .
this is because the sharp lyman edge in the intergalactic radiation fields implies that gamma - ray photons have only a limited redshift interval in which to pair - produce . as they redshift to lower energies , they require photons with @xmath195ev to pair - produce , so the universe becomes optically thin .
the shape of the attenuation curve is primarily sensitive to the overall number density of photons longward of the lyman edge at high redshifts : the higher this number density , the lower the gamma - ray photon energy at which pair - production opacity sets in .
this makes it a useful test of the overall level of star formation and ambient uv radiation fields present at high redshift .
ly@xmath2 photons provide an important contribution to this pair production opacity , and indeed may be the dominant source of opacity if sources with a relatively lower fluxes longward of the lyman edge ( such as quasars or low metallicity stars ) are abundant .
finally , measurements of differential absorption between blazars at the same observed energies will allow us to cleanly isolate the increase in opacity due to radiation fields at high redshift .
there are two large uncertainties .
the first is whether glast will be able to see high - redshift blazars at all .
however , in the unified model of agn , the scarcity of ultra - luminous blazars is a geometrical effect ( due to relativistic beaming ) rather than a requirement of extremely large black hole masses .
in fact , the luminosity boost provided by beaming reduces the black hole mass by several orders of magnitude below that demanded by the eddington limit .
so it is at least plausible that high redshift blazars will be detectable .
the second uncertainty is whether absorption seen in a blazar will be internal , rather than due to pair production against photons in the igm .
however , at gev energies we have some physical understanding of the observed egret spectra ( e.g. , ghisellini et al 1998 ) ; opacity to gamma - ray photons due to internal radiation fields can be constrained by time variability arguments and other constraints .
furthermore , glast should assemble an extremely large catalog ( @xmath31 few thousand ) of low redshift blazars , whose spectra can be studied in detail ( and the contribution to opacity due to low redshift star formation can be quantified by other means ) ; provided blazar properties do not evolve too strongly with redshift , we should have a firm handle on the intrinsic unabsorbed blazar spectrum .
i thank david spergel for his advice and encouragement , youjun lu for helpful conversations on blazar properties , bruce draine for helpful comments , and bart pindor for technical assistance .
i thank the institute of theoretical physics , santa barbara for its hospitality during the completion of this work .
this work is supported by the nasa atp grant nag5 - 7154 , and by the national science foundation , grant number phy94 - 07194 . | the next generation of gamma - ray telescopes may be able to observe gamma - ray blazars at high redshift , possibly out to the epoch of reionization .
the spectrum of such sources should exhibit an absorption edge due to pair - production against uv photons along the line of sight .
one expects a sharp drop in the number density of uv photons at the lyman edge @xmath0 .
this implies that the universe becomes transparent after gamma - ray photons redshift below @xmath1 .
thus , there is only a limited redshift interval over which gev photons can pair produce .
this implies that any observed absorption will probe radiation fields in the very early universe , regardless of the subsequent star formation history of the universe .
furthermore , measurements of differential absorption between blazars at different redshifts can cleanly isolate the opacity due to uv emissivity at high redshift .
an observable absorption edge should be present for most reasonable radiation fields with sufficient energy to reionize the universe .
ly@xmath2 photons may provide an important component of the pair - production opacity .
observations of a number of blazars at different redshifts will thus allow us to probe the rise in comoving uv emissivity with time . |
for quite some time , kaons have attracted the attention of nuclear physicists .
the reason is that strangeness , a quantum number conserved in strong interactions , attributes a special role to the kaons among the possible projectiles for investigating nuclear structure .
kaons have two properties which make them unique tools .
firstly , they can transfer a new degree of freedom to the nucleus , and secondly , in contrast to pions they come in two forms , kaons ( @xmath10 ) and antikaons ( @xmath11 ) which differ substantially in their interaction with the nucleus .
kaons with their quark content @xmath12 ( @xmath13 ) have strangeness s=1 ; however , nuclear states involving strangeness can only contain hyperons ( @xmath14 , @xmath15 ) which have the same strangeness ( s=1 ) as antikaons ( quark content @xmath16 or @xmath17 ) .
therefore , the interaction between kaons and nuclei is rather weak , which is demonstrated , _
e.g. _ , by the large @xmath18 nuclear mean free path of about 57 fm for @xmath19 . consequently , the @xmath18 meson is a suitable probe for investigating the interior region of nuclei . on the other hand
, antikaons have s=1 and their absorption in nuclei can easily produce hypernuclei containing @xmath4 or @xmath5hyperons .
since such processes occur with sizeable reaction probability the region of investigation is probably restricted to the nuclear surface .
the successful use of kaons in nuclear structure requires the precise knowledge of the interaction mechanism of kaons with nuclei .
each uncertainty in the theoretical description necessarily leads to uncertainties in the interpretation of empirical results . since each theoretical model for the kaon
nucleus interaction starts from the free kaon
nucleon interaction and then adds medium modifications in one way or another , a precise knowledge of the free interaction is absolutely essential .
recently @xcite we have presented a meson exchange model for the @xmath9 interaction , which provides a reasonable description of the empirical @xmath9 scattering data for laboratory momenta smaller than 0.8 gev / c . for model b of ref .
@xcite the diagrams included are shown in figs .
1(a ) and 1(b ) ; they have been evaluated in time
ordered perturbation theory . by @xmath20
we denote a very short ranged , phenomenological , repulsive contribution , which has the analytical form of scalar @xmath21exchange with opposite sign and with an exchanged mass of 1.2 gev .
this additional repulsion is required if the @xmath22coupling constants @xmath23 , @xmath24 are restricted to their su(6 ) value .
an important ingredient is the contribution arising from @xmath21 and @xmath8exchange : @xmath21exchange provides the dominant part of the intermediate range attraction and @xmath8exchange determines , to a large part , the isospin dependence of the interaction . despite their importance
, however , these pieces have been treated so far in a very rough way . in both cases
a sharp mass has been used which means their appreciable decay widths have not been taken into account .
the @xmath21 meson is a fictitious particle not observed in nature so both its mass and coupling constant are therefore free parameters . in the case of the @xmath8 meson
the coupling strength is obtained as a product of a coupling constant at the @xmath25 vertex , which is taken from the bonn potential @xcite and the coupling constant at the @xmath26 vertex , which is calculated via su(3 ) relations from the ( empirically known ) @xmath27 coupling constant . whether this procedure provides
a reliable result is doubtful since i ) the @xmath8meson coupling constants used in the bonn potential are questionable @xcite and ii ) su(3 ) relations are not necessarily valid for unstable particles
. in this simple model , @xmath21 and @xmath8exchange essentially stand for the correlated @xmath28exchange contribution in the @xmath29 ( @xmath21 ) and @xmath30 ( @xmath8 ) channel , as illustrated in fig .
the purpose of the present paper is to derive this contribution starting from a microscopic model for the t
channel reaction @xmath31 with @xmath32 ( and @xmath33 ) intermediate states and using a dispersion relation over the unitarity cut .
this realistic model of ( effective ) @xmath21 and @xmath8exchange is then used to reconstruct an extended meson exchange model for @xmath34 scattering .
such a microscopic description of correlated @xmath28exchange is essential not only for an adequate judgment of the quantitative role of meson exchange in free @xmath35 scattering , but also for the calculation of kaon nucleus scattering processes .
effects of medium modifications of meson masses @xcite inevitably require an explicit , realistic model for correlated @xmath28exchange .
in sect.2 we outline the basic formalism .
sect.3 contains the essential features of our model for the @xmath31 transition .
in sect.4 we present the results for the correlated @xmath0 contribution in @xmath34 scattering , in terms of suitably defined effective coupling constants .
furthermore , we compare @xmath34 phase shifts and observables derived from the extended model with those obtained before .
sect.5 contains some concluding remarks .
in this section we outline the formalism which we use in order to derive the correlated @xmath0exchange contribution for the @xmath34 interaction .
the procedure is similar to that which was used in @xmath36 scattering @xcite .
the scattering amplitude @xmath39 is related to the @xmath40 matrix by @xmath41 where @xmath42 ( @xmath43 ) is the total four momentum in the initial ( final ) state , @xmath44 , and @xmath45 with @xmath46 ( @xmath47 ) the nucleon ( kaon ) mass . in the s channel ( @xmath48 ) @xmath39 can be written as @xmath49 here , the dirac spinor @xmath50 with the normalization @xmath51 describes a nucleon with helicity @xmath52 and three
momentum @xmath53 , while @xmath54 ( @xmath55 ) is the isospin wave function of a nucleon ( kaon ) .
the operator @xmath56 acts in spin and isospin space and depends on the two independent mandelstam variables @xmath57 and @xmath58 .
the third variable @xmath59 is related to @xmath60 and @xmath61 by @xmath62 .
the scattering operator @xmath63 has the following isospin structure : @xmath64 where @xmath65 ( @xmath66 ) is the isospin operator for the nucleon ( kaon ) and @xmath67 \ , , \label { t - chan}\ ] ] with @xmath68 , @xmath69 , and @xmath70 being the four dimensional unit matrix .
the corresponding t channel ( @xmath2 ) amplitude is given by @xmath71 with @xmath72 , @xmath73 and @xmath74 ( @xmath75 ) the dirac spinor ( isospin state ) of an antinucleon .
the mandelstam hypothesis states now that @xmath63 ( and therefore @xmath76 , @xmath77 in eq.([t - chan ] ) ) is the same function of s and t as in eq.([s - chan ] ) , but in a different kinematical domain , _
i.e. _ for @xmath78 , @xmath79 , and @xmath80 . in order to isolate the @xmath21 and @xmath8 contribution we have to perform a partial wave decomposition of the amplitudes in the t channel : @xmath81 ( the same for @xmath77 ) .
the @xmath82 are the legendre functions and @xmath83 .
the scattering angle in the t
channel , @xmath84 , can be expressed in terms of s and t : @xmath85 conservation of parity and g parity demands that the sum of spin and isospin must be even in case of a @xmath0 intermediate state .
therefore @xmath86 , @xmath87 ( @xmath88 , @xmath89 ) will vanish for even ( odd ) @xmath90 if we only consider @xmath0 intermediate states
. this argument does not hold for an intermediate @xmath1 state ; there @xmath88 , @xmath89 ( @xmath86 , @xmath87 ) are also possible for odd ( even ) @xmath90 .
however in our model these amplitudes turn out to be negligibly small and can be safely neglected .
then the isospin index @xmath91 can be suppressed since @xmath90 determines uniquely the isospin state .
since the @xmath92 and @xmath93 amplitudes contain kinematical singularities , one has to define new amplitudes @xmath94 , which are free of these singularities @xcite . here the index @xmath95 denotes the helicity of the @xmath96 state : @xmath97 , @xmath98 .
( @xmath99 . ) in terms of these amplitudes @xmath92 and @xmath93 can be written as @xmath100 with @xmath101 and @xmath102 ( @xmath103 ) in the c.m .
system of the t channel process .
the @xmath21 ( @xmath8 ) exchange contribution , defined as the correlated @xmath0exchange in the scalar ( vector ) t channel , is identified as the @xmath104 ( @xmath105 ) term of eq.([ab - amp ] ) .
one now can perform the analytic continuation of the @xmath106 amplitudes to physical t values in the s channel ( @xmath107 ) , which requires knowledge of the cut structure in the complex _
t _ plane .
the right hand ( unitarity ) cut runs from @xmath108 to @xmath109 , whereas the left hand cut , determined by @xmath4 exchange runs from @xmath110 to @xmath111 .
this means that the cuts overlap . in this overlap region
the baryon exchange born term of the reaction @xmath2 has an imaginary part . as we are only interested in the correlated @xmath0 contribution ( which we call @xmath112 in the following ) we perform the dispersion relation for @xmath94 only over the right hand cut , leaving out the born term .
then we get for @xmath21exchange the following contributions to the invariant amplitudes in the s channel : @xmath113 ( for convergence reasons , see ref .
@xcite , the dispersion relation has to be performed for @xmath114 , @xmath115 . )
similarly we obtain for @xmath8exchange @xmath116 here s and t have to assume physical values of the s channel , _
i.e. _ , @xmath117 and @xmath107 .
these amplitudes can be interpreted as meson exchange potentials for which the meson mass @xmath118 is distributed over the range from @xmath119 to @xmath109 .
the corresponding coupling constants depend on the mass ; they are proportional to the spectral functions @xmath120 , @xmath121 .
therefore , these spectral functions determine the dynamical behaviour of the exchanges , _
i.e. _ , they characterize the strength as well as the range of these potentials . concerning @xmath8exchange , it was pointed out in refs .
@xcite that there is a considerable uncertainty in the results .
the reason is that eq.([ab - rhof ] ) disperses the helicity amplitudes directly ; alternatively , one ( @xcite ) can first construct combinations @xmath122 corresponding to vector ( @xmath123 ) and tensor ( @xmath124 ) coupling amplitudes , where @xmath125 and then perform the dispersion integral , which yields @xmath126 both methods would be equivalent if the dispersion integrals could be performed over both the complete left hand and unitarity cut .
however , @xmath8exchange is customarily defined via a dispersion integral over the unitarity cut only .
indeed , the additional t dependence in @xmath127 apart from the t dependence provided by the helicity amplitudes @xmath112 causes the results to be quite different , as will be discussed later . based on eqs .
( [ s - chan ] ) ( [ t - chan ] ) the correlated @xmath0exchange contributions to the @xmath34 scattering amplitude can then be written as @xmath128 u(\vec p , \lambda ) \ } \vec \tau_n \cdot \vec \tau_k \ ; , \label { t - sigrho}\end{aligned}\ ] ] omitting isospin states for convenience . in the c.m .
system of the @xmath48 reaction @xmath129 , @xmath130 , and @xmath131 . since we will treat @xmath132 later as potentials to be iterated in the scattering equation belonging to time
ordered perturbation theory , we apply an off - shell extrapolation of the dispersion relations in the following way : we first replace the denominator @xmath133 in analogy to the time ordered propagator @xmath134 ( @xmath135^{\frac { 1}{2}}$ ] , @xmath136 is the total c.m .
energy ) , which is motivated by the fact that the dispersion integral sums over exchanges of particles with mass @xmath137 . for the additional t dependence in eqs.([ab - rhof],[ab - rhog
] ) we keep @xmath138 and @xmath139 .
note that this does not change the on shell result .
in addition we have to add phenomenological cutoffs in order to generate sufficient convergence in the scattering equation .
again we interpret the correlated @xmath0 potentials as generated by exchange of particles with mass @xmath118 and define a formfactor @xmath140 which is squared under the dispersion integral .
this is analogous to the monopole type form factor we use at each vertex of the ordinary meson and baryon exchange potentials .
we use @xmath141 mev for the model where we perform the dispersion integral for @xmath142 ( @xmath143 ) .
one should realize that this procedure modifies the original on shell result somewhat .
however , it lies in the range of uncertainties which are inherent in the whole procedure , which are discussed further in sect.4 . with the above extensions the amplitudes in eq.([t - sigrho ] ) have a well defined off
shell behaviour with a sufficient fall off for high momenta .
corresponding potential matrix elements , @xmath144 , acquire an additional factor @xmath145 , _
i.e. _ @xmath146 .
unitarization then leads to the scattering amplitude , _
i.e. _ @xmath147 where @xmath148 contains contributions from the diagrams shown in fig . 1 , except that now the @xmath21 and @xmath8 exchange potentials are replaced by the correlated @xmath28 exchange potentials discussed here
in the last section we have outlined a method of obtaining the correlated @xmath0exchange contribution to the @xmath48 scattering amplitude from the @xmath2 partial wave helicity amplitudes @xmath94 . since , unlike for the @xmath149 case ( ref .
@xcite ) , we can not rely on quasiempirical information , we have to provide a field theoretic model for the @xmath2 amplitudes .
anyhow , such a dynamical model has definite advantages when medium modifications of the @xmath34 interaction are considered since it facilitates future investigation of not only possible medium effects due to changes in the kaon and nucleon propagators , but also in the @xmath2 interaction itself .
we will generate the amplitude for the process of fig .
1(c ) in the t channel by solving the scattering equation in the blankenbecler sugar ( bbs ) @xcite reduction scheme : @xmath150 where @xmath151 here @xmath152 is the transition interaction from @xmath96 to @xmath153 , @xmath154 are transition amplitudes from @xmath155 and @xmath1 to @xmath1 , and @xmath156 is the free two particle green s function for the _ aa _ intermediate state .
the ingredients of the dynamical model for the transition interactions @xmath157 and @xmath158 are shown in fig . 2 .
the potential @xmath159 ( @xmath158 ) consists of @xmath160 and @xmath3 ( @xmath4 and @xmath5 ) exchange terms plus @xmath8 meson pole diagrams . @xmath161 and @xmath162 are obtained from the driving terms shown in fig .
such a model involving the coupled channels @xmath155 and @xmath1 was constructed by our group @xcite based on time
ordered perturbation theory . here ,
as in a recent study of @xmath36 scattering @xcite , we use a model with essentially the same physical input , which alternatively uses the bbs scheme .
the description of the data turns out to be as successful as in ref .
for more details , the reader is referred to @xcite .
we stress that all parameters are predetermined : @xmath161 and @xmath162 , through the coupled channel calculation , are fixed by @xmath155 while both transition potentials are determined by the quasiempirical @xmath149 information , cf .
@xcite . in the c.m . system and in helicity representation eq.([tchant ] ) after a partial wave expansion becomes @xmath163 with @xmath164 the @xmath2 on shell amplitudes are related to the helicity amplitudes @xmath94 via @xmath165 with @xmath166 fig .
4 shows the results for @xmath167 and @xmath168 in the pseudophysical region ( @xmath169 ) , needed as input for eqs.([ab - sigma ] , [ ab - rhof ] , [ ab - rhog ] ) .
as expected , the spectral function in the @xmath8channel shows a resonant structure with a maximum at about the @xmath8mass . in the @xmath21channel
the spectral function is much broader than for the @xmath8 ; compared to the @xmath170 case ( ref .
@xcite ) it is weaker and the peak is shifted somewhat to a higher mass ( see also the discussion later ) .
furthermore , the inclusion of intermediate @xmath1 states leads to a sizable enhancement of @xmath167 whereas its effect is negligible in the @xmath8channel .
this is due to the fact that the @xmath1 interaction is weak in the vector but strong in the scalar channel .
in fact , as discussed in refs.@xcite and @xcite , the @xmath171 interaction generates a @xmath171 bound state , the @xmath172 meson .
this state clearly has a strong effect on the shape of @xmath167 .
based on the spectral functions of the last section we can now evaluate , in a first step , the ( on shell ) invariant amplitudes with the help of eqs.([ab - sigma ] , [ ab - rhof ] , [ ab - rhog ] ) . in practice , the integrals have been evaluated up to @xmath173 , _
i.e. _ , in a region in which the dynamical model can be trusted .
it is instructive to parametrize the result by sharp mass @xmath21 and @xmath8exchange with appropriate t
dependent coupling constants which can be compared with those used in our former model .
let us start with the @xmath21channel . from the lagrangians used in ref .
@xcite @xmath174 we get for the invariant amplitude arising from @xmath21exchange : @xmath175 with @xmath176 .
if we now parametrize the result of our correlated @xmath0exchange potential ( eq.([ab - sigma ] ) ) in this form ( allowing @xmath177 to be t
dependent ) we will get for the effective coupling constant @xmath178 the result ( with @xmath179 gev ) is shown in fig . 5 , together with the result of an updated model of ref .
@xcite , hereafter referred to as model i , based on the diagrams in figs .
1(a ) and 1(b ) , with the same value of @xmath180 .
( details of this model will be given below . ) obviously our model predictions for the correlated @xmath0exchange in the scalar channel is in rough agreement with the @xmath21strength used before , which was phenomenologically adjusted to empirical @xmath35 data .
however , our new result has a sizable t dependence , which demonstrates clearly that it can not be well approximated by sharp mass @xmath21exchange with @xmath179gev .
since it grows with @xmath181 , the contribution is shorter ranged , in complete consistency with the behaviour of the spectral function in fig .
therefore , the effective mass exchanged should be higher . indeed , if we use @xmath182gev in eq.([coup - sig ] ) the result for @xmath183 , shown in the dashed curve of fig .
5 , has almost no t dependence .
note that the analogous effective @xmath21mass for the @xmath36 system @xcite is @xmath184gev .
let us now go to the @xmath8channel .
starting from the lagrangians for sharp mass @xmath8exchange @xmath185 we get for the invariant amplitudes : @xmath186 with @xmath187 , @xmath188 .
parametrizing again our correlated @xmath0exchange result ( eq.([ab - rhof ] ) resp .
eq.([ab - rhog ] ) ) in this form we obtain for the effective @xmath8 coupling constants @xmath189 ( note that according to eqs.([ab - rhof ] , [ ab - rhog ] ) the s - dependence drops out . ) the results for @xmath190 and @xmath191 based on @xmath192 mev are shown in fig . 6 , again together with values used in model i.
first we observe that there is a remarkable difference between the two alternatives for doing the dispersion relation in the @xmath8channel , _ i.e. _ , eq.([ab - rhof ] ) on one hand and eq.([ab - rhog ] ) on the other : the second method provides larger tensor ( @xmath191 ) but smaller vector coupling ( @xmath190 ) .
consequently the ratio @xmath193 characterizing the tensor to vector coupling ratio of the ( effective ) @xmath8 to the nucleon , @xmath194 is much larger ( @xmath195 ) for the second than for the first choice ( @xmath196 ) .
note also that the results based on the second choice are almost t independent ; therefore , the result can be well identified with an exchange of an ( effective ) @xmath8 meson with the empirical mass . on the other hand ,
the first method yields a result with a non
negligible t dependence , with opposite behaviour for @xmath190 and @xmath191 .
therefore , given the lagrangians in eq.([l - rho ] ) , a common mass can not be assigned to the @xmath8channel result derived from the first choice .
the situation is similar to the @xmath36 case ( cf .
fig . 9 of ref .
the only difference is that there all results scale by a factor of ( roughly ) 3 , in distinct disagreement with the su(3 ) value of 2 for the @xmath197 ratio .
( note that there is an additional factor 2 due to the factor @xmath198 in the lagrangian used in @xcite . )
this discrepancy should be of no surprise since there is no reason to expect that such effective exchanges generated by correlated @xmath0 exchange should fulfill the symmetry relations for exchanges of genuine particles . in the sharp mass @xmath8model i
the @xmath193 ratio has been taken from the bonn @xmath199 model potential @xcite to be 6.1 , only slightly larger than the correlated result based on eq.([ab - rhog ] ) . on the other hand ,
the absolute values are much larger , by about a factor of 2 .
note however , that in the following calculation of @xmath34 phase shifts and observables this discrepancy in the physical t region is much reduced since in this model a @xmath200 ( and @xmath201 ) formfactor of monopole type with a cutoff mass of only about twice the @xmath8mass are introduced , which suppress the @xmath8potentials by about a factor of 2 at t=0 . in this section
we confront our correlated @xmath0exchange model with the experimental @xmath9 data .
our starting point is model b of ref .
@xcite consisting of the diagrams shown in figs .
1(a ) and 1(b ) .
the baryon exchange diagrams in fig .
1(b ) are now treated in an improved way .
first , both time orderings are included instead of only one ( cf .
1(b ) of ref .
@xcite ) . second , pseudovector coupling is used at the nyk vertex , _
i.e. _ @xmath202 third , in case of @xmath203exchange , an extended spin3/2 propagator is taken .
finally monopole formfactors ( eq.(2.17 ) of ref .
@xcite ) are used throughout , except for @xmath204 and @xmath205 vertices , where a dipole formfactor is used .
the new expressions for the potential matrix elements are given in the appendix .
in addition , we change @xmath206 to its experimental value @xmath207 , instead of its quark model value used before .
correspondingly , via su(6 ) relations , @xmath208 now becomes 32.95 .
as in ref .
@xcite , most parameters ( coupling constants , cutoff masses ) are predetermined : coupling constants ( with the exception of @xmath206 and @xmath209 ) and cutoff masses belonging to @xmath199 and @xmath210 vertices are taken to be precisely the same as those of the ( full ) bonn @xmath199 potential @xcite . coupling constants at the vertices involving strange baryons ( @xmath211 , @xmath212 , @xmath213 )
have been related by the assumption of su(6 ) symmetry to the empirical @xmath214 coupling , as in our hyperon nucleon model @xcite .
the three meson coupling constants have been determined from the empirical @xmath215 coupling , assuming the same symmetry scheme and ideal mixing .
the value of @xmath216 and some cutoff masses have been slightly readjusted to the empirical @xmath9 phase shifts below pion production threshold .
this defines our model i. the values of parameters used in this model are given in table 1 ; the phase shifts which it yields are shown in the dotted curves of fig . 7 .
obviously , the model based on phenomenological sharp mass @xmath21 and @xmath8exchange provides a fair description of the empirical situation , with slight improvements in the @xmath217 and @xmath218 phase shifts compared to ref .
@xcite .
we now replace the sharp mass @xmath21 and @xmath8 contributions by the correlated @xmath0exchange potentials based on fig .
1(c ) ( model ii , a = eq.([ab - rhof ] ) , b = eq.([ab - rhog ] ) ) evaluated off shell using eq.([prop ] ) and including formfactors ( eq.([formfac ] ) ) . in order to avoid
double counting we have then to omit the box diagrams in fig .
1(b ) involving two pions since they are already included in fig .
1(c ) . after a slight readjustment of some parameters , cf .
table 2 , we obtain phase shift results shown likewise in fig .
7 . as expected from figs . 5 and 6
, some discrepancies occur between the various models .
the main point however is that @xmath9 interactions based on a microscopic evaluation of correlated @xmath0 exchange are able to provide a reasonable description of empirical phase shifts .
since the existing phase shift analyses have large error bars and are , in some cases , even contradictory it is instructive to examine the experimental observables directly .
8 shows our model predictions for the elastic cross sections in the relevant momentum range , while figs .
911 show the differential cross sections and polarizations at some selected momenta .
all models are in good agreement with experimental data .
again , there are slight differences between the various model results .
the differential cross sections for @xmath219 suggest an almost complete absence of p waves , which is best realized in the model involving correlated @xmath0exchange evaluated according to eq.([ab - rhof ] ) .
finally , table 3 presents scattering lengths @xmath220 ( i=0,1 ) and effective range @xmath221 of our models , which are in overall agreement with the empirical values .
in this paper we have presented a microscopic model for correlated @xmath0 ( and @xmath1)exchange between kaon and nucleon , in the scalar
isoscalar ( @xmath21 ) and vector isovector ( @xmath8 ) channels .
we first constructed a model for the reaction @xmath2 with intermediate @xmath0 and @xmath1 states , based on a transition in terms of baryon ( @xmath222 ) exchange and a realistic coupled channel @xmath223 , @xmath224 and @xmath225 amplitude .
the contribution in the s channel is then obtained by performing a dispersion relation over the unitarity cut . in the @xmath21channel , the result can be suitably represented by an exchange of a scalar particle with a mass of 0.75 gev .
the strength turns out to be in rough agreement with the strength of phenomenological @xmath21exchange used before @xcite , which has been adjusted , together with other diagrams , to empirical @xmath9 data . in the @xmath8channel ,
considerable ambiguities exist , of the same structure as in the @xmath36 case , depending on how the dispersion integral is performed . in terms of effective coupling constants ,
the results differ strongly from the values used before in our phenomenological @xmath8exchange amplitude , which were determined from the bonn @xmath199 potential and su(3 ) relations .
however , the @xmath8amplitudes actually used are quite similar since the formfactor applied in the phenomenological @xmath8exchange of model i brings the contribution close to the results of the dispersion
theoretic results .
this model for correlated @xmath0exchange has been suitably extrapolated off
shell , and supplemented by short range terms ( generated partly by conventional @xmath22exchange ) and box diagrams involving @xmath226 and @xmath227exchange developed before .
a satisfactory description of the empirical situation is achieved , of the same overall quality as obtained before using phenomenological sharp mass @xmath21 and @xmath8exchange .
such an explicit model for correlated @xmath0exchange has not only conceptual advantages compared to a phenomenological treatment in terms of @xmath21 , @xmath8exchange , but also offers the possibility to study medium modifications of the @xmath34 amplitude in a well defined way a topic of high current interest .
* acknowledgments * one of the authors ( k. h. ) wishes to acknowledge the ministerio da educacion y ciencia for financial support ( sab940218 ) making his stay in valencia possible .
bcp is grateful for the support of the australian research council .
part of the work has been done at the institut de physique nucleaire duniversit claude bernard lyon i , supported by the daad program procope .
m. h. wants to thank guy chanfray for his hospitality .
this work was also partially supported by the daad program hspii / aufe .
for the baryon exchange diagrams we get the following potential matrix elements : @xmath228 where the isospin factors are @xmath229 and @xmath230 . in case of @xmath203exchange
we get @xmath231 with @xmath232 . by @xmath233
we denote the four momentum of the ingoing ( outgoing ) nucleon , by @xmath234 the four - momentum of the ingoing ( outgoing ) kaon , whereas @xmath235 stands for the four momentum of the exchanged hyperon .
we choose @xmath236 with @xmath237 , @xmath238 . | a dynamical model for s and p wave correlated @xmath0 ( and @xmath1 ) exchange between a kaon and a nucleon is presented , starting from corresponding @xmath2 amplitudes in the pseudophysical region , which have been constructed from nucleon , @xmath3isobar and hyperon ( @xmath4 , @xmath5 ) exchange born terms and a realistic meson exchange model of the @xmath6 and @xmath7 amplitude .
the contribution in the s channel is then obtained by performing a dispersion relation over the unitarity cut . in the @xmath8channel ,
considerable ambiguities exist , depending on how the dispersion integral is performed .
our model , supplemented by short range interaction terms , is able to describe empirical @xmath9 data below pion production threshold in a satisfactory way . |
one major challenge in nonequilibrium many - body physics is to identify situations with a sufficient degree of universality , i.e. phenomena that occur independently of a precise microscopic realization and are largely insensitive to the specific choice of initial conditions .
several situations have been discussed in the literature , where many - body systems evolve in time towards nonequilibrium steady states .
the most prominent example in condensed matter is certainly the electron gas exposed to a bias voltage @xcite . in this context , truly many - body properties , such as the effect of nonequilibrium conditions on quantum critical points , have been investigated @xcite .
further implementations of nonequilibrium many - body systems are discussed in the context of exciton - polariton bose - einstein condensates @xcite , or more recently using driven noisy systems of trapped ions or dipolar atomic gases @xcite .
as far as the time evolution of many - body systems is concerned , for closed systems we have seen a plethora of studies of quench dynamics @xcite , thermalization @xcite , and pre - thermalization @xcite , as well as transport @xcite .
further dynamical studies involve situations of crossing quantum critical points in a finite time , in the spirit of the kibble - zurek mechanism @xcite , or investigations of the many - body landau - zener effect @xcite .
the quantum optics toolbox for the manipulation of cold atomic systems @xcite not only offers the possibility to tailor the coherent dynamics of a closed many - body system by hamiltonian engineering , but also to implement the dissipative dynamics generated by a liouville operator @xcite .
of particular interest are liouville operators characterized by the existence of a unique dark state , i.e. a sink state whereto the system is asymptotically driven , independently of the initial configuration . in this case , it has been shown theoretically that the action of the sole dissipation allows to reliably prepare the system in peculiar states with interesting quantum mechanical features , such as phase coherence in bosonic systems @xcite , entanglement in spin systems @xcite , or delocalized pairing with arbitrary symmetry in fermionic systems @xcite . since a dark state can be reached by dissipative dynamics but not left , such processes featuring a dark states clearly violate the principle of detailed balance and
hence may drive the system into a steady state quite different from thermodynamic equilibrium .
the existence of a dark state with known properties nevertheless promises a sufficiently universal nonequilibrium situation with a well - defined steady state .
furthermore , since dissipation needs not be a small perturbation , but on the contrary can be made the dominant contribution to the dynamics , it is natural to investigate the effects of the interplay between unitary and dissipative dynamics . in thermodynamic equilibrium
, the competition of two non - commuting microscopic operators leads to a phase transition if the ground states have different symmetries when one or the other operator dominates @xcite .
a seminal example in the context of cold atoms in optical lattices is the superfluid - mott insulator transition in the bose - hubbard model @xcite . by analogy with equilibrium physics
, one may similarly expect a phase transition in the steady state of the system , resulting from the competition of unitary and dissipative dynamics . in this paper
, we demonstrate a phase transition in the steady state of a driven - dissipative bosonic system on a lattice and discuss its properties in detail .
we point out several features of the phase transition that make it a genuine nonequilibrium phenomenon , not fitting into the well - developed theory of equilibrium quantum phase transitions .
however , we recognize some properties that may hold universally , i.e. that may be shared by entire classes of similar systems , in the nonequilibrium steady state or in the time evolution .
the profound difference between the phenomena that we consider here and the equilibrium quantum phase transition manifests itself , at a purely technical level , in the absence of a known variational principle for the steady state .
the lack of such formal tool makes it necessary to evaluate the real - time evolution of the system , which is rather impractical in the many - body case .
here we develop a mean field method to treat bosonic lattice systems , subject to unitary and dissipative dynamics , which generalizes the gutzwiller ansatz and allows us to describe mixed states . within this approximation scheme , we deal with an effective master equation that acts on a reduced hilbert space , is nonlinear in the quantum state , and is amenable of a workable numerical solution .
moreover , we identify an analytically tractable low - density limit where the most relevant information on the quantum state is encoded in a few correlation functions , which obey coupled equations of motion . the solution of the equations of motion , which we provide for various dynamical regimes in this approximation , explains qualitatively the behavior of the system at arbitrary density .
the method may also be used for semi - quantitative estimates in other problems where unitary and dissipative dynamics appear on equal footing .
the paper is organized as follows . in sec .
[ sec : themodel ] we introduce the microscopic model and provide a heuristic argument to justify the existence of a nonequilibrium phase transition in the model . in sec .
[ sec : meanfield ] we introduce the generalized gutzwiller mean field approximation , discuss the most prominent properties of the homogeneous phase diagram resulting from it , and point out profound differences with respect to equilibrium phase transitions in dissipative environments @xcite .
[ sec : smalldens ] specifies the gutzwiller approximation to the case of low particle density and complements the numerical findings of sec .
[ sec : meanfield ] with analytical results . sec .
[ sec : instability ] is devoted to the analysis of a dynamical instability that arises spontaneously in the nonequilibrium phase diagram .
this completes the discussion of the steady state of the driven dissipative system . in sec .
[ sec : revival ] we focus on aspects of nonlinear dissipative dynamics that occur in mesoscopic settings .
we draw our conclusions in sec . [
sec : conclusions ] .
the appendices [ app : meanfield][app : eqmotion ] provide technical details on the derivation of the mean field theory and on the numerical procedures used to solve the resulting nonlinear equations .
this paper refines the findings that we already presented in ref .
here we provide a more detailed presentation of the material , including the discussion of the complete equations of motion and the explanation of numerical techniques .
new results include the exact solution of the dynamics for a small lattice ( sec .
[ sec : themodel ] ) , the solution of the dynamically unstable equations of motion beyond linear response ( sec .
[ ssec : nonlineresponse ] ) , as well as the investigation of the nonlinear dissipative dynamics following phase quenches ( sec . [ sec : revival ] ) .
in this work we consider a many - body system described by a master equation @xmath0 + { \cal l}[\rho(t)]~.\ ] ] the master equation defines the time evolution of the density matrix @xmath1 of an ensemble of bosonic atoms in an optical lattice .
it consists of a unitary part , described by the hamiltonian @xmath2 , and a dissipative part , represented by the liouvillian @xmath3 .
the liouville operator arises from the coupling of the system to a bath , which is eliminated in the markov approximation to yield the effective evolution ( [ eq : masterequation ] ) .
the validity of the markov approximation , which neglects retardation effects in the bath , is rooted in the possibility of having strong separation of time scales for such quantum optical systems .
we emphasize that this gives rise to a microscopic model which is _ local in time_. therewith , the possibility of an accurate microscopic modeling for many - body systems in terms of a few parameters one of the key features of cold atomic many - body systems is extended from the hamilton operator to the dissipative dynamics as well .
the unitary dynamics for the bosonic atoms in a @xmath4-dimensional lattice is described by the familiar bose - hubbard hamiltonian @xmath5 where @xmath6 is the hopping amplitude , @xmath7 the interaction strength , and @xmath8 the chemical potential ( see sec . [
ssec : chempot ] and app .
[ app : lingeneral ] for a discussion of its role in the time - evolution of the system ) .
the operator @xmath9 ( @xmath10 ) annihilates ( creates ) a boson in the @xmath11th lattice site and @xmath12 indicates that the summation is restricted to nearest neighboring sites .
( @xmath11 is a vector of integers with the same dimension as the lattice . )
the liouvillian @xmath3 that we consider here was derived in ref .
@xcite . written in lindblad form
, it reads @xmath13 = \frac{1}{2}\kappa\ , \sum_{\langle \ell\ell'\rangle } [ 2\hat{c}_{\ell\ell'}\rho \hat{c}_{\ell\ell'}^{\dag } - \hat{c}_{\ell\ell'}^{\dag}\hat{c}_{\ell\ell'}\rho-\rho \hat{c}_{\ell\ell'}^{\dag } \hat{c}_{\ell\ell'}]~,\ ] ] where @xmath14 is the dissipation rate and the quantum jump operators @xmath15 , acting on pairs of neighboring lattice sites , are given by @xmath16 with @xmath17 .
we specialize to the case that the phase difference along the @xmath18 primitive directions @xmath19 of the lattice is a constant @xmath20 .
the quantum jump operators are composed of an annihilation part , which destroys a particular superposition of the bosonic wavefunction , and a creation part , which recycles the bosons after interacting with the bath .
while the choice of the phase in the annihilation part is crucial to the asymptotic behavior of the system ( see below ) , the phase in the creation part is a matter of convention and could also be omitted .
one possible implementation of these jump operators in optical superlattices is discussed in ref .
the key point of such choice for the jump operators is that their nullspace consists of a unique element @xmath21 for a fixed number of particles @xmath22 .
entry of the single - particle correlation matrix in momentum space as the interaction strength is varied , obtained from the solution of the exact eoms ( [ eq : masterequation ] ) for @xmath23 , @xmath24 , @xmath25 , and @xmath26 , at the time @xmath27 .
the inset shows the entire diagonal of the correlation matrix for @xmath28 ( shaded bars ) and @xmath29 ( empty bars ) . ]
it follows that @xmath30 is the unique dark state for the whole liouville operator and that the system is attracted to this state independently of the initial conditions , if no competition arises from the unitary part of the evolution . in refs .
@xcite it was shown that the state @xmath31 [ a bose - einstein condensate ( bec ) with momentum @xmath32 in lattice units ] is the unique dark state of the liouvillian ( [ eq : liouvillian ] ) .
the uniqueness can be easily understood in momentum space , where the annihilation part of @xmath33 reads @xmath34\ } \hat{b}_{\bf q}$ ] . since the creation part of @xmath15
does not have any eigenvalue on the hilbert space of @xmath35 particles , @xmath33 features a unique dissipative zero mode at the given momentum .
the existence of a dark state depends here on the jump operators being not hermitian .
it is important to remark , anyway , that the liouville operator conserves the total particle number @xmath36 , as the jump operators just redistribute a particle s superposition and thus fulfill @xmath37 .
however , the non - hermitian nature of the jump operators entails that the dissipative process represented by the liouvillian ( [ eq : liouvillian ] ) does not respect the principle of detailed balance , because the probability that the system is projected out of the dark state is zero .
the dissipative dynamics produced by the engineered liouvillian is necessarily very different from a thermalization process and hence the phase portrait of the system can not be understood in terms of a modified or perturbed equilibrium picture , but requires a genuinely new approach .
an alternative physical picture of the dynamics in the presence of the dark state is the following . on two neighboring sites along the @xmath38 direction ,
the component of the wave function that is not in the nullspace of @xmath39 is mapped onto the whole hilbert space , including the nullspace , while the nullspace is mapped onto itself .
asymptotically , the weight of the wave function outside the nullspace vanishes and the phase of the wave function between the neighboring sites is locked to @xmath40 . since the phase locking takes place on each pair of sites , the many - body state is asymptotically projected into the bec state , i.e. a state with long - range phase coherence .
the dissipative evolution purifies the initial state of the system and , in this sense , pumping into the bec bears strong analogies to laser cooling of single atoms @xcite .
the analogy to optical pumping is quite tight , because the implementation of the dissipative process involves a coherent ( laser ) driving element , such that the system is indeed driven - dissipative . ) for the same parameters as fig .
[ fig : exactcorr ] .
the inset shows the same data as the main panel ( crosses ) logarithmically rescaled with respect to the asymptotic value @xmath41 and compared to the power - law @xmath42 for @xmath43 ( solid line ) . ]
we start out by considering the effect of strong interactions on the steady state of the system . in the equilibrium phase diagram of the bose - einstein hamiltonian ( [ eq : bosehubbard ] )
, the quantum phase transition between the mott insulator and the superfluid is understood in terms of the competition between the kinetic term proportional to @xmath6 , that tends to delocalize the bosons to decrease the energy , and the interaction term proportional to @xmath44 , that is responsible for an increase of the energy when particles tunnel between sites and repel each other . to understand the interplay of the energy scale in our present setup , a novel point of view is required , because the time - evolution of the system does not minimize the energy .
the absence of a variational principle to characterize the steady state of the system is indeed a noteworthy difference with respect to the equilibrium theory .
still , we identify the existence of a competition between the energy scales of the interaction @xmath44 and the dissipation @xmath14 . to this end
, we proceed by adopting a rotating frame that eliminates the interaction from the hamiltonian .
the local operator that defines the new frame on each site is @xmath45}$ ] .
the annihilation operator in the new frame reads @xmath46 } \hat{b}_{\ell } = \sum_{n } \exp
{ [ i n u t ] } |n\rangle_{\ell}\langle n| \hat{b}_{\ell}~.\ ] ] the effect of the interaction is to rotate the phase of each fock state differently , destroying the phase coherence of the order parameter . hence we argue that increasing the interaction leads to a substantial depletion of the condensate and eventually to the disappearance of the long - range order in the system . before developing a suitable mean field approximation that allows us to tackle the question of the behavior of the system in the thermodynamic limit , we provide signatures of the competition between interaction and dissipation from exact numerical evaluation in a small system . for this purpose , we integrate numerically the equations of motion ( eoms ) ( [ eq : masterequation ] ) , in the hilbert space for @xmath23 particles in @xmath47 lattice sites with periodic boundary conditions , using a fixed - stepsize fourth - order runge - kutta method .
we use the final density matrix @xmath48 to compute the single - particle correlation matrix @xmath49 and the entropy @xmath50 $ ] , where @xmath51,@xmath52 are momenta in the discrete brillouin zone of the finite lattice .
[ fig : exactcorr ] shows that , as the interaction strength is increased , the occupation of the zero - momentum state decreases monotonically .
this is a signature , in the finite system , of the depletion of the condensate fraction that we expect in the thermodynamic limit .
we also see that the distribution of the momenta is localized in the absence of interactions and broadens substantially for @xmath53 , analogously to the broadening of the condensate peak produced by a finite temperature .
such broadening is not due to a rearrangement of the single - particle states , e.g. the momentum states being not the appropriate eigenbasis for the system , because , as we show in fig .
[ fig : exactentropy ] , the entropy steadily increases with the interaction strength .
this means that , in the presence of interactions , the liouvillian is no longer able to drive the system towards the pure dark state .
it is interesting to note that the entropy converges towards a definite constant value as the interaction strength grows , as is clear from the analysis in the inset of fig .
[ fig : exactentropy ] .
this fact suggests that the crossover exhibited by the finite system may in fact turn into a transition from the condensed state to a well - defined new phase . in the following section we validate this picture and provide a detailed analysis of the phase diagram
in order to capture the physics of the above depicted nonequilibrium phase transition on a semi - quantitative level , we develop a formalism which allows us to describe the physics of both the well - controlled limits of weak and strong interactions . in the weakly interacting limit @xmath54
we must recover the condensate phase with long - range order found in ref .
@xcite , while in the strong coupling limit @xmath55 we expect a disordered phase with diagonal density matrix . to this end
, we tackle the solution of eq .
( [ eq : masterequation ] ) within a mean field approximation , in the form of a gutzwiller - like ansatz where the system density matrix is factorized in position space @xmath56 the reduced density matrix @xmath57 on the site @xmath11 is obtained by tracing out all the the other degrees of freedom but those of the @xmath11th site .
the equation of motion for the reduced density matrix then reads @xmath58 + { \rm
tr}_{\neq \ell}{\cal l}[\rho(t)]~.\ ] ] the computation of the traces is reported in app .
[ app : meanfield ] .
the hamiltonian part results in a local commutator @xmath59 = -i[\hat{h}_{\ell}(t ) , \rho_{\ell}(t)]~,\ ] ] with the effective reduced local hamiltonian given by @xmath60 [ cf . also eq .
( [ eq : localhamiltonian ] ) ] .
this is precisely the form obtained in the ordinary gutzwiller approximation for a mean field ground state calculation on the bose - hubbard model .
note , however , that due to the restriction of the ground state , the state must be pure and thus the onsite density matrices in the latter case are of the form @xmath61 . in our case
we do not have an argument in favor of the purity of the steady state for strong interactions so we allow for mixed steady states as well .
the trace of the liouvillian part , assuming an axial geometry in the system ( see app .
[ app : meanfield ] ) , produces the structure @xmath62 & = & \kappa \sum_{r , s } [ 2 a_{\ell}^{r } \rho_{\ell } a_{\ell}^{s \dag } - a_{\ell}^{s \dag } a_{\ell}^{r } \rho_{\ell } - \rho_{\ell } a_{\ell}^{s \dag } a_{\ell}^{r } ] \nonumber \\ & \times & [ \gamma_{\ell , \sigma = + 1}^{r s } + \gamma_{\ell , \sigma = -1}^{r s } + ( z-2 ) \gamma_{\ell , \sigma = 0}^{r s}]~ , \end{aligned}\ ] ] where @xmath63 is the coordination number of the lattice , the vectors of operators @xmath64 are given in eq .
( [ eq : liouvdecomp ] ) , and the matrix of coefficients @xmath65 in eq .
( [ eq : motionlattice ] ) . at this point
we discuss the validity of the factorization approximation ( [ eq : factorization ] ) in real space .
first we note that the dimensionality of the system enters the eoms only through the coordination number @xmath63 of the lattice . in general , one expects higher reliability of mean field approximations in higher spatial dimensions .
for the ( fermionic ) hubbard model , arguments based on the central limit theorem for spatial dimension @xmath66 lead to the observation that mean field theory becomes exact in this limit @xcite .
the statistical nature of the argument suggest that a similar reasoning could be applied also here .
the final results of the analysis do not depend qualitatively on the precise value of @xmath63 hence , when the choice of a definite dimensionality is necessary , we consider for simplicity a one - dimensional ( @xmath67 ) system .
moreover we observe that the system is represented only locally as a quantum state , i.e. a vector or a density matrix in the local bosonic fock state of each site @xmath11 . any correlation function @xmath68 that involves the evaluation of operators on different sites simplifies to the product @xmath69 .
the connected part of the correlation function between different sites is entirely neglected .
however , the onsite correlations are evaluated exactly and thus this method is quite suitable to extract informations on a wide range of variation of @xmath70 , from the weak to the strong coupling limit . in the limit of weak coupling , additional arguments in favor of our approximation scheme
are available ( see sec . [
ssec : linresponse ] ) .
technically , the gutzwiller approximation transforms the full many - body system into a collection of coupled single - site systems , that are amenable of a much simpler analytical treatment .
moreover , in a numerical solution , the total dimension of the state vector of the system decreases from the full size of the hilbert space to @xmath71 times the size of the local hilbert space .
this makes much larger arrays numerically tractable . on the technical side ,
the price to pay for the decrease in size of the hilbert space is that the equation of motion becomes nonlinear in the state vector , contrary to the schrdinger and von neumann equations .
more precisely , the coefficients that enter the effective local hamiltonian @xmath72 and the matrix @xmath73 depend on a set of local correlation functions @xmath74 evaluated with the density matrix of the sites @xmath11 neighbors of @xmath75 .
the appearance of a nonlinear equation from an originally linear equation for the system density operator in the framework of a mean field approximation is actually familiar from e.g. the gross - pitaevski mean field theory for weakly interacting bose gases .
there , a nonlinear classical field equation for the condensate expectation value results from a mean field approximation on the original linear @xmath22-body schrdinger equation .
finally , the mean field equations conserve the average value @xmath76 of the total particle density but do not commute with the number operator .
we reiterate that the treatment of the correlation functions ( [ eq : corrset ] ) on each site is fully quantum mechanical , i.e. there is no approximation on the value of the average of any local operator . in particular , this procedure yields a theory that contains much more information than a gross - pitaevski equation [ see eq .
( [ eq : dissgpe ] ) and the following discussion ] , where also the product of local operators ( [ eq : corrset ] ) is factorized , so that all available information is encoded in @xmath77 .
the effect of the phase @xmath78 in the liouvillian is easily understood for a homogeneous state by considering the change of frame of reference @xmath79 .
in the new frame , the liouvillian maintains the same form , with @xmath80 and @xmath81 . in the effective hamiltonian , instead
, it is necessary to apply @xmath80 but also @xmath82 .
hence , we can reduce to considering the homogeneous states of the system with @xmath83 only , as the general case amounts to a mere reduction of @xmath6 if @xmath84 . for @xmath85
, however , the dark state of the liouvillian corresponds to a boosted condensate that is known to be dynamically unstable @xcite and consequently does not feature a stable homogeneous steady state . in this section
we study the steady states of eq .
( [ eq : motiontrace ] ) and validate the picture , suggested in sec .
[ ssec : competition ] , of an interaction - driven nonequilibrium phase transition that destroys the condensation in the system .
we first identify analytically two exact steady states , for vanishing and large interactions , and then provide numerical results that interpolate between these two cases .
let us consider first the pure coherent state @xcite @xmath86 is the index in the local fock space ) and the corresponding density matrix @xmath87 .
the liouvillian ( [ eq : liouvtrace ] ) vanishes exactly when applied to the homogeneous coherent density matrix , that provides hence a consistent dark state solution for the dissipative dynamics and pinpoints an exact steady state solution of the model at @xmath88 .
moreover , for this density matrix , all onsite correlation functions factorize in fock space . as a consequence
, all the information of the theory is encoded in the simplest one , namely the one - point function @xmath89 , which is precisely the order parameter ( [ eq : corrset ] ) of the condensate .
the equation of motion @xmath90 $ ] for the order parameter reads explicitly @xmath91 this equation constitutes a dissipative gross - pitaevski equation ( gpe ) and , indeed , the unitary part is identical to the standard lattice gpe .
if we choose a homogeneous state @xmath92 [ analogous to the bec state ( [ eq : becstate ] ) ] , the dissipative terms vanish and the equation reduces to @xmath93 , i.e. the system undergoes a global phase rotation that can be removed with a precise choice of the chemical potential @xmath94 . more generally , we will constantly take advantage of the possibility to fix the chemical potential and remove residual phase rotations , ensuring a time - independent steady state as illustrated in this simple case .
the role of the chemical potential in our nonequilibrium analysis is further discussed below in sec . [
ssec : chempot ] .
the coherent state is not a general solution of eq .
( [ eq : motiontrace ] ) precisely because a different phase frequency appears for each different correlation functions but only one can be removed with the chemical potential .
anyway , the similarity of the eq .
( [ eq : dissgpe ] ) to the gpe suggests that small values of the interaction strength @xmath44 do not change qualitatively the dynamics of the system , which is the result rigorously derived in ref .
@xcite within the bogoliubov approximation . on the sites @xmath95 of the lattice , according to the mean field eom ( [ eq : motiontrace ] ) , for @xmath96 , @xmath97 , @xmath98 , and @xmath83 .
the color of the surface represents the phase difference @xmath99 between neighboring sites ( white for @xmath100 , black for @xmath101 ) . ]
let us consider now the opposite limit of a diagonal density matrix state , in which the coherences are totally absent . according to the heuristic argument presented in sec .
[ ssec : competition ] , this situation is expected for large values of the interaction strength . in this case , the kinetic term of the hamiltonian drops out , since no off - diagonal order is present and thus @xmath102 .
the interaction term drops out as well , because the operator @xmath103 commutes with any state that is diagonal in the fock basis .
the only remaining energy scale in the system is the dissipation strength @xmath14 . finally , the matrix of coefficients in eq .
( [ eq : liouvtrace ] ) reads @xmath104 and the eom for a homogeneous system reads @xmath105 the latter equation of motion describes precisely a bosonic mode @xmath9 coupled to a bath with thermal occupation @xmath106 @xcite .
it is intriguing that in this case @xmath76 is a property of the system itself and hence we may say that the system acts as its own bath .
in particular , this effective thermal bath , which drives the dynamics , is not related to the physical bath that is necessary to implement the jump operators ( [ eq : jumpoperators ] ) . in the steady state @xmath107
the remaining energy ( rate ) scale @xmath14 drops out , and the only physical scale is the average particle number @xmath76 . the steady state solution is readily seen to be a thermal - like state @xmath108 we emphasize that this state emerges _ despite _ the fact that the system is driven out of thermodynamic equilibrium , not as a consequence of the latter . and @xmath83 .
( a ) stroboscopic plot of the time evolution of the condensate fraction for @xmath109 , starting from an initially fully condensed state .
( b ) comparison of the full numerical solution of the homogeneous system for @xmath110 to the analytical low density result .
( c ) critical point in the homogeneous phase diagram as a function of density for different values of @xmath97 , @xmath111 , @xmath112 , @xmath113 . ]
finally , we note that the nonlinearity of the physically motivated mean field approximation allows us to describe two distinct dynamical fixed points , namely one ordered state with finite off - diagonal expectation values such as @xmath89 , and one disorder phase with vanishing coherences , in some analogy to @xmath114-theory for weakly interacting boson systems .
we also emphasize that the breaking of the @xmath115 phase symmetry ( associated to the conservation of particle number ) takes place spontaneously in the ordered phase , since both hamiltonian and liouville operator in the original equations ( [ eq : masterequation][eq : liouvillian ] ) conserve the number of particles .
we turn now to the numerical integration of the eom ( [ eq : motiontrace ] ) , specified to a lattice with @xmath71 sites and periodic boundary conditions .
the integration is performed with a fourth - order runge - kutta algorithm , using a fixed stepsize typically taken @xmath116 @xcite .
the local hilbert space is truncated to a maximum particle number @xmath117 , substantially larger than the typical particle densities @xmath118 that we consider here . due to the periodicity of the lattice ,
the momenta are confined to a discretized brillouin zone such that the phase @xmath78 assumes the values @xmath119 , with @xmath120 .
for each site @xmath11 in the lattice , the execution of the integration returns the local correlation functions . to measure the degree of coherence along the array , we compute the phase differences @xmath99 between the first - order correlation functions on neighboring sites @xmath121 a state with definite momentum is such that @xmath99 is integer and constant along the array . a typical output of the numerical integration
is displayed in fig .
[ fig : condensation3d ] , where we show the condensate fraction @xmath122 as a function of time on a lattice with @xmath123 .
the array is initialized as a collection of thermal states ( [ eq : thrmstate ] ) , slightly perturbed with tiny random off - diagonal fluctuations .
the figure shows that the system is driven out of its initial thermal state , develops a uniform macroscopic condensate fraction @xmath124 and , at the same time , the fluctuations of the local correlations vanish and a definite phase difference @xmath125 is established throughout the array .
the remarkable property of the liouvillian ( [ eq : liouvillian ] ) , namely that local phase synchronization results in the establishing of long - range order @xcite , is visually clear in this picture . as the interaction strength @xmath44 is increased , for @xmath97 , @xmath83 .
the dashed line in the main panel corresponds to @xmath126 ; the dot - dashed line corresponds to @xmath127 .
the final time of the time - evolution is @xmath128 in the main panel and @xmath129 , @xmath130 , @xmath131 , @xmath132 in the inset , from the bottom to the top curve .
the arrow marks the critical point @xmath133 . ]
we use now the numerical integration of the eom ( [ eq : motiontrace ] ) to compute the asymptotic value of the order parameter as the interaction strength is increased . since we investigate the properties of the homogeneous steady state , it is sufficient to consider a single site ( @xmath134 ) .
we verify that at the end of the time evolution the infinity - norm @xmath135 converges towards zero . in panel
( a ) of fig . [ fig : depletion ] we plot the condensate expectation value as a function of increasing interaction strength for different , equally spaced times , starting the evolution from an initial fully condensed state ( @xmath136 ) .
we show that the asymptotic values of the order parameter interpolate continuously between the expected values @xmath137 for @xmath136 ( coherent state ) and @xmath138 ( thermal state ) .
similarly , the asymptotic values of the entropy @xmath139 $ ] shown in fig .
[ fig : entropy ] interpolate between @xmath140 for the pure state at @xmath136 and the value @xmath141 that corresponds to a thermal state with unitary average density . between the two extreme values ,
both quantities present a non - analyticity that develops in time , as the system converges better and better to the steady state .
this effect is clearly a qualitative modification over the smooth crossover that appears in the finite systems of fig .
[ fig : exactcorr ] and [ fig : exactentropy ] .
we may conclude that , in the thermodynamic limit described by the gutzwiller ansatz , the competition of phase - enhancing dissipation and dephasing due to the interaction leads to non - analytic behavior , i.e. a nonequilibrium phase transition . to reinforce the previous considerations
, we analyze the convergence of the system to the steady state in the proximity of the non - analyticity .
we verify in fig .
[ fig : critical ] that , in this case , the convergence to zero takes place with a power law @xmath142 , in contrast to the exponential convergence generically granted by the dissipative dynamics . ) for @xmath97 , @xmath143 , @xmath96 in linear ( main panel ) and logarithmic ( inset ) scale .
the dashed line shows a power - law decay @xmath144 . ] to be more precise , in the gutzwiller approach spatial correlations are neglected by construction and we correspondingly observe exponentially decaying correlations , except for the critical point @xcite . to introduce a relation between the time - evolution and the criticality of the system ,
we note that time sets an energy scale @xmath145 for the system , which may be viewed as an irrelevant coupling , corresponding to an attractive direction of the fixed point of the system tuned to criticality by choice of the relevant coupling @xmath70 .
indeed , we observe that the non - analyticity corresponding to the critical point develops only in the limit where this scale is removed . in the vicinity of the critical point
the condensate amplitude scales as @xmath146 , where @xmath147 is the real part of the largest eigenvalue of the system , vanishing at criticality .
there , the system shows polynomial behavior with universal critical exponent , defined as @xmath148 which evaluates to @xmath149 as deduced from fig .
[ fig : critical ] , independently of @xmath150 or the mean filling @xmath76 .
this is the behavior expected within mean field theory and it does not depend on the dimensionality of the system . while the mean field critical behavior is another clear signature of a second order phase transition taking place in our nonequilibrium system , it is not sufficient to characterize the true universality class of the system . for this purpose , the long wavelength
spatial fluctuations would have to be included in an appropriate way , which is not possible within a site decoupling gutzwiller approach . and app .
[ app : lingeneral ] for @xmath96 ( @xmath151 ) .
the red line is the analytical result in the limit @xmath152 and is repeated in the inset with a different choice of the coordinates for the axes . ] finally , we provide in fig . [
fig : phasediagram ] the position of the critical point as the hopping amplitude @xmath6 and the interaction strength @xmath44 are varied , for some values of the density @xmath76 . to draw the phase border with good accuracy ,
it is rather awkward to integrate eq .
( [ eq : motiontrace ] ) repeatedly for long times .
we resort instead to the analysis of the eoms linearized around the homogeneous thermal state .
the key observations here is that the thermal - like state ( [ eq : thrmstate ] ) , contrary to the coherent state ( [ eq : cohestate ] ) , is always an exact steady state of eq .
( [ eq : motiontrace ] ) .
its stability property , of course , are different in the two phases and this must appear in the spectrum of the linearized equations . in the thermal phase , the thermal state is an attractive fixed point , while it is at least a saddle point in the condensed state . around the stable fixed point
our approach analyzes the linear response of the system to small perturbations , while around a generic fixed point it assesses the existence of unstable directions in the phase space . in generic nonlinear equations
, two or more ( meta-)stable steady states could occur .
this does not seem to be the case in the present model based on our numerical findings , and is also reasonable given the fact that the evolution of the density matrix elements @xmath153 is generated by a quadratic form .
( one could also conceive a steady state that is unstable , but still is stable at any order in the taylor series : this again does not happen here . )
the derivation of the linearized eoms is deferred to app .
[ app : linthermal ] , but we mention here that this method allows us to test the stability of several hundred pixels in the phase diagram , in a short computational time , truncating the hilbert space to a maximum number of particles @xmath154 larger than @xmath155
. the latter condition is particularly important because the diagonal elements of the thermal state decrease only as a power law with the fock index @xmath156 .
quite nicely , this approach allows us also to obtain the analytical form of the phase boundary in the limit @xmath152 of small particle filling . to this end , we consider the linearized equations ( [ eq : instabeq ] ) around the thermal state with onsite fock spaces truncated at occupation number @xmath157 , appropriate for sufficiently low filling but still capable of describing the effect of interactions
. then we have a @xmath158 linear system where one eigenvalue is @xmath159 and corresponds to the @xmath160 state .
the other two eigenvalues @xmath161 , @xmath162 can be computed straightforwardly . keeping only the first order in the average density , one eigenvalue @xmath161 has negative real part while for the other @xmath163 = 4 n \kappa ( 4 j^{2 } - u^{2 } + 4 j u + 32 \kappa^{2 } ) / [ ( 2 j + u)^{2 } + 64 \kappa^{2}]$ ] .
the curve in the @xmath164 plane where @xmath165 defines the phase border , i.e. where the thermal state changes from stable to unstable . in the current approximation
it reads @xmath166 with @xmath167 the critical point at @xmath168 ( @xmath169 ) .
this expression is shown as a red solid line in the main panel and in the inset of fig .
[ fig : phasediagram ] . below in sec .
[ sec : smalldens ] we point out an alternative way of deriving this result . by reducing the analysis of the equations of motion to the first three fock states it is also possible to extrapolate the position of the critical point for variable density , as shown in the panel ( c ) of fig .
[ fig : depletion ] .
two features of the phase diagram of fig .
[ fig : phasediagram ] are dramatic manifestations of the system being out of equilibrium .
first , even in the limit of vanishing dissipation the system presents a transition from a thermal - like to a condensed steady state at @xmath170 .
this picture does not connect to the phase transition between the mott state and the superfluid state that characterizes the phase diagram of the equilibrium bose - hubbard system .
formally , such difference is possible because the limits @xmath171 and @xmath172 do not commute and defeats the naive expectation that a small dissipation applied to a system does not wash out entirely its zero - temperature phase diagram .
a more physical explanation is the fact that the engineered dissipation in the liouvillian ( [ eq : liouvillian ] ) does not respect the principle of detailed balance , as discussed above , and hence its action is radically different from a bath that induces thermalization .
a notable example of the latter case is a josephson junction array coupled to a dissipative bath @xcite , which suitably chosen can stabilize the superconducting ordered phase .
the profound difference with respect to our setup stems from the fact that there the system plus bath is in global thermodynamic equilibrium and the associated phase transitions ( thermal or quantum ) are equilibrium ones .
second , we note the absence of commensurability effects . at the mean field level and at zero temperature ,
one expects that any mechanism suppressing superfluidity leads to a mott state for commensurate filling .
in fact , given the density matrix @xmath173 without off - diagonal order ( which represents long - range order in the mean field approximation ) , the only solution for the steady state is a mott state @xmath174 with quantized particle number @xmath175 , because the conditions @xmath176 ( normalization ) and @xmath177 ( purity at zero temperature ) must hold simultaneously on each site of the lattice .
in contrast , in our case the dephasing of correlations induced by the interaction ( see discussion in sec . [
ssec : competition ] ) leads to a diagonal steady state that does not obey an additional purity constraint .
the dephasing of the coherent initial state makes the density matrix more and more diagonal , but does not lead to an additional localization in fock space @xcite .
we also observe directly that the phase diagram in fig .
[ fig : phasediagram ] does not change qualitatively from @xmath178 to @xmath109 . in conclusion ,
the nonequilibrium quantum phase transition presented here shares features of equilibrium quantum phase transitions in that it is interaction driven , and of classical phase transitions in that the ordered phase terminates in a thermal state . in our formulation of the many - body nonequilibrium problem ,
we make use of a chemical potential , which at first sight is a concept tightly bound to thermodynamic equilibrium . in order to better understand its meaning and practical use in the present setting ,
we would like to compare our situation to the thermodynamic equilibrium of an interacting bose gas , say in the continuum , in which the temperature is low enough that spontaneous symmetry breaking takes place . here
, the role of the chemical potential is twofold : first , it conceptually fixes the ( average ) particle number ; second , its actual value is chosen to fulfill an equilibrium condition , namely to remove homogeneous terms which are linear in the fluctuation operators which occur upon expansion about a condensed ground state , a condition which in turn is equivalent to the existence of a gapless goldstone mode . in our nonequilibrium case
, there is no coincidence of these two roles of the chemical potential ; its choice is only connected to the second one .
indeed , the average particle number is an exactly conserved quantity of our mean field master equation .
hence , the average particle number is fixed via the initial state .
however , the introduction of a chemical potential term is needed to ensure the existence of a steady state , in which all time - local correlation functions become time independent .
mathematically , this choice can be made by requiring that the equation of motion for the spatially homogeneous fluctuation @xmath179 experiences no `` driving '' for all @xmath11 , i.e. there is no constant term generating its evolution .
this condition on the equation of motion is precisely analogous to the condition of the vanishing of the coefficient of the term linear in the fluctuation in a hamiltonian . ) , for @xmath180 , @xmath96 , @xmath97 , and @xmath181 .
the main panel shows the fluctuations @xmath182 of the local density @xmath183 with respect to the average value ( short dash , long dash , solid , from the earlier to the later time ) .
the inset shows the absolute values of the fluctuations in logarithmic scale ( solid ) and the prediction of the growth rate obtained with the numerical linearization of the equations of motion around the steady state ( dashed ) . ]
having classified the homogeneous steady states of the system and presented the phase border in fig .
[ fig : phasediagram ] , we consider now the stability of the condensed phase with respect to inhomogeneous perturbations .
although in general we expect that an initial inhomogeneous system is driven towards the homogeneous dark state defined by the liouvillian ( as discussed in sec .
[ sec : themodel ] and shown for a paradigmatic case in fig . [ fig : condensation3d ] ) we can not discard the possibility that the competition between unitary and dissipative dynamics destabilizes the convergence process in some range of parameters . the stability of the homogeneous steady state can be tested numerically by subjecting the steady state to a small perturbation and monitoring the dynamics following the perturbation . due to the nonlinearity of the dynamics described by eq .
( [ eq : motiontrace ] ) , the results of this analysis depend on the kind of perturbation that is applied to the system . from the theory of the linear response of many - body systems ,
we know that the spectrum of the collective excitations is obtained by studying the density - density response function , hence we perturb the density matrix of the system with the addition of a coherent component @xmath184 that oscillates periodically in space @xmath185 with the maximum wavelength allowed by the size of the lattice . in fig .
( [ fig : densitywave ] ) we show that , for specific choices of the parameters , the magnitude of the perturbation @xmath182 is not damped , but increases exponentially in time as @xmath186 , with @xmath187 , preserving its shape . the robustness of the spatial profile @xmath188 and the fact that it increases in time with a well - defined instability exponent @xmath189 classifies this phenomenon as an unstable mode .
the system then manifests a dynamical instability , characterized by the wavelength of the charge density wave show in fig .
[ fig : densitywave ] .
a more thorough analytical understanding of the origin of this instability is deferred to sec .
[ sec : instability ] , after a suitable approximation scheme is developed . to delimitate systematically the range of parameters where the dynamical instability takes place , we linearize the eom ( [ eq : motiontrace ] ) around the steady state and study the spectrum of eigenvalues , analogously to the procedure performed in the previous section to find the border of stability for the thermal state . in this case , however , the substantial analytical simplifications deduced in the previous case can not be applied because the form of the condensed steady state is not known in general and , in any case , the eoms for the entries of the density matrix have a much more intricate structure of couplings .
hence we resort to the numerical procedure outlined in app .
[ app : lingeneral ] , which provides the shape of the unstable domains shown in fig .
[ fig : phasediagram ] with dashed lines .
we see that an unstable domain exists , that includes the whole @xmath168 axis where the thermal state is not stable , and is larger as the interaction strength increases .
the phase transition from the thermal to the condensed state is then substituted by a transition from the thermal to the unstable phase for small enough values of @xmath6 .
the border between the condensed and the unstable regime in the proximity of the origin is characterized by a fixed ratio @xmath190 , as can be clearly seen in the inset of fig .
[ fig : phasediagram ] . to interpret this fact
, we note that the condensed eigenstate of the liouvillian ( [ eq : liouvtrace ] ) is also an eigenstate of the kinetic term in eq .
( [ eq : localhamiltonian ] ) , so @xmath6 is the strength of an unitary term `` compatible '' with the engineered dissipation . on the contrary ,
interaction @xmath44 and dissipation compete strongly , as discussed in detail above .
we may argue then that , in general , the steady state of the liouvillian can be made more robust by increasing the compatible energy scales with respect to the competing energy scales in the hamiltonian .
) and the eom for small density ( solid and dashed lines , respectively ) for @xmath151 , @xmath191 , and @xmath192 from bottom to top in a homogeneous system ( @xmath134 ) .
the inset shows in a bilogarithmic scale the difference @xmath193 between the results at the final time @xmath194 for the three values of the density . ]
we present now several analytical arguments that confirm the nonequilibrium phase diagram presented in fig .
[ fig : phasediagram ] .
we consider the equations of motion for the most general normal ordered correlation functions of our system @xmath195 . in general , these correlation functions form an infinite nonlinear coupled hierarchy . however , we find that the _ a priori _ infinite hierarchy of mean field equations for the onsite correlation functions exhibits a finite subset which decouples from the rest in the limit @xmath196 .
importantly , we find that all qualitative effects of the nonequilibrium phase diagram are contained in this reduced system of equations of motion , making this limit an important tool for an analytical understanding of the physics of our system . in fig .
[ fig : reduced ] we show that the results obtained with the numerical integration of the reduced equations compares quantitatively very well with the full equations for average densities even larger than @xmath110 , i.e. a realistic figure for typical fillings in optical lattices . for the full system of correlation functions , the general form for the equation of motion @xmath197 in closed form
is reported in eq .
( [ eq : motionanycorr ] ) .
we observe that @xmath198 couples to the correlation functions @xmath199 and @xmath200 , thereby not allowing for a truncation to small subset of correlation functions _
a priori_. however , progress can be made by introducing a power counting @xmath201 .
this gives an upper bound on the true powers of the operators , as correlation functions in the state without phase symmetry breaking with @xmath202 vanish exactly .
the leading power for the time evolution of a correlation function is consequently given by @xmath203 .
writing down the system of equations of motions ( [ eq : corrset ] ) , it is then obvious that this system closes if we truncate the equations of motion for @xmath204 , @xmath205 each of the equations at their leading power , thereby physically corresponding to a low density expansion .
we use this result first to obtain analytically the critical exponent @xmath189 discussed above . for a homogeneous system with @xmath206 the equations of motion
read @xmath207 with the chemical potential @xmath208 resulting from the steady state condition , cf .
the discussion following eq .
( [ eq : dissgpe ] ) @xcite . the exponential convergence of the system to its steady state
is described by the linear contribution , which corresponds to a `` mass '' or `` gap '' in the jargon of statistical mechanics .
we are interested in the critical regime of the dynamics , in which the exponential convergence is suppressed due to the vanishing of the gap term and the system is driven to its fixed point by the residual nonlinear contributions to the equations of motion .
we focus on the linear equations for @xmath89 and @xmath205 , with eigenvalues @xmath209 , @xmath210 and we consider @xmath204 as a nonlinear driving term . in the critical regime ,
in which @xmath211 , the secular equation for the eigenvalue fixes the critical value of the interaction strength @xmath212 [ that coincides with @xmath133 defined after eq .
( [ eq : phaseborder ] ) for @xmath152 ] .
the eigenvector corresponding to @xmath209 is @xmath214 and hence the eom for the quantity @xmath215 does not feature linear terms and reads @xmath216 . a similar quantity @xmath217 built with the eigenvector corresponding to @xmath210 , instead , features a regular exponential decay in time .
the time - evolution of @xmath89 is given in general by a linear combination @xmath218 and @xmath217 but , for long times , the exponentially decaying contribution of the latter is negligible and hence we can substitute @xmath219 into the eom for @xmath218 . at this point
we make an adiabatic approximation by solving the equation @xmath220 for @xmath204 and substituting the result into the eom for @xmath218 , that yields @xmath221 . from the solution @xmath222 , using the expressions above , we finally demonstrate the critical behavior @xmath223 .
we focus now on the asymptotic behavior of the system to demonstrate the depletion of the condensate fraction shown in fig .
[ fig : depletion ] . to this end , it is more convenient to introduce the fluctuation operators @xmath224 where @xmath225 is the ( unknown ) order parameter in the homogeneous steady state .
the eoms for the `` connected '' correlation functions built with the fluctuation operator ( see app .
[ app : eqmotion ] ) decouple naturally into linear equations that involve the connected parts of the mean fields ( [ eq : corrset ] ) , the density fluctuation @xmath226 , and @xmath225 as a parameter . in the homogeneous steady state , in which @xmath227 ( see app .
[ app : eqmotion ] ) we have @xmath228 the nonlinearity of the theory is now entirely contained in the identity @xmath229 , whence the value of the order parameter is obtained once the other equations have been solved .
practically , in the solution we can assume that the order parameter is real without loss of generality and fix the chemical potential to ensure that the fluctuation of the density has vanishing imaginary part .
the full result for the depletion then reads @xmath230}~,\ ] ] with @xmath231 and @xmath232 . for @xmath206 ( @xmath169 )
this reduces to the simple inverted parabolic profile @xmath233 these analytical results are compared to the numerical solution of the equations of motion in the panel ( b ) of fig .
[ fig : depletion ] , demonstrating an excellent agreement between the two computations .
the quadratic behavior of the depletion may be contrasted with the thermal depletion in a free equilibrium bose - einstein condensate in three dimensions , where @xmath234 with @xmath235 with @xmath236 the particle s mass and @xmath237 the riemann zeta function .
the quadratic behavior is reminiscent of a harmonically trapped bose - einstein in two dimensions .
it is noteworthy that the phase border that one obtains by requiring @xmath238 in eq .
( [ eq : analytdepletion ] ) coincides exactly with the small - density result in eq .
( [ eq : phaseborder ] ) in the linear response strategy of sec .
[ ssec : phases ] .
we are now in the position of providing a clear - cut explanation of the dynamical instability unveiled in sec .
[ ssec : stability ] , transposing at the level of the correlation functions the instability analysis around the steady state used to delineate the unstable domains in fig .
[ fig : phasediagram ] .
the starting point are the eoms for the connected correlation functions outlined in app .
[ app : eqmotion ] , and the knowledge of the homogeneous steady state @xmath239 demonstrated in the previous section .
then we expand the density matrix of the system as @xmath240 and consider the fluctuations of the connected correlation functions @xmath241
+ { \rm tr } [ \delta \hat{b}_{\ell}^{\dag p } \delta \hat{b}_{\ell}^{q } \delta \rho_{\ell}(t ) ] ~. \end{aligned}\ ] ] the eoms are then linearized in the fluctuations @xmath242 and the values of @xmath243 , given by eq .
( [ eq : corrsteadystate ] ) , are such that all the driving ( i.e. constant ) terms vanish upon the correct choice of the chemical potential outlined above .
we remark that the fluctuation @xmath244 does not vanish as the system is not in its steady state ( see app .
[ app : eqmotion ] ) and that local fluctuations @xmath245 of the density modify the flat distribution @xmath246 of the steady state .
then we take the fourier transform of the eoms and rewrite the linear system in terms of the connected correlation functions in momentum space @xmath247 .
it is important to notice that the fourier transform of the correlation functions is not simply related to the correlation functions of the operators in momentum space , except for the first order correlation where it holds @xmath248 and @xmath249 .
( we may denote such correlations @xmath250 and @xmath251 , respectively , because the fluctuations of the order parameter vanish on the steady state by construction . ) since the instability shown in fig .
[ fig : densitywave ] takes place at small momenta , in performing the fourier transform we focus on the central region of the brillouin zone and substitute the occurrences of the discrete laplacian @xmath252 with the parabolic dispersion @xmath253 . of the linearized eom for @xmath206 , @xmath151 , and @xmath254 . ]
the linear system of eoms takes the form of a @xmath255 matrix ( the @xmath256 complex correlation functions , their complex conjugates , and the real density fluctuation ) whose eigenvalues @xmath257 give the @xmath51-dependent spectrum of the system .
the eigenvectors of the system correspond to modes that evolve as @xmath258 , which are stable ( unstable ) if @xmath259 ( @xmath187 ) . the real part of the spectrum for a typical choices of parameters within the unstable domain is shown in fig .
[ fig : spectrum ] .
the spectrum features : i ) two doubly - degenerate strongly decaying modes ( @xmath260 ) that project mainly on the third - order correlation functions ; ii ) one decaying mode ( @xmath261 ) that projects mainly on the density fluctuation ; iii ) two low - lying modes generated by an admixture of the first - order correlation functions .
the latter modes are magnified in the inset of fig .
[ fig : spectrum ] .
the lower mode gives @xmath187 in a small interval around @xmath262 , hence proving the existence of unstable modes with well - defined momentum .
the domain where an unstable mode exists in this approximation is delimited in fig .
[ fig : phasediagram ] by a red dashed line .
in general , the decay rate of the modes i ) and ii ) is proportional to @xmath263 and @xmath264 , respectively , as it appears from inspection of the linearized eoms .
the clear separation of the dissipative part of the spectrum into groups of modes that have largely different decay rate at low momenta suggests that an adiabatic elimination of the fastest modes can be performed to bring the @xmath255 linear system in a more compact form . in this way
we obtain a renormalized spectrum of the weakly dissipative single particle modes , where the instability is encoded . in general , the adiabatic elimination in a system @xmath265 , \quad \partial_{t } u_{\rm s } = g[u_{\rm f } , u_{\rm s}]~,\ ] ] with fast ( @xmath266 ) and slow ( @xmath267 ) modes consists of solving @xmath268 = 0 $ ] for @xmath266 and using the result into the second equation that becomes @xmath269 , u_{\rm s}]$ ] .
the application of the procedure introduces new terms @xmath270 $ ] that renormalize the equation of the slow modes .
we apply the procedure once to eliminate the @xmath271 modes and then again to eliminate the @xmath272 modes . since the border of the unstable domain extends to the origin of the phase diagram in @xmath6 , @xmath44 ( see fig .
[ fig : phasediagram ] ) , to understand the phenomenon underlying the instability it is enough to perform the algebraic manipulations to the first order in @xmath6 and @xmath44 .
we obtain a renormalized @xmath273 linear system for the time derivative @xmath274 of the fluctuations of the order parameter in time , which reads @xmath275 here , @xmath276 is the `` bare '' quadratic decay rate that follows from the analysis of the linear correlations only , for small interaction and nonzero hopping amplitude , and is shown as a black solid line in the inset of fig . [
fig : spectrum ] .
@xmath277 is the low momentum kinetic energy . finally , @xmath278 and @xmath279 are the terms that renormalize the slow modes obtained by the adiabatic elimination . without the renormalizing terms
the @xmath273 system displays the structure of a bogoliubov equation for the condensate modes , with diagonal dissipation @xmath280 .
we point out that a standard quadratic theory can reproduce the bogoliubov - type eom but necessarily misses the renormalizing terms that are due to third - order local correlations , and thereby the entire physics of the dynamical instability .
the latter is thus a clear fluctuation - induced beyond - mean - field effect .
the eigenvalue of the linear system , which approximates the lower mode in the inset of fig .
[ fig : spectrum ] , reads explicitly @xmath281
if @xmath282 we can identify @xmath283 as the speed of sound @xmath284 of the dissipatively - created condensate and we also find a modified decay rate for the modes that is quadratic in the momentum .
however , as @xmath6 increases , the square root becomes imaginary , the contribution of the dispersion to @xmath285 vanishes and the decay rate of the modes is modified by a non - analytic term @xmath286 , which is positive and dominates over the contribution @xmath287 at low momentum .
the resulting physical picture for the dynamical instability in the dissipatively - driven condensate is then an interplay of short time quantum and long time classical fluctuations .
purely local , i.e. temporal ( or quantum ) fluctuations captured in the gutzwiller approach renormalize the complex spectrum at short times .
this prepares the ground for spatial ( or classical ) fluctuations taking over at later times and responsible for the dynamical instability .
we remark that the gutzwiller approximation that we adopt is suited to describe both i ) the onsite fluctuations that are described without approximation in a fully quantum local hilbert space ; ii ) long - wavelength fluctuations of the coherence that receive the main contribution from the disconnected part , which is computed solely in terms of the order parameter .
it is important to realize that the instability arises at weak coupling already , where the system is well described by the inhomogeneous gutzwiller mean field theory . without referring to any approximation
, it is clear that the instability is due to a renormalization of the single particle ( complex ) excitation spectrum , and thus encoded in the evolution of @xmath288 .
the exact equation of motion for these quantities is a nonlinear equation , with in general nonlocal spatial correlations .
the gutzwiller approximation factorizes the correlations functions in real space , but treats onsite correlations exactly .
the factorization in real space is justified at weak coupling ( large condensate ) because the dominant scattering processes are those for opposite momenta off the macroscopically occupied condensate .
however , the factorization of the onsite correlation functions ( as in the gpe discussed in sec .
[ ssec : phases ] ) would not allow to capture the physics of the dynamical instability . in summary
, because the dynamical instability is a weak coupling phenomenon where our approximations can be rigorously justified , we can rule out the possibility of the effect being an artifact of approximations . of course , the above analysis has the status of a linear response theory , and thus is not able to shed light the fate of the system in the dynamically unstable regime ; we will address this question in sec .
[ ssec : nonlineresponse ] .
we note now that the exact liouville operator ( [ eq : liouvillian ] ) features the dark state also in the case of a single particle on the lattice , corresponding to the limit @xmath196 with @xmath289 and @xmath290 .
our approximation yields correctly the existence of the dark state at @xmath262 , since the expression for the bare quadratic decay rate reads @xmath291 in the limit @xmath196 . on the contrary , the bogoliubov approximation used in ref .
@xcite gives @xmath292 and hence , in the limit @xmath196 , does not yield the correct unique dark state .
the discrepancy between the two methods lies in the fact that , in the bogoliubov approximation , it holds @xmath293 leading to @xmath294 .
the additional occurrence of unity that is neglected in the bogoliubov approximation is indeed essential to the existence of the dark state and is correctly captured by the gutzwiller approximation .
finally , we remark that , in contrast to the dissipative part , the bare theory for the unitary part of the time evolution , where two - point correlations are not included , coincides with the gpe augmented with linear fluctuations .
this is rooted in the precise nature of the nonlinearities occurring in the dissipative and unitary part . because of the dynamical instability demonstrated above , the assumption of a homogeneous steady state is not always compatible with the physics of this nonequilibrium system .
the linear response method applied to the onset of the instability suggests that a charge density wave ( cdw ) , with well - defined wavelength ( see fig . [
fig : densitywave ] ) , sets in at later times .
to confirm this picture , we proceed now to the numerical solution of the nonlinear eoms in the limit of small density . in this case
the computational complexity of the problem is greatly reduced , as each site is described by four numbers only .
the eoms can be integrated quickly and reliably on very large lattices with @xmath295 and for extremely long times @xmath296 .
the spatial profile of the particle density is shown in fig .
[ fig : fullinstability ] for a typical time - evolution .
we clearly observe that a stationary cdw establishes at large times .
the wavelength @xmath297 of the periodic modulation is the same as the most unstable mode in the lowest - lying branch of the dissipative spectrum ( the minima in the inset of fig .
[ fig : spectrum ] ) .
in other words , the fastest - growing mode in the continuum of unstable modes becomes dominant at large times @xcite , its wavelength is transmuted into the wavelength of the cdw , and , arguably , the remaining modes are eventually suppressed by nonlinear effects that are not described by the linear response approach used in sec .
[ ssec : linresponse ] .
this dynamically stable feature , with a relative density modulation @xmath298 in fig .
[ fig : fullinstability ] , gives an unambiguous , robust experimental signature of the fluctuation induced dynamical instability .
typical optical lattice setups are smaller than @xmath299 sites considered here , so it may be difficult in practice to obtain a periodic modulation if the most unstable mode has a very small momentum . for
a lattice with e.g. @xmath300 sites , one would find a cdw where only a single wavelength fits into the optical lattice , and the measurement of the average density on the two halves of the lattice would be enough to detect the density fluctuation .
the reason to focus on larger lattices in our theoretical considerations is to exclude that we deal with a finite size effect . indeed , since only a small interval of momenta is dynamically unstable , it may happen for a lattice with @xmath301@xmath302 sites that only the lowest momentum @xmath303 ( in lattice units ) in the discretized brillouin zone is unstable ( we reiterate that the dark state with @xmath262 is always stable ) .
so , reducing to small lattices only , one can not rule out that @xmath304 , which would lead to a vanishing effect in the thermodynamic limit @xmath305 .
however , fig .
[ fig : fullinstability ] demonstrates that the spatial scale of the density modulation remains finite in the thermodynamic limit and coincides with the wavelength of the most unstable mode . of the system , according to the reduced eoms for small density , for @xmath299 , @xmath110 , @xmath97 , and @xmath98 .
the system is initially prepared in a fully - condensed state and converges to a homogeneous steady state , which becomes unstable at @xmath306 .
a persistent oscillation in the density profile is visible for later times .
the double arrow shows the wavelength @xmath307 ( in lattice units ) of the instability as computed from the theory in sec .
[ ssec : linresponse ] . ] intriguingly , the density profile presented in fig .
[ fig : fullinstability ] suggests that the system in the steady state spontaneously breaks the translational symmetry , which both hamilton and liouville operator present .
this adds to the breaking of the global phase symmetry due to the presence of the condensate .
a state where both symmetries are broken is often termed a `` supersolid '' although , in the equilibrium case , the breaking of the translational symmetry is stronger , with a spatial modulation on the order of the lattice constant @xcite .
the spontaneous breaking of the translational symmetry discriminates the dynamical instability considered here from the dynamical instability suffered by an ensemble of interacting bosons moving with constant momentum through a lattice @xcite . in the latter case ,
if the momentum of the atomic ensemble is larger than an interaction - dependent threshold , the atomic current decays and the order in momentum space is lost .
the origin of this instability lies in the single - particle dispersion , that becomes negative in the outer regions of the brillouin zone . in our case , instead , the system occupies a momentum state @xmath262 in which the single - particle dispersion is positive but the long - wavelength fluctuations of a collective ( condensate ) mode are destabilized by local fluctuations .
moreover , in the case of ref .
@xcite the instability is purely classical because it stems from spatial fluctuations . on the contrary , the local fluctuations that play the crucial role in our case are genuinely quantum , i.e. temporal fluctuations . finally ,
while the order in momentum space is lost in the unstable atomic current , the external drive in our case allows the persistence of a ( spontaneously ) ordered pattern , in analogy to synergetic phenomena @xcite .
, @xmath308 , and @xmath309 .
the initial state is a collection of local coherent states with random phases . the phase @xmath78 is changed abruptly from @xmath310 to @xmath311 at @xmath312 , after the system has reached the steady state .
the inset shows the probability density @xmath313 for the time @xmath314 at which the revival of the condensate fraction on the first site of the lattice reaches @xmath111 ( arrow in the main panel ) , obtained with @xmath315 evolutions with random initial phases . ] in the previous sections we have considered the dynamics of the system in the presence of a stationary dissipative drive . in the spirit of linear response theory , this has allowed us to study the dynamics of the system in terms of the instability of a collective mode . in this section
we focus on the response of the condensate to an abrupt change in the conditions of the drive .
the necessity to adapt to the widely different external conditions triggers a global response in the condensate and we observe intriguing nonlinear dissipative dynamics that can be understood in terms of collective variables . here
we concentrate on the characteristics of this dynamics in the absence of unitary contributions to the time evolution , i.e. @xmath316 , with homogeneous density profile .
we first let the system converge to the steady state corresponding to @xmath83 , i.e. a homogeneous condensate in the zero momentum mode , and then we change abruptly the phase @xmath78 in the liouville operator at some finite time . in the presence of a finite value of @xmath78 ,
the dark state is a condensate with finite lattice momentum @xmath317 and macroscopic inhomogeneous wavefunction @xmath318 .
we study the equilibration dynamics towards the new dark state .
our considerations apply to the mesoscopic dynamics @xcite in periodic lattices , as implemented e.g. in ref .
@xcite for purely coherent dynamics .
the dynamical effects that we identify are related to the finite size of these systems .
the time evolution is reported in fig .
[ fig : revival ] for a typical case .
the overall dynamics of the system consists of the rearrangement of the condensate fraction from the zero momentum of the initial state to the momentum @xmath317 imprinted by the dissipative mechanism after the quench of the phase .
the rearrangement consists of an initial quick collapse of the condensate fraction and a subsequent revival with different momentum .
it is remarkable that the revival takes place after a substantial but well - defined time interval . indeed , in the particular case shown in the inset of fig .
[ fig : revival ] , the fluctuation of the time scale for the revival is of the order of @xmath319 . to understand the dynamical processes that take place during the transient of the rearrangement is the goal of the present section . of the phases of the fluctuations @xmath320 , for the same evolution shown in fig .
[ fig : revival ] .
the fourier spectrum is normalized to its maximum at each distinct value of the time .
since the distribution of phases @xmath321 is real , a symmetric peak appears in the negative part of the @xmath51 axis and is not shown in the plot . ] immediately after the quench , the condensate fraction decreases exponentially towards negligible values . to describe the exponential decay we consider the eoms ( [ eq : corrhom ] ) and we keep only the terms that are linear in the powers of bosonic operators , that are dominant in the regime in which the correlations become negligible , obtaining @xmath322 \langle \hat{b}_{\ell } \rangle~.\ ] ] we see that the initial state is depleted with a rate that increases as the final phase @xmath78
is quenched further from the initial value @xmath83 , where the initial state is stable . to describe the revival we consider again the eoms ( [ eq : corrhom ] ) and it is convenient to use the moving frame of reference introduced in sec .
[ sec : meanfield ] .
the steady state of the system is a homogeneous condensate with respect to the correlations measured by the transformed operators @xmath323 . from the analysis of the instability performed in sec .
[ sec : instability ] we know that the effect of higher - order correlations is not necessarily negligible during the growth of a condensed mode . still , we can neglect the product of correlation functions , that are much smaller than the correlations themselves before the revival has taken place .
taking the fourier transform of the correlation functions with such prescription gives the coupled linear equations @xmath324 computing the eigenvalues of the matrix that defines the linear system we see that the growing mode that originates the revival can be located only at @xmath325 , that in the original frame is a small interval centered around @xmath317 .
we note that reducing the equations of motion to the terms proportional to @xmath9 only , as we did to describe the collapse of the condensate , is not appropriate in this case since the @xmath326 mode would result marginally stable , but not unstable .
hence , we identify a dynamical selection mechanism @xcite that promotes the fluctuations in a certain interval of finite width
. it would be difficult to imagine a scenario in which a single mode is unstable , due to the continuity in @xmath51 of the equations of motion , which can not give discontinuous results at any finite time .
the modes different from the dark state will be damped subsequently due to the action of the liouvillian .
we turn now to the more interesting question , how the mode with lattice momentum @xmath317 is initially populated during the time interval in which the condensate is depleted . in other words
, we are interested in the dynamics of the condensate fraction @xmath327 , which exhibits tiny fluctuations around the fully - collapsed value @xmath328 and which reorganizes at @xmath329 into the new reviving mode .
one naive expectation is that the phases @xmath321 of the coherent fluctuations @xmath330 randomize and that the component @xmath331 with momentum @xmath326 increases independently due to the process of dynamical selection outlined above . on the contrary , it is visible from the fourier spectra shown in fig .
[ fig : tunneling ] that the phases of the fluctuations exhibit a very organized dynamical phenomenon , featuring metastable steady states and rapid jumps into new configurations .
the fourier spectrum has always a very clear peak that samples all the discrete points in the brillouin zone , from the initial @xmath262 to the final @xmath326 .
this contrasts indeed with the simpler organization of the phases that takes place starting from a random distribution , as can be seen in fig .
[ fig : condensation3d ] or for very short times in fig .
[ fig : tunneling ] . in the following
we argue that the momentum drift in discrete jumps shown in fig . [
fig : tunneling ] can be understood as a cascade of macroscopic quantum tunneling ( mqt ) events between unstable steady states of the time - evolution .
such picture is a qualitative modification of the well - known tunneling from a false vacuum into a global ground state that takes place in a system at quasi - equilibrium @xcite . to justify our picture
, we study the dynamics of the phases @xmath321 only , given that the absolute value of the coherences @xmath332 is physically negligible .
the equation of motion for the phases reads , in general , @xmath333 and we specify this expression by keeping only the following particular term in the eom for @xmath332 @xmath334 the reason behind this choice is that , to the extent of identifying the metastable states of the phase dynamics , the linear approximation is enough .
indeed , the linear term in the equations of motion is responsible for the exponential damping of the variables as we discussed in sec . [
sec : smalldens ] in the context of critical dynamics , or at the beginning of the present section where the collapse of the condensate fraction is demonstrated .
the equation of motion for the phases then takes the form @xmath335 where @xmath336 is the discrete laplace operator .
the structure of this equation is reminiscent of an overdamped string displaced by @xmath321 at the position @xmath11 .
the kinetic term @xmath337 is absent from the equation and the first derivative represents a friction term .
we conjecture that including the two neglected degrees of freedom ( i.e. the phases of @xmath205 and @xmath204 ) back into eq .
( [ eq : psilaplacian ] ) would reinstate a kinetic term into eq .
( [ eq : phasemotion ] ) and couple the phases to an effectively fluctuating field , although a more thoroughly designed approach ( e.g. within the keldysh action formalism ) would be necessary to maintain this assertion . following ref .
@xcite , let us parameterize the spatial dependence of the phases in terms of an `` instanton '' or phase slip configuration , @xmath338 , to distinguish the local dynamics @xmath339 from the global variable @xmath340 describing the winding of the phases in the circular geometry .
it is important to notice that the difference @xmath341 between phases on neighboring sites is defined in @xmath342 $ ] , to minimize the total variation of the phase profile . with this convention for the difference between phases , the winding number reads @xmath343 $ ] .
we first consider the case that the global variable is constant and study the dynamics of the spatial fluctuations , that move according to @xmath344 } \
, \delta_{\ell}\delta\theta$ ] with @xmath345 .
this is exactly the equation of an overdamped string , slightly displaced from its equilibrium configuration , hence all the fluctuations @xmath346 converge to zero .
we consider now the dynamics of the global variable @xmath340 .
if @xmath347 , due to the boundary conditions we have that @xmath348 . the equation of motion for @xmath340 reads in this case @xmath349 \ , q$ ] and hence @xmath350 decreases towards zero ( we neglect terms that go to zero for large lattices ) . if @xmath351 , the boundary condition now is such that @xmath352 , that entails @xmath353 \simeq - \cos[\phi ] [ 2\pi - q(t)]$ ] and hence @xmath350 increases towards @xmath354 .
the same procedure applies to the other periodic replicas of the interval @xmath355 and shows that the global variable @xmath340 tends to the closer multiple of @xmath356 . in conclusion
, we have shown that the state with @xmath350 multiple of @xmath354 ( corresponding to a definite momentum in the distribution of the phases ) is stable both with respect to local fluctuation of phases and to a global change of momentum .
we could hence think of a momentum state as a local minimum , whereto the system converges by an overdamped motion
. having clarified the nature of the metastable configurations seen in fig . [
fig : tunneling ] , we improve our conjecture on the effect of the terms that have been neglected in the equation of motion for @xmath321 .
they necessarily provide a mechanism for the system to tunnel out of such minimum towards a nearby momentum configuration .
the picture of a sequence of tunneling events , each with a well - defined lifetime , explains the stability of the revival time over many configurations ( see inset of fig . [
fig : revival ] ) and justifies a bell - shaped probability distribution for this quantity on the basis of the central limit theorem . , for the same evolution shown in fig .
[ fig : revival ] . at each time , the winding number is given by the number of abrupt jumps in the average phase .
the locations of the phase jumps remain stationary in space , between the phase slip events , during which the winding number increases by one . ] to further clarify the phase dynamics , we follow ref . @xcite and compute the average phase difference @xmath357 between sites at distance @xmath358 . for the initial configuration with zero momentum
this quantity is zero everywhere .
we see that the initial phase configuration remains stable for a noteworthy time lag @xmath359 following the phase quench at @xmath360 .
then a fluctuation in the phases takes place and culminates in a phase slip at @xmath361 , that increases the winding number of the phase distribution by one .
another phase slip takes place at @xmath362 and the final one at @xmath363 , corresponding to the abrupt jumps in the peaks of the fourier spectrum shown in fig .
[ fig : tunneling ] .
we notice that the dynamics in momentum space takes place abruptly while it has a smooth appearance in real space : this discrepancy is reminiscent of the instanton description of the mqt phenomenon , in which the change of quantum numbers takes place instantaneously in imaginary time @xcite . ,
@xmath110 , and @xmath309 .
the phase @xmath78 is changed abruptly from @xmath310 to @xmath364 at @xmath365 . for @xmath366
the system stabilizes around the momentum @xmath367 and develops a reduced metastable condensate fraction .
nonetheless , convergence to the steady state with momentum @xmath364 is expected to take place at times larger than @xmath368 .
the inset shows the condensate fraction at @xmath369 as the density @xmath76 is varied . for @xmath370
the system reaches the steady state and the condensate revives fully ( see also fig . [
fig : revival ] ) . for @xmath371
the time - evolution is similar to the main panel , and the size of the condensate fraction depends strongly on the density . for @xmath372 the system develops a metastable condensate fraction with momentum @xmath373 . ] we point out that the strong suppression of the condensate fraction , which takes place for very small densities , is not a necessary condition for the metastable behavior investigated above . in fig .
15 we report a typical time - evolution for larger densities ( still well described by the reduced eom for small density , see fig .
[ fig : reduced ] ) in which the condensate fraction features a partial revival that persists as long as the true steady state is not reached .
while the explanation of the collapse of the condensate still holds , we do not have at present a dynamical model for the generation and persistence of a reduced condensate fraction . as the density in the lattice
is increased , the momentum of the metastable state changes and , at the transition point , a discontinuity in the fraction of revived condensate takes place . to reproduce a curve like that shown in the inset of fig .
[ fig : reduced ] it is only necessary , in principle , to measure the condensate fraction at time @xmath374 .
hence , these features provide quite clear qualitative observable signatures of the underlying phase dynamics .
we have discussed various phenomena in driven - dissipative many - body boson systems governed by many - body master equations with a dissipative zero mode .
this includes both a semi - quantitative characterization of the steady state phase diagram resulting from the competition of the driven - dissipative dynamics with a hamiltonian , as well as a first characterization of mesoscopic nonlinear dissipative equilibration dynamics following phase quenches in the liouville operator . for this purpose ,
we have developed a generalized gutzwiller approach which accounts for density matrices corresponding to mixed states .
we have introduced a low density limit which exhibits all the qualitative features of the phase diagram and is amenable to an analytic discussion due to the decoupling of a finite subset of correlation functions .
several features are direct consequences of the fact that our system is far out of thermodynamic equilibrium .
most prominently , the phase transition in steady state shares features of a quantum phase transition in that it is interaction driven , and of a classical phase transition in that the ordered phase terminates in a thermal state , in stark contrast to dissipative quantum phase transitions in global thermodynamic equilibrium , where the constraint of zero temperature inhibits such behavior .
furthermore , in the limit of vanishing dissipation , the nonequilibrium phase diagram for the steady state does not connect smoothly to the phase diagram of the equilibrium bose - hubbard model this point is a manifestation of the non - commutativity of the limits @xmath171 and @xmath172 .
another unconventional signature is the existence of a dynamical instability , which manifests itself in a spontaneous breaking of translational symmetry .
finally , we have identified interesting nonlinear dissipative dynamics following a quench from one dark state to the other in mesoscopic systems , which hints at an interesting interplay of fluctuation and dissipation for effective macroscopic topological variables @xcite .
we believe that the driven - dissipative many - body systems bear substantial potential for intriguing many - body physics far away from thermodynamic equilibrium .
in particular , interesting questions concern the exploration of the critical behavior close to the phase transition and a more analytical understanding of the mesoscopic dissipative dynamics in terms of collective variables .
we acknowledge collaboration with rosario fazio and andrea micheli in the early stages of this work .
we acknowledge the european union through ip aqute , and the austrian science fund ( fwf ) through sfb f40 foqus .
in this section we compute the right - hand side of eq .
( [ eq : motiontrace ] ) .
let us start from the local interaction term in the hamiltonian @xmath375 \nonumber \\ & = & { \rm tr}_{\neq \ell } \sum_{\ell ' } \bigotimes_{\ell''\neq \ell ' } \rho_{\ell '' } [ u \hat{n}_{\ell'}(\hat{n}_{\ell ' } - 1),\rho_{\ell'}]~.\end{aligned}\ ] ] in the summation above we separate the term with @xmath376 from the others . in the isolated term
the trace applies only to the tensor product of density matrices and the commutator remains unaltered . in the remaining terms , instead , the trace applies also to the commutator part and hence the contribution vanishes .
hence , we obtain @xmath377 = [ u \hat{n}_{\ell}(\hat{n}_{\ell } - 1),\rho_{\ell } ] ~.\ ] ] the contribution of the chemical potential , similarly , is the commutator with the term @xmath378 . to treat a generic nonlocal term @xmath379 in the hamiltonian or in the lindblad structure
we introduce the decomposition @xmath380 .
the hopping term @xmath381 on the @xmath11the site reads in this form @xmath382 with @xmath11 and @xmath75 nearest neighbors and @xmath383 the kronecker symbol .
the trace of the commutator gives @xmath384 \nonumber \\ & = & { \rm tr}_{\neq \ell } \sum_{\ell ' } \bigotimes_{\ell'''\neq\ell',\ell '' } \rho_{\ell ' '' } [ \sum_{rr ' } \gamma_{\ell'\ell''}^{rr ' } a_{\ell'}^{r } \otimes b_{\ell''}^{r'},\rho_{\ell'}\otimes \rho_{\ell '' } ] \nonumber \\\end{aligned}\ ] ] breaking the sum over @xmath75 as described above we obtain @xmath385 = -j \sum_{\langle \ell ' | \ell \rangle}[\hat{b}_{\ell } \langle \hat{b}_{\ell ' } \rangle + \hat{b}_{\ell } \langle \hat{b}_{\ell ' } \rangle,\rho_{\ell}]~,\ ] ] where @xmath386 indicates the sum over all the nearest neighbors @xmath75 of @xmath11 .
finally , the contribution of the hamiltonian is @xmath387 = [ \hat{h}_{\ell}(t ) , \rho_{\ell}(t)]$ ] , with the effective locally - acting hamiltonian @xmath388 in the liouvillian term @xmath389 $ ] in eq .
( [ eq : motiontrace ] ) we insert the identity @xmath390 that holds since @xmath391 and so @xmath392 for every value of @xmath11 .
the summation breaks into three parts @xmath393~.\ ] ] the first term gives zero , since the expression @xmath394 $ ] in this case is equivalent to @xmath395 $ ] that vanishes for every couple of @xmath396 and @xmath397 because of the form of the lindblad term . using that @xmath398 , the two remaining terms give @xmath399~.\ ] ] to proceed further , we assume a cubic square lattice geometry , with the coefficients @xmath400 that vanish on each plane orthogonal to the axial direction @xmath401 and are equal to a constant @xmath78 along @xmath401 .
then the phase difference between adjacent sites in the lattice is @xmath402 , with @xmath403 along the @xmath401 axis and @xmath404 in the orthogonal plane .
the jump operators in the decomposition above read @xmath405 @xmath406 the contribution of a single neighbor in eq .
( [ eq : traceliouv1 ] ) is @xmath407 \gamma_{\ell+1,\sigma}^{rs}~,\ ] ] with the matrix @xmath408 of correlation functions @xmath409 the contribution of @xmath410 and @xmath411 appears once , while @xmath404 appears @xmath412 times , where @xmath63 is the coordination number of the lattice ( @xmath413 for the two - dimensional square , @xmath414 for the three - dimensional cube ) .
we begin the analysis with the definition of the density matrix @xmath415 linearized around the thermal state and we keep only the first order in the entries of the perturbation @xmath184 .
the density matrix enters both the hamiltonian @xmath416 and the liouvillian @xmath417 in the computation of the correlations . to the first order
, the eom for the perturbation matrix reads @xmath418 - { \cal l}_{\ell}^{(0)}[\delta \rho_{\ell } ( t ) ] = - i[\delta \hat{h}_{\ell } , \rho_{\ell}^{\rm ( th ) } ] + \delta { \cal l}_{\ell } [ \rho_{\ell}^{\rm ( th)}]~,\ ] ] where the left - hand side is linear in the perturbation matrix , while the right - hand side acts as an effective driving .
the right - hand side depends on the perturbation density matrices @xmath419 on the neighboring sites of @xmath11 via the correlation functions that enter the hamiltonian and the liouvillian .
the terms in eq .
( [ eq : linliouvmf ] ) read @xmath420 & = & \kappa \sum_{rs } [ 2 \hat{a}_{\ell}^{r } \rho \hat{a}_{\ell}^{s \dag } -\hat{a}_{\ell}^{s \dag } \hat{a}_{\ell}^{r } \rho - \rho \hat{a}_{\ell}^{s \dag } \hat{a}_{\ell}^{r } ] \sum_{\sigma = \pm 1 } \gamma_{\ell + \sigma}^{(0)\ , rs}~ , \nonumber \\ \delta { \cal l}_{\ell}[\rho ] & = & \kappa \sum_{rs } [ 2 \hat{a}_{\ell}^{r } \rho \hat{a}_{\ell}^{s \dag } -\hat{a}_{\ell}^{s \dag } \hat{a}_{\ell}^{r } \rho - \rho \hat{a}_{\ell}^{s \dag } \hat{a}_{\ell}^{r } ] \sum_{\sigma = \pm 1 } \delta \gamma_{\ell + \sigma}^{rs}~ , \nonumber\end{aligned}\ ] ] where we have reserved the possibility for the perturbation to be inhomogeneous in real space .
the unperturbed matrix of the zeroth - order correlations , computed on the thermal state , reads @xmath421 the matrix of the first - order correlations is @xmath422 with the averages @xmath423 $ ] .
we verified with the numerical integrations of the nonlinear eom ( [ eq : motiontrace ] ) that @xmath424 is much smaller than the other correlation functions and hence can be neglected .
the structure of the eom for @xmath425 is determined by the right - hand side of eq .
( [ eq : linliouvmf ] ) , that is tridiagonal in the fock space .
it follows that only the entries of @xmath184 on the upper , central , and lower diagonal need to be taken different from zero .
moreover , the equations for the entries in each diagonal constitute a closed system , so that we have only to consider the variables @xmath426 , where @xmath427 is the index in the fock space .
we derive a set of rules to project the eom onto the lower diagonal . for tridiagonal matrices ( such as @xmath184 )
it holds that @xmath428 and for diagonal matrices ( such as @xmath429 ) it also holds that @xmath430 we apply these rules to the eom ( [ eq : linliouvmf ] ) for the perturbation matrix and we obtain the eom for the variables @xmath431 @xmath432 x_{\nu } \nonumber \\ & + & 2 n \sqrt{\nu ( \nu - 1 ) } x_{\nu - 1 } + 2 ( n + 1 ) \sqrt{\nu ( \nu + 1 ) } x_{\nu + 1 } \nonumber \\ & + & \sqrt{\nu } ( \delta \langle \hat{b}_{\ell - 1 } \rangle + \delta \langle \hat{b}_{\ell + 1 } \rangle ) ( \rho_{\ell ; \nu - 1 , \nu - 1}^{\rm ( th ) } + \rho_{\ell ; \nu , \nu}^{\rm ( th ) } ) \nonumber \\ & + & \frac{1}{2}\sqrt{\nu } [ -\delta \langle \hat{b}_{\ell - 1 } \rangle ( 2\nu -i j / \kappa ) \nonumber \\ & - & \delta \langle \hat{b}_{\ell + 1 } \rangle ( 2 \nu -i j / \kappa ) \nonumber \\ & + & 2 ( \delta \langle \hat{b}_{\ell - 1 } \hat{n}_{\ell - 1 } \rangle + \delta \langle \hat{b}_{\ell + 1 } \hat{n}_{\ell + 1 } \rangle ) ] \nonumber \\ & \times & ( \rho_{\ell ; \nu - 1 , \nu - 1}^{\rm ( th ) } - \rho_{\ell ; \nu , \nu}^{\rm ( th)})~.\end{aligned}\ ] ] we see that in the eoms the subdiagonal entries of the perturbation matrix on the site @xmath11 are coupled to some correlation functions computed on the neighboring sites . to allow for any possible spatial configuration of the correlations
makes the problem analytically unmanageable .
moreover , to compute the linearized equations on a finite number of lattice sites induces a discretization of the brillouin zone such that the numerical effort to solve the linear equations scales quadratically with the size of the lattice .
at this point it is necessary to introduce some physical hypothesis on the form of the instability , which allows us to consider only a finite number of degrees of freedom and to easily scan the brillouin zone of the lattice .
here we make the hypothesis @xmath433 using eq .
( [ eq : ansatzlatticeinstab ] ) we reduce the set of variables in the eom ( [ eq : eqmotionsubdiag ] ) to the sole @xmath434 .
the eom is rewritten in the form @xmath435 , where @xmath436 \delta_{\nu,\nu ' } + 2 ( n + 1 ) \sqrt{\nu ( \nu + 1 ) } \delta_{\nu ' , \nu + 1 } \nonumber \\ & + & \sqrt{\nu \nu ' } [ \cos{(\phi_{0 } ) } 2 ( \nu ' - \nu ) +
i \frac{j}{\kappa } \cos{\phi_{0 } } ] \nonumber \\ & \times & ( \rho_{\ell ; \nu - 1 , \nu - 1}^{(0 ) } - \rho_{\ell ; \nu , \nu}^{(0 ) } ) \nonumber \\ & + & 2 \cos{(\phi_{0 } ) } \sqrt{\nu ' \nu } ( \rho_{\ell ; \nu - 1 , \nu - 1}^{(0 ) } + \rho_{\ell ; \nu , \nu}^{(0)})~,\end{aligned}\ ] ] as can be verified with the formal substitutions @xmath437 the general solution of the eom is given by @xmath438 , with @xmath38 an eigenvalue of the matrix @xmath236 .
if @xmath439 < 0 $ ] for any @xmath38 , the perturbation @xmath184 damps out and the thermal state is stable . on the contrary , if there is at least one eigenvalue @xmath38 with a positive real part @xmath440 $ ] the thermal state is unstable .
the corresponding eigenvector is the unstable mode in the channel labelled by @xmath441 . to diagonalize the matrix
@xmath236 it is necessary to introduce a truncation @xmath442 in the set of variables @xmath434 , corresponding to a truncation in the fock space of the system .
we use @xmath443 and we have verified in some typical cases that the higher part of the spectrum remains unchanged when @xmath442 is increased .
in this section we outline the computation of the stability spectrum of a generic mean field homogeneous steady state with local density matrix @xmath239 .
this method is used in sec .
[ ssec : stability ] to compute the phase border between the stable and the unstable condensate region . the mean field eoms ( [ eq : motiontrace ] ) can be rewritten in the form @xmath444 $ ] , with @xmath350 a bilinear operator .
let us characterize now the action of the operator @xmath350 on the steady - state @xmath445 of the system .
we require the steady state to be such that @xmath446 , where @xmath156 and @xmath447 are indices in the fock space of the @xmath11th site .
it is easy to show that a sufficient and necessary condition for this to happen is @xmath448 , with @xmath449 .
we also require that the absolute value of a generic correlation function @xmath450 is constant in time , and hence @xmath451 .
the stationarity of the absolute values of the correlation functions can be easily checked on the numerical solution of the equations of motion .
still , we can not forbid a residual phase precession both in the off - diagonal elements of the density matrix and in the correlation functions .
the first point raises a problem for the linearization of the equations of motion because we do not obtain numerically a state that nullifies exactly the bilinear operator @xmath350 , i.e. we do not have a good starting point for the instability analysis .
it turns out that a chemical potential @xmath8 can always be fixed into the operator @xmath350 ( that we denote @xmath452 after the choice is made ) such that @xmath453 = 0 $ ] exactly . to explain why this is the case
, we consider the explicit form of the correlation functions @xmath454 , with @xmath455 . from the conditions above it follows that @xmath456 and
hence @xmath457 can only depend on the difference of its arguments and be of the form @xmath458 .
now we may argue that @xmath457 should have the linear form @xmath459 since we want the phase frequency of @xmath460 be twice that of @xmath89 in the condensate limit .
finally , @xmath457 should be a linear function of time if we require that we can remove the phase precession by a frame transformation of the basis vectors .
the result is that , in the steady state , the residual time - evolution of the density matrix is of the form @xmath461 .
this is precisely the contribution to the equation of motion produced by a term @xmath462 in the hamiltonian and can thus be eliminated by the appropriate choice of a chemical potential @xmath463 . with such choice of chemical potential , the equation of motion for the fluctuation
@xmath464 of the density matrix reads @xmath465 + q_{\mu}[\delta \rho_{\ell}(t ) , \rho_{\infty}]$ ] .
then , as we point out in sec . [ ssec : stability ] as well , some hypotheses on the spatial dependence of the unstable mode has to be made to relate the density matrices on different sites as @xmath466 $ ] , where we restrict to a one - dimensional geometry . according to the results of the numerical solution of the eom ( [ eq : motiontrace ] ) , here we consider the stability of the modes defined by @xmath467_{\nu\nu ' } = \cos{(\phi_{0 } ) } [ \rho_{\ell}]_{\nu\nu'}$ ] , that correspond to the generation of a standing wave in the lattice with wavelength @xmath468 .
the eom for the fluctuation matrix then reads @xmath469 $ ] with the linear operator @xmath71 given by @xmath470 \circ ( t^{(+)}[\cdot , \phi_{0 } ] + t^{(-)}[\cdot , \phi_{0 } ] ) + q_{\mu}[\cdot , \rho_{\ell}(t_{0})]$ ] .
the existence of a positive real eigenvalue of @xmath471 $ ] means that the steady state @xmath239 is unstable in the particular mode labeled by @xmath441 .
in this section we provide the eoms of the time - dependent correlation functions @xmath472 that follow from the mean field eom ( [ eq : motiontrace ] ) for the density matrix . for definiteness , in the following we assume a one - dimensional geometry so that the neighbors @xmath386 of the sites @xmath11 have indices @xmath473 and @xmath474 .
the eom of a generic correlation function is @xmath475 $ ] .
substituting the time derivative of the density matrix @xmath476 one obtains products of averages of operators on the site @xmath11 by the mean fields ( [ eq : corrset ] ) on the neighboring sites .
normal ordering the operators acting on the site @xmath11 , one obtains the following expression , which is linear in the correlation functions on the site @xmath11 but contains products between correlation functions on neighboring sites @xmath477 c_{p , q,\ell } \nonumber \\ & & + i u ( p - q ) c_{p+1,q+1,\ell } + 2 \kappa p q ( c_{1,1,\ell-1}+ c_{1,1,\ell+1 } ) c_{p-1,q-1,\ell } \nonumber \\ & & + \left [ - i j p ( c_{1,0,\ell-1 } + c_{1,0,\ell+1 } ) + 2 \kappa p ( 1 - q ) ( e^{i \phi } c_{1,0,\ell-1 } + e^{-i \phi } c_{1,0,\ell+1 } ) + 2 \kappa p ( e^{i \phi } c_{2,1,\ell-1 } + e^{-i \phi } c_{2,1,\ell+1 } ) \right ] c_{p-1,q,\ell } \nonumber \\ & & + \left [ i j q ( c_{0,1,\ell-1 } + c_{0,1,\ell+1 } ) + 2 \kappa q ( 1 - p ) ( e^{-i \phi } c_{0,1,\ell-1 } + e^{i \phi } c_{0,1,\ell+1 } ) + 2 \kappa q ( e^{-i \phi } c_{1,2,\ell-1 } + e^{i \phi } c_{1,2,\ell+1 } ) \right ] c_{p , q-1,\ell } \nonumber \\ & & - 2 q \kappa ( e^{i \phi } c_{1,0,\ell-1 } + e^{-i \phi } c_{1,0,\ell+1 } ) c_{p , q+1,\ell } - 2 p \kappa ( e^{-i \phi } c_{0,1,\ell-1 } + e^{i \phi } c_{0,1,\ell+1 } ) c_{p+1,q,\ell } \nonumber \\ & & + \kappa p(p-1 ) ( e^{2 i \phi } c_{2,0,\ell-1 } + e^{-2 i \phi } c_{2,0,\ell+1 } ) c_{p-2,q,\ell } + \kappa q(q-1 ) ( e^{-2 i \phi } c_{0,2,\ell-1 } + e^{2 i \phi } c_{0,2,\ell+1 } ) c_{p , q-2,\ell}~,\end{aligned}\ ] ] where it is understood that @xmath478 whenever @xmath479 and @xmath480 by the unitarity of the trace of @xmath476 .
the above expression shows that the eoms of the correlation functions are organized as an infinite hierarchy , since @xmath198 is also coupled to @xmath481 ( due to the local interaction ) and to @xmath200 , @xmath482 ( due to the dissipation ) .
hence we do not find a subset of eoms that decouples exactly from the others and that can be solved in a finite number of steps . in sec .
( [ sec : smalldens ] ) we use the different set of `` connected '' correlation functions @xmath483 , built with the fluctuation operator @xmath484 .
we remark that we always work in the schrdinger picture here : the operators @xmath485 are constant in time and depend parametrically on the choice of the quantity @xmath332 .
this set is totally equivalent from a mathematical point of view , but is motivated from a physical perspective if we interpret @xmath332 as the order parameter of a bose - einstein condensate and ( [ eq : corrset ] ) holds .
the new set of correlation functions hence describes the fluctuations about the condensate and , by construction @xmath227 .
we reiterate that the vanishing of the first - order correlation takes place only when the average is taken with respect to some density matrix @xmath486 such that ( [ eq : corrset ] ) is valid .
once we let the density matrix evolve in time , as we do in sec .
[ sec : instability ] , the operators @xmath485 do not change but the average @xmath487 = { \rm tr}[\hat{b}_{\ell } \rho_{\ell } ( t ) ] - \psi_{\ell}$ ] is in general different from zero . to obtain the eom for the correlation functions @xmath485 we expand the product of @xmath9 operators and @xmath332 @xmath488-numbers in the average , obtaining a sum of correlation functions @xmath198 .
then we substitute the eom ( [ eq : motionanycorr ] ) of each @xmath198 and again we expand the products of operators in the averages in terms of @xmath485 and @xmath332 .
this approach does not yield the general eom , analogous to ( [ eq : motionanycorr ] ) , because each choice of @xmath489 , @xmath51 produces a different set of correlations , but can be applied straightforwardly .
the eoms for the connected correlation functions are organized as an infinite hierarchy as well , and can not be reduced analytically to a closed subset .
moskalenko and d.w .
snoke , _ bose - einstein condensation of excitons and biexcitons _
( cambridge univ . press ,
cambridge , 2000 ) ; j. keeling , f.m .
marchetti , m.h .
szymanska , and p.b .
littlewood , semicond .
sci . technol .
* 22 * , r1 ( 2007 ) ; j. kasprzak _
_ , nature * 443 * , 409 ( 2006 ) ; m. wouters and i. carusotto , arxiv:1001.0660 . p. calabrese and j. cardy , phys .
* 96 * , 136801 ( 2006 ) ; c. kollath , a.m. luchli , and e. altman , _ ibid . _ * 98 * , 180601 ( 2007 ) ; a. silva , _ ibid .
_ * 101 * , 120603 ( 2008 ) ; m. mckel and s. kehrein , _ ibid .
_ * 100 * , 175702 ( 2008 ) .
m. greiner , o. mandel , t.w .
hnsch , and i. bloch , nature * 419 * , 51 ( 2002 ) ; b. paredes _ et al .
_ , _ ibid . _ * 429 * , 277 ( 2004 ) ; l.e .
sadler , j.m .
higbie , s.r .
leslie , m. vengalattore , and d.m .
stamper - kurn , _ ibid .
_ * 443 * , 312 ( 2006 ) .
t. gasenzer , j. berges , m.g .
schmidt , and m. seco , phys .
a * 72 * , 063604 ( 2005 ) ; m. cramer , c.m .
dawson , j. eisert , and t.j .
osborne , phys .
lett . * 100 * , 030602 ( 2008 ) ; m. rigol , v. dunjko , and m. olshanii , nature * 452 * , 854 ( 2008 ) ; g. roux , phys .
a * 79 * , 021608(r ) ( 2009 ) ; l.c .
venuti and p. zanardi , phys . rev .
a * 81 * , 032113 ( 2010 ) ; m. znidaric , t. prosen , g. benenti , g. casati , and d. rossini , phys . rev .
e * 81 * , 051135 ( 2010 ) .
k. sengupta , s. powell , and s. sachdev , phys .
a * 69 * , 053616 ( 2004 ) ; w.h .
zurek , u. dorner , and p. zoller , phys .
lett . * 95 * , 105701 ( 2005 ) ; t. prosen and i. pizorn , _ ibid .
_ * 101 * , 105701 ( 2008 ) ; c. de grandi , v. gritsev and a. polkovnikov , phys . rev .
b * 81 * , 012303 ( 2010 ) ; r.a .
barankov , arxiv:0910.0255 .
a. schmid , phys .
lett . * 51 * , 1506 ( 1983 ) ; s. chakravarty , g .-
ingold , s. kivelson , and a. luther , _ ibid . _ * 56 * , 2303 ( 1986 ) ; a. kampf and g. schn , phys . rev .
b * 36 * , 3651 ( 1987 ) ; s. chakravarty , s. kivelson , g.t .
zimanyi , and b.i .
halperin , _ ibid .
_ * 35 * , 7256 ( 1987 ) ; r. fazio and h. van der zant , phys
. rep . * 355 * , 235 ( 2001 ) . in ref .
@xcite polynomial behavior was found for observables like the condensate depletion in the weakly interacting regime well below the critical point .
such behavior is rooted in the momentum continuum corresponding to spatial fluctuations as obtained in the framework of bogoliubov theory . each onsite correlation ( or single momentum mode correlation ) decays at an exponential rate away from the critical point .
we note that the thermal equation ( [ eq : thevol ] ) admits solutions which are quasi - local in fock space , @xmath492 , but we do not observe that they are reached in a time evolution starting from an initial coherent state .
note the subtlety that , while a bogoliubov treatment with factorized higher order correlation functions yields @xmath493 ( @xmath494 the condensate density , cf . sec .
[ ssec : phases ] ) , the solution of our coupled low density equations in steady state yields the result @xmath495 with full average density .
scarola and s. das sarma , phys .
95 * , 033003 ( 2005 ) ; v.w .
scarola , e. demler , and s. das sarma , phys .
a * 73 * , 051601(r ) ( 2006 ) ; c. trefzger , c. menotti , and m. lewenstein , phys . rev . lett . * 103 * , 035304 ( 2009 ) ; s. diehl , m.a .
baranov , a.j .
daley , and p. zoller , phys .
lett * 104 * , 165301 ( 2010 ) .
cataliotti , s. burger , c. fort , p. maddaloni , f. minardi , a. trombettoni , a. smerzi , m. inguscio , science * 293 * , 843 ( 2001 ) ; m. albiez , r. gati , j. flling , s. hunsmann , m. cristiani and m.k .
oberthaler , phys .
rev . lett . * 95 * , 010402 ( 2005 ) ; o. morsch and m. oberthaler , rev .
* 78 * , 179 ( 2006 ) ; c. sias , a. zenesini , h. lignier , s. wimberger , d. ciampini , o. morsch , and e. arimondo , phys .
lett . * 98 * , 120403 ( 2007 ) ; s. levy , e. lahoud , i. shomroni , and j. steinhauer , nature * 449 * , 579 ( 2007 ) . | we study the nonequilibrium dynamics of a many - body bosonic system on a lattice , subject to driving and dissipation .
the time - evolution is described by a master equation , which we treat within a generalized gutzwiller mean field approximation for density matrices .
the dissipative processes are engineered such that the system , in the absence of interaction between the bosons , is driven into a homogeneous steady state with off - diagonal long range order .
we investigate how the coherent interaction affects qualitatively the properties of the steady state of the system and derive a nonequilibrium phase diagram featuring a phase transition into a steady state without long range order .
the phase diagram exhibits also an extended domain where an instability of the homogeneous steady state gives rise to a persistent density pattern with spontaneously broken translational symmetry . in the limit of small particle density
, we provide a precise analytical description of the time - evolution during the instability .
moreover , we investigate the transient following a quantum quench of the dissipative processes and we elucidate the prominent role played by collective topological variables in this regime . |
the k1v star hd 189733 hosts the transiting gas giant planet hd 189733b , at an orbital distance of 0.03 au @xcite .
in some ways this system is nearly optimally suited for study with transmission spectroscopy .
its short planetary orbital period ( 2.2 days ) , its great stellar brightness ( k = 5.5 mag ) , and its large planet - to - star radius ratio ( 0.15 ) combine for convenient observations of high sensitivity and high contrast . on the other hand ,
a significant negative is its intrinsic photometric variability , @xmath4% peak to valley , due to rotational modulation of star spots @xcite . as such , it is photometrically more variable than @xmath395% of stars in the kepler sample ( @xcite ; @xcite ; cf .
also figure 5 of @xcite ) . as
this paper will discuss , star spots complicate the interpretation of transmission spectra , including those of hd 189733b .
transmission spectra provide a powerful means to detect molecular species in exoplanetary atmospheres due to the long path length of the stellar light through the planetary atmosphere ( e.g. @xcite ; @xcite ; @xcite ) .
the differential transit depth of a molecular feature is directly proportional to the scale height of the atmosphere . for hot jupiters , with high temperatures and low mean molecular weight ( i.e. h@xmath5-rich ) ,
the differential transit depth in specific molecular bands is large enough to be observable with existing instruments . in particular , water vapor has bands with large cross sections observable in the near infrared .
upper limits or marginal detections have been reported of water in transmission for exoplanets hd 209458b @xcite , xo-1b ( @xcite ; @xcite ) , and xo-2b @xcite . from _
irac transit photometry at 3.6 @xmath6 m , 5.8 @xmath6 m , and 8 @xmath6 m , @xcite reported evidence of water vapor in the atmosphere of hd 189733b ( but cf .
the near - infrared transmission spectrum of hd 189733b in particular has garnered considerable attention , with the same nicmos data being re - analyzed multiple times , with @xcite being the latest .
the latter meta - analysis of the transmission spectra of @xcite , @xcite , and @xcite , predicted the amplitude of hd 189733b s 1.4 water vapor feature to be 300 ppm to 400 ppm .
other meta - analyses have been performed also ( e.g. @xcite ) .
similarly , from ground - based _ irtf _ near - infrared transmission spectra of hd 189733b , @xcite reported water vapor features including ones at 1.15 and 1.4 with amplitudes at least twice those reported here .
significant improvement has come with the _ hst _ instrument wfc3 .
recently , @xcite reported wfc3 observations of transits of wasp-12b , wasp-17b , and wasp-19b .
@xcite reported wfc3 observations of wasp-19b alone .
made under _
program 12181 ( p.i .
deming ) , those observations preceded the development of spatial scanning and like prior observations , produced marginal detections or upper limits on the 1.4 @xmath6 m water vapor feature .
similar wfc3 spectra have been reported for wasp-12b observed in staring mode under _ hst _
programs 12230 ( p.i .
swain ; @xcite ) and 12473 ( p.i .
sing ; @xcite ) .
coupling wfc3 to the new technique of spatially scanning the hubble space telescope ( _ hst _ ) has enabled detections of water vapor in transmission spectra for hd 209458b , xo-1b , and hat - p-1b under _ hst _ program 12181 ( p. i. deming ) and now hd 189733b ( program 12881 ; p. i. mccullough ) .
for some planets , clouds and/or haze can greatly affect our view of water vapor even if it exists in their atmospheres .
such is the case apparently for the smaller exoplanets gj 1214b @xcite and gj 436b @xcite that both exhibit featureless , flat spectra .
planets with no atmospheres , or insufficiently large scale heights of their atmospheres , would also present undetectably weak diagnostic spectral features in their atmospheres @xcite .
even if molecular features are observed in a clear atmosphere , a degeneracy exists between the composition and the characteristic temperature of the region probed by the data , since the scale height is proportional to temperature .
high resolution ( r@xmath7 ) , high sensitivity data can place joint constraints on the composition and temperature simultaneously ( e.g. @xcite ) .
due to their geometry of light rays grazing the terminator region @xcite , transmission spectra are more sensitive to clouds / haze and/or small atmospheric scale heights than day - side emission spectra , which have their own limitations .
@xcite attributed to water a 10 absorption feature in the emission spectrum of hd 189733b , and recently , @xcite have done so at 3.2 using phase curves rather than observations at secondary eclipse .
such measurements indicate potential for comparison of spectra obtained by the complimentary techniques of transits , eclipses , and phase curves ( cf .
@xcite and @xcite ) .
the spectral observations reported here were obtained in spatial scanning mode @xcite with the _ hst _ in order to rapidly collect a large number of photons from hd 189733 without it also saturating the detector . in scanning mode ,
the _ hst _
scans across a small segment of a great circle on the sky while the detector is integrating photo - generated charge from objects as they trail across the detector . since the first such observation of a transiting exoplanet , gj 1214b on april 18 , 2011 ( program 12325 , p.i . mackenty ; @xcite ) , spatial scanning has become the nominal method of spectroscopic observation of bright exoplanet host stars .
@xcite scanned one transit each of hd 209458b and xo-1b ; @xcite and @xcite each scanned multiple transits of gj 1214b and gj 436b respectively .
@xcite scanned one transit of hat - p-1b .
_ hst _ program 12181 ( p.i .
deming ) recorded one transit of hd 189733b using spatial scans .
however , the earth occulted the telescope s view of the star during the main part of the transit event between contacts 2 and 3 .
@xcite observed transits of hd 189733b in the near - infrared with wfc3 , under program 11740 ( p.i .
pont ) , although they did so prior to the advent of spatial scanning on _ hst_. in those observations , the very bright star saturated the detector even before each exposure began , because the detector does not have a mechanical shutter .
they salvaged light curves from the insensitive blue and red ends of the spectra .
based upon comparison of those measurements at @xmath31.1 and @xmath31.6 with measures by @xcite between 0.5 and 1.0 , gibson et al . tentatively suggested that haze dominates the transmission spectrum of hd 189733b both in the visible and the near - infrared .
however , they concluded that future observations of the sort reported here would provide the evidence to resolve the issue of whether exoplanetary hazes or clouds obscure spectral features in the near - infrared .
this paper is organized as follows . in section
[ sec : obs ] we describe the observations , which we analyze in section [ sec : analysis ] .
results are in section [ sec : results ] .
we separately discuss star spots ( [ sec : spots ] ) and the exoplanet s atmospheric spectrum ( [ sec : planet ] ) , and the combination of the two ( [ sec : combined ] ) .
we draw conclusions and summarize in section [ sec : summary ] .
_ hst _ program 12881 ( mccullough , p.i . ) was allocated five hst orbits to observe the transit of hd 189733b and five orbits to observe its eclipse .
the eclipse observations are reported by @xcite . in both cases ,
the event ( transit or eclipse ) occurs in the fourth of the five hst orbits . by design and common practice ,
the data from the first hst orbit is considered unreliable , as the telescope settles to its new thermal environment associated with a new orientation to the sun after slewing to the target , and charge traps in the hgcdte detector equilibrate to the unusual voltage associated with strong illumination .
discarding the data from the first orbit would leave two reliable hst orbits pre - transit , one in - transit , and one post - transit . ) , the n. c. one discarded data from the entire first orbit whereas the d. d. one discarded only the first eight scans of the first orbit . ]
we observed hd 189733b in transit with hst wfc3 on june 5 , 2013 ( table [ tbl-1 ] ) .
we used the g141 grism , to obtain slitless spectroscopy with spectral coverage from 1.1 to 1.7 at a resolution r = @xmath8130 and a dispersion of 4.7 nm pixel@xmath9 @xcite . in each hst orbit of the visit
, we obtained a set of thirty - two 5.97-s exposures , each with the rapid sample sequence of seven samples , and each with a 512 pixel by 512 pixel subarray .
as noted in the introduction , the spectra were obtained in spatial scanning mode .
we interleaved scanning first forward and then reversing direction , obtaining sixteen forward - reverse pairs of exposures in each orbit .
we had intended to locate the target on the detector such that its first - order spectrum would be entirely within a region read from a single amplifier ; however we miscalculated and the 1st order spectrum crosses an amplifier boundary ( at physical column 512 ) corresponding to @xmath10 .
a similar miscalculation caused the spectrum to scan off the top of the subarray and hence some of the samples were lost .
in the forward direction , samples 1 through 5 occur when hd 189733 s spectrum is on the subarray but during samples 6 , 7 , and 8 the spectrum is off the subarray and hence only the first four differences between samples are useful . in the reverse direction ,
samples 3 through 8 recorded hd 189733 s spectrum , and samples 1 , and 2 were lost ( figure [ fig1 ] ) .
the loss of 5 differences between samples of 14 total implies a loss in potential , poisson - limited , signal - to - noise ratio of 20% .
we also obtained a direct image as the first exposure of the visit ( figure [ fig1 ] ) ; the purpose of the direct image is to provide a high - resolution image to verify the location of the target with respect to any potentially contaminating stars in the same field of view .
hd 189733 is a double star ; the secondary is 4.0 magnitudes fainter in h band and is 11.2 away from the primary at a pa of 246 @xcite .
there are a few stars closer to hd 189733a in projection than hd 189733b , but they are much fainter ; in h band each is 1% to 2% of the brightness of hd 189733b , which itself is 2.5% of the brightness of hd 189733a .
because the nearby stars are so faint , and because our analysis technique isolates the spectrum of the target star in space and references it in time by dividing the in - transit spectrum by the out - of - transit spectrum , we expect any contributions of nearby stars to the transmission spectrum of the planet hd 189733b to be entirely negligible .
we needed to scan at @xmath32s@xmath9 to properly expose the detector and to avoid saturation .
also we wanted samples often enough to avoid overlapping the two spectra of the primary star hd 189733a and the secondary star hd 189733b , 11.2 apart .
combined , those requirements imply sampling every @xmath11 s , or faster .
of multiaccum sequences currently available for wfc3 , only the rapid sequence sufficed .
the observational data are interleaved in time ; 7 forward scan differences , each a 0.85-s exposure , followed by a 51 second gap as the scanning telescope reverses and the detector is prepared for the next exposure , followed by the 7 reverse scan - differences each one also 0.85-s long , and then a 51 second gap to return to the beginning of the pattern and to prepare the detector for the next forward - scan exposure . with two 6-s exposures
every 114 s , the observational duty cycle would be 10.5% , although because only 9 of the 14 multiaccum differences were useful , the duty cycle was 6.7% .
in total , we obtained @xmath12-s of exposure in transit , which is only 3% of the transit s duration .
the latter 3% is approximately half of the 6.7% observational duty cycle , because the earth occulted the target during the second half of the transit .
rather unsatisfactorily , a 3% duty cycle with the 2.4-m _ hst _ would be equivalent to observing at 100% duty cycle with a 0.4-m diameter telescope of the same end - to - end throughput . further improving wfc3 s duty cycle by any practical operational changes
would be beneficial for observations of stars such as hd 189733b .
two of us ( n.c . and d.d . ) analyzed the data nearly independently .
as planned from the outset to maintain independence , no computer code was shared between the two efforts , but both analysts had read or written the relevant literature , particularly descriptions by @xcite @xcite , @xcite , @xcite , and @xcite . the two resulting exoplanetary transmission spectra were compared and found to be very similar .
after discussions each analyst slightly revised his algorithms , which resulted in two final spectra . in this section
, we described one analysis , that of n.c . , the other method has been described already in its application to hd 209458 and xo-1 @xcite .
table [ tbl - lc ] lists the transit light curve of the n.c . analysis .
we describe the differences between the two analyses at the end of this section and list the difference between the two resulting spectra in the last column of table [ tbl - spectra ] . the r.m.s . of the differences , 52 ppm ,
is 0.75 times the median poisson noise associated with the 4-column binning , indicating that the differences in the resulting binned spectra are statistically less significant than the physical limit imposed by poisson statistics .
we begin the analysis with the set of intermediate multiaccum flexible image transport system ( ima.fits ) files , which are produced by the calwfc3 pipeline at stsci and have a number of corrections applied , including dark current subtraction and nonlinearity correction @xcite .
as described in section 2 , we discard the data from the visit s first hst orbit .
we sort the files into two sets , forward scans ( keyword postarg2 = 54.9 ) and reverse scans ( postarg2 = -2.0 ) , and analyze each separately until combining them near the end of our analysis .
the separate analyses of forward- and reverse - scans are very similar , differing only in the parameters defining the useful readouts and the pixels of interest , both of which differ slightly due to the forward and reverse scans not overlapping exactly
. from each 3-d multi - extension ima.fits file , we extract the difference images @xmath13 corresponding to the @xmath14th difference of consecutive nondestructive reads of the detector subarray .
we divide the 2-d difference images by a 2-d flat field obtained through the f139 m filter . because the in - transit spectrum of hd 189733 will be divided by its out - of - transit spectrum to form the planetary transmission spectrum , i.e. because the flat field is self - calibrated
, the choice of particular flat - field is unimportant .
flat - fielding makes bad pixels and cosmic ray hits easier to identify .
we perform such identification and accordingly correct the images by interpolation as described by @xcite .
we define fixed regions on the 2-d images corresponding to the target s signal area and a corresponding sky area .
we subtract the sky s value from each image .
next we sum each column over the stellar region , yielding a 1-d spectrum of hd 189733 for each difference image .
we average those into in - transit spectra ( in , from the fourth _ hst _ orbit ) and out - of - transit spectra ( out , from three other _ hst _ orbits ) .
we form each of nine planetary transit spectra , ( out - in)/out , associated with the nine difference images ( four _ forward _ and five _ reverse _ ) , which we then average to form a single planetary spectrum , subtract the wavelength - integrated mean , and bin every four columns to produce the final , differential planetary spectrum ( column 2 of table [ tbl - spectra ] ) .
we calibrated the wavelength as a function of detector column by comparing the ( unbinned ) wfc3 spectrum of hd 189733 to the spectrum of a k1 v comparison star @xcite.spex / irtf_spectral_library / data / k1v_hd10476.txt ] .
we adjusted the comparison star s spectrum to match hd 189733 s spectrum by multiplying the comparison spectrum by the telescope and instrument throughput @xcite , transforming it from energy units to photo - electron units , smoothing it to 4.7 nm ( 1 column ) resolution , high - pass filtering it and the wfc3 spectrum to emphasize weak spectral features in each , and finally , match the two spectra by linearly transforming the comparison star spectrum with an offset and dispersion that together constitute our wavelength calibration , which we estimate to be accurate to better than 4.7 nm ( 1 column ) .
additional details of this procedure are in @xcite . the position of the spectrum shifts slightly throughout each exposure in response to feedback from the fine guidance sensors .
variations in the scan rate perpendicular to dispersion introduce apparent photometric variations due to each row having a slightly shorter ( longer ) exposure time because the telescope s scan rate is faster ( slower ) than nominal ( cf .
figure 1 of @xcite ) . because such variations are achromatic
, they do not affect the differential exoplanetary spectrum .
we also examined the shifts in telescope pointing parallel to dispersion by methods similar to those described in @xcite .
the peak - to - valley variation of the shifts throughout the visit of program 12881 was typical for _ hst _
@xcite , @xmath30.1 column ( @xmath30.014 ) , similar to that for xo-1 and much less than the @xmath31.0 column shift of hd 209458b , both observed in program 12181 ( cf .
figure 4 of @xcite ) . adjusting the spectra of hd 189733 for its small ( @xmath15 column ) shifts made negligible difference in the final exoplanet spectrum
compared to not making the adjustments .
an important difference between the two analyses is their smoothing .
the analysis of d. d. smoothed its spectrum with a gaussian of fhwm=4 columns whereas the other one , of n. c. , used a box - car smoothing of full width 4 columns .
purely as a consequence of the choice of smoothing kernels , for an idealized , featureless , flat power spectrum of noise in the unbinned spectrum , the ratio of the noise of the resulting smoothed spectra should be @xmath16 .
stated another way , in terms of noise power , the d. d. spectrum has an effective bin width of 6 columns , or 1.5 times the n. c. spectrum s 4-column bin width .
the uncertainties in table [ tbl - spectra ] match this expectation .
we subdivided the available wfc3 data into nine separate wavelength - integrated `` white light '' curves , corresponding to the nine differences between multiaccum samples in which hd 189733 was on the subarray .
figure [ fig2 ] illustrates one example of the nine available .
we removed trends in each light curve as follows , for each of the nine multiaccum differences separately .
we fit a straight line by least - squares to the first points of the second , third , and fifth _ hst _
orbits , and divided the first point of each _ hst _ orbit ( including the in - transit fourth orbit ) by the fit .
we repeated the procedure for the second point , and so on , for all thirty - two scans per hst orbit .
table [ tbl - lc ] lists the nine light curves , both before and after de - trending . from the light curve of hd 189733b s transit
we obtain the mean level of the transit depth @xmath17 , which is important for comparison with measurements at other wavelengths from other instruments . otherwise , however , we are not as interested in the light curve as the differential transit spectrum described in the next section .
we adopted hd 189733b s ephemeris , orbital inclination , and normalized semi - major axis from @xcite .
we fit the light curve with the procedure of @xcite , using quadratic limb darkening coefficients @xmath18 and @xmath19 , which we derived by fitting two parameters of a quadratic law to the three - parameter limb - darkening law adopted by @xcite for @xmath20 . the resulting mean transit depth , appropriate for the integrated wfc3 bandpass , is @xmath21 , where the estimated uncertainty ( 100 ppm ) is the standard deviation of the fits to the nine light curves .
the observed value of @xmath17 is an upper bound in so far as it neglects unocculted star spots ( cf .
[ sec : spots ] ) .
the exoplanet transmission spectrum is the transit depth @xmath17 as a function of wavelength . in column 2 of table [ tbl - spectra ]
we report the differential transmission spectrum as ( in - out)/out , with its mean value subtracted , where in is the mean in - transit spectrum ( from the fourth _ hst _ orbit ) and out is the mean out - of - transit spectrum ( from the other _ hst _ orbits , neglecting the first one ) .
figure [ fig3 ] illustrates the spectrum extending from 1.13 to 1.64 in 28 bins .
each bin corresponds to 19 nm or four detector columns .
the simple ( in - out)/out estimate neglects limb darkening .
our more complicated analysis @xcite involves subtracting a scaled template spectrum from each exposure s 1-d spectrum and then fitting a transit light curve to the residuals to derive the differential transit depth at each wavelength . in the latter method ,
limb darkening can be included or not , and we experimented with either choice . although the mean `` white light '' transit depth is sensitive to the limb - darkening correction , applying a wavelength - dependent limb - darkening correction , or not , made negligible difference to the differential transit depth . to emphasize the latter point
, we computed one differential spectrum with no limb - darkening correction at all ( column 2 of table [ tbl - spectra ] ) , although we did model limb - darkening for the white - light transit depth in the previous section .
we estimated the uncertainties in column 3 of table [ tbl - spectra ] as follows .
we formed an individual 1-d stellar spectrum from each difference image .
we form a template spectrum from data taken out - of - transit .
we then scale and subtract the template from each of the individual spectra in order to minimize the r.m.s . of the resulting residuals .
for each spectral channel , we group the residuals into two sets , one in - transit and one out - of - transit , divide by the mean flux , and evaluate their respective standard deviations @xmath22 and @xmath23 .
we combine the latter in quadrature , accounting for the number of measurements in - transit or out - of - transit respectively , in order to estimate the uncertainties of each spectral channel , @xmath24 .
in figure [ fig4 ] we bin the data from column 2 of table [ tbl - spectra ] into seven points , with a channel width of 75 nm ( 16 detector columns ) . with that smoothing ,
the local maxima of the transit depths in the 1.15 and 1.4 water vapor features occur at the first and fourth data points , with a local minimum , the third point , in between . from those
, we estimate the amplitudes of the 1.15 and 1.4 features respectively are @xmath1 ppm and @xmath2 ppm greater than the local minimum at 1.3 .
the stated uncertainties are the quadrature sum of the uncertainties of the two points in each case .
the 1.4 feature in particular is detected at much greater statistical significance than @xmath25 ,
i.e. 200/47 , because the feature extends over two 75-nm wide points and the adjacent baseline extends over two or four points instead of just one . by averaging over four points ,
two inside the 1.4 feature and the two adjacent to it , the @xmath25 result becomes @xmath26 , commensurate with the feature s appearance in figures [ fig4 ] and [ fig5 ] .
the peak - to - valley amplitude of the 1.4 feature in figure [ fig3 ] , which has spectral resolution of 30 nm fwhm in the model or 19 nm ( 4 columns ) in the binned data , is @xmath3400 ppm or @xmath3350 ppm , respectively .
the peak - to - valley measurement of a noisy spectrum depends on the smoothing that reduces noise and hence reduces the height of the peak and the depth of the valley .
the similarity of the model to the data in figure [ fig3 ] illustrates that the data are consistent with a clear - atmosphere of solar - composition , with a volume mixing ratio for water of @xmath27 and a pressure scale height commensurate with a temperature of 700 k. the @xmath27 mixing ratio has been used in prior models as well ( e.g. @xcite , @xcite , @xcite ) and is consistent with the broad range of values presented in figure 3 of @xcite , @xmath28 to @xmath29 .
however , inspection of figure [ fig3 ] shows that slightly reducing our model s amplitude would improve its fit to the data .
hence , either the volume mixing ratio or the temperature should be smaller than the assumed values . @xcite explore the fitting of these data in greater detail and conclude that the h@xmath5o volume mixing ratio is sub - solar .
hd189733 is an active star ( boisse et al .
2009 ) , and transit observations exhibit ubiquitous star spot crossings ( pont et al . 2007 ) .
our wfc3 observations have only partial coverage during transit because of occultation by the earth , and no spot crossings are obvious in our data . in any case , because hd 189733 is a spotted star , we must address two effects _ unocculted _ star spots may have on the transmission spectrum . first , unocculted star spots can make the radius of a transiting planet appear larger in the blue than in the red , because unocculted spots make the star darker and redder than it would appear otherwise .
if in fact a star is _ darker _ away from the transit chord than a model assumes , that model will overestimate the size of the transiting planet . if in fact the star is also _ redder _ away from the transit chord than the model assumes , then the overestimate of the planet s size will be even greater in the blue than in the red .
second , at @xmath30k , hd 189733 s photosphere is too hot for water to exist , but in very cool star spots , water vapor could exist . water vapor absorption in unocculted star spots will make the star spots darker in spectral features of water than in the continuum , and hence make the radius of the transiting planet appear larger in water - vapor features than it would in the continuum @xcite .
quantitatively and specifically for hd 189733 , we show in this section that the first effect is likely to be significant but the second effect is not .
@xcite analyze star spots on hd 189733 and their potential effects on the transmission spectrum . in their analysis , they show that star spots modulate the stellar flux primarily due to stellar rotation , and also due to the changing set of spots and their characteristics
. however , latitudinal _ bands _ of star spots would not modulate the flux much , because as one spot rotates into view , another one rotates out of view .
if hd 189733 has a large polar spot , or one or more latitudinal band(s ) of spots beyond the latitudes at which the transit chord passes , then those unocculted spots will contribute to the apparent rise in the `` planetary '' transmission spectrum seen in the visible with acs @xcite and in the ultraviolet with stis @xcite .
@xcite were aware of such possibilities but chose to interpret that rise in the spectrum as evidence of rayleigh - scattering dust in the planet s upper atmosphere instead of spots on the star . in this paper
we match all of the transit data assembled by @xcite and the wfc3 spectrum derived here with a combination of a clear planetary atmosphere and a spotted stellar photosphere . as in prior work ( e.g. , pont et al .
2008 , 2013 ; sing et al .
2011 ) , we model star spots in aggregate and ignore limb darkening . in the spotted - star model , a fraction @xmath31 of the star s projected surface area is covered with spots , emitting with radiance @xmath32 , less than that of the photosphere , @xmath33 .
such a model predicts an apparent transit depth , @xmath34 where the first factor on the right is the transit depth that would be measured in the absence of star spots .
@xcite used simulated , rotating spotted stars modeled on hd 189733 to estimate the product of its spots area coverage @xmath31 and ( 1-contrast ) , @xmath35 where @xmath36 is the difference in the stellar flux from maximum to minimum , and @xmath37 is the scatter in the light curve , also due to rotational modulation ( not observational uncertainty ) . from six years of photometry at @xmath38 of hd 189733
illustrated in figure 3 of @xcite , we estimate @xmath39 and hence @xmath40 can be comparable to , or larger than , @xmath41 . because @xmath40 represents the product , spot area times ( 1-contrast ) , and contrast must be less than unity , in general , the fraction @xmath31 of the star s disk covered with spots must be greater than @xmath40 .
thus , for hd 189733 , we expect @xmath31 at times is greater than approximately @xmath41 .
a caveat worth noting is that the analysis of @xcite is most appropriate for stars for which a single active region dominates the photometric variability .
also , as noted by @xcite , geometric foreshortening de - emphasizes the contribution to the rotational modulation of spots near the stellar limb . with those caveats in mind ,
we admit the possibility that equation [ eq - f ] underestimates @xmath40 and hence @xmath31 could be even larger than 0.04 , if for instance hd 189733 has many active regions spread uniformly in longitude , i.e. bands of star spots .
the values of @xcite , which we use as a starting point in our analysis , have been corrected already for unocculted star spots that modulate the light curve , with time - dependent `` ac '' values for @xmath31 corresponding to flux changes in the optical between @xmath42% and @xmath43% ( table 3 of @xcite ) .
the most important corrections are to the acs and stis data , because the transmission spectrum at wavelengths longer than @xmath44 are less affected by the unocculted spot correction and/or by the rayleigh - scattering dust model ( figures [ fig4 ] and [ fig5 ] ) .
the unocculted spot corrections derived by the procedure of @xcite produced a discontinuity between the stis and acs transmission spectra in the region of overlap at @xmath30.5 . to eliminate the discontinuity , they applied a smaller correction than their method had indicated for unocculted spots at the epoch of the stis observations . while @xcite have corrected for an ac component
, here we investigate an underlying `` dc '' component and its potential effects on the transmission spectrum .
the dc component is an additional , hypothetical set of unocculted spots in bands or a polar cap .
with either such configuration , they would not modulate the light curve or precision radial velocities either .
polar spots have been deduced from doppler imaging of rapidly rotating stars , e.g. the k0 dwarf ab dor has a rotation period of 0.5 days and exhibits a prominent , long - lived polar spot @xcite , but we know little about the spots of more slowly - rotating stars such as hd 189733 . before proceeding to the more realistic phoenix atmosphere models , we begin with the approximation of black bodies for the stellar photosphere and the star spots , at temperatures @xmath45 and @xmath46 respectively . in this approximation ,
equation [ eq - r - tilde ] becomes @xmath47 where @xmath48 is planck s constant , @xmath49 is boltzmann s constant , and @xmath50 is the observing frequency .
the spots multiplicative effect on the apparent transit depth as a function of wavelength asymptotically approaches one value in the wien limit ( in the ultraviolet ) and another in the rayleigh - jeans limit ( in the far infrared ) : @xmath51 the difference between the two limits is @xmath52 where the first expression is generic and the second is specific for hd 189733b s nominal transit depth of 2.40% .
intuitively , the greater the fractional area of star spots , @xmath31 , the greater the increase in apparent transit depth in the ultraviolet relative to the thermal infrared .
counter - intuitively ( perhaps ) , the cooler the spots , the less their effect on the difference in apparent transit depth @xmath53 .
this is because in the wien limit , the radiance of cool spots is arbitrarily smaller than that of the photosphere , i.e. the spots are essentially black silhouettes , whereas in the rayleigh - jeans limit , radiance is proportional to temperature , so the _ difference _ in radiances ( rayleigh - jeans minus wien ) is proportional to the ratio of temperatures , and then so is the difference in apparent transit depth .
these characteristics of blackbodies inform our expectations for the effect of unocculted spots on the apparent transit depth .
pont et al .
( 2008 ) estimate that @xmath544000k , and @xcite find @xmath55 k for hd 189733 .
if the typical spot covering fraction @xmath31 is approximately 4% , then equation [ eq - delta ] predicts the difference in apparent transit depth between the ultraviolet and the infrared should be @xmath56 ppm , entirely due to unocculted star spots and not due to any variation in planetary radius with wavelength . from table 6 of @xcite ,
the observed difference in transit depth between 0.33 and 24 is 1100 ppm .
evidently , a majority fraction of the increased apparent transit depth in the ultraviolet compared to the infrared could be due to _ more _ unocculted star spots than that assumed by @xcite .
if so , then there would be no need for rayleigh - scattering , non - absorbing dust in the atmosphere of hd 189733b in order to match the slope in the apparent transit depth with wavelength from 0.3 to 1.0 as in the interpretation of @xcite . to achieve greater fidelity than the black body models permit , we have modeled the effect of unocculted star spots using phoenix nextgen model atmospheres ( hauschildt et al .
we use a 5000/4.5/0.0 ( teff / log(g)/[m / h ] ) model for the star , and for the spots we use nextgen models whose temperature we vary between 3000 and 5000k , in steps of 50 k via interpolation in the nextgen grid .
we constrain the log(g ) and metallicity of the spots to be the same as the stellar model , and we ignore center - to - limb effects on the star spot spectrum .
this procedure is consistent with similar previous work ( e.g. , pont et al .
2008 , 2013 ; sing et al . 2011 ) . at each adopted spot temperature , we vary the assumed fractional area covered by the spots , and we calculate the effect on the transit depth as a function of wavelength ( equation [ eq - r - tilde ] ) .
we calculate the @xmath57 of the difference between the modeled transit depth as a function of wavelength , using the observations in table 5 of @xcite . to start
, we adopt the limiting - case in which all of the observed planetary radius increase in the blue and uv is due to unocculted star spots and the spot - free radius ratio @xmath58 is 0.15459 , from the longest wavelength transit measurement at 24@xmath6 m ( @xcite ; @xcite ) . in that case
we find a best - fit spot temperature of 4250k and a best - fit area coverage of 4.3% .
this fitting process is well - posed in the sense that the wavelength of the upward bend in the curve is determined by the temperature adopted for the star spots ( the cooler the spots , the longer the wavelength ) , and the difference in apparent transit depth from the rayleigh - jeans end to the wien end of the curves is proportional to the product of the spots temperature and area coverage ( equation [ eq - delta ] ) .
our best - fit temperature agrees well with spot temperatures inferred from the amplitudes of spot crossings by this planet ; @xcite find that @xmath59 k , and @xcite find @xmath60k .
our best - fit model predicts @xmath61 , similar to the value @xmath62 that we estimated from the spot modulation of hd 189733 s light curve at @xmath63 m . in this case , the contamination of the 1.4 water - vapor feature is negligible , as we now discuss .
the model exhibits no spectral feature due to water absorption in the unocculted star spots , since 4250k is too warm for significant water absorption to be present .
hence even the worst - case assumption that all of the transit radius measurements are due to star spots does not produce significant contamination of the wfc3 water spectrum . that remains true if a much cooler spot temperature ( 3200k ) is adopted , for example .
such a spot is much cooler than the estimates noted in the previous paragraph , and is also cooler than much larger spots seen on other stars ( oneal et al .
1998 ) . moreover
, even spots of t = 3200k do not produce sufficient water absorption to contaminate our wfc3 measurement ; the predicted amplitude of false water absorption is @xmath340 ppm - much less than the @xmath3200 ppm feature at 1.4 that we measure .
we conclude that our wfc3 measurement of water absorption is due to the planetary atmosphere , not to an effect of unocculted star spots .
we constrain the h@xmath5o abundance from our observed transmission spectra of hd 189733b using model atmospheric spectra of its terminator .
we use the atmospheric modeling and retrieval technique of madhusudhan & seager ( 2009 ) and madhusudhan ( 2012 ) .
the model involves computing line - by - line radiative transfer in a 1-d plane - parallel atmosphere , assuming hydrostatic equilibrium .
the temperature profile and chemical composition are free parameters in the model , and , as such , has no constraints of radiative or chemical equilibrium .
this modeling approach allows one to compute large ensembles of models ( @xmath64 ) , and explore the parameter space of molecular compositions and temperature structure in search of the best - fitting models .
typically , the model includes opacity contributions from the major molecular species expected in hot jupiter atmospheres ( i.e. h@xmath5-h@xmath5 collision - induced absorption and line absorption due to h@xmath5o , co , ch@xmath65 , and co@xmath5 ) , though h@xmath5o is the most dominant molecule spectroscopically in the wfc3 bandpass .
the transmission spectrum is less sensitive to the detailed temperature structure , compared to a thermal emission spectrum , but does depend on the characteristic temperature in the upper - atmosphere above the infrared photosphere when viewed through the terminator . in the present work ,
we assume a characteristic 1-d temperature profile that has been derived from thermal emission measurements in the past ( madhusudhan & seager 2009 ) , and explore the range of chemical compositions that explain the transit data .
we find that our wfc3 transmission spectrum of hd 189733b can be explained very well by h@xmath5o absorption at the terminator .
the wfc3 spectrum exhibits an absorption feature at 1.4 with an amplitude of @xmath66 ppm that coincides with the expected features of gaseous h@xmath5o in the same band . as shown in fig .
[ fig3 ] , the data can be fit well by a solar composition atmosphere in chemical equilibrium .
such an atmosphere has a h@xmath5o mixing ratio of @xmath27 , where the mixing ratio is defined as number density relative to h@xmath5 . at the pressures probed by the observations ( @xmath67 bar ) ,
the temperature profile has typical isothermal temperature of @xmath3700 k , slightly cooler than prior p - t curves used to model nicmos transmission spectra of hd 189733b s terminator and its day - side emission spectrum from _ spitzer _ ( c.f . figures 5 and 8 of @xcite ) . in principle , allowing for higher temperatures ( e.g. 1000 k ) leads to larger scale heights thereby constraining the h@xmath5o abundance to be slightly sub - solar ( @xmath68 , @xcite ) . nevertheless , the evidence of an absorption feature in the h@xmath5o band is unambiguous , despite the degeneracy between the characteristic temperature and the h@xmath5o mixing ratio .
a featureless flat spectrum does not fit the data , ruling out the possibility of thick clouds fully obscuring the atmosphere observable along the planet s terminator . on the other hand ,
a fully - obscuring cloud deck at lower altitudes ( i.e. higher pressures , @xmath69 bar ) in the atmosphere is not ruled out by these wfc3 data .
combining our wfc3 spectrum of hd 189733b with previously reported spectra in other bandpasses reveal new constraints on its atmospheric composition .
the sum - total of previous data have led to an interpretation of hazes or clouds in the atmosphere of the planet ( pont et al .
2008 ; sing et al . 2011
; pont et al .
in figure [ fig5 ] we present a model that combines the effects a clear planetary atmosphere and unocculted star spots . for the gas giant planet , we model a clear atmosphere of solar composition , a mixing ratio for water of @xmath70 , and zero alkali metal lines ( na and k ) .
the rayleigh scattering of h@xmath5 in the planetary atmosphere contributes significantly at wavelengths less than 0.5 microns .
incidentally , that scattering may be responsible for the higher geometric albedo for the day side of hd 189733b observed at short wavelengths , @xmath71 , than at longer wavelengths , @xmath72 @xcite .
because the stis spectrum ( @xmath73 ) has a rayleigh - scattering slope , the star spot contribution must saturate in order that the model spectrum not rise too quickly into the ultraviolet .
that is accomplished with a cool , 3700 k spot temperature .
the star s model is a phoenix atmosphere model for the 5000 k stellar photosphere with unocculted spots of temperature @xmath74 k and spot fractional area @xmath75 .
although it fits the available data well , the combined model presented here is not unique : it is one example of many possibilities . as another example that could fit the transmission spectrum of hd 189733b
is an `` enhanced rayleigh '' model proposed for wasp-12b by @xcite in which the planet s atmosphere has both rayleigh scattering from aerosols and water vapor absorption .
a main purpose of this paper is to emphasize that fitting any transmission spectrum with a planetary atmosphere model must also consider the potential contribution and complication of unocculted star spots . in prior work ,
the case for haze / clouds in the atmosphere of the planet has been based on three key arguments : ( 1 ) the monotonic rise in the spectrum from the near - infrared to the uv consistent with strong rayleigh and mie scattering , ( 2 ) the lack of strong na or k features in the optical spectrum , i.e. presumably masked by a thick haze layer , and ( 3 ) the lack of striking @xmath76o absorption in previous datasets .
we are able to explain the monotonic blue - ward rise in the transmission spectrum by considering the effect of star spots on the measured transit depths in the relevant wavelengths ( section [ sec : spots ] ) .
secondly , the lack of na and k features in optical transmission spectra could be explained by either a low metallicity in the atmosphere @xcite and/or by condensation of na and k either in the low - temperature upper atmosphere or on the night side .
the latter model is a distinct possibility , given that hd 189733b is one of the least irradiated hot jupiters ( @xmath77 k ) . @xcite and
@xcite show the condensation curves of various species in pressure - temperature phase space . at @xmath78 bar ,
condensation temperatures of kcl and na@xmath5s are @xmath56k and @xmath79k , respectively , which can be higher than the temperatures in hd 189733b at the corresponding pressure on the terminator .
consequently , na and k could be condensed out of the observable atmosphere by arbitrarily large factors into their corresponding compounds ( e.g. kcl , na@xmath5s ) .
finally , the @xmath66 ppm amplitude of the 1.4 feature reported here ( and also in other exoplanets ; see 1 ) explains why it was not as obvious in previous datasets as had been anticipated .
we report detection of two water vapor features in the transmission spectrum of hd 189733b , a strong one at 1.4 and a weaker one at 1.15 .
their shapes and amplitudes are matched well by a solar - composition planetary atmosphere with a water mixing ratio of @xmath80 .
we investigate the possibility that the water vapor could exist in star spots , but even for very cool spots ( @xmath81 k ) , the amplitude of the predicted 1.4 feature is much too small to produce the observed feature . the ensemble , polychromatic transit spectrum of hd 189733b , from the ultraviolet to the thermal infrared , has been interpreted by @xcite , and references therein , as evidence for raleigh scattering in the planetary atmosphere . that interpretation requires specific adjustments for unocculted star spots in order to match the transmission spectrum across the entirety of the ultraviolet and visible .
we re - interpret the polychromatic transit spectrum using a clear planetary atmosphere and unocculted star spots . in the model presented here ,
rayleigh scattering in a clear planetary atmosphere contributes significantly to the slope in the planetary radius as function of wavelength for 0.3@xmath82 . in the range 0.5 @xmath83 ,
the observed slope is caused by unocculted star spots , not the planetary atmosphere .
since rayleigh scattering by transparent dust grains in the planetary atmosphere has been central to prior interpretations of hd 189733b s transmission spectra , our modeling of a clear planetary atmosphere and a spotted stellar atmosphere is a significant revision in the interpretation .
given the complexities of stitching together data from different instruments observing transits of a variable star at different epochs , and given the potential of omnipresent star spots to further confuse matters , either the star spot interpretation or the rayleigh scattering interpretation , or some combination of the two , remain viable .
prior work has rejected clear planetary atmosphere models based upon the predicted large - and - broad alkali metal lines in transmission spectra , na i ( 0.589 ) and k i ( 0.769 ) @xcite . in our model ,
the strengths of those features are reduced by arbitrarily large amounts because na and k are expected to precipitate out ( @xcite ; @xcite ) at the low temperatures ( @xmath3700 k ) that we used to model the @xmath3200 ppm amplitude of the water vapor feature at 1.4 .
the authors gratefully acknowledge everyone who has contributed to the _ hubble space telescope _ and the wfc3 , and particularly those responsible for implementing the spatial scanning , which was critical to these observations .
we thank in particular john mackenty and merle reinhart .
we acknowledge conversations with suzanne aigrain , david ciardi , suzanne hawley , leslie hebb , veselin kostov , rachel osten , frederic pont , neill reid , and david sing .
this research used nasa s astrophysics data system bibliographic services , and the simbad database , operated at cds , strasbourg , france , and was funded in part by _ hst _
grant go-12881 and origins of solar systems grant nnx10ag30 g .
lc hst program ( p.i . ) & 12881 ( mccullough ) + number of hst orbits & 5 + number of scans per orbit & 16 forward & 16 reverse + duration of scan ( s ) & 5.97 + scan rate ( arcsec s@xmath9)[pixels s@xmath9 ] & ( 2.00)[16.5 ] + peak signal on detector ( electrons per pixel ) & @xmath85 + grism ( @xmath86 ) & g141 ( 1.1@xmath6 m to 1.7@xmath6 m ) + detectory subarray size ( pixels ) & 512x512 + sample sequence & rapid + samples per scan & 8 + start of first scan ( hjd ) & 2456448.983024 + corresponding planetary orbital phase & -0.09948 + start of last scan ( hjd ) & 2456449.272364 + corresponding planetary orbital phase & 0.03094 + cccrcccc expstart mjd & expstart hjd & orbit & scan & sample & flag & counts & flux + 2456448.541735 & 2456449.044294 & 2 & 1 & 1 & 0 & 55426193 & 0.99996 + 2456448.541745 & 2456449.044304 & 2 & 1 & 2 & 0 & 55452872 & 1.00014 + 2456448.541755 & 2456449.044314 & 2 & 1 & 3 & 0 & 55434284 & 1.00011 + 2456448.541765 & 2456449.044324 & 2 & 1 & 4 & 0 & 55335761 & 0.99988 + 2456448.542414 & 2456449.044973 & 2 & -1 & 3 & 0 & 61415662 & 0.99990 + 2456448.542424 & 2456449.044983 & 2 & -1 & 4 & 0 & 61552322 & 1.00005 + 2456448.542434 & 2456449.044993 & 2 & -1 & 5 & 0 & 61495951 & 0.99995 + 2456448.542444 & 2456449.045003 & 2 & -1 & 6 & 0 & 61571866 & 1.00017 + 2456448.542454 & 2456449.045013 & 2 & -1 & 7 & 0 & 61610879 & 1.00067 + 2456448.543054 & 2456449.045613 & 2 & 1 & 1 & 0 & 55466554 & 0.99990 + 2456448.543064 & 2456449.045623 & 2 & 1 & 2 & 0 & 55489200 & 0.99993 + notes : the printed table is a truncated version of the electronic table , to illustrate the format .
columns , left to right , are modified julian date of the start of the exposure , the associated heliocentric julian date , the hst orbit in the visit , the scan direction ( 1 = forward ; -1 = reverse ) , the multiaccum sample , a data - analysis flag , the total number of photoelectrons from hd 189733 , and the associated normalized flux after detrending .
orbit 1 was not detrended and was ignored .
rrrrrrr @xmath87 & @xmath88 & @xmath37 & column & @xmath88 & @xmath37 & @xmath89 + ( @xmath6 m ) & ( ppm ) & ( ppm ) & & ( ppm ) & ( ppm ) & ( ppm ) + 1.1279 & -70 & 73 & 171.5 & -58 & 55 & -12 + 1.1467 & 15 & 67 & 175.5 & -25 & 54 & 41 + 1.1655 & 46 & 105 & 179.5 & 36 & 53 & 9 + 1.1843 & 3 & 87 & 183.5 & 13 & 52 & -9 + 1.2031 & -71 & 80 & 187.5 & -36 & 51 & -35 + 1.2218 & -77 & 70 & 191.5 & -63 & 50 & -14 + 1.2406 & -148 & 56 & 195.5 & -136 & 50 & -11 + 1.2594 & -32 & 62 & 199.5 & -91 & 50 & 59 + 1.2782 & -169 & 61 & 203.5 & -108 & 50 & -61 + 1.2970 & -45 & 69 & 207.5 & -21 & 49 & -23 + 1.3157 & -71 & 60 & 211.5 & -86 & 49 & 15 + 1.3345 & -50 & 66 & 215.5 & 20 & 49 & -71 + 1.3533 & 102 & 55 & 219.5 & 106 & 49 & -4 + 1.3721 & 117 & 61 & 223.5 & 75 & 49 & 42 + 1.3909 & 59 & 63 & 227.5 & 53 & 49 & 5 + 1.4096 & 183 & 77 & 231.5 & 195 & 49 & -11 + 1.4284 & 167 & 64 & 235.5 & 165 & 50 & 2 + 1.4472 & 76 & 71 & 239.5 & 73 & 50 & 2 + 1.4660 & -14 & 67 & 243.5 & -41 & 50 & 26 + 1.4847 & 156 & 75 & 247.5 & 85 & 51 & 70 + 1.5035 & -91 & 62 & 251.5 & 32 & 65 & -123 + 1.5223 & 65 & 61 & 255.5 & 60 & 51 & 4 + 1.5411 & -30 & 62 & 259.5 & -53 & 52 & 22 + 1.5599 & -22 & 72 & 263.5 & -27 & 52 & 5 + 1.5786 & 68 & 87 & 267.5 & 61 & 53 & 7 + 1.5974 & -69 & 75 & 271.5 & 12 & 54 & -82 + 1.6162 & -116 & 98 & 275.5 & -111 & 55 & -5 + 1.6350 & 30 & 84 & 279.5 & -131 & 56 & 162 + notes .
units are as indicated ; parts per million is abbreviated ppm .
the tabulated uncertainties apply to the differential transit depths ; an additional uncertainty applies to the overall depth - see text .
the first three columns refer to the analysis of n. c. ; columns 5 and 6 refer to the analysis of d. d. ; the last column contains the difference of the differential spectra , column 2 minus column 5 . | we report near - infrared spectroscopy of the gas giant planet hd 189733b in transit .
we used the _ hubble space telescope _ wide field camera 3 ( _ hst _
wfc3 ) with its g141 grism covering 1.1 to 1.7 and spatially scanned the image across the detector at 2@xmath0 . when smoothed to 75 nm bins ,
the local maxima of the transit depths in the 1.15 and 1.4 water vapor features respectively are @xmath1 ppm and @xmath2 ppm greater than the local minimum at 1.3 .
we compare the wfc3 spectrum with the composite transit spectrum of hd 189733b assembled by @xcite , extending from 0.3 to 24 .
although the water vapor features in the wfc3 spectrum are compatible with the model of non - absorbing , rayleigh - scattering dust in the planetary atmosphere @xcite , we also re - interpret the available data with a clear planetary atmosphere . in the latter interpretation
, the slope of increasing transit depth with shorter wavelengths from the near infrared , through the visible and into the ultraviolet is caused by unocculted star spots , with a smaller contribution of rayleigh scattering by molecular hydrogen in the planet s atmosphere . at relevant pressures along the terminator
, our model planetary atmosphere s temperature is @xmath3700 k , which is below the condensation temperatures of sodium- and potassium - bearing molecules , causing the broad wings of the spectral lines of na i and k i at 0.589 and 0.769 to be weak . |
flocking refers to a family of behaviours regularly seen in nature , including schooling of fish@xcite and flocking of birds and moths@xcite . whilst the features that are truly essential to flocks remain open to speculation@xcite
it is clear that the variety of biological@xcite and engineered@xcite systems in which flocking behaviour does , or could , prove useful is vast .
statistical mechanics has recently played an important role in uncovering and determining the source of many flocking phenomena@xcite .
one of the pioneers in flocking research was reynolds@xcite . within his work
he determined three behaviours characteristic of flocking : the desire to match the velocity of flockmates ( alignment ) , the desire to stay close to flockmates ( centring ) , and the desire to avoid collisions ( separation ) .
although reynolds original research was motivated by applications in computer graphics , his definitions have been adopted widely and implemented by researchers from a variety of disciplines .
reynolds coined the term _ boids _ to describe the generic self - propelled particles obeying these rules , and we shall continue that convention in this paper . within the physical science literature research has focused primarily on the steady state behaviour of flocking models and the non - equilibrium elements which cause solutions to differ from equilibrium systems .
many models involving only an alignment interaction have been shown to demonstrate a symmetry broken velocity state@xcite .
results have also been found which are strongly non - equilibrium in one and two dimensions , where equilibrium systems do not undergo comparable symmetry breaking phase transitions@xcite .
important work on alignment has included that of vicsek et al@xcite and toner and tu@xcite who shed light on the non - equilibrium processes involved .
a wide variety of models also exist in two or three dimensions incorporating one or several of reynolds rules besides alignment@xcite .
the lattice model of @xcite , which includes separation and alignment , shows a density - dependent symmetry broken state , for example .
similar results have also been derived using a network approach to flocking , whereby boids are nodes within a dynamic graph@xcite .
there are fewer obvious physical realisations of one - dimensional flocking , although analogy may be made with ring shaped aquariums , and with pedestrian dynamics in corridors @xcite .
however , research into one - dimensional systems certainly is interesting from a fundamental viewpoint : in particular , nonequilibrium phase transitions may be exhibited@xcite . in the one dimension alignment model of czirok et al@xcite and the one dimensional lattice model of oloan and evans@xcite novel states
not anticipated by higher dimensional treatments were found . in @xcite a symmetry broken steady state
was reported , whilst @xcite found a steady state which alternated its direction of motion on an unusually short time scale @xmath0 where @xmath1 is the number of boids in the system . in the one dimensional model of levine et al@xcite the effects of centring and separation
were considered and new behaviours were also found . our aim in this work is to introduce and study a one - dimensional lattice model involving all three of reynolds flocking criteria .
we do this by generalising the model of @xcite .
we demonstrate a variety of new flocking regimes : further to the alternating flock state found in @xcite we show that on small systems a homogeneous flock may occur ; also , the inclusion of centring may produce various dipole structures in which the velocities of the boids point towards the centre of the structure .
our generalisation also addresses some deficiencies in the dynamics of @xcite which implied full knowledge of neighbouring boids in the present work the dynamics is motivated from the sampling of a finite number of neighbouring boids and yields a smoother form [ eq .
( [ abias ] ) below ] for the expected update velocity of a boid as a function of the average neighbourhood velocity .
the paper is organised as follows . in section 2
we define and motivate the model to be studied . in section 3
we present numerical simulations of the model and examine the different regimes that may be observed . in section 4
we employ mean - field theory to provide further evidence for these regimes .
we conclude with a discussion in section 5 .
we first review the model of @xcite which consists of @xmath1 boids inhabiting a lattice of @xmath2 sites with unit lattice spacing . the global density , @xmath3 ,
is defined as @xmath4 .
each boid @xmath5 is defined by a position @xmath6 and velocity @xmath7 .
an update of the system consists of the following steps : 1 .
a boid is chosen at random .
a _ preferred direction _ @xmath8 or @xmath9 is determined from the velocity and spatial distribution of neighbouring boids .
@xmath10 is updated to @xmath11 with probability @xmath12 @xmath13 where the constant @xmath14 is a parameter of the model .
4 . the boid moves : @xmath15 is updated to @xmath16 .
we take the neighbourhood of site @xmath17 to be the sites @xmath18 , @xmath17 and @xmath19 . in the model of @xcite @xmath20
is given by the local majority velocity which we now define .
let the number of boids on a site with velocity @xmath21 be @xmath22 , the site density @xmath23 and site momentum @xmath24 .
the majority velocity of neighbouring boids is equivalent to the sign of the neighbourhood ( 3 site ) averaged momentum @xmath25 given by @xmath26 in what follows such a neighbourhood average of a quantity @xmath27 is represented by the hat symbol @xmath28 .
the preferred direction @xmath29 is then determined by @xmath30 thus in the model of @xcite the preferred direction is given deterministically by the local momentum and the model thus includes alignment . following a determination of @xmath20
, @xmath14 is the probability that the velocity is not updated to this value ( and instead to @xmath31 ) .
this probability is independent of the dynamical variables . in this paper
we generalise the model of @xcite at step ( ii ) above to include the effects of centring and separation as well as alignment this will be discussed in section [ caa ] .
we also generalise to consider the case where @xmath20 is itself a stochastic function of local variables : a boid may determine its preferred direction from the local variables as @xmath11 or @xmath9 with some probabilities ( as will be motivated below )
. however , since @xmath20 appears linearly in ( [ wpm ] ) , averaging over its possible values amounts to replacing @xmath20 by its expectation value @xmath32 in ( [ wpm ] ) : @xmath33 \label{wpm2}\end{aligned}\ ] ] where @xmath34 thus the important quantity is @xmath35 , the expectation value of the preferred direction which should be a function of the local density and velocity , for example .
the expectation value of the updated velocity is given by @xmath36 .
in the next subsection we propose an explicit form for @xmath32 and hence @xmath36 .
first it is useful to review the form of ( [ g ] ) .
the quantity @xmath37 is the velocity to which a boid is updated in the absence of noise , whereas @xmath38 is the certainty with which this outcome is attained .
the uncertainty in the outcome comes from two sources , the first is due to random errors , controlled by the parameter @xmath14 , where the boid moves against its preferred direction @xmath20 .
the second is from the stochastic nature of @xmath20 itself , which as we argue in section [ joffvca ] comes from the uncertainty with which a boid perceives the local flock properties and therefore determines its preferred direction . in many models @xcite
it has been shown that alignment alone is sufficient to produce complicated behaviour recognisable as flocking .
alignment is effected in our dynamics through @xmath39 . in this work
we consider @xmath39 to be of the form @xmath40 where @xmath41 is the _ neighbourhood average boid velocity_. the parameter @xmath42 controls how nonlinear @xmath36 is as the function of @xmath43 . in the limit where @xmath44 the majority rule case ( [ originalbias ] ) is in effect . in the limit @xmath45
we obtain a linear function .
the form ( [ abias ] ) addresses several deficiencies in ( [ originalbias ] ) .
firstly ( [ abias ] ) is analytic , whereas ( [ originalbias ] ) is non - analytic .
secondly with ( [ abias ] ) boids become sensitive to increasing majority size , as might be expected in physical systems . in the following section
we will see how this second intuitive feature can arise out of errors in local flock perception .
we now turn to a justification of the form ( [ abias ] ) from microscopic considerations .
we present an argument amounting to a simple algorithm for the determination of @xmath39 based on sampling of neighbours . in the model of @xcite boids
determine @xmath20 with perfect knowledge of all neighbouring boids ( and perfect ignorance of other boids ) ; hence a local majority is determined with certainty .
consider instead the situation where a boid can only make @xmath46 observations of its neighbours , each observation is of one boid selected randomly from neighbouring boids with replacement .
( also note that from the definition of neighbourhood a boid acts as a neighbour to itself . )
the majority rule applied to the sample of @xmath46 boids then requires a fixed observation time @xmath47 regardless of local density .
this is in contrast to the strict majority rule algorithm which requires information about all neighbours ( [ originalbias ] ) , there being no restriction on the potential number of neighbours .
consider a sample of @xmath46 neighbouring boids from the group of @xmath48 neighbours , containing @xmath49 rightward and @xmath50 leftward moving boids . with binomial probability @xmath51 , @xmath52 will be selected from the rightward and @xmath53 from the leftward moving groups :
@xmath54 the preferred direction is determined from the sample as @xmath55 .
the expected updated velocity @xmath56 is , taking @xmath46 is odd , @xmath57\;. \label{ubarm1}\end{aligned}\ ] ] using the substitution @xmath58 @xmath56 can be expanded as an odd polynomial in @xmath59 : @xmath60 a comparison of the two forms of @xmath39 ( [ ubarm ] ) and ( [ abias ] ) shows a qualitative correspondence between increasing @xmath46 and @xmath61 .
for example , we have from ( [ ubarm ] ) @xmath62 whereas expanding ( [ abias ] ) in powers of @xmath59 yields @xmath63 in the two limiting cases @xmath64 and @xmath65 the correspondence is exact .
for intermediate values of @xmath46 it is possible to define an approximate correspondence through a function @xmath66 using one of several schemes . anticipating our mean field treatment below ( section [ mf ] ) we choose to match the gradient of @xmath39 with respect to @xmath21 at the origin : @xmath67 this yields , for example , @xmath68 and @xmath69 .
to summarise we have argued that the form ( [ abias ] ) , parametrised by @xmath61 , for the expected updated velocity qualitatively corresponds to ( but has a simpler form than ) the result of taking a sample of @xmath46 neighbours .
the correspondence @xmath66 is quantified in ( [ betam ] ) .
centring , the tendency to move towards the local centroid @xmath70 , defined as @xmath71 can be introduced through @xmath39 in an analogous way to alignment .
if a relative importance of @xmath72 is assigned to the effect of centring over alignment in any observation then ( [ abias ] ) can be generalised to @xmath73 the desire for separation , typically through hard core repulsion , is very restrictive in one dimension . if instead we consider separation as some cost associated with moving through a dense region then it amounts to a tendency to move away from the centroid ( opposite to centring ) . in order to include these effects together we introduce a capacity , @xmath74 , which is the capacity at which the relative strengths of these two effects neutralise . when the local density @xmath75 is greater than @xmath74 the desire for separation exceeds the desire to centre , and vice versa .
thus , our most general form for @xmath39 is chosen as @xmath76 \right\ } \label{acabias}\;.\ ] ] note that ( [ acbias ] ) is the limit @xmath77 of ( [ acabias ] ) . at this stage
it is useful to summarise the parameters of the model with our choice of @xmath39 : the capacity @xmath74 is the density where the separation tendency surpasses the centring tendency ; @xmath72 determines the relative strengths of spatial density ( centring and separation ) and alignment effects .
thus @xmath72 and @xmath74 determine the relative importance of reynolds three flocking criteria .
the parameters @xmath14 and @xmath61 may both be considered as sources of noise acting through @xmath39 : @xmath61 introduces sensitivity to different field strengths while @xmath14 introduces uncorrelated errors . with reference to section [ joffvca ]
it is useful to consider the noise from @xmath61 as arising out of stochastically sampling the local flock to determine the boid s preferred direction whereas @xmath14 encompasses additional error sources once the preferred direction has been selected .
numerical simulations were undertaken for a variety of noise levels ( @xmath61 and @xmath14 ) , and flocking parameters ( @xmath72 and @xmath74 ) .
run time constraints limited our investigations to systems of size upto @xmath78 sites and times upto @xmath79 timesteps , where one timestep is sufficient time for each boid to be updated once on average ( @xmath1 updates ) .
we found four characteristic and robust behaviours which we now discuss , alongside a myriad of intermediate behaviours .
the first characteristic behaviour ( figure [ fig1]a ) is the _
disordered state_. this state has homogeneous density and a global velocity of zero , global velocity being the mean velocity of a boid in the system @xmath80 this state persists at noisy parameter sets ( high @xmath14 , low @xmath61 ) especially where reynolds effects exist at equivalent strengths or where separation dominates ( @xmath81 and @xmath82 ) .
the _ alternating flock _ is the second regime ( figure [ fig1]b ) .
this system undergoes a repeating pattern of cohesive traversals interspersed by rapid reversals , similar to that of @xcite . the _ homogeneous flock _ as shown in figure [ fig1]c is a flocking state which complements the alternating flock .
the homogeneous flock has a homogeneous density and fixed non - zero global velocity , @xmath83 .
this regime is found to be unstable where noise is high or where global density is low , and is resilient only to small levels of centring or separation .
the homogeneous flock is often established following an alternating transient as shown in the lower part of figure [ fig1]c .
the homogeneous flock can also be observed in the model of @xcite but at larger densities , or smaller system sizes than those considered there the final regime observed is a _ dipole state _
( figure [ fig1]d ) , which exists in systems where centring is the dominating reynolds effect .
a dipole is a dense , localised structure . as we shall discuss in detail ( in section [ dr ] ) a dipole
has a well defined centre with @xmath39 pointing inwards trapping boids , and a sharply decaying density profile at its edges .
a system of one or many dipoles can exist on the array , either as direct neighbours or separated by depopulated domains . as can be seen in ( figure [ fig1]d ) a slow coarsening process
is observed in a system of many dipoles , whereby small dipoles are eliminated and the boids are absorbed by larger dipoles .
the dipole states exist for large values of @xmath72 and @xmath74 i.e. relatively weak alignment and separation .
these four regimes all have readily characterised behaviours and occupy characteristic regions of parameter space as shown in figure [ fig2 ] which describes the case @xmath77 ( absence of separation ) .
figure [ fig2 ] may be thought of ( and we shall refer to it ) as a phase diagram .
however , the different regimes are not truly `` phases '' as the boundaries between them depend on system size and , as we shall see below , the homogeneous flock regime actually disappears in the limit of large system size . in the following subsection we shall discuss in more detail the flocking and dipole regimes .
the alternating flock was originally studied in @xcite where only alignment was present .
here we find that it persists in systems with all three reynolds effects present provided alignment dominates ( @xmath84 ) and @xmath61 remains finite . in its traversing stage
a single dense group of boids ( the flock ) , confined within a non - spanning section of the lattice , slowly flows across the system at a consistent non - zero velocity and diffuses . at a later time
the _ reversal mechanism _ begins : this is initiated by a fluctuation near the front of the flock causing a local reversal in momentum .
the fluctuation moves against the direction of the flock and grows in density and momentum until it passes entirely through the flock inverting its velocity . at this time
the traversing behaviour of a flock with opposite velocity is established .
initiation of reversals can occur either internally or by collision with boids external to the flock .
while not all large momentum fluctuations initiate a full reversal , they occur often enough to reverse the flock before it spreads out to occupy the whole lattice through diffusion .
the reversals occur stochastically with no fixed period but there is a well defined mean time between reversals
. in the alternating flock instabilities occur at the front edge of the flock .
in @xcite it was established that the reversal timescale for alternating flocks is @xmath0 where @xmath1 is the number of boids .
the argument is that this is the timescale on which the front edge becomes susceptible to _ relatively _ large momentum fluctuations , due to its low density . in our model
there is a new form for @xmath39 , but logarithmic timescales appear to persist for the alternating regime as shown in figure [ fig3n]b . as was shown in figure [ fig1]c , from many initial conditions a homogeneous flock develops after an initial transient alternating flock .
however many homogeneous flocks are temporarily unstable towards an alternating flock profile even after a long time in the homogeneous regime .
such a process occurs for the homogeneous flock of figure [ fig1]c after 58000 timesteps . in the breakdown process
a fluctuation in momentum , initially confined to a few sites , creates a disturbance which moves through the flock in a manner similar to the reversal mechanism in alternating flocks .
figures [ fig4n]a , b , c show the progression of the fluctuation through the flock to a point where global velocity is completely inverted .
before the flock returns to a homogeneous flocking regime there are several further reversals .
the localisation of the boids and high reversal rate during this period characterise a transient alternating flock state ( see figures [ fig4n]d , e ) .
it appears that the temporary failure of the homogeneous flock arises out of a large local fluctuation in momentum . during flock reversal
the initial fluctuation acquires boids from the oncoming flock and continues to grow in momentum and stability .
although it appears that the size and type of disturbance required might vary for different systems , the rate at which such disturbances occur over the system should be proportional to the number of nucleation sites . within the homogeneous flock
, there is a homogeneous density so the number of nucleation sites is proportional to @xmath2 , thus we expect the breakdown time for the flock , @xmath85 to behave as @xmath86
. in figure [ fig3n]a we take the mean time a homogeneous flock of positive momentum first reverses and attains a negative global velocity as the flock breakdown time .
we fitted the data by @xmath87 , where we expect @xmath88 to be approximately @xmath89 .
the constant @xmath90 is our approximation to the time for the fluctuation to pass through half the flock .
such a prediction appears to closely match computer simulations as illustrated in figure [ fig3n]a .
the previous paragraph implies that for large systems the homogeneous flock will always become unstable due to nucleation of momentum fluctuations . to quantify this ,
consider how the breakdown of the homogeneous flock to an alternating regime is initiated by a sequence of successive microscopic flips against the preferred direction sufficient to invert the momentum .
the number of flips required for this to occur will be proportional to the local density , @xmath3 , and number of sites over which @xmath39 is determined ( range , @xmath91 in our simulations ) .
given that a flip occurs with probability @xmath92 , flock breakdown will be initiated with probability @xmath93 in any timestep .
therefore one expects the typical time @xmath94 for the flock to flip to be @xmath95 .
this implies that for the homogeneous flock to be stable over times @xmath96 where @xmath97 , a density @xmath98 would be required .
we conclude that for finite range and density no homogeneous flocks will be stable in the thermodynamic limit ( @xmath99 )
. however , as we have seen in figure [ fig2 ] on intermediate size systems there are well defined regions of parameter space where homogeneous flocks do exist .
finally we examine the impact of varying the strength of centring and separation on flocking states .
the alternating flock persists even when centring is a relatively strong effect , whereas the homogeneous flock typically becomes unstable towards an alternating flock under such conditions .
thus in the phase diagram in figure [ fig2 ] , for large enough @xmath61 , increasing @xmath72 leads to transitions from the homogeneous to the alternating flock and thence to the dipole state .
the phase diagram with the inclusion of separation ( finite @xmath74 ) is qualitatively similar . perhaps surprisingly
, decreasing @xmath74 does not substantially affect the alternating flock .
the dipole regime is characterised by single or multiple localised structures .
these may be stationary or slowly moving .
the density is maximal at the centre of the dipole and the density profile is symmetric about the centre .
however , to the left of the centre @xmath39 is positive but to the right of the centre @xmath39 is negative .
thus the structure traps particles in such a way that the net flux of boids across the centre of the dipole is zero . in figure [ fig5 ] we illustrate two systems , one at high density and the other at densities typical of our other simulations ( approximately 1 ) . in the case of high density
it can be seen that the system is dominated by adjacent large dipoles .
each of these dipoles has an exponentially decaying density to a boundary with another dipole , and a steady value for @xmath39 trapping boids . in this system
there is slow redistribution of boids across boundaries towards larger dipoles . in the second system density
is lower and the exponential density profiles are not obvious ; instead there is one large dipole and several smaller dipoles .
the dipoles span a very small fraction of space and are not in direct contact with others .
the gaps between the dipoles are essentially homogeneous regions of very low density but small dipoles ( of size 25 boids in the figure ) regularly form and evaporate away .
occasionally a small dipole arising out of the low density region can grow to supersede larger dipoles . in figure [ fig6 ] details of a typical coarsening process ,
whereby small dipoles are eliminated and large dipoles grow , is illustrated . figure [ fig6 ] shows the development of a system , after an initial transient period when many dipoles form , through an intricate coarsening process .
figure [ fig6]a shows the system coarsening into two large dipoles and eventually ( at a later time not included in [ fig6]a ) one large dipole emerges .
this may be seen from figures [ fig6]b , c which plot the statistics of the sizes of the dipoles remaining in the system .
the regions in between the dipoles is a low density domain containing independent boids which combine sporadically to form small transient dipoles . in the fully coarsened system
there is a relatively small probability of a boid being elsewhere than the single dominant dipole
all dipoles of intermediate size disappear
. in the low density regime ( as in [ fig5]b ) the exchange of boids between dipoles occurs through the homogeneous low density domains .
dipoles may lose boids through stochastic fluctuations at their boundaries .
if two adjacent dipoles have such noise - induced boid loss at different rates , there will be a net flux across this domain towards the more stable dipole .
numerically we have observed that in fact smaller dipoles have a higher loss rate than larger ones
. therefore the rate of boid loss depends on the fine details of the dipole profile which depend on the dipole size .
( an assumption of a pure exponential density profile for all dipoles would imply no dependence of the rate of boid loss on dipole size . )
other effects might also be important , for example the greater mobility of smaller dipoles .
dipoles are created and sustained where centring is the dominant effect . within dipole systems increasing alignment ( decreasing @xmath72 )
is most noticable in the resulting greater mobility or _ wandering_@xcite of dipoles ( and independent boids ) ; a significant process since it hastens the coarsening of the system , especially in the initial stages .
wandering can occur in any size dipole but is most prominent in smaller dipoles . typically wandering is initiated by a net flux of boids across the centre of the dipole causing the centre to shift .
this motion can be induced or enhanced by alignment , since the matching of velocities in such a process favours the collective motion . with alignment sufficiently strong , this process can be self - reinforcing to the extent that a sequence of movements or sustained translations ( flocking ) become possible .
dipole systems are characterised by their central region of high density .
however , we might expect such a profile to be suppressed by separation .
the inclusion of separation in dipole systems leads to a dramatic alteration of behaviour : where alignment is a small effect , and @xmath100 , a system of _ fixed - density dipoles _ can be observed .
fixed - density dipoles have reduced density at the centre , but the density is constant within an extended central region and only near the boundaries does one see the sharply decaying density profile , thus the density profile has a flat top . within fixed - density dipoles alignment
plays a much more important role than in comparably sized dipoles in the absence of separation , since at the centre of fixed - density dipoles the separation and centring tendencies tend to cancel allowing alignment to dominate .
the relative stability of large and small fixed - density dipoles is left as an open question .
fixed - density dipoles can , for example , split from their centre and wander more substantially .
these processes are appreciable in figure [ fig7 ] where the mobility of the large fixed - density dipoles is second only to that of solitary boids .
in order to understand some of the regimes observed numerically above , we now develop , from a mean field treatment of the full dynamics , a continuum model in which the density @xmath101 and neighbourhood velocity @xmath21 are continuous functions . in
so doing we are able to identify a linearly stable flocking solution comparable to the homogeneous flock and a piecewise continuous solution comparable to the dipole regime .
furthermore we show that these solutions exist in numerical iterations of the mean - field equations .
the equations satisfied by the mean - field density , @xmath101 , and mean - field neighbourhood velocity , @xmath21 , are derived in appendix a and read @xmath102 \label{continuumv } + \frac{1}{2\rho}\left[- v\frac{\partial^2}{\partial x^2 } \rho+\frac{\partial^2}{\partial x^2 } ( \rho g)\right]\end{aligned}\ ] ] where @xmath103 \label{mfu}\ ] ] and @xmath104 equation ( [ continuumrho ] ) is rather easy to understand : note that , since @xmath39 is the mean velocity of boids , @xmath105 is a mass current , thus the equation is a continuity equation with a current and a diffusive term .
the equation for the neighbourhood velocity @xmath21 ( [ continuumv ] ) , on the other hand is more complicated .
we now look for steady state solutions of the above equations , for which we set the lhs of ( [ continuumrho],[continuumv ] ) to zero .
further , we look for homogeneous solutions where spatial derivatives are zero
. the global density may be freely chosen , in these solutions @xmath106 , whereas velocities are determined by ( [ continuumv ] ) as @xmath107 which yields @xmath108 thus , there are either one or three solutions to the mean field theory which are homogeneous in density
. for all parameter sets one solution to this is @xmath109 , the disordered solution , whilst for certain values of @xmath61 , @xmath72 and @xmath14 two _ flocking solutions _
@xmath110 exist , corresponding to the homogeneous flock .
the existence of the flocking solutions is determined by the gradient in @xmath39 with respect to @xmath21 at the origin . since @xmath39 is a concave function of @xmath111 a flocking solution of ( [ transcendental1 ] ) requires that @xmath112 .
thus the existence of this flocking solution is dependent on the following criterion being satisfied : @xmath113 this is possible in systems with low @xmath14 , but strong alignment ( so that @xmath114 is large ) .
we show in this section how it is possible to determine a further steady - state solution of ( [ continuumrho],[continuumv ] ) where the density is an exponentially increasing or decreasing function of @xmath115 and @xmath21 is a positive or negative constant respectively .
domains of these solutions may be pieced together so that the density is continuous to form dipole structures .
first , we assume a fixed value of @xmath116 implying by definition ( [ omegamf ] ) a density profile @xmath117 where @xmath118 is a constant .
next we propose a solution whereby within any domain @xmath39 and hence velocity are fixed .
then , from ( [ continuumrho ] ) we have @xmath119 which implies @xmath120 the velocity @xmath21 must similarly be steady and ( [ continuumv ] ) implies @xmath121 which after substitution of @xmath39 from ( [ gomega ] ) simplifies to @xmath122\;. \label{v(omega)}\ ] ] the condition required for a non - homogeneous solution is the consistency of ( [ gomega ] ) and ( [ mfu ] ) , eliminating @xmath39 we obtain @xmath123\right\}\label{dipoleomega}\\\ ] ] alongside the homogeneous density solutions , @xmath124 , there are potentially dipole solutions of ( [ dipoleomega ] ) @xmath125 . correspondingly @xmath21 takes values @xmath126 via ( [ v(omega ) ] )
. one of these has rightward @xmath39 and exponentially increasing density , the other a leftward @xmath39 and exponentially decreasing density .
a dipole is a localised structure consisting of a left hand region of exponentially increasing density and a right hand region of decreasing density . to satisfy periodic boundary conditions
it is necessary that the system consists of complementary domains .
consider a boundary between the dipole domains : the two exponentially decaying or exponentially growing domains meet with equal density at such a point .
the density gradient and other qualities are locally antisymmetric about the boundary ; hence the net flux of boids will be zero and the boundary stationary ( this would not be true for a dipole domain in contact with a homogeneous domain ) . within each domain
the presence of a boundary is not felt , and so domains of differing width and height can neighbour one another . hence non - symmetric dipoles of differing sizes can be neighbours .
our solutions are therefore the set of coupled non - overlapping dipole systems ( domain pairs @xmath17 ) , each defined uniquely by the set of dipole centre positions , heights and widths .
these solutions are comparable to the full stochastic system at very high density ( see figure [ fig5]a ) . note that the larger dipoles in the sparser system ( figure [ fig5]b ) also approximate a pair of domains with the same fixed @xmath39 and @xmath116 . in numerical iterations of the mean field equations ( see next subsection )
the dipole structures were rounded at the cusps which form at domain boundaries .
one expects this rounding and the physics at the cusps to be described by higher order terms in the lattice spacing @xmath127 which have been ignored ( see appendix a ) .
numerical iteration of the dynamical mean field equations ( [ continuumrho],[continuumv ] ) allows further analysis of these states and comparison to full system dynamics .
the iteration was carried out by rediscretising ( [ continuumrho],[continuumv ] ) onto a lattice .
we found the flocking ( @xmath128 ) , non - flocking ( @xmath129 ) and dipole solutions ( @xmath130 ) to be present , in addition to systems comparable to the fixed - density dipole .
the alternating flock however was never observed , nor were systems comparable to sparser dipole regimes . where a flocking solution is predicted , and not made unstable by the presence of centring ( see section [ stab ] ) a homogeneous flock with the anticipated global velocity ( [ transcendental1 ] ) was always observed to emerge from any initial conditions . in dipole regimes we found that the system iterated to a dipole system consisting of many dipoles .
however the mean field dipole system does not exhibit a further coarsening process comparable to the full stochastic model .
this is likely due to the absence of noise driven fluctuations which drive the coarsening process . on the other hand , where the initial boundary between domains does not fall perfectly on a lattice space boundary or centre a very slow redistribution of boids
does occur in the mean - field system , but apparently at only an exponentially slow rate .
this redistribution was not observed to cause the disappearance of any dipoles , nor did it appear to favour transfer to larger dipoles ( is the case in the full stochastic system ) .
we have developed a linear stability analysis of the disordered and flocking solutions to the mean field equations .
this is done by linearising ( [ continuumrho ] , [ continuumv ] ) , computing the evolution of the fourier components then checking whether positive ( unstable ) eigenvalues exist . as the details
are rather involved we do not present them here but quote some results . in the case
@xmath131 , when the flocking solutions exist they are linearly stable and the disordered solution is unstable ; when the flocking solution does not exist the disordered solution is stable . however , for sufficiently large @xmath72 an instability will occur in both flocking and disordered solutions .
this instability is expected to be with respect to stable dipole solutions , although we did not explicitly check the stability of dipole solutions .
thus the equality in ( [ solc ] ) defines a mean - field phase boundary between the disordered and possible flocking states .
comparing to figure [ fig2 ] this curve is in approximate agreement to the boundary between the disordered and alternating flock regimes .
however for other parameter sets the boundary is less accurately predicted .
generally the boundary becomes increasingly more accurate as the density increases . as well as
the disordered / flocking phase boundary , the instability of the flocking solutions at large @xmath72 is also consistent with figure [ fig2 ] where for large @xmath72 dipole solutions appear .
the only discrepancy remains the absence of alternating flocking solutions within the mean field theory : in regions where one sees an alternating flock in the full stochastic system the mean field theory has a homogeneous flock solution . as noted above the alternating flock
was not expressed within the mean field theory , even in the time dependent form ( [ continuumrho ] , [ continuumv ] ) .
as we shall now discuss , the mean field theory is not capable of describing the alternating flock which is driven by fluctuations .
however the flocking solution of the mean field theory does appear to correctly predict the velocity of the alternating flock between reversals i.e. the solutions of ( [ transcendental1 ] ) proves to be in good agreement with transient results from the full stochastic system .
we also note that we found numerically that the mean field theory correctly describes the evolution of the profile of an alternating flock as it spreads between reversals .
however , whereas an alternating flock in the full stochastic system will inevitably reverse due to some fluctuation in momentum ( as described section [ fr ] ) the mean - field profile will keep evolving until the homogeneous flock is attained .
similarly , due to the absence of fluctuations the homogeneous flock will not reverse for any system size . clearly , to reproduce the alternating flock ,
a noise term should be added to the mean field equations ( [ continuumrho ] , [ continuumv ] ) thus turning them into langevin equations .
this would be done , following the proposal of @xcite , by adding a noise term to the velocity equation ( [ continuumv ] ) .
the equations ( [ continuumrho ] , [ continuumv ] ) would then be similar to those of @xcite although our velocity equation , which we derived from the microscopic dynamics , is rather more complicated and contains additional terms .
in this work we have extended the flocking model of @xcite to include all three of reynolds three effects alignment , centring and separation implemented in combination or independently . within this model we demonstrated the robustness of the alternating flock highlighted in @xcite to the addition of other reynolds effects .
also we showed the existence of two new regimes : the homogeneous flock and dipole structures .
these are consistent with results from a mean field treatment which , in particular , correctly predicts the form of the dipole solutions .
furthermore , we investigated the coarsening process in the low density dipole regime .
we also derived from microscopic considerations the proposed form for the function @xmath39 ( [ abias ] ) .
the algorithm presented in section [ joffvca ] is not only intuitive but significantly is a fixed time algorithm , by which it is meant that the algorithm would involve sampling a fixed number of neighbouring boids whatever the density .
such algorithms are aspired to in flocking applications @xcite and are presumed to underly natural boid decision making @xcite . several ways to extend the algorithm could be to include a density , velocity or historical dependence in determining the sample , anisotropic spatial sampling or considering a direction selection rule other than majority .
for example boids could require unanimity@xcite amongst their observed neighbours to believe that any orientation is a worthy choice .
reynolds three effects were defined as required behaviours for simulating real flocks . in this paper
we have shown several states which may characterise real systems .
firstly the _ homogeneous flock _ , which has many analogies in two or three dimensions , can be realised in some systems approximating one dimension : for example skaters confined to the outside of a rink , or fish in a ring shaped tank ( as in the boston aquarium ) .
fixed density dipole _ is characterised by sharp edges and fixed internal density , and the ability to wander either slowly , or with the inclusion of alignment , quickly and with a sustained orientation .
these qualities , as well as as the ability to split from within , could describe a number of biological systems . finally consider the reversal mechanism present in the alternating flock ; such sharp reversals of direction are characteristic of the manoeuvres seen in many flocks , though they are poorly described by many models .
the numerical results obtained for the full stochastic system have some noteworthy similarities with previous studies .
a flocking state is observed in the one dimensional continuous space model of @xcite which appears to resemble an alternating flock .
dipole - like regimes have been observed in @xcite , where dense states are supported by a centring interaction . as with our dipoles
the density at the centre of these systems grows rapidly with boid number .
this effect is suppressed by a hard core repulsion , producing states similar to our fixed density dipoles . in @xcite
two dimensional systems also display wandering and oscillating ( circling ) behaviour . in @xcite
a two dimensional cellular automaton demonstrates a density - dependent state with symmetry - broken velocity .
this density dependence arises from a _ hard _ capacity being placed on the number of boids which can occupy a single site ; in effect the flocking state is destroyed by separation at high density .
our mean field treatment has similarities with the continuum model in @xcite .
not only are the same states demonstrated , but the terms within the equations are comparable also .
special note is made in @xcite of a term of the form @xmath132 which is present in our equation ( [ continuumv ] ) .
this term relates to domain competition , and aids in an understanding of the reversal mechanism and breakdown of the homogeneous flock . as ( [ continuumv ] )
describes very well the shape of a spreading flock , and it is believed that this shape allows certain fluctuations to cause the flock reversal mechanism @xcite , it might be hoped that the addition of noise into ( [ continuumrho],[continuumv ] ) would allow the alternating flock to be further analysed .
the coarsening process in low and high density dipole regimes may also be observed by allowing fluctuations to enter the dynamics at the boundaries of the dipoles .
finally we mention one further variation of the model of @xcite that we have studied @xcite . to imitate the effects of inertia we modified step ( iii ) of our dynamics such that boids which decide to change direction must proceed through a temporary zero velocity state . numerical simulations and a mean field treatment ( which we do not present here )
indicate that under such a scheme flocking solutions become more prominent and have higher global velocities .
although the flipping mechanism proceeds in much the same manner as the two velocity case , the rate of reversal for both the alternating and homogeneous flock is reduced this is not surprising since more unfavoured flips are required to create momentum fluctuation leading to a sustainable reversal of the velocity .
+ * acknowledgements * we thank mike cates and jamie wood for carefully reading the manuscript and for helpful discussions .
the work was supported in part by epsrc programme grant gr / s10377/01 .
1 e. bonabeau , l. dagorn , p. freon _ j. phys . a : math . gen _ * 31 * l731-l736 ( 1998 ) p. d. lorch , d. t. gwynne _ naturwissenschaffen _ * 87 * 370 - 372 ( 2000 ) c. w. reynolds , _
computer graphics _ * 21 * 4 ( 1987 ) a. jadbabaie , j. lin , and a. s. morse _ ieee transactions on automatic control _ * 48 * 988 ( 2003 ) r. olfati - saber , r.m .
murray _ ieee transactions on automatic control _ * 49 * 1520 ( 2004 ) y. liu , k. k. m. passino , m. polycarpou _ ieee transactions on automatic control _ * 48 * 1848 ( 2003 ) s. hubbard , p. bank , s. sigurdsson and k. g. magnussen _ ecological modelling _ * 174 * 359 - 374 ( 2004 ) w. j. crowther , _ proc .
engrs part g ( journal of aerospace engineering ) _ * 218 * 111 - 124 ( 2004 ) j. toner , y. tu and s. ramaswamy _ annals of physics _ * 318 * 170 ( 2005 ) t. vicsek , a. czirok , e. ben - jacob , i. cohen , and o. shochet , 1226 ( 1995 ) j. toner and y. tu , 4326 ( 1998 ) j. toner and y. tu , 4828 ( 1998 ) n. shimoyama , k. sugawara , t. mizuguchi , y. hayakawa , m. sano 20 ( 1996 ) h. j. bussemaker , a. deutsch , and e. geigant 26 ( 1997 ) m. aldana and c. huepe , _ j. stat .
phys . _ * 112 * 135 ( 2003 ) t. vicsek , a. czirok , i. j. farkas and d. helbing , 182 - 189 ( 1999 ) m. r. evans _ braz . j. phys . _
* 30 * 42 ( 2000 ) a. czirok , a. barabasi , t. vicsek , 209 ( 1999 ) o. j. oloan and m. r. evans , _ j. phys .
a. _ * 32 * l99 ( 1999 ) j. r. raymond ( masters thesis , university of edinburgh , unpublished ) , h. levine , w j. rappel and i. cohen 017101 ( 2001 )
in our mean field theory we shall approximate the density @xmath133 ( the number of particles at site @xmath115 and time @xmath134 ) and the momentum @xmath135 ( the number of right going minus the number of left moving particles at site @xmath115 and time @xmath134 ) by continuous functions and derive the evolution of these funcions by ignoring certain correlations .
we begin with the dynamics of site @xmath115 in a single update , as defined by @xmath136 and @xmath137 .
consider @xmath133 , in a single update this can increase if a boid selected from a neighbouring site moves into site @xmath115 , or can decrease if a boid is selected from site @xmath115 , with an associated probability : @xmath138 further we assume @xmath133 and @xmath141 vary smoothly on the time and length scales effective in a single update .
with these assumptions we expand about site @xmath115 to second order ( in lattice spacing @xmath127 ) and about @xmath134 to first order ( @xmath142 ) .
@xmath143 by introducing @xmath39 ( [ g ] ) the following equation can be determined . in this equation
there is a diffusion term arising from the choice of random stochastic updates , and a current term controlled by the mass current @xmath105 .
@xmath144 in a similar way one can develop an equation for @xmath141 beginning with the full dynamics at @xmath145 , @xmath146 taking the expectation value and expanding to second order in @xmath127 and first order in @xmath134 : @xmath147 using ( [ g ] ) this can be simplified to its final form @xmath148 defining the ( neighbourhood ) velocity as @xmath149 yields using ( [ rhoeq],[phieq ] ) a velocity equation : @xmath150 we now write @xmath36 , the expected update velocity , as a function of @xmath151 and @xmath133 only . to simplify we keep only leading order terms in @xmath127 and obtain @xmath152\right ) } { \tanh\beta}\ ] ] where @xmath153 finally , setting @xmath127 to one in ( [ rhoeq],[phieq ] , [ omegaa ] ) yields ( [ continuumrho][omegamf ] ) . | we study a one - dimensional lattice flocking model incorporating all three of the flocking criteria proposed by reynolds [ _ computer graphics _ * 21 * 4 ( 1987 ) ] : alignment , centring and separation .
the model generalises that introduced by o. j. o loan and m. r. evans _ j. phys .
a. _ * 32 * l99 ( 1999 ) .
we motivate the dynamical rules by microscopic sampling considerations .
the model exhibits various flocking regimes : the alternating flock , the homogeneous flock and dipole structures .
we investigate these regimes numerically and within a continuum mean - field theory . *
pacs : * 05.70.fh 64.60.cn 87.10e |
hopping conduction in solids governed by strong electric fields is in the focus of intensive theoretical and experimental study since several decades [ see , for instance , chapter 7 in ref . and references therein ] . in recent years
, particular interest to this research area has been caused by growing device applications of amorphous organic and inorganic materials in which the incoherent hopping transitions of charge carriers between spatially and energetically distributed localized states dominate the optoelectronic phenomena [ see , for instance , ref . and references therein ] .
one of the mostly discussed topics is whether the differential negative conductivity ( ndc ) , i.e. , the decreasing conductivity with increasing electric field , is possible in the hopping regime .
the discussion was , to much extent , provoked by the reports on the apparent decrease of the drift mobility with rising electric field at relatively high temperatures and low field strengths in disordered organic materials.@xcite this apparent decrease of the mobility with increasing electric field was reported to be succeeded by the increase of the mobility at higher field strengths .
however , the self - consistent effective - medium theory for drift and diffusion at low electric fields@xcite does not show any decrease of the mobility with increasing field .
furthermore , it has been shown experimentally@xcite and theoretically@xcite that the apparent decrease of the mobility with rising field at low field strengths is an artifact .
the experimental data were obtained by the time - of - flight technique , in which charge carriers are created close to one surface of a sample with a given thickness @xmath0 and the transient time @xmath1 is measured , which is needed for charge carriers to reach the opposite surface of the sample at a particular strength of the applied electric field @xmath2 . then the drift mobility is calculated as @xmath3 . however , at high temperatures and low electric fields the current transients in the time - of - flight experiments are determined mostly by diffusion of charge carriers rather than by their drift . therefore , using the drift formula one strongly overestimates the mobility .
it is the presence of the field strength in the denominator that leads to the apparent `` increase '' of the mobility at decreasing @xmath2.@xcite if one uses at low fields and high temperatures the diffusion formulas instead of the drift ones , then no decrease of the mobility with increasing field can be claimed at low electric fields.@xcite this result does not exclude , however , the possibility of the ndc in the hopping regime .
bttger and bryksin@xcite and shklovskii et al.@xcite have suggested analytical theories for the mobility and conductivity decreasing with increasing electric field in various disordered materials .
remarkably , this effect of the negative differential conductivity is to be expected at high field strengths .
this regime succeeds the very strong increase of the mobility with rising field,@xcite and does not precede it at lower fields as claimed on the basis of the drift equations.@xcite the decreasing conductivity with increasing electric field at high field strengths has been observed experimentally for hopping transport in lightly doped and weakly compensated crystalline silicon .
@xcite the hopping transport mode in such systems at low electric fields had been described theoretically in all detail , @xcite which made these systems particularly attractive for studying the new non - ohmic effects .
shklovskii et al .
@xcite developed an analytical theory , which predicted the ndc effect in the lightly doped weakly compensated semiconductors .
the experimental observations in lightly doped and weakly compensated crystalline silicon appear in qualitative agreement with his theoretical predictions .
furthermore computer simulations of levin et al.@xcite confirmed qualitatively the existence of the ndc effect , though no quantitative comparison with the analytical theory @xcite has been attempted .
recent interest in the ndc effect has been caused by its importance for construction of memory devices .
these devices typically contain conducting particles embedded into a nonconductive material .
for such devices , made from both inorganic @xcite and organic @xcite materials , ndc and switching phenomena have been reported . since electrical conduction in the materials , which are currently being tried for device applications , is dominated by hopping of charge carriers , it is necessary to study the possibility of the ndc in this regime in more detail . in the present paper we report on the theoretical study of hopping transport in high electric fields . in section [ sec - model ]
we describe the theoretical model and briefly outline the analytical approach suggested by nguyen and shklovskii@xcite for the limit of extremely high electric fields . in section [ sec - mc - inf ]
we present our results obtained by straightforward monte carlo computer simulations and show the range of applicability for the analytical theory of nguyen and shklovskii . in section
[ sec - our - theory ] we further develop the analytical theory whereby we improve its agreement with the results of computer simulations .
in particular , the analytical dependence of the drift mobility on the concentration of charge carriers comes in better agreement with the simulation results .
section [ sec - finite ] is dedicated to the ndc .
a new numerical algorithm has been developed to study the ndc effect theoretically .
numerical results obtained in the framework of this algorithm confirm qualitatively the conclusion of nguyen and shklovskii on the possibility of the ndc in the hopping regime .
however , the field dependence of the drift velocity of charge carriers obtained numerically differs essentially from the one predicted so far .
@xcite we suggest in section [ sec - finite ] a further development of the analytical theory , improving essentially its agreement with numerical results .
concluding remarks are gathered in section [ sec - conclusions ] .
aiming to clarify whether the ndc effect is inherent for the hopping transport regime , we consider first , following nguyen and shklovskii,@xcite the simplest possible model a three - dimensional array of isoenergetic sites with a random spatial distribution with the concentration @xmath4 .
each site can be either empty or occupied by a single electron .
energies of electrons are equal on all sites so that no energy disorder and no electron - electron interactions between different sites are taken into account . only in the final part of section [ sec - finite ] we study the effect of the energy disorder on the ndc .
an electric field @xmath5 is put along the negative direction of the axis @xmath6 , so that the drift velocity of the negatively charged electrons is directed along the @xmath6 axis .
conduction takes place due to tunnelling hops of electrons between the localization sites .
the rate @xmath7 for an electron hop from site @xmath8 to site @xmath9 is determined as @xmath10 where @xmath11 is the distance between the sites , @xmath12 is the localization length , @xmath13 is the elementary charge , @xmath14 is the boltzmann constant , @xmath15 is the temperature , and @xmath16 are the occupation factors of the sites , @xmath17 .
the function @xmath18 is related to energy the gain or the energy loss during the jump : @xmath19 in the limit of infinite electric field , the factor @xmath20 $ ] reduces to the heaviside s function @xmath21 . below we will assume that @xmath22 . as a measure of length , we introduce the typical distance between the neighboring sites @xmath23 . using this simple model , nguyen and shklovskii@xcite
have shown analytically that the effect of the ndc is inherent for hopping transport .
let us consider briefly their arguments starting from the case of infinitely high fields @xmath2 and extremely small electron concentrations @xmath24 . under such circumstances
each electron can be treated independently from the others and electrons can move only toward the increasing values of their @xmath25 coordinate . in fig .
[ fig - traps ] this is the direction to the right . at each jump ,
an electron moves along the axis @xmath6 to a distance @xmath26 , so that its drift velocity can be estimated as @xmath27 , where @xmath28 is an average time between jumps ( dwell time ) .
a dwell time @xmath29 for hopping from the site @xmath8 is of the order of @xmath30 , where @xmath31 is the distance from site @xmath8 to its nearest neighbor `` to the right '' , i.e. with co - ordinate @xmath25 larger than @xmath32 .
in other words , @xmath31 is the maximum radius of a hemisphere centered at the site @xmath8 that does not contain any other sites ( fig .
[ fig - traps](a ) ) .
if @xmath31 is much larger than @xmath33 ( the typical distance between the neighboring sites ) , such an empty hemisphere can be considered as a trap for electrons .
the contribution of traps with radii in the range @xmath34 $ ] to the average dwell time @xmath28 is proportional to @xmath35 , and also to the probability of the corresponding configuration of sites , @xmath36 : @xmath37 this integral is easy to evaluate , taking into account that the integrand has a sharp maximum at @xmath38 .
consequently one obtains for the current density @xmath39 @xmath40 therefore , one can conclude that in the limit @xmath41 the current is determined by hemispherical traps ( fig .
[ fig - traps]a ) with an `` optimal '' radius @xmath42 . in the case of finite electric fields ,
a hemispherical trap is not an efficient one , because an electron has a possibility to move in the energetically unfavored directions , and thus to escape the trap ( for example , along the dotted arrows in fig .
[ fig - traps](a ) ) . according to nguyen and
shklovskii,@xcite an `` optimal '' trap for an electron in large though finite electric fields @xmath2 consists of a hemisphere to the right and of a cone to the left of the site on which an electron is captured , with a chain of sites along the @xmath6 axis that provides an easy path into the trap ( fig .
[ fig - traps]b ) .
the height @xmath43 of the cone is chosen so that it is equally hard to escape the trap in all directions taking the chain along the @xmath6 axes into account : @xmath44 .
therefore , the smaller is the field , the larger is the volume of a trap with the same dwell time , and consequently the smaller is the probability @xmath45 of finding such a trap .
it means that the average dwell time @xmath46 decreases with decreasing field strength , and concomitantly the current density @xmath47 increases with decreasing field .
this is the essence of the physical mechanism that causes the ndc effect.@xcite to obtain an expression for the current density , one can substitute the volume of the trap shown in fig . [
fig - traps]b , @xmath48 , instead of the hemispherical trap volume , @xmath49 , into the integral ( [ eq - shklovskii4 ] ) .
the result reads @xmath50.\ ] ] this is the mathematical expression for the ndc .
the approach leading to this expression is applicable only for fields @xmath51 . in smaller fields ,
the assumption that almost every jump is directed along the axis @xmath6 is violated .
therefore one should expect that eq .
( [ eq - shklovskii10 ] ) overestimates the current density for @xmath52 .
equations ( [ eq - shklovskii5 ] ) and ( [ eq - shklovskii10 ] ) are valid only if the concentration of electrons @xmath24 is small as compared to the concentration of `` optimal '' traps , @xmath53 $ ] . in the opposite case , @xmath54 , the `` optimal '' traps are almost always occupied and play a negligible role . in such a case the most important traps , which determine the drift velocity of electrons , are the ones whose concentration is equal to @xmath24 .
one can estimate the electron drift velocity as @xmath55 , where @xmath56 is a radius of the most important traps , @xmath57 is their dwell time , @xmath58 is their capture crosssection .
assuming @xmath59 , one obtains for the current flow @xmath60 .
for an infinitely large field , the radius @xmath56 is defined via @xmath61 that gives @xmath62 .
consequently , the current density is@xcite @xmath63.\ ] ] the corresponding expression for the concentration range @xmath64 in the case of finite electric fields was also obtained by nguyen and shklovskii [ see eq .
( 11 ) in ref . ] .
the case of almost filled sites , @xmath65 , is similar to the case of almost empty sites , @xmath66 due to electron - hole symmetry .
the current density is a symmetrical function of the electron concentration : @xmath67 .
nguyen and shklovskii@xcite also emphasized that a special consideration is needed for the case of half - filled system , @xmath68 .
they have shown that the concept of directed percolation can be used to obtain the current density at infinitely high electric fields . in the half - filled system
the trapping of electrons does not play any role , because ( due to the electron - hole symmetry ) it does not change the electron concentration on the infinite cluster which is responsible for the current .
current is determined by electron jumps to distances @xmath69 $ ] , where @xmath70 is the percolation threshold of a directed percolation problem .
the number of pairs of sites with distances @xmath69 $ ] in the infinite cluster per unit area is @xmath71 , where @xmath72 is a transversal correlation length of the percolation cluster , and @xmath73 is a critical index.@xcite the current density is equal to@xcite @xmath74 nguyen and shklovskii@xcite have also obtained the value of the percolation threshold @xmath75 and that of the correlation length index @xmath76 .
the above arguments of nguyen and shklovskii@xcite provide an analytical theory of non - ohmic hopping conduction , based on the concept of the trapping - determined transport .
the theory is valid for the case of large electric fields in two concentration ranges : @xmath77 and @xmath78 .
most remarkably , this theory predicts the effect of the ndc .
also a theory for the case of the half - filled system @xmath79 for infinitely high electric fields @xmath80 has been suggested based on the directed - percolation - approach.@xcite below we present our numerical study of the field - dependent hopping conductivity .
it shows the range of validity for the analytical theory of nguyen and shklovskii.@xcite furthermore , the analytical theory is developed below in order to improve the agreement between the analytical and numerical results .
in order to calculate the electron drift velocity and the current density at high fields , we used a monte carlo approach . in the limit of infinitely high fields the direction of the electron motion
is prescribed .
therefore it was possible to simulate by a monte carlo algorithm the motion of an electron in an infinite medium along the field direction and therefore to avoid any size effects . without loosing generality one can restrict the maximal length of electron transitions involved into the algorithm by a reasonably large value @xmath81 . in order to simulate the @xmath14-th monte carlo step in the electron motion
, one has to store information only about sites inside a layer @xmath82 , where @xmath83 is the electron coordinate before the @xmath14-th step .
we have chosen @xmath84 , which provides a possibility to hop to @xmath85 neighbors in average .
for all sets of parameters used in the simulation the size of the optimal trap @xmath86 considered by nguyen and shklovskii was essentially less than @xmath81 .
therefore , the restriction imposed by @xmath81 did not lead to any loss of generality . before making the next step ,
the computer can forget all the information about sites in the layer @xmath87 , but it has to get information about new sites in a layer @xmath88 . as these `` new '' sites did not affect the calculation at all previous steps , they can be created at random .
therefore , each monte carlo step includes not only the choice of a jump , but also a generation of some `` new '' sites and deleting some `` old '' sites . to make their number finite , one should restrict the system size in the directions perpendicular to the field , i.e. in the plane @xmath89 .
a calculation domain @xmath90 , @xmath91 with periodical boundary conditions in the plane @xmath89 was used .
the motion of a single electron was simulated within the described algorithm in order to evaluate the drift velocity in the limit @xmath92 . for finite electron concentrations , we perform simulations in a cubic domain with size @xmath93 and with periodic boundary conditions for all three axes
. the rates of all possible jumps are calculated before starting the monte carlo steps , but without the factor @xmath94 related to occupation .
this factor determines which jumps are allowed and which are forbidden .
information about allowed and forbidden jumps is updated at each step .
we used a binary - tree data structure for storing the jump rates that gives the possibility to `` switch on '' and `` off '' jumps efficiently . in the limit of infinitely large electric field and small electron concentration , as a function of localization length @xmath12 ; ( b ) the same data divided by the velocity value predicted in ref . ( eq . ( [ eq - shklovskii5])).,width=326 ]
a routine monte carlo procedure has been used . in each monte carlo
step the final site for electron hops was calculated via the probabilities proportional to the hopping rates to different sites and the time @xmath95 spent to jump was calculated via the reciprocal of the sum of rates of all possible jumps .
hops from an occupied site were possible to any empty one in the direction of increasing coordinate @xmath25 with the restriction that the hop distance is less than @xmath96 .
at each hopping event the increment @xmath97 in electron @xmath25-coordinate is calculated .
an outcome of the simulation is either an average velocity of an electron , @xmath98 in the case of single electron hopping , or a flow of electrons , @xmath99 in the case of finite electron concentrations ( where @xmath100 is the volume of the calculation domain ) .
the summation was carried out over all sequential monte carlo steps . for simulations of the behavior of a single electron in an empty system
, @xmath101 monte carlo steps were used for @xmath102 , @xmath103 steps for @xmath104 and @xmath105 and @xmath106 steps for @xmath107 . as a result , for @xmath108 convergence was not worse than @xmath109 . for finite electron concentration ,
@xmath110 monte carlo steps were used , this gave a convergence not worse than @xmath109 for a given realization . at @xmath111 , there were sometimes essential differences between current densities in different realizations .
the scatter is shown by error bars in the figures .
simulation results for the electron drift velocity @xmath112 in the limit @xmath92 are shown in fig .
[ fig - zero - n](a ) by dots as a function of the localization length . the analytical result of nguyen and shklovskii ( eq . ( [ eq - shklovskii5 ] ) )
is shown by the solid line .
one can see that eq .
( [ eq - shklovskii5 ] ) correctly describes the dependence of the drift velocity on the localization length and , furthermore , it correctly estimates the magnitude of the velocity . the concept of nguyen and shklovskii on the hopping drift velocity controlled by hemispherical traps is herewith confirmed . however , there is some deviation of the simulation data from the analytical results . to make this deviation more transparent , we plot the ratio of the simulated drift velocity to its analytical prediction ( eq .
( [ eq - shklovskii5 ] ) ) in fig .
[ fig - zero - n](b ) .
it is seen that eq .
( [ eq - shklovskii5 ] ) overestimates the electron velocity by a factor of two to five . in section [ sec - our - theory]a , some reasons for this mismatch will be considered .
the analytical theory is further developed there to give a better agreement with the simulation data .
the result of the improved theory for the drift velocity ( eq . ( [ eq - v - improved ] ) ) is also shown in fig .
[ fig - zero - n ] by the dashed line . as a function of electron concentration @xmath24 for values of the localization length @xmath113 , @xmath114 , and @xmath115 ( from top to bottom ) .
the electric field is infinitely large.,width=326 ] the dependence of the current density on the electron concentration is shown in figs .
[ fig - n - dep ] and [ fig - near - half ] .
[ fig - n - dep ] shows this dependence in a wide concentration range , in comparison with the analytical results for small ( @xmath116 , eq . ( [ eq - shklovskii5 ] ) , solid line ) and intermediate ( @xmath117 , eq . ( [ eq - shklovskii7 ] ) , dashed line ) concentrations .
one can see that the simulated concentration dependence can be roughly divided into three parts : for very low concentrations ( @xmath118 ) the dependence is linear , only for @xmath119 , because for @xmath120 and @xmath115 the value @xmath121 is less than @xmath122 . ] in accordance with eq .
( [ eq - shklovskii5 ] ) ; then , for @xmath123 , it becomes superlinear , as described by eq .
( [ eq - shklovskii7 ] ) ; and finally , for @xmath124 , this dependence is sublinear and is not described by the theory based on the transport controlled by traps . in section
[ sec - our - theory]b , we will present an analytical approach valid for the range of parameters covering the ranges of applicability of eqs .
( [ eq - shklovskii5 ] ) and ( [ eq - shklovskii7 ] ) .
the result of this developed approach is eq .
( [ eq - n - integral ] ) shown by dashed lines in fig .
[ fig - n - dep ] . ) is evaluated numerically .
] one can see that it provides an accurate description of the current density for any concentration less than @xmath125 . for @xmath124 ,
the simulated values of the current density are smaller than those predicted by the analytical theory due to the following reason . at sufficiently large electron concentrations , the conducting paths are not almost empty , as is assumed in the theory . moreover , there are `` bottlenecks '' for the current , where the electron concentration is much larger than the mean concentration @xmath24 . in these places , the factor of @xmath126 in eq .
( [ eq - rate ] ) turns out to be important , and due to this factor the current density is suppressed . as a function of electron concentration @xmath24 , normalized on its maximum value @xmath127 , at different localization lengths .
the electric field is infinitely large.,width=326 ] in fig .
[ fig - near - half ] , the simulation results are shown for the whole range of carrier concentrations . for convenience ,
all values of current density are divided by the maximum value for the given localization length . for large localization lengths ( @xmath128 ) , the concentration dependence of the current density @xmath9 obeys approximately a parabolic law : @xmath129 at concentrations in the vicinity of the half filling .
one can interpret this behavior in terms of the hopping rates , namely , by substituting the mean occupancy @xmath130 instead of @xmath131 and @xmath132 into eq .
( [ eq - rate ] ) .
concomitantly , one obtains that the contribution of each pair of sites is proportional to @xmath133 .
the same concentration dependence is expected then for the current density . as a function of localization length @xmath12 for the half - filled system @xmath79 .
the electric field is infinitely large.,width=326 ] fig .
[ fig - half - fill ] shows the simulated dependence of the current density on the localization length ( dots ) in comparison with the analytical theory based on the concept of directed percolation ( eq . ( [ eq - shklovskii3 ] ) , solid line ) for @xmath68 .
apparently , the theory of nguyen and shklovskii@xcite correctly describes this dependence within the range of current densities of almost 15 orders of magnitude .
however , the theory underestimates the magnitude of the current density by approximately a factor of 30 .
further research is necessary to clarify the reasons of this discrepancy .
our numerical studies show that although the analytical description of hopping conduction in very strong electric field by nguyen and shklovskii is qualitatively correct the quantitative predictions differ sometimes by more than an order of magnitude from the numerical results . in this section ,
we show how to improve the accuracy of the analytical theory . for low electron concentrations @xmath77 , where @xmath135 $ ] is the concentration of the `` optimal '' traps , the prediction of ref . for the electron drift velocity
is expressed by eq .
( [ eq - shklovskii5 ] ) .
now we discuss several corrections to this equation .
\1 ) there is a numeric factor of @xmath136 in @xmath137 , arising from the evaluation of the integral ( [ eq - shklovskii4 ] ) that should be taken into account .
it gives a factor of @xmath138 for the drift velocity .
\2 ) the mean electron displacement along the @xmath6 axis , @xmath139 , is taken equal to @xmath33 in ref . .
we performed monte carlo calculations for @xmath139 as a function of the localization length @xmath12 and get the following fitting expression : @xmath140 ( the accuracy of fitting is not worse than 0.3 % in the range @xmath141 ) .
therefore , the drift velocity @xmath142 gets an additional factor equal approximately to @xmath143 .
\3 ) the dwell time @xmath144 of a trap with a radius @xmath145 is in fact somewhat less than the value @xmath146 used in ref . because an electron can escape the trap by moving not only to the nearest site to the right , but also to a more distant site .
a contribution @xmath147 of these distant sites to the escaping rate is @xmath148 then , the dwell time @xmath144 is a reciprocal of the sum @xmath149 , where @xmath150 is the rate of a jump to the nearest neighbor : @xmath151 for @xmath152 , @xmath144 is approximately half of @xmath146 , that results in a factor of two in the drift velocity .
\4 ) the geometrical crosssections of larger traps have larger capture crosssections for electrons than the smaller ones .
this results in different probabilities for carriers to be captured by traps with different radii .
the probability @xmath153 that the next visited site will be a trap with radius in the range @xmath154 is @xmath155 where @xmath156 , @xmath157 is a capture crosssection of a trap with radius @xmath145 , and @xmath158 is a mean crosssection . below we will use a notation @xmath159 for a `` relative crosssection '' @xmath160 . then , instead of eq .
( [ eq - shklovskii4 ] ) we get @xmath161 we calculated the relative crosssections with monte carlo method as ratios @xmath162/(n_j p(r)\delta r)$ ] , where @xmath162 $ ] is a number of traps with radii in the specified range visited by an electron , and @xmath163 is a total number of electron jumps .
we used @xmath164 and @xmath165 .
the results are presented in fig .
[ fig - cross ] .
the relative crosssection is almost independent of the localization length for @xmath166 .
its dependence on the trap radius is described by the quadratic function : @xmath167 for the `` optimal '' traps with @xmath168 we get @xmath169 in the limit @xmath170 . according to eq .
( [ eq - dwelltime - s ] ) , it results in a factor of @xmath171 for the mean dwell time @xmath137 and , consequently , in a factor of @xmath172 for the drift velocity .
now we can improve eq .
( [ eq - shklovskii5 ] ) of nguyen and shklovskii , starting from eq .
( [ eq - dwelltime - s ] ) .
since the integrand has a sharp maximum at @xmath173 , we can estimate the integral approximately as @xmath174 then , using eqs .
( [ eq - deltax ] ) , ( [ eq - tau ] ) , and ( [ eq - s - rel ] ) , we get the following expression for the drift velocity @xmath175 : @xmath176 } { ( 4\pi)^{1/4 } \left(0.81 + 0.36\frac{r}{\pi a}\right ) } \,v_{\mathrm ns}.\ ] ] here @xmath177 is the drift velocity corresponding to eq .
( [ eq - shklovskii5 ] ) . finally , the latter expression can be fitted ( with accuracy of about 3 % for @xmath178 ) by a simple formula , @xmath179 this expression is to be considered as a corrected analytical form for the drift velocity at infinitely high fields in the limit of small electron concentration . a comparison of eq .
( [ eq - v - improved ] ) with the values of the drift velocity obtained by the monte carlo method is shown in fig .
[ fig - zero - n ] .
the difference between the analytical and simulated results does not exceed 20 % .
we believe that the main source of this small difference is some inaccuracy in determining @xmath144 by eq .
( [ eq - tau ] ) .
in fact , for a given trap radius there is some distribution of the dwell times . the quantity @xmath144 contributing to eq .
( [ eq - dwelltime - s ] ) is the mean dwell time for radius @xmath145 .
however , eq . ( [ eq - tau ] ) gives the reciprocal value of the mean escaping rate that is slightly smaller than @xmath144 .
for this reason , eq .
( [ eq - v - improved ] ) can slightly overestimate the drift velocity .
let us now try to improve the analytical approach at finite , though small electron concentration , @xmath180 . in this case ,
electron flow can be considered as a homogeneous one on the scale of distances between the traps that determine the transport .
hence one can express the frequency @xmath181 of electron capture by a trap as @xmath182 , where @xmath9 is the current density , @xmath58 is the trap capture crosssection , and @xmath183 is its mean occupancy . under the steady - state conditions , @xmath184 , where @xmath185 is a frequency of emission of electrons from the trap ,
@xmath186 is a dwell time . from this equation
one can get @xmath183 : @xmath187 since in a snapshot of the system almost all electrons are captured by traps , the total electron concentration @xmath24 is @xmath188 the dwell time @xmath144 can be estimated by eq .
( [ eq - tau ] ) . in order to find the crosssection @xmath157
, one should note that the mean crosssection @xmath189 is equal to @xmath190 .
consequently , @xmath191 equation ( [ eq - n - integral ] ) with @xmath144 and @xmath157 determined by eqs .
( [ eq - tau ] ) and ( [ eq - s ] ) , respectively , gives a functional dependence between the electron concentration and the current density for any @xmath180 .
[ fig - n - dep ] evidences a good agreement between eq .
( [ eq - n - integral ] ) and the monte carlo calculations for @xmath192 .
although there is probably no simple way to resolve eq .
( [ eq - n - integral ] ) with respect to @xmath9 analytically in the general case , it is possible to simplify this equation in some limiting cases . for small @xmath24 and @xmath9 @xmath193 , the unity term in the denominator of eq .
( [ eq - n - integral ] ) can be dropped , and we get @xmath194 , where the drift velocity @xmath195 is determined by eq .
( [ eq - v - improved ] ) . in the opposite limit @xmath196
, one can evaluate eq .
( [ eq - n - integral ] ) as @xmath197 where a cutting parameter @xmath56 is given by condition @xmath198 .
therefore , @xmath199 with @xmath200 equation ( [ eq - j - upper - limit ] ) , with parameters determined by eqs .
( [ eq - tau ] ) , ( [ eq - s ] ) and ( [ eq - r - n ] ) is the improved version of eq .
( [ eq - shklovskii7 ] ) by nguyen and shklovskii for the concentration range @xmath201 .
so far we have considered the limiting case of infinitely high electric fields .
let us now turn to the field dependence of the charge carriers velocity in order to reveal the possibility of the negative differential conductivity predicted by nguyen and shklovskii.@xcite eq .
( [ eq - shklovskii10 ] ) predicts a decreasing drift velocity with increasing electric field , provided the field is strong enough . on the other hand , for very small fields
, ohmic transport can be expected , i.e. , the drift velocity should depend linearly on the field . in order to simulate hopping transport at finite electric fields
, we solved a system of balance equations instead of using a direct mc simulation . in the following subsection
a we describe the details of the numerical procedure , while the results are presented in subsection b. we consider a cubic system ( side length @xmath0 ) with randomly placed sites .
periodic boundary conditions are used in all directions . the balance equation for the occupation probability @xmath202 of a site @xmath8
has the form @xcite
@xmath203 if all occupation probabilities @xmath204 are small , i.e. the charge carrier concentration is low , the balance equation can be linearized : @xmath205 these equations are solved by defining @xmath206 where @xmath207 is the rate of jumping out of site @xmath8 .
the equation is then @xmath208 , which we solve numerically .
the matrix @xmath209 defined in this way is singular , which makes a direct solution rather difficult . by replacing one of the balance equations with the normalization
@xmath210 the matrix becomes nonsingular , and the solution can be obtained more efficiently . additionally , the solution obtained in this way is correctly normalized .
after this replacement , the equation has the form : @xmath211 as in section [ sec - mc - inf ] , the rates for jumps longer than @xmath212 are assumed to be zero .
hence , it is efficient to use a sparse storage scheme for the matrix , where only the non - zero elements are stored .
we obtained the most accurate results in the shortest time by solving the equation by lu factorization ( in matlab or octave with the @xmath213 operator ) .
this method demands much memory , and did not work for @xmath0 above about @xmath214 on a 32-bit computer .
when the steady - state occupation probabilities are known , the average velocity of a charge carrier along the field direction is given by @xmath215 and the mobility is then @xmath216 .
is @xmath217 . ]
the simulated dependence of the drift velocity @xmath218 on the electric field @xmath2 is presented in fig .
[ fig - loclen ] .
simulations are performed for @xmath219 sites in a cubic domain with the size @xmath220 , in the limit of infinitely small electron concentration .
different symbols refer to different localization lengths and/or different realizations of the distribution of sites in the domain .
the size of the simulated system was 10 times larger than that in the simulations of levin et al.,@xcite whose computer simulations for the first time confirmed the existence of the ndc effect for hopping transport . corresponding to the maximum of the drift velocity as a function of the localization length @xmath12 .
the linear fit is given by @xmath221 . ] in the limit of small electric field , @xmath222 , simulations show an ohmic conductivity , i.e. , @xmath218 is proportional to @xmath2 , in accordance to the miller
abrahams concept of the resistance network.@xcite with increasing field , the drift velocity reaches a maximum value .
the field strength @xmath223 corresponding to the maximum of the velocity appears to be nearly proportional to the localization length @xmath12 within the range @xmath224 ( see fig .
[ fig - peakpos ] ) . at field strengths
@xmath225 the ndc appears , i. e. the drift velocity drops with increasing field .
simulations show the presence of the ndc for localization lengths up to @xmath226 ; when the localization length is decreased , the ndc effect becomes more pronounced . ) and eq .
( [ eq - shklovskii10-mod ] ) , scaled to approach eq .
( [ eq - v - improved ] ) in the limit of large fields .
the system size @xmath0 is @xmath217 . ]
[ fig - loclen - log ] shows the comparison between the simulation results ( symbols ) and the predictions of nguyen and shklovskii@xcite ( eq . ( [ eq - shklovskii10 ] ) , dashed lines ) . for better agreement between the theory and the simulation at @xmath227 ,
we take into account the @xmath2-independent correction ( [ eq - v - improved ] ) to eq .
( [ eq - shklovskii10 ] ) .
some discrepancies between the simulated and predicted drift velocities remain at large fields for @xmath228 and @xmath229 .
we believe that these discrepancies are due to the small size of the simulated system .
in fact , the simulated system must be large enough to contain a reasonable number of optimal traps .
the concentration of optimal traps decreases sharply with decreasing localization length , so that at smaller localization lengths larger systems are needed .
thus , for small localization length ( @xmath230 and @xmath231 ) , only the shape of the simulated filed dependence should be taken as representative , but not the values of the calculated velocities themselves . the range of applicability of eq .
( [ eq - shklovskii10 ] ) is determined by the condition @xmath232 .
one can see nevertheless that even within this range the field dependence of the velocity is much weaker than the one predicted by eq .
( [ eq - shklovskii10 ] ) .
this result forced us to consider another possible optimal trap shape for the case of a finite field as compared to the one considered in ref . .
the essential feature of the optimal trap proposed by nguyen and shklovskii ( fig . [ fig - traps]b ) is the chain of sites along the axis of the cone .
this chain was introduced in order to provide an easy path for an electron into a trap .
the chain affects the trap shape and volume , as it serves also as a channel for escaping of an electron from the trap .
we suggest that at moderate localization lengths ( @xmath233 ) traps without such a chain can also play a significant role .
our next aim is to consider the shape of traps without a chain of sites and to estimate their influence on the electron conduction .
a sketch of such a trap is shown in fig .
[ fig - newtrap ] .
its shape is defined by the condition that the rate of jumping from the central site to any point of the trap s surface is the same . from eq .
( [ eq - rate ] ) one can see that in the positive direction along the axis @xmath6 the trap is bounded by a hemisphere , and in the negative direction by a surface defined by an equation @xmath234 where @xmath145 is the radius of the hemisphere , and the origin is placed at the central site of the trap . the surface determined by eq .
( [ eq - trap - surface ] ) is a quadric surface ( an ellipsoid , a paraboloid , or a hyperboloid , depending on the values of parameters ) .
the volume of the trap , @xmath235 is @xmath236 where @xmath237 . to obtain the drift velocity @xmath218 ( or the current density @xmath238 ) in the assumption that the most important traps are those shown in fig .
[ fig - newtrap ] , one can proceed in the same way as the one applied in section [ sec - model ] to get eq .
( [ eq - shklovskii10 ] ) ; the only difference is using @xmath239 instead of nguyen and shklovskii s trap volume @xmath240 .
the result is @xmath241\!.\ ] ] for the optimal trap radius one gets @xmath242 since @xmath243 ,
the new trap shape gives a weaker field dependence of the drift velocity , and a better agreement with the data from simulations , as one can see in fig .
[ fig - loclen - log ] .
however , the simulated field dependence appears even weaker than the one expressed by eq .
( [ eq - shklovskii10-mod ] ) .
it leads to the assumption that an actual optimal trap has a shape different from both fig .
[ fig - traps]b and fig . [ fig - newtrap ] , and hence has a different volume . .
the position of a charge carrier ( at the origin ) is pointed out by a cross .
boundaries of optimal traps predicted in ref .
( see fig .
[ fig - traps ] ) are depicted by solid lines , the boundary of the trap sketched in fig .
[ fig - newtrap ] by the dotted line .
spatial coordinates are in units of @xmath33 .
the value @xmath244 for @xmath245 corresponds to the maximum of the drift velocity . ]
to further investigate the shape of the most efficient traps , we collect information about the trap shape from the simulations . fig .
[ fig - trapdistr ] show the time - average of the density of sites around the charge carrier .
to calculate this density @xmath246 , the space was divided into small elements of equal volume @xmath247 ; then @xmath246 was evaluated as a sum over pairs of sites : @xmath248 where @xmath204 is an occupation probability of the @xmath8-th site ; @xmath249 if the vector @xmath250 belongs to the same spatial element as the vector @xmath251 ; otherwise @xmath252 .
finally , values of @xmath246 were averaged over several realizations of site distributions . since the carrier spends the most time in the efficient traps ,
the density distribution @xmath246 directly reflects the shape of these traps . at high fields ( fig .
[ fig - trapdistr]a ) the hemispherical shape and the size of the trap are in an excellent agreement with the nguyen and shklovskii s theory . however , at moderate fields , in the region of the ndc ( fig .
[ fig - trapdistr](b ) ) , neither fig .
[ fig - traps](b ) nor fig .
[ fig - newtrap ] describe the simulated optimal trap .
the optimal trap consists in such a case of a hemisphere in the spatial region @xmath253 , and of a toroidal `` barrier '' in the region @xmath254 , adjoining to a periphery of the hemisphere . the volume of the optimal trap turns out to be smaller than the one predicted by both eq .
( [ eq - shklovskii10 ] ) and eq .
( [ eq - shklovskii10-mod ] ) , in accordance with the result that the simulated ndc effect is weaker than the predicted one .
we would like to emphasize that the numerically obtained ndc has exactly the origin predicted by nguyen and shklovskii,@xcite consisting in spreading of the optimal trap into the region @xmath254 at finite fields and consequently in the increase of the trap volume with decreasing the field strength .
the possibility of traps in the form of clusters instead of single chains of sites has been considered in ref . .
we interpret our numerical result as a confirmation of that idea .
a further decrease of the electric field @xmath2 results in the washing out the empty region in the density of sites @xmath246 , as shown in fig .
[ fig - trapdistr](c ) for @xmath255 . finally , at small @xmath2 the trap almost disappeared ( fig .
[ fig - trapdistr](d ) ) , which points to a negligible role of the trapping effect in the regime of ohmic conduction . versus electric field @xmath2 for gaussian disorder in site energies .
disorder is characterized by a standard deviation @xmath256 of site energies from the reference energy . the localization length @xmath257 .
] materials studied experimentally usually have disorder not only in the spatial distribution of localized states , but also in the site energies.@xcite it is therefore necessary to check how stable the ndc effect is with respect to energetic disorder . in order to study the role of the energetic disorder for the ndc effect
, we repeated the simulation in a system with a gaussian distribution of site energies , characterized by the standard deviation @xmath256 .
[ fig - disorder ] shows that introducing a random energy for each site ( with a gaussian distribution ) decreases the drift velocity and it also decreases the height of the peak of the velocity as a function of the electric field as compared to systems with only spatial disorder .
this weakening of the ndc effect with the increase of the energetic disorder ( or with the decrease of temperature ) is in agreement with experimental observations.@xcite generally , the ndc effect is confirmed herewith also for systems with the energetic disorder .
however , at extremely large energetic disorder ( parameter @xmath256 ) , the peak in the field dependence of the drift velocity disappears completely .
the effect of energetic disorder becomes smaller at larger fields , as expected from the fact that in the limit @xmath227 the hopping rates do not depend on site energies .
numerical studies of the field - dependent drift velocity of charge carriers in the hopping regime at high electric fields confirm the prediction of the existing analytical theories@xcite that the negative differential conductivity is inherent for this transport mode .
however , the shape of the field dependence on the current density obtained numerically differs essentially from the one predicted so far.@xcite the analytical theory has been improved to give a much better agreement with the numerical results . in the limit of the infinitely high electric fields , the predictions of the analytical theory of nguyen and shklovskii@xcite are to much extent confirmed by our straightforward monte carlo simulations .
the authors are indebted to prof .
boris shklovskii for numerous valuable comments .
financial support from the academy of finland project 116995 and the tekes namu project , from the deutsche forschungsgemeinschaft and that of the fonds der chemischen industrie is gratefully acknowledged .
a. v. n. thanks the russian foundation for basic research ( project 06 - 02 - 16988 ) and the dynasty foundation for financial support .
the authors thank oleg rubel and kakhaber jandieri for stimulating discussions . | effects of strong electric fields on hopping conductivity are studied theoretically .
monte - carlo computer simulations show that the analytical theory of nguyen and shklovskii [ solid state commun .
38 , 99 ( 1981 ) ] provides an accurate description of hopping transport in the limit of very high electric fields and low concentrations of charge carriers as compared to the concentration of localization sites and also at the relative concentration of carriers equal to 0.5 . at intermediate concentrations of carriers between 0.1 and 0.5 computer simulations evidence essential deviations from the results of the existing analytical theories .
the theory of nguyen and shklovskii also predicts a negative differential hopping conductivity at high electric fields .
our numerical calculations confirm this prediction qualitatively .
however the field dependence of the drift velocity of charge carriers obtained numerically differs essentially from the one predicted so far .
analytical theory is further developed so that its agreement with numerical results is essentially improved . |
the relativistic binary pulsar b1913 + 16 , discovered by hulse & taylor ( 1975 ) , has now been observed for twenty seven years . after a several year hiatus due to the arecibo upgrading project
, we have resumed regular timing and profile measurements . in this paper , we describe the results of our relativistic analysis of these data .
in order to optimize timing data acquisition , we and colleagues have developed a number of specialized backends for recording binary and millisecond pulsar data , including a sweeping local oscillator system at 430 mhz ( mcculloch , taylor , & weisberg 1979 ) and a succession of wideband samplers and signal averagers for use at 21 cm ( taylor & weisberg 1982 ; rawley 1986 ; stinebring et al 1992 ) .
most of our highest quality timing data were acquired with the princeton mark iii backend ( stinebring et al 1992 ) , which typically achieved time of arrival uncertainties @xmath0toa @xmath1s in five minute integrations of 40 mhz total bandwidth .
we have found that the quality of data acquired starting in 1981 are so much higher than earlier data that the previous measurements contribute negligibly to our fits .
conseqently , we use 5083 toas measured between 1981 and 2001 in what follows .
these toas were fed into program tempo ( taylor & weisberg 1989 ; http://pulsar.princeton.edu/tempo/ ) which fitted for 18 parameters according to the phenomenological relativistic timing model of damour & deruelle ( 1985 , 1986 ) . aside from pulsar spin and astrometric parameters , we fitted for five `` keplerian '' parameters and three relativistic quantities : projected semimajor axis of the pulsar orbit @xmath2 , eccentricity @xmath3 , reference epoch @xmath4 , binary orbital period @xmath5 , argument of periastron @xmath6 , mean advance of periastron rate @xmath7 , time dilation / gravitational redshift amplitude @xmath8 , and orbital period change rate @xmath9 .
lll fitted parameter & value & uncertainty + @xmath10 i ( s ) & 2.341774 & 0.000001 + e & 0.6171338 & 0.0000004 + t@xmath11 ( mjd ) & 46443.99588317 & 0.00000003 + p@xmath12 ( d ) & 0.322997462727 & 0.000000000005 + @xmath6 ( deg ) & 226.57518 & 0.00004 + @xmath7 ( deg / yr ) & 4.226607 & 0.000007 + @xmath8 ( s ) & 0.004294 & 0.000001 + @xmath9 ( @xmath13 s / s ) & -2.4211 & 0.0014 + the results for these eight fitted parameters , and their formal 1-@xmath14 uncertainties , are listed in table 1 . only seven are required in order to specify completely all the orbital parameters and component masses except for an unknown rotation about the line of sight .
for example the relativistic measurables @xmath7 and @xmath8 depend on the keplerian parameters and the companion masses @xmath15 and @xmath16 in the following fashion : @xmath17 and @xmath18 consequently measurement of the keplerian paramaters and @xmath7 and @xmath8 permits us to solve these two equations simultaneously for component masses , yielding @xmath19 , where the uncertainty in the newtonian gravitational constant @xmath20 is not included .
a graphical solution for the masses , also using equations ( 1 ) and ( 2 ) and the same measurables , is shown in figure 1 .
these seven measured quantities and the parameters derived from them , such as @xmath15 and @xmath16 , are then adequate to derive all other interesting orbital parameters , such as the sine of the orbital inclination @xmath21 i and the various semimajor axes @xmath22 , and @xmath23 : @xmath24 @xmath25 @xmath26 and @xmath27 in earlier datasets we also solved for the two parameters of the shapiro gravitational delay , @xmath28 and @xmath29 , and we were marginally able to constrain them ( taylor and weisberg 1989 ; taylor et al 1992 ) .
as the orbit has precessed , however , the geometry has become less favorable for measuring this phenomenon .
continuing orbital precession is now causing the signature to begin to grow again .
the emission of gravitational radiation should lead to orbital energy loss and orbital decay .
this will be observable as an orbital period change , @xmath30 . according to general relativity ,
the orbital period change due to gravitational radiation damping will depend on the system parameters as follows : @xmath31 inserting our measured values of the parameters given in table 1 into equation ( 7 ) , we calculate that the general relativistic prediction for orbital period decay rate is @xmath32 s / s . the galactic acceleration of the system and the sun also cause an orbital period change , @xmath33 ( damour & taylor 1991 ) .
using their equation ( 2.12 ) and their parameters for @xmath34 [ except for the following updated parameters : solar galactocentric distance @xmath35 kpc ( reid 1993 ) ; pulsar distance @xmath36 kpc ( cordes & lazio 2002 ) ; solar galactocentric circular velocity @xmath37 km / s ( reid et al 1999 , backer & sramek 1999 ) ; pulsar proper motion @xmath38 mas / yr ] , we find that @xmath39 s / s .
the galactic quantities will be improved in the next several years with additional interferometric observations of sgr a * and infrared measurements of stars orbiting it ( salim & gould 1999 ; ghez et al 2000 ; schodel et al 2002 ) . correcting the observed orbital period derivative for the galactic acceleration term ,
@xmath40 we find @xmath41 s / s , in good agreement with the theoretical general relativistic result @xmath42 given above .
the comparison is also shown graphically in figure 2 , where we plot the observed and general relativistically predicted shift of periastron time resulting from gravitational radiation damping .
this result provides strong confirmation of the existence of gravitational radiation as predicted by general relativity .
weisberg & taylor ( 1981 ) show that it rules out numerous other relativistic theories of gravitation .
this result plus relativistic measurements of other binary pulsars also constrain the parameters of viable tensor scalar theories ( taylor et al 1992 ; damour & esposito - farese 2001 ) .
the spinning , orbiting pulsar should also experience geodetic precession , which is caused by general relativistic gravitomagnetic spin
orbit coupling ( damour & ruffini 1974 ; barker & oconnell 1975a , b ) .
this will lead to a change in the observed pulseshape as the earth
pulsar line of sight cuts across different parts of the precessing beam .
weisberg , romani , & taylor ( 1989 ) noted the first pulseshape changes ascribed to geodetic precession .
the pattern became clearer when kramer ( 1998 ) detected a secular narrowing of the separation between the two principal pulse components , which he was able to successfully model .
weisberg & taylor ( 2002 ) measured shape changes throughout the pulse in twenty years of arecibo 21 cm data .
the pulse profile was divided into even and odd components to separate overall beam structure from patchy structure , respectively .
we found that the outer skirts of the even profile barely changed while the peaks drew closer and the central saddle filled in ( figure 3a ) .
our model that best fits these data indicates an hourglass - shaped beam elongated in the latitude direction ( figure 3b ) .
interestingly , link & epstein ( 2001 ) found a similar beam shape for a ( nonrelativistically ) precessing pulsar , b1828 - 11 .
see also kramer ( this volume ) for additional observations and modeling .
backer , d. c. & sramek , r. a. 1999 , , 524 , 805 barker , b. m. & oconnell , r. f. 1975a , , 12 , 329 barker , b. m. & oconnell , r. f. 1975b , , 199 , l25 cordes , j. m. & lazio , t. j. w. 2002 , astro - ph 0207156 damour , t. & ruffini , r. 1974 , academie des sciences paris comptes rendus serie sciences mathematiques , 279 , 971 damour , t. & deruelle , n. 1985 , ann .
h. poincare ( phys
. theorique ) , 43 , 107 damour , t. & deruelle , n. 1986 , ann .
h. poincare ( phys .
theorique ) , 44 , 263 damour , t. & taylor , j. h. 1991 , , 366 , 501 damour , t. & esposito - farese , g. 2001 , , 58 , 042001 ghez , a. m. , morris , m. , becklin , e. e. , tanner , a. , & kremenek , t. 2000 , nature , 407 , 349 hulse , r. a. & taylor , j. h. 1975 , , 195 , l51 kramer , m. 1998 , , 509 , 856 link , b. & epstein , r. i. 2001 , , 556 , 392 mcculloch , p. m. , taylor , j. h. , & weisberg , j. m. 1979 , , 227 , l133 rawley , l. a. 1986 , ph .
d. thesis , princeton u. reid , m. j. 1993 , , 31 , 345 reid , m. j. , readhead , a. c. s. , vermeulen , r. c. , & treuhaft , r. n. 1999 , , 524 , 816 salim , s. & gould , a. 1999 , , 523 , 633 schodel , r. , et al 2002 , nature , 694 stinebring , d. r. , kaspi , v. m. , nice , d. j. , ryba , m. f. , taylor , j. h. , thorsett , s. e. , & hankins , t. h. 1992 , review of scientific instruments , 63 , 3551 taylor , j. h. & weisberg , j. m. 1982 , , 253 , 908 taylor , j. h. , & weisberg , j. m. 1989 , , 345 , 434 taylor , j. h. , wolszczan , a. , damour , t. , & weisberg , j. m. 1992 , nature , 355 , 132 weisberg , j. m. & taylor , j. h. 1981 , gen .
rel . & grav . , 13 , 1 weisberg , j. m. , romani , r. w. , & taylor , j. h. 1989 , , 347 , 1030 weisberg , j. m. & taylor , j. h. 2002 , , 576 , 942 | we describe the results of a relativistic analysis of our observations of binary pulsar b1913 + 16 , up to the latest measurements in 2001 august . # 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
carbon nanotubes belong to the most promising new materials for the future molecular electronics , they are believed to potentially replace in the near future the silicon - based conventional electronics . to illustrate the enormous scientific and technological progress that has been made since carbon nanotubes
were discovered it is worth to mention new concepts such as : the room temperature single electron transistor,@xcite the ballistic carbon nanotube field - effect transistor@xcite or the non - volatile random access memory for molecular computing.@xcite recently several both experimental@xcite and theoretical@xcite papers have been published on spin - dependent electrical transport in ferromagnetically contacted carbon nanotubes in an attempt to test their ability to operate as spintronic devices .
it has been found that carbon nanotubes though intrinsically nonmagnetic reveal quite a considerable gmr effect .
however , it should be stated in this context that the experimental results are very diverse , and reflect to a great extent sample - specific features .
the poor reproducibility of experiments on carbon nanotubes , or more generally on molecular systems , is due to hardly controllable interface conditions between the molecule and external electrodes .
incidentally , judging from the conductance value , this difficulty is sometimes successfully overcome but , to our knowledge , only if nonmagnetic - electrodes are used .
the examples proving this statement are numerous as far as single wall carbon nanotubes ( swcnts ) are concerned,@xcite and much more seldom in the case of multiwall carbon nanotubes ( mwcnts ) @xcite . despite intensive studies,@xcite fundamental problems such as ( _
i _ ) the internal structure of nanotubes ( especially multiwall ones ) , ( _ ii _ ) the aforementioned nature of the electrode / nanotube coupling , and ( _ iii _ ) the energy band line - up at the interfaces are still far from being well understood . motivated by these facts
, we study a minimal geometrical model of the mwcnt , which enables us to define in a unique way the coupling of the mwcnt to the electrodes . having done that , we can restrict ourselves to the most crucial issues addressed in the present studies , namely : the inter - wall interactions , and the effect of the breakage of the contact between the inner wall and one of the transversally coupled ( burying ) electrodes .
so far tangential contact geometries have been mainly studied in connection with stm experiments.@xcite however in experiments with ferromagnetic - electrodes , transition metals are usually deposited by evaporation on top of the carbon nanotube cutting off a segment of the latter , so that the nanotube finds itself buried in the electrode and the contact geometry can be regarded as transversal .
( color online ) view of the ( 2,2)@(6,6 ) carbon nanotube sandwiched between two fcc(111 ) leads and detail of the contact region .
what is shown consists of a few ferromagnetic - electrode atoms with the nanotube forming the so - called extended molecule .
the other parts of the electrodes ( not shown ) are infinite in all the directions . ]
( color online ) the total conductance for the parallel ( pa ) and antiparallel ( ap ) alignments , as well as the gmr _ vs. _ energy .
the outer nanotube length equals that of the inner one ( @xmath0=@xmath1=41 carbon rings ) , the fermi energy is equal to @xmath2 and @xmath3
. left panel : there are no inter - wall interactions ( @xmath4 ) .
middle and right panels : the inter - wall interactions are included , @xmath5 and @xmath6 , respectively . ]
our system consists of a double - wall carbon nanotube ( dwcnt ) perfectly contacted to semiinfinite electrodes having an fcc(111 ) geometry : every carbon atom at the interface has exactly three nearest neighbors in the electrode ( see fig . [
fig : fig1 ] ) .
such a construction is possible in the case of the ( 6,6 ) armchair swcnt@xcite due to the almost perfect fitting of the armchair nanotube lattice constant ( @xmath7 ) to the interatomic distances at the electrode close packed surfaces ( _ e.g. _ 2.51 , 2.49 and 2.55 for co , ni and cu respectively ) .
this coincidence makes it possible to put the interface ( contact ) ring of carbon atoms on top of the electrode substrate in such a manner that all carbon atoms sit in the geometrical centers of three adjacent electrode - atoms , and additionally the perimeter of the contact carbon ring equals , within a few percent , the standard value for the ( 6,6 ) swcnt , _
i.e. _ @xmath8 . on geometrical grounds ,
the only other armchair structure which can be constructed likewise is the ( 2,2 ) one ( see fig . [
fig : fig1 ] ) .
so we will use it in our studies as the inner tube , in spite of the fact that it is known to be energetically unstable.@xcite as a matter of fact , since we focus here on electrical transport properties rather than stability problems , we are convinced that the present model gives a qualitative insight into the role of the inter - wall interaction in real mwcnts , depending on whether or not the inner tubes are fully contacted to the electrodes . in this very context , the present model is advantageous in view of its small number of atoms within the unit cell , ideal geometrical fitting to the electrode surface atoms , and its well - defined inter - wall coordination zones .
the last feature enables us to define inter - wall hopping integrals in such a way that each inner - tube carbon atom has just two outer - tube atoms to interact with .
thus , again no ambiguity exists concerning the interaction range , and no hidden cutoff parameters are needed .
1 corresponds to an odd number of carbon rings ( inverse - symmetric system ) .
the case of even numbers of rings makes no serious complication@xcite , but the drain electrode has to be rotated by @xmath9 with respect to the source one if the perfect geometrical matching at the drain interface is also required .
( color online ) unlike in the previous figures , here the inner tube is out of contact to the right electrode ( see the upper right diagram ) .
a drastic drop in the conductance ( at @xmath10 ) , accompanied by a positive gmr , takes place .
no inter - wall interactions ( left panel ) , and @xmath11 and @xmath6 ( middle and right panels , respectively ) . ]
the device under consideration is described by means of a single - band tight - binding hamiltonian , assuming @xmath12 and @xmath13-electrons in the double - wall carbon nanotube and ferromagnetic - electrodes , respectively .
the total hamiltonian can be compactly written as @xmath14 with vanishing initial on - site potentials , @xmath15 , in the dwcnt , whereas in the electrodes , they are spin - dependent and chosen so as to give the required magnetization . to solve the green function problem
, we use the partitioning technique,@xcite treating the whole device as a left - electrode dwcnt
right - electrode system ( shorthand notation l - c - r ) .
here c stands for the central carbon - based part , which incorporates not only the dwcnt but also the first two closest atomic layers ( with @xmath16 atoms ) from both the l- and r - electrode , implementing the extended molecule concept ( as shown in fig .
[ fig : fig1 ] ) .
the green function of the extended molecule reads @xmath17 whereas the density matrix and the conductance ( per spin ) are given by : ( color online ) as fig.2 but for 40 carbon rings in each nanotube . ]
@xmath18 @xmath19 with @xmath20 and @xmath21 being the fermi - dirac distribution function ( @xmath22 l , r ) .
the @xmath23-th electrode surface green function , @xmath24 , is calculated as described in ref . .
while back fourier transforming to the real space , the integration over the two - dimensional brillouin zone has been performed by the _ special - k - points method_.@xcite the giant magnetoresistance is defined as gmr@xmath25 , where the arrows denote the aligned and antialigned magnetic configuration . to parameterize the hamiltonian ,
we have put the polarization of the electrodes equal to 50% ( with the number of electrons per atom being 0.75 and 0.25 for majority and minority bands , respectively ) .
the energy scale is given by the hopping integral @xmath26 ev and the nanotube lattice constant is @xmath7 , fixing the energy and length units , respectively . in order to minimize the number of free parameters the nearest neighbor hopping integral is kept constant in the entire device . as regards the inter - tube interaction we treat the hopping integral as a free parameter of the order of @xmath27 ( as for the interlayer distance in graphite@xcite ) ,
although in general @xmath28 depends exponentially on distance and also on angles between the @xmath12 orbitals of the involved atoms.@xcite it should be also admitted that another complexity , not captured by our simple structural model ( two armchair tubes ) , is a possible additional off - diagonal disorder due to incommensurability between shells ( cf . ref . ) . + ( color online ) the conductance in the case of paramagnetic - electrodes . the breakage of the inner - tube contact to one of the electrodes results in a drastic drop of the conductance ( the lower curves ) .
the inter - wall hopping integrals , @xmath29 , are equal to @xmath30 and @xmath6 ( solid , dashed and dotted curves , respectively ) . ]
our energy band line - up procedure goes as follows : first , @xmath10 is fixed at 0 , next the on - site potentials in the electrodes are set so as to yield the required number of electrons per atom ( @xmath31 ) and magnetic moment ( 0.5 and 0 in the ferromagnetic and paramagnetic cases , respectively ) . the so - determined fermi energy is kept fixed at the charge neutrality point and no doping effects are taken into account . finally , all the on - site potentials within the extended molecule are self - consistently computed from the global charge neutrality condition and from eq .
( [ eq : rho ] ) . rather than restrict ourselves to the fermi energy
, we present our results in an energy window , accessible e.g. by applying a gate potential ( see e.g ref . ) .
we have first considered a perfectly contacted case , when both walls are in touch with the source and drain electrodes .
the results , presented in fig . [ fig : fig2 ] , show that for the mwcnt length equal to 20.5_a _ ( 41 carbon rings ) the gmr is negative regardless of whether inter - wall interactions are included or not .
next , the inner tube has been allowed to be out of contact to , say , the drain electrode , see fig .
[ fig : fig3 ] .
it turns out that then , not only the conductance of the whole system decreases roughly twice , but also the gmr becomes positive .
the effect of the inter - wall interactions is moderate ( at least up to @xmath32 ) , they happen to increase the conductance , always keeping the gmr positive .
the inverse gmr effect in the case of full end - contacted nanotubes , and @xmath29 not exceeding a length - dependent threshold value , has been also found for the nanotube lengths of 40- and 39-rings , and is depicted in fig .
[ fig : fig4 ] for @xmath33=40 . as it is seen , the gmr is now far less robust than for @xmath33=41 .
judging from the conductance behavior at @xmath10 for 3 consecutive nanotube lengths , one can anticipate what happens for an arbitrary length , because the conductance is a quasi - periodic function with a period equal to 3-rings spacing .
thus , for dwcnts which are @xmath34 rings long we predict a quite pronounced negative gmr effect ( @xmath33 is an integer ) .
the @xmath34 length rule ( in a/2 units ) was numerically found@xcite , with these lengths an ideal nanotube has got maximum conductance ( on - resonance device ) .
disorder at the interfaces makes the maximum of the conductance shift beyond @xmath10 , but the quasi - periodicity is still preserved .
the aforementioned quasi - periodicity originates from the fact that ideally the swcnt armchair energy spectrum crosses the [ @xmath35-point for particular lengths only , and additionally conductance is a measure of the squared electron wave function . as a test illustrating the quality of the present approach , we have completed our studies by additional computations for dwcnts sandwiched between _
paramagnetic _ electrodes with the same number of electrons per lattice site ( @xmath36 ) as in the ferromagnetic calculations .
our attention has been again focused both on the effect of the inner nanotube contact to one of the electrodes , and the importance of the inter - wall interactions .
the latter are now allowed to take values up to @xmath37 .
[ fig : fig5 ] shows that the conductance in the vicinity of @xmath10 does not reach the maximum theoretical value of @xmath38 for a dwcnt ( @xmath39 contribution for every wall obtained for infinite mwcnts ) .
the conductance suppression in mwcnts below the expected value for the infinite homogeneous tube has been experimentally well documented.@xcite in our model system charge - transfer induced changes in the band structure of the dwcnt lead to a reduction ( roughly to @xmath40 ) of the conductance -as one can see in the upper curves in fig .
[ fig : fig5 ] .
this finding is in line with the well - known scenario , that the reduction of the conductance is due to the interface - induced suppression of one of the two transport channels.@xcite obviously , if a finite armchair nanotube is contacted to external electrodes it can no longer be viewed as a periodic one with the repeat unit consisting of two carbon rings , so the band degeneracy is lifted and the two channels couple very differently to their counterparts in the leads .
it is noteworthy that recent _ ab initio _
results on single wall carbon nanotubes end - contacted to paramagnetic electrodes also show that the interface mismatch may result in roughly halving the conductance.@xcite the lower part of fig .
[ fig : fig5 ] shows the effect of the inner - tube contact breakage .
the conductance gets reduced by a factor close to two , very much like in the ferromagnetic computations .
more importantly however , this behavior is in qualitative agreement with experimental results@xcite if we associate the lower and upper bunch of curves in fig .
[ fig : fig5 ] with the major plateaux of the conductance of mwcnt immersed into hg .
incidentally , the peculiar peak in fig .
[ fig : fig5 ] more than 0.15 @xmath41 ( ca . 0.4 ev ) away from the charge neutrality point , is due to model - specific features of no universal nature . both the gmr and the conductance are oscillatory , non - monotonic functions of the inter - wall coupling , and reveal a quasi - periodic behaviour with the dwcnt length .
it is also noteworthy that the conductances in the ferromagnetic case are always clearly smaller ( regardless of the electrode magnetization alignment ) than their paramagnetic counterparts .
we attribute this behavior to the increased energy band mismatch at the interfaces in the ferromagnetic case .
in conclusion , we have investigated the effect of inter - wall interactions on the giant magnetoresistance of double wall carbon nanotubes , depending on whether or not the inner wall is fully contacted to the electrodes .
we have found that the gmr of perfectly contacted dwcnts is very sensitive to the inter - wall coupling strength , and show a tendency to become negative . on the contrary , dwcnts with inner tubes being out of contact with one of the electrodes reveal ( close to the fermi energy ) a robust positive gmr only weakly dependent on the inter - wall coupling .
these observations may be related to the experimental results showing either positive or negative ( sample - dependent ) gmr of comparable value ( @xmath42),@xcite as well as recently reported negative gmr values in mwcnts strongly contacted to permalloy electrodes.@xcite moreover , our complementary calculations for the case with paramagnetic electrodes show that the conductance of the perfectly contacted dwcnt is suppressed by a factor two with respect to the case of a homogeneous infinite dwcnt ( that is @xmath40 instead of @xmath43 ) .
consequently , for the case in which the inner wall gets disconnected , we observe conductance values close to @xmath44 instead of @xmath40 .
this is related to the charge transfer induced energy band rearrangements in the extended molecule .
the present approach may be regarded as a reference for further generalizations .
the diameters and the number of walls forming the mwcnt may be easily increased and the respective stable positions of the interface atoms can be readily determined by a simple relaxation procedure involving just a few variational parameters ( displacements and rotations ) .
we are indebted to gerrit e. w. bauer , uwe krey and nitesh ranjan for fruitful discussions .
sk acknowledges the support from the daad foundation and the kbn research project pbz - kbn-044/p03 - 2001 .
this work was partially funded by the volkswagen stiftung and the eu centre of excellence contract g5ma - ct-2002 - 04049 . | we report on the giant magnetoresistance ( gmr ) of multiwall carbon nanotubes with ultra small diameters .
in particular , we consider the effect of the inter - wall interactions and the lead / nanotube coupling .
comparative studies have been performed to show that in the case when all walls are well coupled to the electrodes , the so - called inverse gmr can appear .
the tendency towards a negative gmr depends on the inter - wall interaction and on the nanotube length .
if , however , the inner nanotubes are out of contact with one of the electrodes , the gmr remains positive even for relatively strong inter - wall interactions regardless of the outer nanotube length .
these results shed additional light on recently reported experimental data , where an inverse gmr was found in some multiwall carbon nanotube samples . |
a large variety of natural systems exhibit irregular and complex behavior which at first looks erratic , but in fact possesses scale invariant structure ( e.g. , @xcite ) .
a process @xmath9 is called self - similar@xcite if for some @xmath10 , @xmath11 where the symbol of equality refers here to all finite - dimensional distributions of the process on the left and the right , and the parameter @xmath12 is called self - similarity index or exponent . equation ( [ eq1 ] )
means a `` scale invariance '' of the finite - dimensional distributions of @xmath13 , which does _ not _ imply , in stochastic processes , the same for the sample paths ( e.g.,@xcite ) .
examples of self - similar processes are brownian , fractional brownian ( fbm ) , lvy stable and fractional lvy stable motion ( flsm ) .
lvy stable distributions ( which are followed by many natural processes , e.g. , @xcite ) differ greatly from the gaussian ones because they have heavy tails and their variance is infinite ( e.g.,@xcite ) .
an important point in analyzing data from natural systems that exhibit scale invariant structure , is the following : in several systems this nontrivial structure points to long - range _ temporal _ correlations ; in other words , the self - similarity results from the process memory only ( e.g. , the case of fbm ) .
alternatively , the self - similarity may solely result from the process increments infinite variance , e.g. , lvy stable motion .
( note , that in distributions that are applicable to a large variety of problems , extreme events have to be _ truncated _ for physical reasons , e.g. , finite size effects , -when there is no infinity@xcite- and this is why we write hereafter `` infinite '' . ) in general , however , the self - similarity may result from both these origins ( e.g. , flsm ) .
it is the main aim of this paper to discuss how a distinction of the two origins of self - similarity ( i.e. , process memory , process increments `` infinite '' variance ) can be in principle achieved by employing the natural time analysis . in a time series comprising @xmath14 events , the _ natural time _
@xmath15 serves as an index@xcite for the occurrence of the @xmath16-th event .
the evolution of the pair ( @xmath17 ) is considered@xcite , where @xmath18 denotes in general a quantity proportional to the energy released in the @xmath16-th event .
for example , for dichotomous signals @xmath18 stands for the duration of the @xmath16-th pulse while for the seismicity @xmath18 is proportional to the seismic energy released during the @xmath16-th earthquake@xcite ( which is proportional to the seismic moment @xmath19 ) .
the normalized power spectrum @xmath20 was introduced@xcite : @xmath21 where @xmath22 and @xmath23 ; @xmath24 stands for the _
natural frequency_. when the system enters into the _ critical _ stage , the following relation holds@xcite : @xmath25 for @xmath26 , eq.([fasma ] ) leads to@xcite @xmath27 which reflects@xcite that the variance of @xmath0 is given by @xmath28 , where @xmath29 .
it has been argued@xcite that in the case of earthquakes , @xmath30 for @xmath31 , can be considered as an order parameter and the corresponding probability density distribution function ( pdf ) is designated by @xmath32 $ ] . since , at @xmath33 , @xmath4 is linearly related to @xmath30 ( because eq.([eq3 ] ) leads to @xmath34 for @xmath31 ) one can study , instead of @xmath32 $ ] , the pdf of @xmath4 , i.e. , @xmath6 . this will be used here .
the entropy @xmath35 in the natural time - domain is defined as@xcite @xmath36 , which depends on the sequential order of events@xcite and for infinitely ranged temporal correlations its value is smaller@xcite than the value @xmath37 ) of a `` uniform '' distribution ( defined in refs .
@xcite , e.g. when all @xmath38 are equal ) , i.e.,@xmath39 .
the value of the entropy obtained@xcite upon considering the time reversal @xmath40 , i.e. , @xmath41 , is labelled by @xmath42 .
this paper is organized as follows : in section ii , we treat the case when the self - similarity solely results from the process memory . section iii deals with the self - similarity resulting from the process increments infinite variance by restricting ourselves to slowly driven systems that emit energy bursts obeying power law distributions .
a brief discussion follows in section iv , while section v presents the main conclusion .
we consider , as an example , seismic electric signals ( ses ) activities which exhibit infinitely ranged temporal correlations@xcite .
figure [ fg1](a ) shows a recent ses activity recorded at a station located in central greece ( close to patras city , pat ) on february 13 , 2006 .
it comprises 37 pulses , the durations @xmath18 of which vary between 1s and 40s ( see fig .
[ fg1](b ) ) . the natural time representation of this ses activity can be seen in fig.[fg1](b ) and the computation of @xmath4 , @xmath35 and @xmath42 leads to the following values : @xmath43 , @xmath44 , @xmath45 .
these values obey the conditions @xmath46 and @xmath47 that have already been found@xcite to be obeyed for other ses activities .
if we repeat the computation for surrogate data obtained by _ shuffling _
the durations @xmath48 randomly ( and hence their distribution is conserved ) , the corresponding quantities , designated by adding a subscript `` shuf '' , have the following values : @xmath49 and @xmath50 with standard deviations 0.008 and 0.011 respectively .
they are almost equal to the corresponding values of a `` uniform '' distribution , i.e. , @xmath51 and @xmath52 , although the experimental error in this case is large due to the small number of pulses . by applying the same procedure to other ses activities reported earlier@xcite , we find ( see table [ tab0 ] ) that actually @xmath53 and @xmath54 .
this points to the conclusion that the self - similarity of ses activity results from the process memory only , which agrees with an independent analysis of ref.@xcite .
( a)an ses activity recently recorded at pat station(sampling rate @xmath55hz ) . the actual electric field * e * is 6 mv / km , but here the signal is presented in normalized units , i.e. , by subtracting the mean value @xmath56 and dividing by the standard deviation @xmath57 .
( b ) how the ses activity in ( a ) is read in natural time . ] ( color online )
the values of @xmath58 as a function of @xmath59 for the case of the fbm simulations of ref.@xcite(crosses , lower scale ) or as a function of @xmath3 for power law distributed data ( asterisks , upper scale ) . they have been estimated by following the procedure described in the appendix b of ref.@xcite .
the corresponding lines have been drawn as a guide to the eye . ]
( color online ) the probability density function @xmath6 versus @xmath4 for several values of @xmath3 ( see the text and @xcite ) . ]
( color online)the pdfs of @xmath4 when using either the actual seismic catalogue of japan(solid ) treated in ref.@xcite or the same data in random order ( dashed ) . ]
( color online ) the calculated values of the differential entropy @xmath60 versus the exponent @xmath3 .
two window lenghts are used and their results are almost the same . ] [ cols="^,^,^,^,^,^",options="header " , ] in addition , the detrended fluctuation analysis ( dfa)@xcite in natural time of the ses activity depicted in fig.[fg1 ] , leads to an exponent @xmath61 , which agrees with the earlier finding @xmath62 in several other ses activities@xcite .
interestingly , the values of the quantities @xmath63 and @xmath59 are consistent with the results deduced from a numerical simulation in fbm time series described in ref.@xcite ; the latter showed that when @xmath62 the corresponding values are : @xmath46 , @xmath64 .
figure [ fg2 ] depicts the most probable value @xmath58 of @xmath4 versus @xmath59 resulting from such a numerical simulation .
we now study a case of self - similarity resulting from the process increments `` infinite '' variance . here , we restrict ourselves to slowly driven systems that emit energy bursts obeying power law distribution @xmath65 where @xmath3 is constant . in a large variety of such systems , in diverse fields , an inspection of the recent experimental data reveals that the @xmath3 exponent lies in a narrow range , i.e. , @xmath66 ( and mostly even within narrower bounds , i.e. , @xmath67 1.5 to 1.8 ) . to realize the diversity of the phenomena that exhibit the aforementioned property
, we compile some indicative examples in table [ tab1 ] , which are the following .
first , crystalline materials subjected to an external stress , display bursts of activity owing the nucleation and motion of dislocations .
these sudden local changes produce acoustic emission waves which reveal that a large number of dislocations move cooperatively in an intermittent fashion ( e.g. , see @xcite and references therein ) . as a precise example , we include in table [ tab1 ] the results of acoustic emission experiments on stressed single crystals of ice under viscoelastic deformation ( creep ) , which show that the probability distribution of energy bursts intensities obey a power - law distribution with @xmath67 1.6 spanning many decades ( see fig.[fg1 ] of @xcite ) .
second , we consider the case of solar flares that represent impulsive energy releases in the solar corona ( e.g. see ref . @xcite and references therein ; see also ref.@xcite in which it is concluded that earthquakes and solar flares exhibit the same distributions of sizes , interoccurrence times , and the same temporal clustering ) .
this energy release is observed in various forms : thermal , soft and hard x - ray emissions , accelerated particles etc .
the statistical analysis of these impulsive events show that the energy distribution exhibit , over several orders of magnitude , a power - law with exponents @xmath3 ranging from 1.5 to approximately 2.1 ( depending on the experimental procedure and the geometrical assumptions adopted in the analysis ) .
other examples are : acoustic emission from microfractures before the breakup of heterogeneous materials ( wood , fiberglass ) , icequakes and earthquakes .
concerning the latter , the best known scaling relation is the gutenberg - richter law@xcite , which states that the ( cumulative ) number of earthquakes with magnitude greater than m occurring in a specified area and time is given by @xmath68 where @xmath8 is a constant , which varies only slightly from region to region ( cf .
eq.([eq10 ] ) holds both _ regionally _ and _ globally _ ) being generally in the range @xmath69 ( see @xcite and references therein ) .
considering that the seismic energy @xmath1 released during an earthquake is related@xcite to the magnitude through @xmath70 -where @xmath71 is around 1.5- eq.([eq10 ] ) turns to eq.([eq9 ] ) , where @xmath72 .
hence , @xmath73 means that the exponent @xmath3 is around @xmath3=1.6 to 1.7 .
the following procedure is now applied : we generate(see also@xcite ) a large amount of artificial data obeying eq.([eq10 ] ) for a certain @xmath3 value .
these _ randomized _ ( `` shuffled''@xcite ) data are subsequently analyzed , in the natural time domain , for each @xmath3 value , with the following procedure@xcite : first , calculation of the variance @xmath4 is made for an event taking time windows for 6 to 40 consecutive events ( the choice of the precise value of the upper limit is not found decisive , because practically the same results are obtained even if the number of consecutive events was changed from 6 - 40 to 6 - 100 , as it will be further discussed below ) . and second , this process was performed for all the events by scanning the whole dataset . in fig.[fg3 ] we plot the quantity @xmath6 versus @xmath4 for several @xmath3-values .
the most probable value @xmath58 ( for @xmath3 = constant ) is also plotted in fig.[fg2 ] versus the corresponding @xmath3-value .
this curve interrelates @xmath4 and @xmath3 for the shuffled data ( thus an eventual process memory is here destroyed@xcite ) and hence the plotted @xmath58 values ( which differ markedly from @xmath74 ) correspond to the self - similarity resulting from the heavy - tailed distribution only . in order to identify the origin of self - similarity in a real data set , let us consider here the example of earthquakes . using the japan catalogue mentioned in ref.@xcite , we give in fig.4 the two curves @xmath6 versus @xmath4 that result when the aforementioned calculation is made by means of a window of 6 - 40 concecutive events sliding through either the original catalogue or a shuffled one . comparing the resulting @xmath58 values ( both of which markedly differ from @xmath74 )
we see that the value of the surrogate data ( @xmath75 ) differs slightly from the one ( @xmath76 ) corresponding to the original data .
this reflects that the self - similarity mainly originates from the process increments `` infinite '' variance .
note , however , that the @xmath58 value of the original data is comparable to the value @xmath4=0.070 that was found in infinitely ranged _ temporal _ correlations . this merits further investigation .
here , we discuss a challenging point that emerges from a further elaboration of the results depicted in fig.3 . first , note that upon increasing the @xmath3 value from @xmath3 = 1.3 to 2.0 , the feature of the curve changes significantly , becoming bimodal at intermediate @xmath3-values .
second , we calculate , for each @xmath3-value studied , the so called _ differential entropy _ , defined as @xmath77 which is the shannon information entropy of a continuous probability distribution , e.g. , see @xcite .
( note that , the shannon information entropy is _
static _ entropy and _ not _ a dynamic one @xcite . )
finally , we investigate the resulting @xmath60-values versus @xmath3 .
such a plot is given in fig.5 , whose inspection reveals that @xmath60 maximizes at a value of @xmath3 lying between @xmath3 = 1.6 and @xmath3=1.7 , which is more or less comparable with the experimental values , see table [ tab1 ] .
( this value is not practically affected by the window length ( @xmath78 ) chosen ; in reality , upon increasing @xmath78 from @xmath78=10 to @xmath78=1000 , we find that , @xmath3-value at which @xmath60 maximizes in fig.[fg2 ] , decreases only slightly from @xmath3=1.70 to @xmath3=1.63 . ) in particular for the case of earthquakes this @xmath3-value corresponds to @xmath73 , thus agreeing with the experimental findings mentioned above .
does it mean that the @xmath8 or @xmath3 value can be determined just by applying the maximum entropy principle in the sense developed by jaynes@xcite , who suggested to look statistical mechanics as a form of _ statistical inference _ and _ start _ statistical physics from the principle of maximum entropy inference ( maxent ) ?
this is not yet clear , because a widely accepted formalism for _ non - equilibrium _ statistical mechanics is still lacking . finally
, the fact that in some experiments the resulting @xmath3-values differ slightly from @xmath3=1.6 to 1.7 predicted from fig.[fg5 ] could be attributed to the following : figure [ fg5 ] is based on randomized data , while the actual data may also exhibit temporal correlations ( e.g. , the case of aftershocks ) .
in addition , finite size effects@xcite might play a significant role .
in summary , the origin of self - similarity may be distinguished as follows : if self - similarity exclusively results from the process memory , the @xmath4 value should _ change _ to @xmath74=0.0833 ( and the values of @xmath79 to @xmath80=0.0966 ) for the surrogate data . on the other hand ,
if the self - similarity results from process increments `` infinite '' variance only , the @xmath58 values should be the same ( but differing from @xmath74 ) for the original and surrogate data .
when studying the differential entropy associated with the pdf of @xmath4 it maximizes when the exponent @xmath3 lies in the narrow range 1.6 to 1.7 , in agreement with the experimental findings in diverse fields .
this , for the case of earthquakes , immediately reflects that the @xmath8-value in the gutenberg - richter law is @xmath73 , as actually observed . | self - similarity may originate from two origins , i.e. , the process memory and the process increments `` infinite '' variance . a distinction is attempted by employing the natural time @xmath0 . concerning the first origin , we analyze recent data on seismic electric signals , which support the view that they exhibit infinitely ranged _ temporal _ correlations . concerning the second , slowly driven systems that emit bursts of various energies @xmath1
obeying power - law distribution , i.e. , @xmath2 , are studied .
an interrelation between the exponent @xmath3 and the variance @xmath4(@xmath5 ) is obtained for the shuffled ( randomized ) data . in the latter ,
the most probable value of @xmath4 is approximately equal to that of the original data .
finally , it is found that the differential entropy associated with the probability @xmath6 maximizes for @xmath3 around @xmath7 1.6 to 1.7 , which is comparable to the value determined experimentally in diverse phenomena , e.g. , solar flares , icequakes , dislocation glide in stressed single crystals of ice e.t.c .
it also agrees with the @xmath8-value in the gutenberg - richter law of earthquakes . |
let @xmath3 be an oriented closed surface of genus @xmath4 the @xmath5-representation space @xmath6 is the space of group homomorphisms @xmath7 from the fundamental group of @xmath3 into @xmath8 endowed with the compact open topology .
the euler class @xmath9 of @xmath1 is the euler class of the associated @xmath10-bundle on @xmath11 which satisfies the milnor - wood inequality @xmath12 in @xcite , goldman proved that equality holds if and only if @xmath1 is fuchsian , i.e. , discrete and faithful ; and in @xcite , he proved that the connected components of @xmath6 are indexed by the euler classes .
i.e. , for each integer @xmath13 with @xmath14 the representations of euler class @xmath13 exist and form a connected component of @xmath15 the lie group @xmath5 acts on @xmath6 by conjugation , and the quotient space @xmath16 is the character space of @xmath17 since the euler classes are preserved by conjugation , the connected components of @xmath18 are also indexed by the euler classes , i.e. , @xmath19 where @xmath20 is the space of conjugacy classes of representations of euler class @xmath13 . by the results of goldman@xcite , the extremal components @xmath21 are respectively identified with the teichmller space of @xmath3 and that of @xmath3 endowed with the opposite orientation .
the mapping class group @xmath22 of @xmath3 is the group of isotopy classes of orientation preserving self - diffeomorphisms of @xmath17 by the dehn - nielsen theorem , @xmath22 is naturally isomorphic to the group of positive outer - automorphisms @xmath23 which acts on @xmath18 preserving the euler classes .
therefore , @xmath22 acts on each connected component of @xmath24 it is well known ( see e.g. fricke@xcite ) that the @xmath22-action is properly discontinuous on the extremal components @xmath25 i.e. , the teichmller spaces , and the quotients are the riemann moduli spaces of complex structures on @xmath17 on the non - extremal components @xmath26 @xmath27 goldman conjectured in @xcite that the @xmath22-action is ergodic with respect to the measure induced by the goldman symplectic form @xcite .
closely related to goldman s conjecture is a question of bowditch@xcite , question c , that whether for each non - elementary and non - extremal ( i.e. non - fuchsian ) representation @xmath1 in @xmath28 there exists a simple closed curve @xmath29 on @xmath3 such that @xmath30)$ ] is an elliptic or a parabolic element of @xmath31 recall that a representation is non - elementary if its image is zariski - dense in @xmath31 recently , march - wolff@xcite show that an affirmative answer to bowditch s question implies that goldman s conjecture is true .
more precisely , they show that for @xmath32 @xmath22 acts ergodically on the subset of @xmath20 consisting of representations that send some simple closed curve on @xmath3 to an elliptic or parabolic element .
therefore , when @xmath32 the @xmath22-action on @xmath20 is ergodic if and only if the subset above has full measure in @xmath33 in the same work , they answer bowditch s question affirmatively for the genus @xmath34 surface @xmath35 implying the ergodicity of the @xmath36-action on @xmath37 for the action on the component @xmath38 they find two @xmath36-invariant open subsets due to the existence of the hyper - elliptic involution , and show that on each of them the @xmath36-action is ergodic . for higher genus surfaces
@xmath11 @xmath39 souto @xcite recently gives an affirmative answer to bowditch s question for the euler class @xmath40 representations , proving the ergodicity of the @xmath22-action on @xmath41 for @xmath42 and @xmath43 both bowditch s question and goldman s conjecture are still open .
+ bowditch s question was originally asked for the type - preserving representations of punctured surface groups .
recall that a punctured surface @xmath44 of genus @xmath45 with @xmath46 punctures is a closed surface @xmath3 with @xmath46 points removed . through out this paper , we required that the euler characteristic of @xmath44 is negative .
a _ peripheral element _ of @xmath47 is an element that is represented by a curve freely homotopic to a circle that goes around a single puncture of @xmath48 a representation @xmath49 is called _ type - preserving _ if it sends every peripheral element of @xmath47 to a parabolic element of @xmath31 in @xcite , question c asks whether it is true that if a non - elementary type - preserving representation of a punctured surface group sends every non - peripheral simple closed curve to a hyperbolic element , then @xmath1 must be fuchsian .
the main result of this paper gives counterexamples to this question . to state the result , we recall that there is the notion of relative euler class @xmath9 of a type - preserving representation @xmath1 that satisfies the milnor - wood inequality @xmath50 and equality holds if and only if @xmath1 is fuchsian ( see @xcite and also proposition [ mf ] ) . [ main1 ] there are uncountably many non - elementary type - preserving representations @xmath51 with relative euler class @xmath52 that send every non - peripheral simple closed curve to a hyperbolic element . in particular
, these representations are not fuchsian .
our method is to use penner s lengths coordinates for the decorated character space defined by kashaev in @xcite .
briefly speaking , decorated character space of a punctured surface is the space of conjugacy classes of _ decorated representations _ , namely , non - elementary type - preserving representations together with an assignment of _ horocycles _ to the punctures .
the lengths coordinate of a decorated representation depend on a choice of an ideal triangulation of the surface , and consists of the _ @xmath53-lengths _ of the edges determined by the horocycles , and of the _ signs _ of the ideal triangles determined by the representation .
the decorated teichmller space is a connected component of the decorated character space , and the restriction of the lengths coordinates to this component coincide with penner s lengths coordinates .
( see @xcite or section [ sec:2 ] for more details . ) a key ingredient in the proof is formula ( [ trace ] ) of the traces of closed curves in the lengths coordinates , found in @xcite by sun and the author . with a careful choice of an ideal triangulation of the four - punctured sphere , called a _
tetrahedral triangulation _
, we show that the traces of three _ distinguished simple closed curves _ are greater than @xmath34 in the absolute value if and only if the @xmath53-lengths of edges in this triangulation satisfy certain anti - triangular inequalities .
we then show that each simple closed curve is distinguished in some tetrahedral triangulation , and all tetrahedral triangulations are related by a sequence of moves , called the _ simultaneous diagonal switches_. by the change of @xmath53-lengths formula ( proposition [ change ] ) , we show that the anti - triangular inequalities are preserved by the simultaneous diagonal switches .
therefore , if the three distinguished simple closed curves are hyperbolic , then all the simple closed curves are hyperbolic .
finally , we show that there are uncountably many choices of the @xmath53-lengths that satisfy the anti - triangular inequalities .
a consequence of formula ( [ trace ] ) is theorem [ dominance ] that each non - fuchsian type - preserving representation is dominated by a fuchsian one , in the sense that the traces of the simple closed curves of the former representation are less than or equal to those of the later in the absolute value .
this is a counterpart of the result of gueritaud - kassel - wolff@xcite and deroin - tholozan@xcite , where they consider dominance of closed surface group representations . using the same technique
, we also give an affirmative answer to bowditch s question for the relative euler class @xmath40 type - preserving representations of the four - punctured sphere group .
[ main2 ] every non - elementary type - preserving representation @xmath51 with relative euler class @xmath54 sends some non - peripheral simple closed curve to an elliptic or parabolic element .
in contrast with the connected components of the character space of a closed surface , those of a punctured surface are more subtle to describe . for @xmath44 with @xmath55 denote by @xmath56 be the space of conjugacy classes of type - preserving representations with relative euler class @xmath57 as explained in @xcite , @xmath56 can be either empty or non - connected .
for example , @xmath58 the non - connectedness of @xmath56 comes from the existence of two distinct conjugacy classes of parabolic elements of @xmath31 more precisely , each parabolic element of @xmath5 is up to @xmath59 conjugate to an upper triangular matrix with trace @xmath60 and its conjugacy class is distinguished by whether the sign of the non - zero off diagonal element is positive or negative .
therefore , two type - preserving representations of the same relative euler class which respectively send the same peripheral element into different conjugacy classes of parabolic elements can not be in the same connected components . throughout this paper ,
we respectively call the two conjugacy class of parabolic elements the _ positive _ and the _ negative _ conjugacy classes . for a type - preserving representation @xmath61
we say that the _ sign _ of a puncture @xmath62 is _ positive _ , denoted by @xmath63 if @xmath1 sends a peripheral element around this puncture into the positive conjugacy class of parabolic elements , and is _ negative _ , denoted by @xmath64 if otherwise .
for an @xmath65 we denote by @xmath66 the space of conjugacy classes of type - preserving representations with relative euler class @xmath13 and signs of the punctures @xmath67 it is conjectured in @xcite that each @xmath68 if non - empty , is connected .
the result confirms this for the four - punctured sphere .
[ main3 ] let @xmath69 then 1 .
@xmath70 is non - empty if and only if @xmath71 contains exactly two @xmath72 and two @xmath73 2 .
@xmath74 is non - empty if and only if @xmath71 contains at most one @xmath75 3 .
@xmath76 is non - empty if and only if @xmath71 contains at most one @xmath73 and 4 .
all the non - empty spaces above are connected . as a consequence of theorem [ main3 ] , @xmath77 has six connected components and each of @xmath78 has five connected components .
the main tool we use in the proof is still the lengths coordinates ; and we hope that the technique could be used for the other punctured surfaces . the mapping class group @xmath79 of a punctured surface @xmath44 is the group of relative isotopy classes of orientation preserving self - diffeomorphisms of @xmath44 that fix the punctures . by the dehn - nielsen theorem ,
@xmath79 is isomorphic to the group of positive outer - automorphisms @xmath80 that preserve the cyclic subgroups of @xmath47 generated by the peripheral elements , and hence acts on @xmath81 preserving the relative euler classes and the signs of the punctures .
therefore , for any integer @xmath13 with @xmath82 and for any @xmath83 @xmath79 acts on @xmath84 for the four - punctured sphere , we have [ main4 ] the @xmath85-action on each non - extremal connected component of @xmath86 is ergodic . by march - wolff@xcite , it is not surprising that the @xmath85-action is ergodic on the connected components of @xmath86 where bowditch s question has an affirmative answer .
a new and unexpected phenomenon theorem [ main4 ] reveals here is that , for punctured surfaces , the action of the mapping class group can still be ergodic when the answer to bowditch s question is negative .
evidenced by theorem [ main4 ] , we make the following the @xmath79-action is ergodic on each non - extremal connected component of @xmath87 the paper is organized as follows . in section [ sec:2 ] ,
we recall kashaev s decorated character spaces and the lengths coordinates , in section [ sec:3 ] , we obtain a formula of the traces of closed curves in the lengths coordinates , and in section [ sec:4 ] , we introduce tetrahedral triangulations , distinguished simple closed curves and simultaneous diagonal switches
. then we prove theorems [ main1 ] and [ main2 ] , theorem [ main3 ] and theorem[main4 ] respectively in sections [ sec:5 ] , section [ sec:6 ] and section [ sec:7 ] .
it is pointed out by the anonymous referee that the results concerning representations of relative euler class @xmath2 can be deduced more directly from the results of goldman in @xcite , where we present his argument in appendix [ b ] for the interested readers .
+ * acknowledgments : * the author is grateful to the referee for bringing his attention to the relationship of this work with a previous work of goldman , and for several valuable suggestions to improve the writing of this paper .
the author also would like to thank steven kerckhoff , feng luo and maryam mirzakhani for helpful discussions and suggestions , ser peow tan , maxime wolff , sara moloni , frdric palesi and zhe sun for discussion and showing interest , julien roger for bringing his attention to kashaev s work on the decorated character variety , ronggang shi for answering his questions on ergodic theory and yang zhou for writing a python program for testing some of the author s ideas .
the author is supported by nsf grant dms-1405066 .
we recall the decorated character spaces and the lengths coordinates in this section .
the readers are recommended to read kashaev s papers @xcite for the original approach and for more details .
let @xmath44 be a punctured surface of genus @xmath45 with @xmath46 punctures , and let @xmath49 be a non - elementary type - preserving representation .
a _ pseudo - developing map _ @xmath88 of @xmath1 is a piecewise smooth @xmath1-equivariant map from the universal cover of @xmath44 to the hyperbolic plane @xmath89 by @xcite , @xmath1 is the holonomy representation of @xmath90 let @xmath91 be the hyperbolic area form of @xmath89 since @xmath88 is @xmath1-equivariant , the pull - back @xmath34-form @xmath92 descends to @xmath48 the relative euler class @xmath9 of @xmath1 could be calculated by @xmath93 an ideal arc @xmath94 on @xmath44 is an arc connecting two ( possibly the same ) punctures .
the image @xmath95 of a lift @xmath96 of @xmath94 is an arc in @xmath97 connecting two ( possibly the same ) points on @xmath98 each of which is the fixed point of the @xmath1-image of certain peripheral element of @xmath99 we call @xmath94 _
@xmath1-admissible _ if the two end points of @xmath95 are distinct .
it is easy to see that @xmath94 being @xmath1-admissible is independent of the choice of the lift @xmath96 and the pseudo - developing map @xmath90 an ideal triangulation @xmath100 of @xmath44 consists of a set of disjoint ideal arcs , called the edges , whose complement is a disjoint union of triangles , called the ideal triangles .
we call @xmath100 _ @xmath1-admissible _ if all the edges of @xmath100 are @xmath1-admissible .
if @xmath101 is conjugate to @xmath102 then it is easy to see that @xmath100 is @xmath101-admissible if and only if it is @xmath1-admissible . in @xcite
, kashaev shows that [ kashaev1 ] for each ideal triangulation @xmath103 the set @xmath104\in\mathcal m(\sigma_{g ,
n})\ \big|\ \mathcal t\text { is } \rho\text{-admissible}\big\}\ ] ] is open and dense in @xmath105 and there exist finitely many ideal triangulations @xmath106 @xmath107 such that @xmath108 let @xmath44 and @xmath1 be as above . a _ decoration _ of @xmath1 is an assignment of horocycles centered at the fixed points of the @xmath1-image of the peripheral elements of @xmath109 one for each , which is invariant under the @xmath110-action . in the case that the fixed points of the @xmath1-image of two peripheral elements coincide , which may happen only when @xmath1 is non - fuchsian , we do not require the corresponding assigned horocycles to be the same .
if @xmath111 is a decoration of @xmath1 and @xmath45 is an element of @xmath8 then the @xmath45-image of the horocycles in @xmath111 form a decoration @xmath112 of the conjugation @xmath113 of @xmath114 we call a pair @xmath115 a _ decorated representation _ , and call two decorated representations @xmath115 and @xmath116 _ equivalent _ if @xmath117 and @xmath118 for some @xmath119 the _ decorated character space _ of @xmath120 denoted by @xmath121 is the space of equivalence classes of decorated representations . in @xcite
, kashaev shows that the projection @xmath122 defined by @xmath123)=[\rho]$ ] is a principle @xmath124-bundle , where @xmath125 is the set of punctures of @xmath120 and the pre - image of the extremal components are isomorphic as principle @xmath124-bundles to penner s decorated teichmller space@xcite . fixing a @xmath1-admissible ideal triangulation @xmath100 and a pseudo - developing map @xmath126 the lengths coordinate of @xmath115 consists of the following two parts : the @xmath53-lengths of the edges and the signs of the ideal triangles .
the @xmath53-length of an edge @xmath127 of @xmath100 is defined as follows . since @xmath127 is @xmath1-admissible , for any lift @xmath128 of @xmath127 to the universal cover the image @xmath129 connects to distinct points @xmath130 and @xmath131 on @xmath132 the decoration @xmath111 assigns two horocycles @xmath133 and @xmath134 respectively centered at @xmath130 and @xmath135 let @xmath136 be the signed hyperbolic distance between the two horocycles , i.e. , @xmath137 if @xmath133 and @xmath134 are disjoint and @xmath138 if otherwise .
then the _ @xmath53-length _ of @xmath127 in the decorated representation @xmath139 is defined by @xmath140 the sign of an ideal triangle @xmath141 of @xmath100 is defined as follows .
let @xmath142 @xmath143 and @xmath144 be the vertices of @xmath141 so that the orientation on @xmath141 determined by the cyclic order @xmath145 coincides with the one induced from the orientation of @xmath48 let @xmath146 be a lift of @xmath141 to the universal cover , and let @xmath147 @xmath148 and @xmath149 be the vertices of @xmath146 so that @xmath150 is a lift of @xmath151 @xmath152 since @xmath100 is @xmath1-admissible , the points @xmath153 @xmath154 and @xmath155 are distinct on @xmath98 and hence determine a hyperbolic ideal triangle @xmath156 in @xmath97 with them as the ideal vertices .
the _ sign _ of @xmath141 is positive , denoted by @xmath157 if the orientation on @xmath156 determined by the cyclic oder @xmath158 coincides with the one induced from the orientation of @xmath89 otherwise , the sign of @xmath141 is negative , and is denoted by @xmath159 from the construction , it is easy to see that the @xmath53-lengths @xmath160 and the signs @xmath161 depend only on the equivalence class of @xmath162 let @xmath163 be the set of ideal triangles of @xmath164 then the integral of the pull - back form @xmath92 over @xmath44 equals @xmath165 and the relative euler class of @xmath1 can be calculated by @xmath166 let @xmath125 be the set of punctures of @xmath44 and let @xmath167 be the set of edges of @xmath164 then there is a principle @xmath168-bundle structure on @xmath169 defined as follows . for @xmath170 and @xmath171
we define @xmath172 by @xmath173 where @xmath174 and @xmath143 are the punctures connected by the edge
@xmath175 [ kashaev2 ] let @xmath122 be the principle @xmath168-bundle , and let @xmath176 the pre - image of @xmath177 then @xmath178 where each @xmath179 is isomorphic as principle @xmath168-bundles to an open subset of @xmath180 the isomorphism is given by the @xmath53-lengths , and the image of @xmath179 is the complement of the zeros of certain rational function coming from the image of the peripheral elements not being the identity matrix . on @xmath81 there is the goldman symplectic form @xmath181 which restricts to the weil - petersson symplectic form on the teichmller component@xcite .
by @xcite , for each ideal triangulation @xmath103 the pull - back of @xmath181 to @xmath176 is expressed in the @xmath53-lengths by @xmath182 where @xmath183 @xmath184 and @xmath185 are the edges of the ideal triangle @xmath141 in the cyclic order induced from the orientation of @xmath48 this formula is first obtained by penner@xcite for the decorated teichmller space . from ( [ wp ] ) , it is easy to see that the measure on each @xmath179 induced by @xmath186 is in the measure class of the pull - back of the lebesgue measure of @xmath180 a diagonal switch at an edge @xmath127 of @xmath100 replaces the edge @xmath127 by the other diagonal of the quadrilateral formed by the union of the two ideal triangles adjacent to @xmath175 by @xcite , any ideal triangulation can be obtained from another by doing a finite sequence of diagonal switches .
let @xmath115 be a decorated representation and let @xmath100 be a @xmath1-admissible ideal triangulation of @xmath48 if @xmath187 is the ideal triangulation of @xmath44 obtained from @xmath100 by doing a diagonal switch at an edge @xmath188 then the @xmath1-admissibility of @xmath187 and the lengths coordinate of @xmath115 in @xmath187 are determined as follows .
let @xmath189 and @xmath190 be the two ideal triangles of @xmath100 adjacent to @xmath188 let @xmath191 be the new edge of @xmath187 and let @xmath192 and @xmath193 be the two ideal triangles in @xmath187 adjacent to @xmath194 we respectively name the edges of the quadrilateral @xmath195 in the way that @xmath196 is adjacent to @xmath189 and @xmath197 @xmath184 is adjacent to @xmath189 and @xmath198 @xmath185 is adjacent to @xmath190 and @xmath193 and @xmath199 is adjacent to @xmath190 and @xmath200 then @xmath196 is opposite to @xmath201 and @xmath184 is opposite to @xmath199 in the quadrilateral .
[ change ] 1 . if the signs @xmath202 then @xmath187 is @xmath1-admissible . in this case ,
@xmath203 and the signs of the common ideal triangles and the @xmath53-lengths of the common edges of @xmath100 and @xmath187 do not change .
2 . if @xmath204 then @xmath187 is @xmath1-admissible if and only if @xmath205 in this case , 1 . if @xmath206 then @xmath207 2 .
if @xmath208 then @xmath209 + and the signs of the common ideal triangles and the @xmath53-lengths of the common edges of @xmath100 and @xmath187 do not change , the rule for the signs in ( 2.1 ) and ( 2.2 ) is that the signs of the ideal triangles adjacent to the shorter edges do not change .
this could be seen as follows .
if , for example , @xmath210 @xmath211 and @xmath206 then the hyperbolic ideal triangle @xmath212 determined by @xmath189 is negatively oriented , the hyperbolic ideal triangle @xmath213 determined by @xmath190 is positively oriented , and the geodesic arcs @xmath214 and @xmath215 determined by @xmath184 and @xmath199 intersect . see figure [ fig1 ] . as a consequence ,
the hyperbolic ideal triangle @xmath216 determined by @xmath192 is negatively oriented and the hyperbolic ideal triangle @xmath217 determined by @xmath193 is positively oriented , hence @xmath218 and @xmath219 the @xmath53-lengths of @xmath191 follows from penner s ptolemy relation@xcite that @xmath220 the other cases could be verified similarly . by theorems [ kashaev1 ] and [ kashaev2 ] ,
the family of open subsets @xmath221 where @xmath100 goes over all the ideal triangulations of @xmath44 together with the @xmath53-lengths functions @xmath222 form a coordinate system of @xmath121 and the transition functions are given by proposition [ change ] .
[ kashaev3 ] for each relative euler class @xmath223
let @xmath224 be the pre - image of @xmath56 under the projection @xmath225 then @xmath226 where the union is over all the ideal triangulations @xmath100 and the disjoint union is over all @xmath227 such that @xmath228 moreover , @xmath224 s are principle @xmath168-bundles , and are disjoint for different @xmath57
throughout this section , we let @xmath100 be an ideal triangulation of @xmath120 and let @xmath167 and @xmath163 respectively be the set of edges and ideal triangles of @xmath164 given the lengths coordinate @xmath229 up to conjugation , the type - preserving representation @xmath49 can be reconstructed up to conjugation as follows .
suppose @xmath127 is an edge of @xmath103 and @xmath189 and @xmath190 are the two ideal triangles adjacent to @xmath175 let @xmath196 and @xmath184 be the other two edges of @xmath189 and let @xmath185 and @xmath199 be the other two edges of @xmath190 so that the cyclic orders @xmath230 and @xmath231 coincide with the one induced from the orientation of @xmath48 define the quantity @xmath232 by @xmath233 note that if @xmath1 is discrete and faithful , then @xmath234 is the shear parameter of the corresponding hyperbolic structure at @xmath175 ( see e.g. @xcite . )
it is well known that each immersed closed curve on @xmath44 is homotopic to a normal one that transversely intersects each ideal triangle in simple arcs that connect different edges of the triangle .
let @xmath29 be an immersed oriented closed normal curve on @xmath48 for each edge @xmath127 intersecting @xmath235 define @xmath236.\ ] ] for each ideal triangle @xmath141 intersecting @xmath235 define @xmath237\ ] ] if @xmath29 makes a left turn in @xmath141 ( figure [ fig2 ] ( a ) ) , and define @xmath238\ ] ] if @xmath29 makes a right turn in @xmath141 ( figure [ fig2 ] ( b ) ) .
[ 3.1 ] let @xmath239 be the edges and let @xmath240 be the ideal triangles of @xmath100 intersecting @xmath29 in the cyclic order induced by the orientation of @xmath29 so that @xmath241 is the common edge of @xmath242 and @xmath243 for each @xmath244 then up to a conjugation by an element of @xmath8 @xmath245)=\pm s(e_{i_1})r(t_{j_1})s(e_{i_2})r(t_{j_2})\dots s(e_{i_m})r(t_{j_m}).\ ] ] the proof is parallel to that of lemma 3 in @xcite .
the idea is to keep track of the image of the unit tangent vector @xmath246 at @xmath247 under @xmath30).$ ] the contributions of each edge @xmath127 and of each ideal triangle @xmath141 intersecting @xmath29 to @xmath30)$ ] are respectively @xmath248 and @xmath249 see also exercise 8.5 - 8.7 and 10.14 in @xcite . for each puncture @xmath62 of @xmath120 let @xmath250 be the simple closed curve going counterclockwise around @xmath62 once . by lemma [ 3.1 ] , the image of @xmath250 is up to conjugation @xmath251)= \pm\left [ \begin{array}{cc } 1 & \psi_{v,\epsilon}(\lambda)\\ 0&1 \end{array } \right],\ ] ] where @xmath252 is a rational function of @xmath53 depending on @xmath253 therefore , @xmath1 is type - preserving if and only if @xmath254 for all punctures @xmath255 the following proposition gives a more precise description of this rational function in theorem [ kashaev2 ] .
[ rational ] let @xmath229 let @xmath125 be the set of punctures of @xmath44 and let @xmath256 be the rational function defined by @xmath257 then @xmath258 defines a type - preserving representation if and only if @xmath259 the following theorem provides a more direct way to calculate the absolute values of the traces of closed curves using the @xmath53-lengths , which is first found by sun and the author in @xcite .
we include a proof here for the reader s convenience .
for each ideal triangle @xmath141 intersecting @xmath235 let @xmath260 be the edge of @xmath141 at which @xmath29 enters , let @xmath261 be the edge of @xmath141 at which @xmath29 leaves and let @xmath262 be the other edge of @xmath263 see figure [ fig2 ] . define @xmath264\ ] ] if @xmath29 makes a left turn in @xmath141 , and define @xmath265\ ] ] if @xmath29 makes a right turn in @xmath263 [ trace formula ] for an immersed closed normal curve @xmath29 on @xmath120 let @xmath239 be the edges and let @xmath240 be the ideal triangles of @xmath100 intersecting @xmath29 in the cyclic order following the orientation of @xmath29 so that @xmath241 is the common edge of @xmath242 and @xmath243 for each @xmath244 then @xmath266)\big|=\frac{\big|tr\big(m(t_{j_1})\dots m(t_{j_m})\big)\big|}{\lambda(e_{i_1})\dots\lambda(e_{i_m})}.\ ] ] for each ideal triangle @xmath141 and an edge @xmath127 of @xmath267 let @xmath191 and @xmath268 be the other two edges of @xmath141 so that the cyclic order @xmath269 coincides with the one induced by the orientation of @xmath48 define the matrix @xmath270.\ ] ] then @xmath271 for each @xmath272 where as a convention @xmath273 a case by case calculation shows that @xmath274 for each @xmath272 where as a convention @xmath275 by lemma [ 3.1 ] , ( [ split ] ) , ( [ local ] ) and the fact that @xmath276 for any two matrices @xmath277 and @xmath278 we have @xmath279)\big|=&\big|tr\big(s(e_{i_1})r(t_{j_1})\dots s(e_{i_m})r(t_{j_m})\big)\big|\\ = & \big|tr\big(s(t_{j_1},e_{i_1})r(t_{j_1})s(t_{j_1},e_{i_2})\dots s(t_{j_m},e_{i_m})r(t_{j_m } ) s(t_{j_m},e_{i_1})\big)\big|\\ = & \frac{\big|tr\big(m(t_{j_1})\dots m(t_{j_m})\big)\big|}{\lambda(e_{i_1})\dots\lambda(e_{i_m})}. \end{split}\ ] ] formula ( [ trace ] ) is first obtained by roger - yang@xcite for decorated hyperbolic surfaces , i.e. , discrete and faithful decorated representations , using the skein relations , where their formula includes both the traces of closed geodesics and the @xmath53-lengths of geodesics arcs connecting the punctures .
it is interesting to know whether there is a similar formula for the @xmath53-lengths of arcs for the non - fuchsian decorated representations . as a consequence of theorem [ trace formula ]
, we have the following [ dominance ] 1 . for every non - fuchsian type - preserving representation @xmath280
there exists a fuchsian type - preserving representation @xmath101 such that @xmath281)\big|\leqslant\big|tr\rho'([\gamma])\big|\ ] ] for each @xmath282\in\pi_1(\sigma_{g , n}),$ ] and the strict equality holds for at least one @xmath283 2 .
conversely , for almost every fuchsian type - preserving representation @xmath284 and for each @xmath13 with @xmath285 and @xmath286 there exists a type - preserving representation @xmath1 with @xmath287 such that @xmath281)\big|\leqslant\big|tr\rho'([\gamma])\big|\ ] ] for each @xmath282\in\pi_1(\sigma_{g , n}),$ ] and the strict equality holds for at least one @xmath283 for ( 1 ) , by theorem [ kashaev1 ] , there exists a @xmath1-admissible ideal triangulation @xmath164 choose arbitrarily a decoration @xmath111 of @xmath102 and let @xmath288 be the decorated representation that has the same @xmath53-lengths of @xmath115 and has the positive signs for all the ideal triangles . by ( [ euler ] ) and goldman s result in @xcite
, @xmath101 is fuchsian .
applying formula ( [ trace ] ) to @xmath289)|$ ] and @xmath290)|,$ ] we see that they have the same summands with different coefficients @xmath291 , and the coefficients for the later are all positive .
since each summand is a product of the @xmath53-lengths , which is positive , the inequality follows .
since @xmath1 is non - fuchsian , by ( [ euler ] ) , there must be an ideal triangle @xmath141 that has negative sign in @xmath162 therefore , if @xmath29 intersects @xmath267 then some of the summands in the expression of @xmath289)|$ ] has negative coefficients , and the inequality for @xmath29 is strict
. for ( 2 ) , choose arbitrarily an ideal triangulation @xmath100 of @xmath120 and let @xmath163 be the set of ideal triangles of @xmath164 by theorems [ kashaev1 ] , [ kashaev2 ] and [ kashaev3 ] , if @xmath286 then there exists @xmath292 such that @xmath293 and the subset @xmath179 is homeomorphic via the lengths coordinate to a full measure open subset @xmath294 of @xmath180 for each @xmath295 let @xmath115 be the decorated representation determined by @xmath296 then @xmath297 on the other hand , @xmath169 is identified with the decorated teichmller space via the lengths coordinate , hence @xmath53 determines a fuchsian type - preserving representation @xmath298 by the same argument in ( 1 ) , the inequality holds for @xmath1 and @xmath299 and is strict for @xmath29 intersecting the ideal triangles @xmath141 with @xmath159
it is very interesting to know if ( 2 ) holds for every fuchsian type - preserving representation . this amounts to ask wether @xmath300 where the union is over all the ideal triangulations @xmath100 of @xmath44 and all @xmath301 that gives the right relative euler class .
a _ tetrahedral triangulation _ of the four - punctured sphere @xmath302 is an ideal triangulation of @xmath302 that is combinatorially equivalent to the boundary of an euclidean tetrahedron ( figure [ fig3](a ) ) .
a pair of edges of a tetrahedral triangulation are called opposite if they are opposite edges of the tetrahedron .
let @xmath303 be the four punctures of @xmath304 in the rest of this paper , for each tetrahedral triangulation @xmath103 we will let @xmath305 be the unique ideal triangle of @xmath100 disjoint from the puncture @xmath306 and let @xmath307 be the unique edge of @xmath100 connecting the punctures @xmath306 and @xmath308 we respectively denote by @xmath309 the pair of opposite edges @xmath310 by @xmath311 the pair @xmath312 and by @xmath313 the pair @xmath314 see figure [ fig3 ] ( b ) .
a non - peripheral simple closed curve on @xmath302 is _ distinguished _ in a tetrahedral triangulation @xmath100 if it is disjoint from a pair of opposite edges of @xmath100 and intersects each of the other four edges at exactly one point . in each tetrahedral triangulation
, there are exactly three distinguished simple closed curves .
we respectively denote by @xmath315 @xmath316 and @xmath317 the distinguished simple closed curves disjoint from the pair of opposite edges @xmath318 @xmath311 and @xmath319 see figure [ fig4 ] . the curves @xmath315 @xmath316 and @xmath317 mutually intersect at exactly two points . on the other hand , for each triple of simple closed curves that mutually intersect at two points , there is a unique tetrahedral triangulation in which these three curves are distinguished . in particular , each non - peripheral simple closed curve on @xmath302 is distinguished in some tetrahedral triangulation . note
that being the @xmath320- , @xmath316- or @xmath317-curve is independent of the choice of the tetrahedral triangulation , since , for example , the curve @xmath320 always separates @xmath321 from @xmath322 in the rest of this paper , we will call a simple closed curve an @xmath320- ( resp .
@xmath316- or @xmath317- ) _ curve _ if it is disjoint from the pair of opposite edges @xmath309 ( resp . @xmath311 or @xmath313 ) of some tetrahedral triangulation . in this way
, we get a tri - coloring of the set of non - peripheral simple closed curves on @xmath304 a _ simultaneous diagonal switch _ at a pair of opposite edges of @xmath100 is an operation that simultaneously does diagonal switches at this pair of edges
. see figure [ fig6 ] ( a ) .
denote respectively by @xmath323 @xmath324 and @xmath325 the simultaneous diagonal switches at the pair of opposite edges @xmath318 @xmath311 and @xmath319 then @xmath326 ( reps . @xmath324 and @xmath325 ) changes the @xmath320- ( resp . @xmath316- and @xmath317- ) curve and leaves the other two distinguished simple closed curves unchanged .
see figure [ fig6 ] ( b ) .
the relationship between tetrahedral triangulations , simultaneous diagonal switches and non - peripheral simple closed curves can be described by ( the dual of ) the farey diagram .
recall that the farey diagram @xmath327 is an ideal triangulation of @xmath97 whose vertices are the extended rational numbers @xmath328 and the dual ferey diagram @xmath329 is a countably infinite trivalent tree properly embedded in @xmath89 each vertex of @xmath327 corresponds to a non - peripheral simple closed curve on @xmath330 each edge of @xmath327 connects two vertices corresponding to two simple closed curves that intersect at exactly two points and each ideal triangle of @xmath327 corresponds to a triple of simple closed curves mutually intersecting at two points .
( see e.g. @xcite . )
therefore , each vertex of the dual graph @xmath329 corresponds to a tetrahedral triangulation of @xmath330 each edge of @xmath329 corresponds to a simultaneous diagonal switch and each connected component of @xmath331 corresponds to a non - peripheral simple closed curves on @xmath304 see figure [ fig7 ] .
since @xmath329 is connected , any tetrahedral triangulation can be obtained from another by doing a finitely sequence of simultaneous diagonal switches .
we close up this section by showing the relationship between simultaneous diagonal switches and the mapping classes of @xmath304 [ evenodd ] a composition of an even number of simultaneous diagonal switches determines an element of @xmath332 conversely , any element of @xmath85 is determined by a composition of an even number of simultaneous diagonal switches .
let @xmath100 be a tetrahedra triangulation of @xmath304 we write @xmath333 if @xmath334 is the self - diffeomorphism of @xmath302 such that the tetrahedral triangulation @xmath335 is obtained from @xmath100 by doing the simultaneous diagonal switch @xmath336 followed by the simultaneous diagonal switch @xmath337 then @xmath338 and @xmath339 see figure [ fig8 ] .
similarly , we have @xmath340 @xmath341 @xmath342 and @xmath343 thus , any composition of an even number of simultaneous diagonal switches determines an element of @xmath332 for the converse statement , we define a cyclic order on the set @xmath344 of paris of opposite edges of @xmath100 as follows .
since each puncture @xmath62 of @xmath302 is adjacent to three edges @xmath188 @xmath191 and @xmath268 with @xmath345 @xmath346 and @xmath347 the orientation of @xmath302 induces a cyclic order on the set @xmath348 around @xmath349 inducing a cyclic order on the set @xmath350 it is easy to check that this cyclic order is independent of the choose of @xmath349 hence is well defined .
we call the _ sign _ of a tetrahedral triangulation @xmath100 positive if the cyclic order @xmath351 coincides with the one induced from the orientation , and negative if otherwise .
it easy to see that a simultaneous diagonal switch changes the sign of @xmath103 and an orientation preserving self - diffeomorphism of @xmath302 preserves the sign of @xmath164 since the dual farey diagram @xmath329 is a connected tree , for any self - diffeomorphism @xmath334 of @xmath330 up to redundancy there is a unique path of @xmath329 connecting the vertices @xmath100 and @xmath352 since @xmath100 and @xmath335 have the same sign , the path consists of an even number of edges , corresponding to an even number of simultaneous diagonal switche @xmath353 then @xmath354 where @xmath355
let @xmath1 be a type - preserving representation of @xmath356 and let @xmath111 be a decoration of @xmath114 suppose @xmath100 is a @xmath1-admissible tetrahedral triangulation of @xmath330 @xmath167 and @xmath163 respectively are the sets of edges and ideal triangles of @xmath103 and @xmath357 is the lengths coordinate of @xmath358\in\mathcal m^d_{\pm1}(\sigma_{0,4}).$ ] let @xmath303 be the punters of @xmath330 let @xmath305 be the ideal triangle of @xmath100 disjoint from @xmath306 and let @xmath307 be the edge of @xmath100 connecting the punctures @xmath306 and @xmath308 define the quantities @xmath359 @xmath360 and @xmath361 the quantities @xmath362 @xmath363 and @xmath364 will play a central role in the rest of this paper .
suppose @xmath365 then by ( [ euler ] ) , there is exactly one ideal triangle , say @xmath366 such that @xmath367 and @xmath368 for @xmath369 as a direct consequence of lemma [ 3.1 ] and theorem [ trace formula ] , we have the following lemmas .
[ 5.1 ] let @xmath370 be the simple closed curve going counterclockwise around the puncture @xmath306 once .
then up to conjugation , the @xmath1-image of the peripheral element @xmath371\in\pi_1(\sigma_{0,4})$ ] is @xmath372,\ ] ] and the @xmath1-image of the other peripheral elements @xmath373,$ ] @xmath374 $ ] and @xmath375 $ ] are respectively @xmath376 , \pm\left [ \begin{array}{cc } 1 & \lambda(x)+\lambda(z)-\lambda(y)\\ 0&1 \end{array } \right ] \text { and } \pm\left [ \begin{array}{cc } 1 & \lambda(x)+\lambda(y)-\lambda(z)\\ 0&1 \end{array } \right].\ ] ] [ 5.2 ] 1 .
the absolute values of the traces of the distinguished simple closed curves @xmath315 @xmath316 and @xmath317 of @xmath100 can be calculated by @xmath377)\big|&=\frac{\big|\lambda(y)^2+\lambda(z)^2-\lambda(x)^2\big|}{\lambda(y)\lambda(z)},\\ \big|tr\rho([y])\big|&=\frac{\big|\lambda(x)^2+\lambda(z)^2-\lambda(y)^2\big|}{\lambda(x)\lambda(z)}\ \text { and}\\ \big|tr\rho([z])\big|&=\frac{\big|\lambda(x)^2+\lambda(y)^2-\lambda(z)^2\big|}{\lambda(x)\lambda(y)}. \end{split}\ ] ] 2 .
the right hand sides of the equations in ( [ 5.1 ] ) are strictly greater than @xmath34 if and only if @xmath362 @xmath363 and @xmath364 satisfy one of the following inequalities @xmath378 note that reversing the directions of the inequalities in ( [ anti - tri ] ) , we get the triangular inequality .
the idea of the proof of ( 2 ) is that if we regard the quantities @xmath362 @xmath363 and @xmath364 as the edge lengths of a euclidean triangle , then the right hand sides of ( [ 5.1 ] ) are twice of the cosine of the corresponding inner angles .
the next lemma shows the rule of the change of the quantities @xmath362 @xmath363 and @xmath364 under a simultaneous diagonal switch .
[ lambda ] suppose @xmath187 is a tetrahedral triangulation of @xmath304 if @xmath187 is @xmath1-admissible , then let @xmath379 be the @xmath53-lengths of @xmath115 in @xmath380 and let @xmath381 @xmath382 and @xmath383 be the corresponding quantities . 1 .
if @xmath187 is obtained from @xmath100 by doing @xmath323 then @xmath187 is @xmath1-admissible if and only if @xmath384 in the case that @xmath187 is @xmath1-admissible , @xmath385 @xmath386 and @xmath387 2 .
if @xmath187 is obtained from @xmath100 by doing @xmath388 then @xmath187 is @xmath1-admissible if and only if @xmath389 in the case that @xmath187 is @xmath1-admissible , @xmath390 @xmath386 and @xmath391 3 .
if @xmath187 is obtained from @xmath100 by doing @xmath392 then @xmath187 is @xmath1-admissible if and only if @xmath393 in the case that @xmath187 is @xmath1-admissible , @xmath390 @xmath394 and @xmath395 for ( 1 ) , we have that the edge @xmath396 is adjacent to the ideal triangle @xmath397 and @xmath398 with @xmath399 and @xmath400 is adjacent to the ideal triangles @xmath189 and @xmath190 with @xmath401 let @xmath402 and @xmath403 respectively be the edges of @xmath187 obtained from diagonal switches at @xmath396 and @xmath404 i.e. , @xmath405 is the edge of @xmath187 connecting the punctures @xmath174 and @xmath143 and @xmath406 is the edge of @xmath187 connecting the punctures @xmath144 and @xmath407 by proposition [ change ] , @xmath187 is @xmath1-admissible if and only if @xmath408 i.e. , @xmath384 by proposition [ change ] again , if @xmath187 is @xmath1-admissible , then @xmath409 for @xmath410 or @xmath411 and @xmath412and @xmath413 therefore , @xmath385 @xmath386 and @xmath414 the proofs of ( 2 ) and ( 3 ) are the similar .
a consequence of lemma [ lambda ] is that the inequalities in ( [ anti - tri ] ) are persevered by the simultaneous diagonal switches .
[ preserve ] suppose @xmath187 is a @xmath1-admissible tetrahedral triangulation of @xmath302 obtained from @xmath100 by doing a simultaneous diagonal switch .
let @xmath379 be the @xmath53-lengths of @xmath115 in @xmath380 and let @xmath381 @xmath382 and @xmath383 be the corresponding quantities .
then @xmath415 , @xmath382 and @xmath383 satisfy one of the inequalities in ( [ anti - tri ] ) if and only if @xmath362 @xmath363 and @xmath364 do . without lost of generality , we assume that @xmath416 is obtained from @xmath100 by doing @xmath417 if @xmath418 then by lemma [ lambda ] , @xmath419 therefore , either @xmath420 or @xmath421 on the other hand , if either @xmath422 or @xmath423 i.e. , @xmath424 then by lemma [ lambda ] , @xmath425 another consequence of lemma [ lambda ] is the following [ 5.5 ] there are uncountably many @xmath426\in\mathcal m_{\pm1}(\sigma_{0,4})$ ] such that all the tetrahedral triangulations of @xmath302 are @xmath1-admissible .
suppose @xmath1 is a typer - preserving representation of @xmath356 with @xmath427 and @xmath111 is a decoration of @xmath114 let @xmath100 be a @xmath1-admissible tetrahedral triangulation of @xmath302 and let @xmath258 be the lengths coordinate of @xmath115 in @xmath164 recall that there is a ono - to - one correspondence between the tetrahedral triangulations of @xmath302 and the vertices of the dual farey diagram @xmath428 which is a countably infinity tree .
therefore , for each tetrahedral triangulation @xmath380 there is up to redundancy a unique path in @xmath329 connecting @xmath100 and @xmath380 which corresponds to a sequence @xmath429 of simultaneous diagonal switches .
let @xmath430 and for each @xmath431 let @xmath432 be the tetrahedral triangulation obtained from @xmath433 by doing @xmath434 suppose @xmath432 is @xmath1-admissible for some @xmath431 and suppose @xmath435 is the @xmath53-lengths of @xmath115 in @xmath436 then by lemma [ lambda ] , @xmath437 is @xmath1-admissible if and only if the laurent polynomial @xmath438 an induction in @xmath439 shows that @xmath440 is @xmath1-admissible if and only if certain laurent polynomial @xmath441 the set of zeros @xmath442 of @xmath443 is a zariski - closed subset of @xmath444 in particular , the lebesgue measure @xmath445 since @xmath329 is a countably infinity tree , there are in total countably many tetrahedral triangulations @xmath100 of @xmath330 and hence @xmath446 therefore , the set @xmath447 has a full measure in @xmath444 in particular , @xmath448 contains uncountable many points .
now each @xmath449 with @xmath450 @xmath451 and @xmath452 determines a type - preserving representation @xmath1 as follows .
take a tetrahedral triangulation @xmath100 of @xmath330 and let @xmath167 and @xmath163 respectively be the set of edges and ideal triangles of @xmath164 choose @xmath292 so that @xmath453 and define @xmath454 by @xmath455 @xmath456 and @xmath457 then @xmath458 @xmath459 and @xmath460 by theorem [ kashaev2 ] , proposition [ rational ] and lemma [ 5.1 ] , @xmath258 determines a decorated representation @xmath115 up to conjugation . in particular , by lemma [ 5.1 ] , @xmath1 is type - preserving . by ( [ euler ] ) , the relative euler class @xmath365 finally , since @xmath461 the laurent polynomial @xmath462 for all tetrahedral triangulation @xmath463 as a consequence , all the tetrahedral triangulations are @xmath1-admissible . by symmetry , there are also uncountably many type - preserving representations @xmath1 with @xmath464 such that all the tetrahedral triangulations of @xmath302 are @xmath1-admissible .
let @xmath100 be a tetrahedral triangulation of @xmath302 and let @xmath448 be the full measure subset of @xmath465 constructed in the proof of proposition [ 5.5 ] .
then each @xmath449 satisfying one of the following identities @xmath466 @xmath467 or @xmath468 determines a decorated representation @xmath115 with @xmath52 such that all the tetrahedral triangulations of @xmath302 are @xmath1-admissible . since elementary representation have relative euler class @xmath40 and @xmath469 @xmath1 is non - elementary . for each tetrahedral triangulation @xmath380 let @xmath379 be the @xmath53-lengths of @xmath115 in @xmath463 since @xmath416
can by obtained from @xmath100 by doing a sequence of simultaneous diagonal switches , by lemma [ preserve ] , the quantities @xmath381 @xmath382 and @xmath383 satisfy one of the inequalities in ( [ anti - tri ] ) . by lemma
[ 5.2 ] , the traces of the distinguished simple closed curves @xmath315 @xmath316 and @xmath317 in @xmath416 are strictly greater than @xmath34 in the absolute value .
since each simple closed curve @xmath29 is distinguished in some tetrahedral triangulation @xmath380 we have @xmath470)\big|>2.$ ] suppose @xmath471 then by ( [ euler ] ) , there are exactly two ideal triangles having the positive sign and two having the negative sign . without loss of generality , we assume that @xmath472 and @xmath473 note that under this assumption , the edges @xmath396 and @xmath400 in the pair @xmath309 are adjacent to ideal triangles having the same sign , and as will be seen later , the @xmath320-curves will play a different role than the @xmath316- and @xmath317- curves do . as a direct consequence of lemma [ 3.1 ]
, we have the following [ 57 ] let @xmath370 be the simple closed going counterclockwise around the puncture @xmath474 then up to conjugation , the @xmath1-image of the peripheral elements @xmath371 $ ] and @xmath373 $ ] of @xmath356 are @xmath376,\ ] ] and the @xmath1-image of the other two peripheral elements @xmath374 $ ] and @xmath375 $ ] are @xmath475.\ ] ] [ 5.8 ] 1 .
the absolute values of the traces of the distinguished simple closed curves @xmath315 @xmath316 and @xmath317 of @xmath100 can be calculated by @xmath476)\big|&=\frac{\big|\lambda(x)^2+\lambda(y)^2+\lambda(z)^2 - 2\lambda(x)\lambda(y)-2\lambda(x)\lambda(z)\big|}{\lambda(y)\lambda(z)},\\ \big|tr\rho([y])\big|&=\frac{\lambda(x)^2+\lambda(y)^2+\lambda(z)^2 + 2\lambda(y)\lambda(z)-2\lambda(x)\lambda(y)}{\lambda(x)\lambda(z)}\ \text { and}\\ \big|tr\rho([z])\big|&=\frac{\lambda(x)^2+\lambda(y)^2+\lambda(z)^2 + 2\lambda(y)\lambda(z)-2\lambda(x)\lambda(z)}{\lambda(x)\lambda(y)}. \end{split}\ ] ] 2 .
the right hand sides of the last two equations in ( [ 5.3 ] ) are always strictly greater than @xmath60 whereas the right hand side of the first equation is less than or equal to @xmath34 if and only if @xmath362 @xmath363 and @xmath364 satisfy the following inequalities @xmath477
\(1 ) is a direct consequence of theorem [ trace formula ] . for ( 2 ) , since @xmath1 is type - preserving , by theorem [ kashaev2 ] , proposition [ rational ] and lemma [ 57 ] , @xmath478 therefore , the right hand side of the second equation of ( [ 5.3 ] ) equals @xmath479 and the right hand side of the third equation equals @xmath480 in the case that @xmath481 the right hand side of the first equation of ( [ 5.3 ] ) equals @xmath482 the quantity also equals @xmath483 which is less than or equal to @xmath34 if and only if the equalities in ( [ tri ] ) are satisfied .
for the case that @xmath484 the proof is similar .
the next lemma shows the rule of the change of the quantities @xmath362 @xmath363 and @xmath364 under a simultaneous diagonal switch .
[ lambda2 ] suppose @xmath187 is a tetrahedral triangulation of @xmath304 if @xmath187 is @xmath1-admissible , then let @xmath379 be the @xmath53-lengths of@xmath115 in @xmath380 and let @xmath381 @xmath382 and @xmath383 be the corresponding quantities . 1 .
if @xmath187 is obtained from @xmath100 by doing @xmath323 then @xmath187 is @xmath1-admissible . in this case , @xmath385 @xmath386 and @xmath485 2 . if @xmath187 is obtained from @xmath100 by doing @xmath388 then @xmath187 is @xmath1-admissible if and only if @xmath389 in the case that @xmath187 is @xmath1-admissible , @xmath390 @xmath386 and @xmath486 3 .
if @xmath187 is obtained from @xmath100 by doing @xmath392 then @xmath187 is @xmath1-admissible if and only if @xmath389 in the case that @xmath100 is @xmath1-admissible , @xmath390 @xmath394 and @xmath487 this is a consequence of proposition [ change ] , and the proof is similar to that of lemma [ lambda ] .
let @xmath1 be a type - preserving representation of @xmath356 with relative euler class @xmath488 and choose arbitrarily a decoration @xmath111 of @xmath114 let @xmath100 be a tetrahedral triangulation of @xmath304 if @xmath100 is not @xmath1-admissible , then there is an edge @xmath127 of @xmath100 that is not @xmath1-admissible , and the element of @xmath356 represented by the distinguished simple closed curve in @xmath100 disjoint from @xmath127 is sent by @xmath1 to a parabolic element of @xmath31 if @xmath100 is @xmath1-admissible , then we let @xmath258 be the lengths coordinate of @xmath139 in @xmath164 if the quantities @xmath362 @xmath363 and @xmath364 satisfy the inequalities in ( [ tri ] ) , then by lemma [ 5.8 ] , the element of @xmath356 represented by one of the distinguished simple closed curves @xmath315 @xmath316 and @xmath317 is sent by @xmath1 to either an elliptic or a parabolic element of @xmath31 therefore , to prove the theorem , it suffices to find a tetrahedral triangulation @xmath187 of @xmath302 such that either @xmath187 is not @xmath1-admissible or @xmath187 is @xmath1-admissible with the quantities @xmath381 @xmath382 and @xmath383 satisfying the inequalities in ( [ tri ] ) .
our strategy of finding @xmath187 is to construct a sequence of tetrahedral triangulations @xmath489 with @xmath490 by the following _ trace reduction algorithm : _ let @xmath491 and suppose that @xmath492 is obtained
. if @xmath492 is not @xmath1-admissible , then we stop .
if @xmath492 is @xmath1-admissible , then we let @xmath493 be the lengths coordinate of @xmath115 in @xmath494 if @xmath495 @xmath496 and @xmath497 satisfy the inequalities in ( [ tri ] ) , then we stop . if otherwise , then there is a unique maximum among @xmath495 @xmath496 and @xmath498 since other wise the inequalities ( [ tri ] ) are satisfied .
suppose @xmath499 is the pair of opposite edges of @xmath492 such that @xmath500 equals the maximum of @xmath495 @xmath496 and @xmath501 then we let @xmath502 be the tetrahedral triangulation obtained from @xmath492 by doing a simultaneous diagonal switch at @xmath307 and @xmath503 by lemma [ finite ] below , the algorithm stops at some @xmath504 [ finite ] the trace reduction algorithm stops in finitely many steps .
for each @xmath505 let @xmath305 be the ideal triangle of @xmath492 disjoint from the puncture @xmath151 and let @xmath307 be the edge of @xmath492 connecting the punctures @xmath306 and @xmath308 without loss of generality , we assume in @xmath100 that @xmath472 and @xmath473 then by proposition [ change ] , @xmath506 and @xmath507 for each @xmath494 for each @xmath505 we let @xmath508 @xmath509 and @xmath510 and let @xmath511 then @xmath495 @xmath496 and @xmath497 satisfy the inequalities in ( [ tri ] ) if and only if @xmath512 assume that the sequence @xmath513 is infinite , i.e. , @xmath514 for all @xmath515 then we will find a contradiction by the following three steps . in step
i we show that @xmath516 is decreasing in @xmath46 by considering two mutually complementary cases , where in one of them ( case 1 ) the gap @xmath517 is bounded below by the minimum of @xmath518 and @xmath519 in step ii we show that there must be a infinite subsequence @xmath520 of @xmath513 such that each @xmath521 is of case 1 of step i , and in step iii we show that for @xmath439 large enough , @xmath522 is increasing .
the three steps together imply that @xmath523 for some @xmath46 large enough , which is a contradiction .
i : we show that @xmath516 is decreasing in @xmath524 there are the following two cases to verify
. case 1 : @xmath525 in this case , by lemma [ lambda2 ] , @xmath526 without loss of generality , we assume that @xmath527 then @xmath528 is the largest among @xmath529 @xmath528 and @xmath530 by a direct calculation and that @xmath531 we have @xmath532 moreover , since @xmath533 and @xmath534 we have @xmath535 and hence @xmath536 therefore , we have @xmath537 case 2 : one of @xmath538 and @xmath539 is strictly greater than @xmath540 in this case , we without loss of generality assume that @xmath541 then by lemma [ lambda2 ] , @xmath542 without loss of generality , we assume that @xmath543 then @xmath544 is the largest among @xmath529 @xmath528 and @xmath530 by a direct calculation and that @xmath545 we have @xmath546 from @xmath547 and @xmath548 we have @xmath549 @xmath550 and hence @xmath551 as a consequence , @xmath552 step ii : we show that there is an infinite subsequence @xmath520 of @xmath513 such that @xmath553 is in case 1 of step i. we use contradiction . for each @xmath554 in case 2 of step i , let @xmath555 and let @xmath556 then @xmath557 by lemma [ lambda2 ] , @xmath558 is in case 1 of step i if and only if @xmath559 now suppose that there is an @xmath560 such that @xmath561 is in case 2 of step i and @xmath562 for all @xmath563 then by lemma [ lambda2 ] , we have @xmath564 and @xmath565 for @xmath563 by induction , @xmath566 and @xmath567 for all @xmath568 which is impossible .
step iii : we show that for @xmath439 large enough , @xmath522 is increasing . in figure [ fig9 ]
below , we let @xmath569 and for each @xmath13 let @xmath570 be the intersection of @xmath156 with the set @xmath571 a direct calculation shows that @xmath570 s are parts of the concentric circles centered at @xmath572 with radii increasing in @xmath223 and that @xmath573 is the inscribed circle of @xmath574 in figure [ fig9 ] ( a ) , let @xmath575 be the intersection of @xmath156 and the set @xmath576 then @xmath575 is a quadratic curve in @xmath156 going through the points @xmath577 and @xmath578 let the line segment @xmath579 be the intersection of @xmath156 and the plane @xmath580 then by ( [ 5.5 ] ) , if @xmath554 is on @xmath575 with @xmath581 then @xmath558 is on @xmath582 denote by @xmath583 the line segment connecting @xmath584 and @xmath585 and by @xmath586 the line segment connecting @xmath577 and @xmath587 let @xmath588 be the region of @xmath156 bounded by @xmath589 @xmath583 and @xmath590 and let @xmath167 be the region in @xmath156 bounded by @xmath591 @xmath583 and @xmath592 in figure [ fig9 ] ( b ) , let @xmath593 be the intersection of @xmath575 and @xmath594 let @xmath301 be the third coordinate of @xmath595 let @xmath596 be the intersection of @xmath156 and the plane @xmath597 and let @xmath598 be the region in @xmath156 bounded by @xmath594 @xmath599 @xmath583 and @xmath592 note that @xmath598 is a subset of @xmath600 now consider the infinite subsequence @xmath520 guaranteed by step ii such that @xmath553 is in case 1 of step i. by ( [ est ] ) , there exists an @xmath601 such that @xmath602 for all @xmath603 since otherwise @xmath604 for @xmath439 large enough , and the algorithm stops . without loss of generality , we assume that @xmath605 and we have the following two claims . claim 1 : if @xmath606 and @xmath581 then @xmath607 @xmath608 and @xmath609 indeed , in this case , @xmath610 by ( [ 5.5 ] ) , @xmath611 and @xmath612 furthermore , we have @xmath613 since otherwise @xmath558 would be in the disk bounded by the circle @xmath614 i.e. , @xmath615 and the algorithm stops .
therefore , @xmath616 and @xmath617 claim 2 : for @xmath618 if @xmath619 and @xmath581 then @xmath620 @xmath608 and @xmath621 indeed , in this case , @xmath610 by ( [ 5.7 ] ) , @xmath558 is in the triangle above @xmath622 @xmath608 and @xmath612 therefore , @xmath616 and @xmath617 furthermore , since @xmath623 we have @xmath624 and by step i , we have @xmath625 as a consequence , @xmath626 since the intersection @xmath627 of the quadratic curve @xmath575 and the circle @xmath573 lies on the line determined by @xmath628 @xmath598 lies on the right half of @xmath629 and hence @xmath630 since @xmath631 and by assumption @xmath632 and @xmath633 we have @xmath634 by an induction and claims 1 and 2 , we have for all @xmath635 that @xmath636 with @xmath637 and @xmath638 with @xmath639 and hence for @xmath640 have @xmath641 similar to the relative euler class @xmath291 case , we have [ 5.11 ] there are uncountably many @xmath426\in\mathcal m_0(\sigma_{0,4})$ ] such that all the tetrahedral triangulations of @xmath302 are @xmath1-admissible . for each such @xmath102 there is a simple closed curve @xmath29 on @xmath302 such that @xmath30)$ ] is an elliptic element in @xmath31 since the functions in lemma [ lambda2 ] are rational in @xmath362 @xmath363 and @xmath642 the argument in the proof of proposition [ 5.5 ] applies here and proves the first part .
the second part is a result of the trace reduction algorithm .
we describe the connected component of the character space @xmath86 in this section .
recall that for a quadruple @xmath71 of positive and negative signs , @xmath643 is the space of conjugacy classes of type - preserving representations with relative euler class @xmath13 and signs of the punctures @xmath67 let @xmath644 be the set of punctures of @xmath304 then theorem [ main3 ] is equivalent to the following [ main3 ] 1 . for @xmath645 let @xmath646 be defined by @xmath647 and @xmath648 then @xmath649 2 . for @xmath650 let @xmath651 be defined by @xmath652 and @xmath653 for @xmath654 and let @xmath655 be defined by @xmath656 for all @xmath657 then @xmath658 3 . for @xmath650 let @xmath659 be defined by @xmath660 and @xmath661 for @xmath654 and let @xmath662 be defined by @xmath663 for all @xmath657 then @xmath664 4 .
all the spaces @xmath665 @xmath666 @xmath667 @xmath668 and @xmath669 are connected .
let @xmath100 be a tetrahedral triangulation of @xmath304 recall that @xmath670 is the space of conjugacy classes of type - preserving representations @xmath1 such that @xmath100 is @xmath1-admissible . by theorem [ kashaev1 ] , @xmath670 is a dense and open subset of @xmath671 let @xmath167 and @xmath163 respectively be the sets of edges and ideal triangles of @xmath103 let @xmath672 be the ideal triangle disjoint from the puncture @xmath306 and let @xmath673 be the edge connecting the punctures @xmath306 and @xmath308 for @xmath171 let @xmath359 @xmath360 and @xmath361 we first show that the quantities @xmath362 @xmath363 and @xmath364 parametrize the components of @xmath674 [ delta ] let @xmath169 be with the principle @xmath168-bundle structure given by @xmath675 and let @xmath465 be with the principle @xmath676-bundle structure defined by @xmath677 then the map @xmath678 sending @xmath679 to @xmath680 induces a diffeomorphism @xmath681 since @xmath682 @xmath683 is well defined , and since @xmath684 for all @xmath685 @xmath683 is surjective . for the injectivity , we suppose that @xmath686 let @xmath687 and let @xmath688 then @xmath689 therefore , @xmath683 is injective .
the differentiability of @xmath683 and @xmath690 follows from the definition of @xmath691 as a consequence of theorem [ kashaev2 ] , lemma [ 5.1 ] , lemma [ 57 ] and lemma [ delta ] , we have [ delta ] let @xmath100 be a tetrahedral triangulation of @xmath302 with the set of ideal triangles @xmath692 then @xmath693 where each @xmath694 is a subset of @xmath695 defined as follows . 1 . for @xmath650 let @xmath696 be given by @xmath697 and @xmath698 for @xmath654 and let @xmath699 be given by @xmath700 and @xmath701 for @xmath702 then @xmath703 2 . for @xmath704 let @xmath705 be given by @xmath706 and @xmath707 then @xmath708 suppose @xmath71 is a quadruple of positive and negative signs and @xmath709
let @xmath100 be a tetrahedral triangulation of @xmath302 . since @xmath670 is dense and open in @xmath710 @xmath711 if and only if @xmath712 for ( 1 ) , by lemma [ 57 ] , the only possibility for @xmath713 is that @xmath714 for some @xmath715 for ( 2 ) , by lemma [ 5.1 ] , the only possibility for @xmath716 is that either @xmath717 in which case @xmath362 @xmath363 and @xmath364 satisfy the triangular inequality , or @xmath718 for some @xmath650 in which case @xmath362 @xmath363 and @xmath364 satisfy one of the inequalities in ( [ anti - tri ] ) .
the proof of ( 3 ) is parallel to that of ( 2 ) . for ( 4 ) , by symmetry , it suffices to prove the connectedness of @xmath719 @xmath720 and @xmath721 we consider the following subsets of @xmath574 let @xmath722 ( resp . @xmath723 and @xmath724 ) be the set of points @xmath725 such that @xmath726 ( resp . @xmath467 and @xmath468 ) ,
let @xmath727 ( resp . @xmath728 and @xmath729 ) be the set of points @xmath725 such that @xmath730 ( resp . @xmath731 and @xmath732 ) and let @xmath733 see figure [ fig10 ] . by theorem [ kashaev2 ] , lemma [ 5.1 ] , lemma [ 57 ] and corollary [ delta ] , we have via the lengths coordinate that 1 .
@xmath734 is diffeomorphic to a dense and open subset of @xmath719 2 .
@xmath735 is diffeomorphic to a dense and open subset of @xmath736 and 3 .
@xmath737 is diffeomorphic to a dense and open subset of @xmath738 in the rest of the proof , we let @xmath187 be the tetrahedral triangulation of @xmath302 obtained from @xmath100 by doing a simultaneous diagonal switch @xmath392 let @xmath739 be the set of ideal triangles of @xmath380 and let @xmath740 and @xmath741 be sign assignments defined in the same way as @xmath742 and @xmath743 for @xmath685 we let @xmath744\doteq(\frac{a}{a+b+c},\frac{b}{a+b+c},\frac{c}{a+b+c})\in\delta.$ ] for the connectedness of @xmath719 since both @xmath745 and @xmath746 are connected , it suffices to choose two points @xmath593 and @xmath747 respectively in @xmath745 and @xmath746 and find a path in @xmath748 connecting @xmath593 and @xmath749 now let @xmath750 and let @xmath751 with @xmath752 by proposition [ change ] and lemma [ lambda2 ] , @xmath593 corresponds to the point @xmath753\in\delta_x(\mathcal t',\epsilon'_{12})$ ] and @xmath747 corresponds to the point @xmath754\in\delta_x(\mathcal t',\epsilon'_{12}).$ ] since @xmath755 is connected , there is a path in @xmath755 connecting @xmath756 and @xmath757 giving a path in @xmath748 connecting @xmath593 and @xmath749 for the connectedness of @xmath736 we let @xmath758 and let @xmath759 by proposition [ change ] , lemma [ lambda ] and lemma [ preserve ] , @xmath593 corresponds to the point @xmath760\in\delta_z(\mathcal t',\epsilon'_{4})$ ] and @xmath747 corresponds to the point @xmath761\in\delta_z(\mathcal t',\epsilon'_{4}).$ ] since @xmath762 is connected , there is a path in @xmath762 connecting @xmath756 and @xmath757 giving a path in @xmath720 connecting @xmath593 and @xmath749 similarly , any pair of points @xmath763 and @xmath764 and any pair of points @xmath765 and @xmath764 can respectively be connected by paths in @xmath766 therefore , @xmath767 is connected . finally , for the connectedness of @xmath768 we let @xmath769 with @xmath770 and let @xmath771 with @xmath772 by proposition [ change ] , lemma [ lambda ] and lemma [ preserve ]
, @xmath593 corresponds to the point @xmath760\in\delta^c(\mathcal t',\epsilon'_{4})$ ] and @xmath747 corresponds to the point @xmath761\in\delta^c(\mathcal t',\epsilon'_{4}).$ ] since @xmath773 is connected , there is a path in @xmath773 connecting @xmath756 and @xmath757 giving a path in @xmath774 connecting @xmath593 and @xmath749 similarly , all the other pieces can be connected by paths in @xmath768 and @xmath775 is connected .
the goal of this section is to prove the ergodicity of the @xmath85-action on the non - external connected components of @xmath671 to use the techniques we used in the previous sections , we need to understand the measure on @xmath86 induced by the goldman symplectic @xmath34-form in terms the quantities @xmath362 @xmath363 and @xmath776 let @xmath100 be a tetrahedral triangulation of @xmath330 and let @xmath163 be the set of ideal triangles of @xmath164 for each @xmath777 let @xmath694 be the subset of @xmath465 defined in corollary [ delta ] . then by equation ( [ wp ] ) , lemma [ delta ] , corollary [ delta ] and a direct calculation , we have the following [ weilp ] for each @xmath777 the @xmath34-form @xmath778 on @xmath694 corresponds to the goldman symplectic @xmath34-form @xmath181 on @xmath710 and the measure induced by @xmath91 is in the same measure class of the lebesgue measure on @xmath779 as a consequence , we have [ dense ] for @xmath780 the set @xmath781 consisting of conjugacy classes of type - preserving representations @xmath1 with the relative euler class @xmath287 such that all the tetrahedra triangulation of @xmath302 are @xmath1-admissible is a full measure subset of @xmath782 and is invariant under the @xmath85-action . by the proof of proposition [ 5.5 ] , @xmath783 is a countable union of lebesgue measure zero subsets , hence is of lebesgue measure zero . then by proposition [ weilp ] , @xmath783 is a null set in the measure induced by the goldman symplectic @xmath34-form . by the similar argument ,
@xmath784 and @xmath785 are respectively full measure subsets of @xmath786 and @xmath787 by proposition [ evenodd ] , since all the simultaneous diagonal switches act on each @xmath788 so does @xmath332 since @xmath789 is dense in @xmath790 and a representation in @xmath791 sends each simple closed curve to a hyperbolic element , by continuity , every representation in @xmath792 sends each simple closed curve to either a hyperbolic or a parabolic element . in @xcite ,
delgado explicitly constructed a family @xmath793 of representations in @xmath794 that send every simple closed curve to either a hyperbolic or a parabolic element , and for each @xmath795 , at least one simple closed curve is sent to a parabolic element .
therefore , the representations @xmath793 are in the measure zero subset @xmath796 let @xmath644 be the set of punctures of @xmath304 for @xmath797 and @xmath798 let @xmath799 by theorem [ main3 ] and proposition [ dense ] , theorem [ main4 ] follows from the following [ e ] 1 .
the @xmath85-action on @xmath800 is ergodic for each @xmath801 2 .
the @xmath85-action on @xmath802 and @xmath803 is ergodic .
3 . the @xmath85-action on @xmath804 and @xmath805 is ergodic for each @xmath806 the ergodicity of the @xmath85-action on the components of @xmath78 was first known to maloni - palesi - tan in @xcite using the markoff triple technique . by symmetry
, it suffices to prove the ergodicity of the @xmath85-action on @xmath807 let @xmath808 and let @xmath809 by theorem [ kashaev2 ] , lemma [ delta ] and corollary [ delta ] , given a tetrahedral triangulation of @xmath330 @xmath810 is diffeomorphic to a dense and open subset of @xmath719 where the diffeomorphism is given by the quantities @xmath362 @xmath363 and @xmath776 consider the embedding of @xmath811 defined by @xmath812 then @xmath813 let @xmath814 be the subset of @xmath815 consisting of the elements coming from @xmath807 as an immediate consequences of lemma [ lambda2 ] , the simultaneous diagonal switches @xmath324 and @xmath325 act on @xmath814 by @xmath816 and @xmath817 [ mm ] let @xmath818 be the cyclic subgroup of @xmath85 generated by the dehn twist @xmath819 along the distinguished simple closed curve @xmath820 then for every @xmath821 the ellipse @xmath822 is invariant under the action of @xmath823 and for almost every @xmath821 the action of @xmath818 on @xmath824 is ergodic .
a direct calculation shows that @xmath824 is invariant under the actions of @xmath324 and @xmath325 for all @xmath825 recall that @xmath826 therefore , the ellipse @xmath824 is invariant under the @xmath818-action . by lemma [ lambda2 ] ,
the action of @xmath324 and @xmath325 respectively move a point @xmath593 on @xmath824 vertically and horizontally . as show in figure [ fig11 ]
, the an affine transformation of @xmath827 sending the ellipse @xmath824 to a circle @xmath570 sends the vertical and the horizontal lines in @xmath827 respectively to two family of parallel lines in @xmath828 as a consequence , for each point @xmath593 on @xmath829 the angle @xmath830 is a constant @xmath831 depending only on @xmath223 and the center angle @xmath832 therefore , @xmath819 acts on @xmath570 by a rotation of angle @xmath833 since @xmath834 is an irrational multiple of @xmath835 for almost every @xmath821 the action of @xmath836 is ergodic .
[ mm2 ] let @xmath837 be the self - diffeomorphism of @xmath302 given by the dehn twist @xmath838 along the distinguished curve @xmath317 in @xmath100 followed by the dehn twist @xmath839 along the distinguished curve @xmath316 in @xmath840 and let @xmath841 be the cyclic subgroup of @xmath85 generated by @xmath842 then for every @xmath843 the quartic curve @xmath844 is invariant under the action of @xmath845 and for almost every @xmath821 the action of @xmath841 on @xmath846 is ergodic .
a direct calculation shows that @xmath846 is the @xmath326-image of @xmath847 since @xmath848 the map @xmath849 is @xmath850-equivariant , where @xmath851 acts on @xmath824 by @xmath819 and acts on @xmath846 by @xmath842 by lemma [ mm ] , @xmath841 acts on @xmath846 for every @xmath821 and the action is ergodic for almost every @xmath825 we show that every @xmath85-invariant measurable function @xmath852 is almost everywhere a constant .
consider the following region @xmath853 in @xmath814 inclosed by the parabola @xmath854 and the line segment @xmath855 we claim that each point @xmath856 in @xmath857 is an intersection an ellipse @xmath858 and a quantic curve @xmath859 for some @xmath860 @xmath861 indeed , we can let @xmath862 and let @xmath863 since @xmath864 @xmath865 and since @xmath866 @xmath861 a direct calculation shows that the intersection of @xmath858 and @xmath859 is transverse at @xmath595 i.e. , the gradients @xmath867 and @xmath868 span the tangent space of @xmath814 at @xmath869 then by lemma [ mm ] and lemma [ mm2 ] , the restriction of @xmath598 to @xmath857 is almost everywhere a constant . for @xmath870 let @xmath871 be the @xmath85-orbit of @xmath869 to show that the @xmath598 is almost everywhere a constant in @xmath872 it suffices to show that @xmath873 for almost every @xmath593 in @xmath874 let @xmath875 be the region in @xmath814 enclosed by parabola @xmath591 i.e.
, @xmath876 then @xmath875 is foliated by the ellipses @xmath877 we note that the parabola @xmath579 is the @xmath439-image of the inscribe circle @xmath573 of @xmath574 then by the trace reduction algorithm , lemma [ finite ] and proposition [ dense ] , for almost every @xmath593 in @xmath872 there is a composition @xmath334 of finitely many , say @xmath878 simultaneous diagonal switches such that @xmath879 by proposition [ evenodd ] , if @xmath880 is even , then @xmath881 and @xmath882 and if @xmath880 is odd , then @xmath883 since @xmath324 keeps invariant an ellipse @xmath884 passing through @xmath885 @xmath886 and hence @xmath887 finally , by lemma [ mm ] , for almost every @xmath593 in @xmath888 there is @xmath46 such that @xmath889 by symmetry , it suffices to prove the ergodicity of the @xmath85-action on @xmath890 the strategy is to find two transversely intersecting families of curves @xmath891 and @xmath892 foliating @xmath775 such that the @xmath818 -action on almost every @xmath893 and the @xmath894-action on almost every @xmath895 is ergodic .
let @xmath100 be an tetrahedral triangulation of @xmath330 and let @xmath163 be the set of ideal triangles of @xmath164 let @xmath808 and for @xmath896 let @xmath897 by theorem [ kashaev2 ] , lemma [ 5.1 ] and corollary [ delta ] , @xmath737 is diffeomorphic to a dense and open subset of @xmath768 where the diffeomorphism is given by the quantities @xmath362 @xmath363 and @xmath776 we define the embedding @xmath898 by @xmath899 for @xmath900 we let @xmath901 be the subset of @xmath902 consisting of the elements coming from @xmath903 and let @xmath904 ( see figure [ fig12 ] . )
[ mm3 ] for every @xmath821 the ellipse @xmath905 is invariant under the action of @xmath823 and for almost every @xmath821 the action of @xmath818 on @xmath893 is ergodic . for @xmath906 let @xmath907 by lemma [ lambda ] , the action of @xmath324 and @xmath325 on @xmath814 satisfies @xmath908 and @xmath909 therefore , we have @xmath910 we claim that @xmath911 if ( [ mm4 ] ) is true , then a direct calculation shows that @xmath912 is on @xmath913 to verify ( [ mm4 ] ) , we let @xmath187 be the tetrahedral triangulation obtained from @xmath100 by doing @xmath388 and let @xmath914 be the tetrahedral triangulation obtained from @xmath187 by doing @xmath915 let @xmath916 and @xmath917 respectively be the signs of @xmath1 assigned to the ideal triangles of @xmath187 and @xmath918 in @xmath103 @xmath187 and @xmath919 we denote uniformly by @xmath305 the ideal triangle disjoint from the puncture @xmath474 if @xmath920 i.e. @xmath921 then we consider the following cases . case 1 : @xmath922 and @xmath923 in this case , since @xmath924 and @xmath925 we have by proposition [ change ] that @xmath926 since @xmath927 by proposition [ change ] again , @xmath928 therefore , @xmath929 and ( [ mm4 ] ) follows .
case 2 : @xmath922 and @xmath930 in this case , by proposition [ change ] , @xmath931 and @xmath932 therefore , @xmath933 and and ( [ mm4 ] ) follows .
case 3 : @xmath934 and @xmath923 in this case , by proposition [ change ] , @xmath935 and @xmath936 therefore , @xmath937 and and ( [ mm4 ] ) follows .
case 4 : @xmath934 and @xmath930 in this case , by proposition [ change ] , @xmath935 and @xmath938 therefore , @xmath939 and and ( [ mm4 ] ) follows .
the verification of ( [ mm4 ] ) for @xmath593 in @xmath940 @xmath941 and @xmath942 is similar , and is left to the readers . by ( [ mm4 ] ) ,
the action of @xmath819 on @xmath893 is a horizontal translation followed by a vertical translation .
see figure [ fig12 ] . by doing a suitable affine transform ,
the @xmath818-action is a rotation of an angle @xmath834 on a circle , where @xmath834 is an irrational multiple of @xmath835 for almost every @xmath57 therefore , for almost every @xmath821 the @xmath818-action on @xmath893 is ergodic .
consider the embedding @xmath943 by @xmath944 for @xmath900 we let @xmath945 be the subset of @xmath946 consisting of the elements coming from @xmath947 and let @xmath948 then we have the following lemma whose proof is similar to that of lemma [ mm3 ] . [ mm5 ] for every @xmath821 the ellipse @xmath949 is invariant under the action of @xmath950 and for almost every @xmath821 the action of @xmath894 on @xmath895 is ergodic .
[ proof of theorem [ e ] ( 2 ) ] a direct calculation shows that the two family of curves @xmath951 and @xmath952 transversely intersect
. then by lemma [ mm3 ] and lemma [ mm5 ] , the @xmath85-action on @xmath775 is ergodic . by symmetry
, it suffices to prove the ergodicity of the @xmath85-action on @xmath953 we let @xmath954 @xmath955 and @xmath956 by theorem [ kashaev2 ] , lemma [ delta ] and corollary [ delta ] , given a tetrahedral triangulation of @xmath330 @xmath957 is diffeomorphic to a dense and open subset of @xmath958 where the diffeomorphism is given by the quantities @xmath362 @xmath363 and @xmath776 let @xmath959 and consider the two - fold covering map @xmath960 defined by @xmath961,\ ] ] where @xmath744\doteq(\frac{a}{a+b+c},\frac{b}{a+b+c},\frac{c}{a+b+c}).$ ] let @xmath962 be the subset of @xmath957 consisting of the elements coming from @xmath963 and let @xmath964 then by proposition [ dense ] , @xmath962 is invariant under the @xmath85-action . recall that @xmath85 is isomorphic to a free group @xmath965 of rank two generated by the dehn twists @xmath819 and @xmath966 ( see e.g. @xcite . ) it is well known that @xmath965 is isomorphic to the quotient group @xmath967 where @xmath968 is the mod-@xmath34 congruence subgroup of @xmath969 and the matrices @xmath970 $ ] and @xmath971 $ ] correspond to the generators of @xmath972 this induces a group homomorphism @xmath973 defined by @xmath974\big)= d_x \quad \text{and}\quad \pi\big(\left[\begin{array}{cc } 1 & 2\\ 0 & 1\\ \end{array}\right]\big)= d_y.\ ] ] by lemma [ lambda ] and a direct calculation , @xmath975 is @xmath976-equivariant , i.e. @xmath977 for all @xmath978 where the @xmath979-action on @xmath980 is the standard linear action . by moore s
ergodicity theorem@xcite , the @xmath979-action on @xmath981 hence on @xmath982 is ergodic .
therefore , the @xmath85-action on @xmath962 is ergodic .
proposition [ mf ] below is a consequence of goldman ( @xcite , theorem d ) and is stated without proof in @xcite .
the purpose of this appendix is to include a proof of it for the readers interest , where the argument is from a discussion with f. palesi and m. wolff .
[ mf ] a type - preserving representation @xmath49 is extremal , i.e. , @xmath983 if and only if @xmath1 is fuchsian . by goldman@xcite , theorem d ,
a representation @xmath1 is maximum if and only if @xmath1 is fuchsian and the quotient @xmath984 is homeomorphic to @xmath48 therefore , to prove the proposition , it suffices to rule out the possibility that @xmath1 is non - maximum , fuchsian and @xmath985 which we will do using a contradiction .
now since @xmath986 there is an isomorphism @xmath987 and since @xmath1 is type - preserving , @xmath988 contains @xmath46 parabolic elements from the primitive peripheral elements of @xmath99 on the other hand , since the only possible parabolic elements of a fuchsian subgroup of @xmath5 are from the peripheral elements , the composition @xmath989 must send the primitive peripheral elements of @xmath47 to the primitive peripheral elements of @xmath990 this is impossible when @xmath991 since @xmath992 has only @xmath993 primitive peripheral elements . for the case
@xmath994 we recall the fact that in the first homology @xmath995 of a punctured surface @xmath996 the full set of vectors represented by the primitive peripheral elements of @xmath997 are linearly dependent , but the vectors in any proper subset of it are linearly independent . therefore , the induced isomorphism @xmath998 sends a set of linearly dependent vectors represented by the primitive peripheral elements of @xmath47 to a set of linearly indecent vectors , which is a contradiction .
in this appendix , we show that the results concerning representations of relative euler class @xmath2 in this paper can also be seen , and more straightforwardly , as consequences of some previous results of goldman ( @xcite , chapter 4 ) , where the argument presented here is due to the anonymous referee . in @xcite , goldman considers @xmath999-representations of the one - puncture torus group @xmath1000 which is the free group of two generators @xmath1001 the character space @xmath1002 of reducible representations @xmath1003 satisfy @xmath1004)=2,$ ] and hence could be described by the equation @xmath1005 where @xmath1006 and @xmath1007 on the other hand , the fundamental group of the four puncture sphere @xmath1008 where the generators are the four primitive peripheral elements corresponding to the four punctures .
if @xmath51 is type - preserving , then @xmath1009 and if @xmath1010 then one can lift @xmath1 to a representation @xmath1011 such that @xmath1012 hence the character spaces @xmath1013 can be described by the equation @xmath1014 where @xmath1015 and @xmath1016 ( see e.g. @xcite for more details . ) comparing ( [ 11 ] ) and ( [ 04 ] ) , it is clear that @xmath1017 moreover , the mapping class group actions are commensurable and the variables @xmath1018 correspond in each case to the traces of simple closed curves on the surface , hence all the results known for @xmath1002 can be translated to the results on @xmath1019 to be more precise , by ( @xcite , chapter 4 ) , @xmath1002 has five connected components , one of which is compact corresponding to @xmath1020 and four of which are non - compact corresponding to @xmath1021 a full measure subset of the characters in the non - compact components have all coordinates @xmath1018 strictly greater than @xmath34 in absolute value .
each coordinate correspond to the trace of the image of a simple closed curve . starting from a representation in one of these components and using the transitivity of the mapping class group action on the set of simple closed curves
, one gets that every simple closed curve is sent to an hyperbolic element .
therefore , a full measure subset of representations in the non - compact components are counterexamples to bowditch s question .
finally , the ergodicity of the @xmath1022-action on the non - compact components is already proved in ( @xcite , chapter 4 ) , implying the @xmath85-action on @xmath1021 , _ mapping class group dynamics on surface group representations _ , problems on mapping class groups and related topics , 189214 , proc .
pure math .
, 74 , amer .
soc . , providence , ri , 2006 . goldman , w. m. and xia , e , _ ergodicity of mapping class group actions on su(2)-character varieties _ , geometry , rigidity , and group actions , 591608 , chicago lectures in math . , univ .
chicago press , chicago , il , 2011 . | we give counterexamples to a question of bowditch that if a non - elementary type - preserving representation @xmath0 of a punctured surface group sends every non - peripheral simple closed curve to a hyperbolic element , then must @xmath1 be fuchsian .
the counterexamples come from relative euler class @xmath2 representations of the four - punctured sphere group .
we also show that the mapping class group action on each non - extremal component of the character space of type - preserving representations of the four - punctured sphere group is ergodic , which confirms a conjecture of goldman for this case .
the main tool we use is kashaev - penner s lengths coordinates of the decorated character spaces . |
the presence of rashba - type @xcite spin - orbit interaction @xcite ( soi ) in semiconducting materials can lead to spin - dependent quantum mechanical phenomena that are of both fundamental and practical interest . in rings and loop geometries
fundamental interference effects can be observed , and they can have important applications as well , mainly due to the fact that the strength of the soi has been proven to be experimentally tuneable @xcite .
the relevance of the physical system motivated extensive studiesbia84,ahiye12,nmt99,fr04,mpv04,fmbp05,szp05,sn05,kmga05,bo07,chr07,bo08,cpc08,kfbp08c , m10 , dmk11,fe12,nmmg11,neea13 of quantum rings or systems of them .
since the strength of the soi that determines the spin sensitive behavior of the devices can be controlled by external gate voltages , the question what effects appear if these voltages are time dependent arise naturally .
this point is to be investigated in the current paper .
transport phenomena in the presence of an alternating field can be strongly inelastic , like in the case of photon - assisted tunneling @xcite .
barriers with oscillating height were investigated in a series of papers , mainly in the context of traversal time and photon assisted transport , see e.g. refs .
theory @xcite was proven to be an efficient tool for the investigation of time - dependent transport in various materials @xcite .
transport related problems with oscillating soi have been studied in refs .
@xcite for a ring , and in ref .
@xcite for a ring - dot system .
additionally , ref .
@xcite demonstrated gate - driven electric dipole spin resonances in a quantum wire subjected to both static and oscillating potentials .
these works mainly focused on spin currents and time averaged transport properties ; in the following we take a different point of view by presenting time - resolved results .
as we shall see , the oscillation of the spin - orbit interaction strength leads to effects that are qualitatively different from the case of constant soi .
most remarkably , we point out the existence of time instants when the simple geometries we consider produce spin - polarized output from a completely unpolarized input .
this finding is in strong contrast with earlier results that considered static soi : as it can be proved ( using strong , symmetry based arguments ) , a two - terminal device utilizing any type of ( static ) soi can never produce spinpolarized output @xcite . in the following
we show that the physical reason for the temporal spin - polarization we report is the spatial interference of various `` floquet channels '' @xcite , i.e. , quantum mechanical plane waves with different wavenumbers . in the current paper
first we analytically determine the floquet spectrum of electrons in quantum wires with oscillating soi , and use this result to develop a numerical method for the related time - dependent transport problem . in sec .
[ resultsec ] , we present time - resolved results showing that the oscillating soi generates spin polarization waves .
the building blocks of the devices we consider are straight , narrow quantum wires ( see fig . [ latticefig ] ) .
the relevant time - dependent hamiltonian can be written @xcite as @xmath0 , \label{ham}\ ] ] where the unit vector @xmath1 points to the chosen positive direction along the wire , and we introduced the characteristic kinetic energy @xmath2 ( with @xmath3 being the relevant length scale , see fig .
[ latticefig ] ) . the length variable ( in units of @xmath3 ) along the wire
is denoted by @xmath4 @xmath5 depends on time via the strength of the soi @xmath6 , where it is the rashba parameter @xmath7 that can be tuned by external electrodes @xcite . in the following
we assume : @xmath8 note that @xmath5 is linear in @xmath9 [ the compact form given by eq .
( [ ham ] ) does not show it explicitly , but the quadratic terms cancel each other ] . introducing dimensionless units
, the time - dependent schrdinger equation reads @xmath10 where @xmath11 and @xmath12 .
the time - dependent part of the soi can be written as @xmath13 where @xmath14 note that for @xmath15 nm , @xmath16 ( gaas ) , @xmath17 is in the experimentally achievable range , and @xmath18 is of the order of @xmath19 hz . for larger samples
according to the scaling discussed above the characteristic frequencies are lower , and therefore the experimentally achievable maximal soi strength corresponds to lower values of @xmath20 since the hamiltonian appearing in eq .
( [ tdse ] ) is periodic in time , @xmath21 with @xmath22 floquet theory @xcite can be applied . using a plane wave basis , the spinors @xcite @xmath23
satisfy @xmath24\vert\varphi^{\pm}\rangle(s),\ ] ] where @xmath25 is the azimuthal angle corresponding to @xmath1 and the wave number @xmath26 is measured in units of @xmath27 since @xmath28=0,$ ] the eigenspinors above can be used to construct solutions to the time - dependent equation ( [ tdse ] ) : @xmath29\vert\varphi^{\pm}\rangle(s ) .
\label{eig}\ ] ] using the explicit form of the oscillating soi , and performing the integral in the exponent , we obtain that the two nonequivalent floquet quasienergies for a fixed wavenumber @xmath26 are given by @xmath30 and the `` time - dependent basis spinors '' read @xmath31 \vert\varphi^{\pm}\rangle(s ) .
\label{tdbase}\ ] ] the frequencies ( dimensionless energies ) that appear in the exponent for a given @xmath26 can be seen most directly by applying an appropriate jacobi - anger identity @xcite , leading to @xmath32 where bessel functions of the first kind @xcite appear in the expansion .
we assume no soi in the leads [ that are denoted by the horizontal ( @xmath33 direction ) black arrows in fig .
[ latticefig ] ] , and consider a monoenergetic input @xmath34 where @xmath35 can be an arbitrary spinor .
in order to obtain a time - dependent solution in the whole domain , the spinor valued wave functions have to be joined at the junctions .
we require the spinor components to be continuous .
additional boundary conditions can be obtained by using the relevant continuity equation @xmath36 where the ( unnormalized ) electron density is given by @xmath37 while the corresponding current density @xcite reads @xmath38 @xmath39 above denotes a solution to the the time - dependent schrdinger equation ( [ tdse ] ) and the inner product and the expectation value appearing in eqs . ( [ density ] ) and ( [ currentdens ] )
are understood in the spinor sense , that is , no spatial integration is involved . as usual , the physical meaning of the continuity equation ( [ continuity ] ) is seen most directly by integrating it over a finite spatial domain : the ( unnormalized ) probability of finding an electron inside the domain changes as a consequence of the currents flowing in / out at the boundaries . focusing on a certain junction , we obtain that the current that enters the junction , should also leave it _ always _
, i.e. , at any time instants .
as we shall see , the boundary conditions @xcite described above can be handled conveniently in the frequency domain .
let @xmath40 denote the ( dimensionless ) frequency of the input .
a spinor valued wave with this frequency partially enters the domain of oscillating soi ( and partially gets reflected ) . according to the previous subsection , whenever a frequency component @xmath41 appears in the solution of the time - dependent schrdinger equation ( [ tdse ] ) , an infinite number of additional `` floquet channels '' @xcite corresponding to frequencies @xmath42 open for transmission ( with @xmath43 being integer ) .
therefore the frequency components given by eq .
( [ freqs ] ) should be taken into account .
however , these frequencies are also sufficient for the complete description of the problem : since the frequency resolved fitting equations are linear and generally nondegenerate , they provide a nonzero result only for nonzero input , i.e. , for the set of frequencies given by eq .
( [ freqs ] ) . according to eq .
( [ tdbase2 ] ) , inside a domain with oscillating soi , the relevant frequencies are members of the set ( [ freqs ] ) , if one of the floquet quasi - energies ( [ quasie ] ) is equal to @xmath44 with an arbitrary integer @xmath43 .
the solutions of the equations @xmath45 are @xmath46 and @xmath47 where the subscripts correspond to the @xmath48 signs in the equations . using these wave vectors , a general solution that contains all the frequencies relevant for the description of the problem with the monoenergetic input ( [ input ] )
can be written as @xmath49 where @xmath50 are unknown coefficients , the space and time dependence of the spinors have been omitted , and @xmath51 denote @xmath52 evaluated at @xmath45 and @xmath53 that is , @xmath54 the relevant spinor valued wave function in the input lead is the sum of @xmath55 and the reflected part : @xmath56 where @xmath57 while @xmath58 and @xmath59 can be arbitrary , provided they are orthogonal in the spinor sense . note that for negative @xmath44 @xmath60 becomes imaginary ; by choosing @xmath61 we can ensure that the corresponding waves decay exponentially towards @xmath62 these `` evanescent '' solutions can play an important role in our description based on floquet states .
( the wave numbers @xmath63 given by eqs .
( [ kap ] ) and ( [ kam ] ) can also be purely imaginary , but in such cases both signs of the square root are allowed , since they contribute to the wave function in a finite domain . ) the transmitted solution in the output lead is analogous to @xmath64 only the signs of the wave numbers are opposite due to the different propagation directions : @xmath65 eqs .
( [ instate ] ) , ( [ reflected ] ) and ( [ transmitted ] ) describe the spinor valued solutions of the time - dependent schrdinger equation in all spatial domains . in order to take boundary conditions into account , first one has to evaluate these solutions and their spatial derivatives at the junctions ( including the internal ones that are not connected to the input / output leads ) .
as one can see easily , griffith s boundary conditions @xcite mean a system of linear equations for the coefficients appearing in eqs .
( [ instate ] ) , ( [ reflected ] ) and ( [ transmitted ] ) .
although in principle we have an infinite number of equations , since the bessel functions @xmath66 for a given argument decrease as a function of their index @xcite , correct numerical solutions can be obtained by limiting ourselves to a finite number of frequencies .
the convergence of the jacobi - anger expansion as well as the obtained wave functions were carefully checked when calculating the results to be presented in the following .
the simple straight geometry shown by fig . [ latticefig](a ) already shows important consequences of the oscillating soi .
additionally , it can be used to determine the parameter ranges to focus on .
although the term `` traverse time '' is difficult to interpret when the input is an infinite wave , the ratio of the length @xmath3 and @xmath67 with a characteristic wave vector @xmath26 can tell us which soi oscillation frequency domain is quasistatic .
accordingly , when @xmath68 ( in dimensionless units ) , soi oscillation related effects are expected to be weak .
furthermore , by inspecting eqs .
( [ kap]-[instate ] ) , one can see that the oscillating part of the soi alone does not induce spin precession . however , when @xmath69 the wave numbers that belong to different eigenspinor directions are not the same , thus the spin direction related to superpositions have a nontrivial spatial dependence .
@xmath70 and @xmath71 the top panel corresponds to @xmath72 while @xmath73 for panel b ) . the thin black line that connects the arrowheads is plotted in order to guide the eyes.,width=226 ] fig .
[ linespinfig ] demonstrates this effect . in order to focus on the spin direction alone ,
the solution @xmath39 ( that stands for @xmath74 , @xmath75 or @xmath76 , depending on the position ) has been divided by the space and time dependent electron density given by eq .
( [ density ] ) ( which happens to be nonzero in this case . ) the change of the spin direction along the wire is represented by plotting @xmath77 where the usual pauli matrices @xmath78 appear .
more precisely , the arrows shown in fig . [ linespinfig ] point from @xmath79 to @xmath80 i.e. , they visualize the spin direction in a local coordinate system .
the input spinor valued wave function is polarized in the positive @xmath81 direction in fig .
[ linespinfig ] .
note that considering any of the eigenspinors ( [ eigen ] ) as input , the spin direction does not change . since any difference of the relevant frequencies ( [ freqs ] ) is an integer multiple of @xmath82
the time evolution is periodic ( @xmath83 ) .
[ linespinfig ] corresponds to two different time instants , @xmath84 ( panel a ) and @xmath73 ( panel b ) .
as we can see , the spin direction has a strong spatial dependence in the region , where soi is present , and when @xmath85 is relatively large , there is also a visible time dependence ( see around the output lead ) .
although it is difficult to observe in fig .
[ linespinfig ] , the oscillating soi can generate waves that propagate away from the source even in the case of a simple straight wire .
the physical reason for the existence of these propagating waves is that soi oscillations pump energy in the system , populate floquet states with various frequencies and wave numbers .
boundary conditions `` transfer '' these populations to regions without soi , and the interference of these states appear as wave propagation .
[ densfig ] shows snapshots of the time evolution of the electron density given by eq .
( [ density ] ) .
the quantum wire is also shown in this figure , and for each point @xmath86 of the wire ( located in the @xmath87 plane ) @xmath88 is plotted as @xmath89 see the solid red and blue lines .
as we can see , density waves arise and propagate even for moderate soi strengths . given by eq .
( [ density ] ) for the @xmath90 and @xmath91 eigenspinor inputs [ eq . ( [ eigen ] ) ] , respectively .
the parameters are @xmath92 @xmath93 @xmath70 and @xmath94 time instants when the snapshots were taken are indicated in units of @xmath95,width=302 ] the figure shows the time evolution of @xmath96 for both eigenspinors given by eq .
( [ eigen ] ) .
let us recall that the spin direction is conserved for these input spinors , i.e. , in contrast to the case shown in fig .
[ linespinfig ] , there are no time - dependent spin rotations . on the other hand , however , the space and time dependence of the probability density is different for the two eigenspinor directions .
let us emphasize that this effect is absent for static soi , when the ( time independent ) transmission probability is the same for any input spin direction for a two terminal device .
this remarkable difference on the level of the equations can be understood by observing that the wavenumbers ( [ kap ] ) and ( [ kam ] ) that correspond to the two input spin directions are different when neither the oscillating , nor the static part of the soi is zero .
consequently , the spatial interference of superpositions of plane waves with these wave numbers produces different patterns for different input eigenspinors .
note that ( as we shall see in the following subsection ) this fact can lead to temporal spin - polarization which is completely absent for the case of static soi .
the related no - go theorem
@xcite for the equilibrium spin currents is based on symmetry - based considerations , like the unitarity of the scattering matrix , that ensures that the sum of the transmission and reflection probabilities is unity .
however , the probability density inside a region with oscillating soi is generally not constant , thus due to the continuity equation ( [ continuity ] ) , the magnitude of the current that flows out of the domain does not need to be equal to the current that flows in at least not at any time instants .
this is the symmetry - related , physical reason for the qualitative difference between the transmission properties of two - terminal devices with static and oscillating soi . as a final comment on this point ,
let us add that although the sum of the reflected and transmitted currents does not need to be always equal to the input current , the periodicity of the problem ensures that the average of these quantities over an oscillation period @xmath97 satisfies the relation @xmath98 the reflected and transmitted current densities appearing in the integrands are to be calculated using eq .
( [ currentdens ] ) with @xmath99 or @xmath100 [ see eqs .
( [ reflected ] ) , ( [ transmitted ] ) ] evaluated at the input and output junctions , respectively . in practice
, we can use these average quantities to check the accuracy of our numerical method : whenever the requirement above is not satisfied within the required numerical precision , more frequency components ( [ freqs ] ) have to be taken into account .
the parameters are @xmath101 @xmath102 @xmath70 and @xmath71 time instants when the snapshots were taken are indicated in units of @xmath95,width=302 ] in fig .
[ spinloopfig ] we can see the spin direction in the output lead of the geometry shown in fig . [ latticefig ] ( b ) for several time instants .
the input spin is polarized in the positive @xmath81 direction ( not shown ) , and the spin direction along the lead is visualized in the same way as in fig .
[ linespinfig ] .
we can clearly identify propagating patterns in this figure . ) ] for the geometry shown in fig .
[ latticefig](b ) .
@xmath103 ( a ) [ eq .
( [ density2 ] ) ] and @xmath104 ( b ) [ eq .
( [ purity ] ) ] are shown for parameters @xmath101 @xmath105 @xmath106 and @xmath94 recall that the density @xmath103 is not normalized ; as a reference , @xmath107 for a single spinor valued plane wave [ like the input given by eq .
( [ input])].,width=302 ] the results presented so far were obtained using a spin - polarized input . however , according to the previous subsection , interesting results can be expected when the incoming electrons polarization is random ( which is usually the case ) , their spin state can be described by a quantum mechanical density operator that is proportional to unity : @xmath108 let us note that for a single plane wave , @xmath109 is constant , but in general it can depend on both space and time .
we can calculate the density operator solution of the scattering problem for the input density operator ( [ pin ] ) simply by obtaining @xmath110 for the spin - up and spin - down inputs separately ( let us denote these solutions by @xmath111 and @xmath112 ) , and add the corresponding projectors : @xmath113 .
\label{pop}\ ] ] note that the natural generalization of eq .
( [ density ] ) in our case is the following : @xmath114 .
\label{density2}\ ] ] the most interesting result related to @xmath109 as a function of space and time is that completely pure states can appear in the waves propagating away from the source . as we have already discussed , this effect is in strong contrast with the case of constant soi , when a loop rotates the input spin direction always in the same way , whatever that direction was @xcite .
the physical reason for the temporal spin - polarization seen in fig . [ splotfig ] becomes most transparent by recalling fig .
[ densfig ] , where the probability densities had different space and time dependent interference patterns for the two eigenspinor inputs . in other words
, there are spacetime points where the interference is completely destructive for one of the eigenspinor directions , but not for the other one .
this results in a completely polarized spinor .
( note that this is formally analogous to the case reported in ref .
@xcite for a loop with _
static _ soi and _ two _ output terminals . )
the results to be presented in this subsection hold for both geometries shown in fig .
[ latticefig ] , but due to the increased number of paths that can interfere
the effects are stronger for the triangle loop [ fig .
[ latticefig](b ) ] .
however , let as emphasize that the appearance of the temporal spin - polarization as a physical effect has a weak dependence on the device geometry , e.g. , we expect it to be present also for quantum rings ( where the technique introduced in the previous section has to be modified ) as well as for various polygon geometries @xcite . in order to quantify the polarization effects , panel b ) of fig .
[ splotfig ] shows @xmath115 ^ 2}\mathrm{tr}\left[p^2(x,\tau)\right ] \label{purity}\ ] ] as a function of time and the @xmath33 coordinate in the output lead .
note that the division by the square of the electron density is just for normalization .
this function measures the `` purity '' of the spin state : its range is @xmath116,$ ] where the minimum and maximum corresponds to completely unpolarized and 100% polarized ( i.e. , pure ) states , respectively . according to fig .
[ splotfig ] , almost perfect polarization can occur in the output lead for moderate soi strengths .
additionally , taking a look at panel a ) of fig .
[ splotfig ] , we can also see that @xmath88 is not zero when the spin is polarized , i.e , there is a finite probability to find the spin - polarized electron at that point .
[ given by eq .
( [ purity ] ) ] is shown at @xmath117 for the linear geometry shown in fig .
[ latticefig](a ) .
the parameters are @xmath118 @xmath105 @xmath106 and @xmath119 the red curve corresponds to the case without scatterers [ @xmath120 in eq .
( [ scatt ] ) ] , while @xmath121 for the black curves .
this relatively strong scattering is present only in the region with oscillation soi for the dotted curve , only in the output lead for the dashed curve , while the dash - dot line corresponds to the case when there are scatterers in both spatial domains . technically , we considered three independent , randomly located scattering centers in both regions.,width=302 ] finally we investigate the question to what extent our findings are modified by the decrease of the mean free path as a consequence of scattering processes . to this end , similarly to ref .
@xcite , we introduce a random potential @xmath122 where @xmath123 denote uniformly distributed random positions and @xmath124 represents a @xmath125 diagonal matrix , with independent random diagonal elements @xmath126 and @xmath127 the argument @xmath128 is the root - mean - square deviation of the corresponding normal distribution the mean of which is zero .
these dirac delta peaks mean spin - dependent random scatterers and provide an effective model for magnetic impurities of various concentrations : @xmath120 corresponds to the ballistic case , while increasing values mean shorter mean free paths . in order to see the physical consequences of the scattering processes , we perform a sufficiently large number of computational runs with different realizations of the random potentials ( [ scatt ] ) and appropriately average the result ( see ref .
@xcite for more details ) . the polarization effect predicted by our model can be destroyed by scattering induced decoherence in two ways , depending on the position of the scatterers : in the region with oscillating soi the _ generation _ of the spin - polarized waves can be hindered , and/or the amplitude of these waves can be decreased by scatterers in the region where they _ propagate _
( i.e. , where there is no soi ) .
note that the input spin state is already completely unpolarized , thus there is no need to consider scatterers in the input lead . as it is shown by fig .
[ scattfig ] according to the expectations
the polarization effect gets definitely weaker when we introduce scatterers .
note that @xmath129 that corresponds to this figure means a relatively strong influence on the transport properties , it increases the reflection probability by roughly a factor of two
. moreover , fig .
[ scattfig ] visualizes the `` worst case '' , since nonmagnetic scatterers [ where @xmath130 in eq .
( [ purity ] ) is diagonal ( this case is not shown in the figure ) ] decrease the conductance by a similar amount , but their influence on spin - polarization is considerably weaker .
this fact emphasizes the importance of spin coherence length in our findings .
the most interesting fact we can see in fig .
[ scattfig ] is that the generation of spin - polarized waves is less sensitive to scattering processes than the propagation of these waves :
the same number of scatterers with the same value of @xmath128 have weaker effect when they are placed in the region where the soi oscillates .
that is , although spin - dependent random scattering decreases the degree of polarization independently from the position of the scatterers , when this process takes place inside the region with oscillating soi , spin - polarization still can build up , at least partially .
this effect can be understood qualitatively : the spin - polarization at the output is stronger when the `` interaction region '' ( where the soi oscillates and polarization is generated ) is longer ( with all other parameters being the same ) . when the spin coherence length decreases below the extension of this region , it defines a new , effective length along which polarization is generated .
thus , realistically , in a sample with long interaction region and disorder , it is the spin coherence length that determines the degree of polarization right at the output , and this is also the length scale that tells us the distance below which the spin - polarized electron waves can be detected in the output lead .
the results of this subsection show that the polarization effect we described earlier in this paper is not extremely sensitive to scattering induced decoherence , thus its experimental observation can be possible .
in the current paper we investigated spin - dependent quantum transport through devices in which the spin - orbit interaction ( soi ) is time dependent , more precisely , it oscillates . by considering a monoenergetic input ,
we have shown the emergence of electron density and spin polarization waves propagating away from their source , i.e. the region with oscillating soi .
additionally , it was demonstrated that simple geometries can produce spin - polarized wave packets even for completely unpolarized input . according to our calculation ,
this dynamical spin polarization effect appears for realistic , experimentally achievable parameter ranges and remains observable when moderately intensive scattering processes are also taken into account . in other words ,
our model suggests a novel source of spin - polarized electrons that can be realized with pure semiconducting materials without the use of external magnetic fields .
this work was supported by the hungarian scientific research fund ( otka ) under contract no .
t81364 as well as by the projects tmop-4.2.2.a-11/1/konv-2012 - 0060 , tmop-4.2.2/b-10/1 - 2010 - 0012 and tmop-4.2.2.c-11/1/konv-2012 - 0010 supported by the european union and co - financed by the european social fund . | ballistic transport through nanoscale devices with time - dependent rashba - type spin - orbit interaction ( soi ) can lead to spin - polarized wave packets that appear even for completely unpolarized input .
the soi that oscillates in a finite domain generates density and spin polarization fluctuations that leave the region as propagating waves .
particularly , spin polarization has space and time dependence even in regions without soi .
our results are based on an analytic solution of the time - dependent schrdinger equation .
the relevant floquet quasi - energies that are obtained appear in the energy spectrum of both the transmitted and reflected waves . |
the physics of low - dimensional magnetic systems continues to attract attention because of the association with peculiar quantum effects.@xcite prominent examples like the spin - peierls transition and the spin ladder compounds have been characterized by ground states of the spin singlet with a finite spin gap.@xcite for a quasi - one - dimensional chain with half - integer spin ( @xmath7 = 1/2 ) , the gap can be opened via either frustration due to next nearest neighbor antiferromagnetic exchange or dimerization due to an alternating coupling to nearest neighbors along the chain.@xcite several @xmath7 = 1/2 chain systems such as ( vo)@xmath1p@xmath1o@xmath8 , bacu@xmath1v@xmath1o@xmath9 , and pbcu@xmath1(po@xmath2)@xmath1 have been reported to possess spin gaps and their characteristics have been interpreted in accordance with these scenarios.@xcite cu@xmath1sc@xmath1ge@xmath2o@xmath3 , which crystallizes in a monoclinic structure with the space group _ p_2@xmath10_/m _ , was recently synthesized by one of the present authors ( g.j.r.).@xcite a schematic picture of the crystal structure is illustrated in fig .
1 . taking into account the known oxidation states of o@xmath11 and ge@xmath12 , the remaining valences are nonmagnetic sc@xmath13 and cu@xmath14 ( @xmath7 = 1/2 ) .
the spin chain in cu@xmath1sc@xmath15ge@xmath2o@xmath3 can be described as an arrangement of spin dimers oriented along the crystallographic _
b_-axis , separated by geo@xmath2 tetrahedra and crankshaft - like chains of sco@xmath16 octahedra ( fig .
each dimer consists of two cu@xmath14 ions in a cu@xmath1o@xmath1 plaquette and weak interdimer interactions could be possible to induce a quantum phase transition from a gapless state into a gapped state for cu@xmath1sc@xmath1ge@xmath2o@xmath3 .
the bulk magnetic susceptibility of cu@xmath1sc@xmath1ge@xmath2o@xmath3 exhibits a broad maximum at around 170 k and decreases rapidly at low temperatures , indicative of spin gap behavior for this material.@xcite also the magnetization data have been fitted well to the dimer chain model , yielding a spin gap of 290 k ( 25 mev)_. _ furthermore , the result of heat capacity has confirmed no magnetic ordering above 2 k.@xcite recently , masuda _ et al . _ have performed a neutron inelastic - scattering experiment on cu@xmath15sc@xmath1ge@xmath2o@xmath3 and found a narrow - band excitation of about 25 mev , which is quite consistent with the value extracted from the susceptibility.@xcite they thus attributed the observation to the spin - gap excitation and proposed cu@xmath1sc@xmath1ge@xmath2o@xmath3 as a new spin - gapped dimer material .
sc@xmath1ge@xmath2o@xmath3.,width=3 ] in order to further identify the existence of a spin gap in cu@xmath1sc@xmath1ge@xmath17o@xmath3 , we carried out a detailed @xmath0sc nuclear magnetic resonance ( nmr ) study invoking nmr shifts and spin - lattice relaxation rates on this compound .
the nmr shift provides a local measurement of the susceptibility which is less sensitive to the presence of impurities and other phases .
the spin - lattice relaxation rate is a sensitive probe for the low - energy spin excitations , yielding direct evidence for the presence of an energy gap .
regarding the title compound cu@xmath1sc@xmath1ge@xmath2o@xmath3 , a transfer of magnetic _
d_-spin from the cu@xmath14 onto sc@xmath13 4_s _ orbital allows us to probe the cu@xmath14 spin dynamics and determine the spin gap through the transfer hyperfine interaction .
the experimental nmr results clearly reveal spin gap behavior in this material .
data analysis using the isolated dimer model is found to give good agreement with the observations .
interestingly , the deduced spin gap value of 275 k is close to the 25 mev peak revealed by the inelastic neutron experiment,@xcite indicating the same energy excitation detected by both measurements .
a polycrystalline cu@xmath1sc@xmath1ge@xmath2o@xmath3 sample was synthesized by a ceramic sintering solid - state reaction technique.@xcite a mixture of cuo , sc@xmath1o@xmath18 and geo@xmath1 , weighted in a predetermined molar ration of cu@xmath15sc@xmath1ge@xmath2o@xmath3 , was carefully ground under alcohol , pressed into pellets , put into an open platinum crucible , and fired under ambient pressure and ambient oxygen fugacity in a temperature range between 900 @xmath19c and 1100 @xmath20c .
after each of six heating cycle , the sample was reground , pressed , and reheated . in initial stages the product consisted of a mixture of the title compound and cugeo@xmath18 .
the amount of cugeo@xmath18 reduced successively as increasing synthesis temperature and heating time . in a final synthesis cycle , the sample was fired at 1150 @xmath19c and a light blue product was thus obtained .
a room - temperature x - ray diffraction confirmed the single phase for cu@xmath1sc@xmath1ge@xmath2o@xmath3 with the lattice parameters @xmath2112.336(2 ) @xmath22 , @xmath238.7034(9 ) @xmath22 , @xmath244.8883(8 ) @xmath22 , and @xmath2595.74(2 ) .
nmr experiments were performed using a varian 300 spectrometer , with a constant field of 7.05 t. a home - built probe was employed for the low - temperature measurements.@xcite the powder specimen was put in a plastic vial that showed no observable @xmath0sc nmr signal .
the nmr spectrum was obtained from spin echo fast fourier transforms using a standard @xmath26 sequence . within the _ p_2@xmath10_/m _ space group for cu@xmath1sc@xmath1ge@xmath2o@xmath3 ,
scandium atoms occupy one crystallographic site having octahedral oxygen coordination , yielding an one - site @xmath0sc nmr powder pattern , as demonstrated in the inset of fig .
the observed nmr line shape is manifested by the anisotropic knight shift as well as the quadrupole effects .
however , no visible quadrupole edges have been detected with our static nmr probe , suggesting that the value of quadrupole frequency @xmath27 is essentially small .
such a result indicates that the sc atoms are located in a symmetric environment , probably at the center of the sco@xmath16 octahedron .
it is of great importance to note that the line shape of the spectrum remains unchanged with temperature .
this phenomenon confirms the nonmagnetic ground state for cu@xmath1sc@xmath1ge@xmath2o@xmath28 and no structural changes above 4 k. sc nmr shift in cu@xmath1sc@xmath1ge@xmath2o@xmath3 .
note that @xmath29 is shown by the negative direction .
dashed curve : fit to the dimer model plus a constant term .
inset : @xmath0sc nmr spectrum for cu@xmath1sc@xmath1ge@xmath2o@xmath3 measured at 24 k.,width=2 ] in fig .
2 , we displayed the observed temperature - dependent nmr shift ( _ k_@xmath30 ) for the sc site .
the shift was taken at the peak position of the resonance line referred to the @xmath0sc resonance frequency of aqueous sccl@xmath31 . here _
k_@xmath30 has different sign from that of the susceptibility since the transfer hyperfine field arising from 3_d _ electrons of cu@xmath32 ions is negative . in spite of the sign difference of _
k_@xmath30 , the whole temperature variation is quite consistent with the bulk susceptibility data , with a broad maximum at around @xmath4 @xmath5 170 k and a rapid decrease as lowering temperature.@xcite in general , _
k_@xmath30 can be decomposed into _
k_@xmath30 = _ k_@xmath33 _ k_@xmath34(_t _ ) . the first term _
k_@xmath35 = 112 ppm , mainly arising from the orbital shift , is independence of temperature . on the other hand ,
the spin shift _ k_@xmath36 , which reflects the cu@xmath14 spin behavior through the transfer hyperfine interaction , is a function of temperature . against 1/_t _ for cu@xmath1sc@xmath1ge@xmath2o@xmath28 .
solid curve : fit to the function described by the dimer model with @xmath37 of 270 k.,width=2 ] as mentioned above , the spin chain configuration of spin dimers could be responsible for the magnetic behavior of cu@xmath1sc@xmath1ge@xmath2o@xmath3 .
accordingly , the temperature dependence of the spin shift will obey the relation _
k_@xmath36 @xmath38 1/@xmath39(@xmath40).@xcite here @xmath41 is the gap energy determined from the nmr shift based on the dimer model . as demonstrated in fig . 3 , _
k_@xmath36 can be fitted well in a fairly wide temperature range to this relation ( solid curve in fig .
it is also found that _
k_@xmath36 vanishes to almost zero at low temperatures , indicative of the gap in the spin excitation spectrum . with this fit
, we extracted @xmath41 = 270@xmath4220 k , consistent with 290 k obtained from the susceptibility measurement.@xcite the spin shift here is related to the magnetic susceptibility @xmath43 by the expression @xmath44 where _
h_@xmath45 is the transfer hyperfine field due to an intermixing of sc and cu spin states , @xmath46 is the avogadro s constant , and @xmath47 is the bohr magneton .
the clogston - jaccarino plot@xcite which shows the observed shift against magnetic susceptibility ( deduced from the data of ref .
10 ) is given in fig .
4 . the linear behavior indicates a unique hyperfine field over the entire temperature range we investigated .
the slope yields a small value of _ h_@xmath45 = @[email protected] koe . versus @xmath49 ( deduced from ref .
10),width=2 ] to gain more insight into the spin gap characteristics of cu@xmath1sc@xmath1ge@xmath2o@xmath3 , we performed spin - lattice relaxation rate ( 1/_t_@xmath10 ) measurements , being sensitive to the low - energy magnetic excitations .
it thus provides direct information about the low - energy spin dynamics and the presence of a spin gap . here
t@xmath10 _ measurement was carried out using the saturation recovery method .
the saturation rf comb with 30 short 2 @xmath50s pulses was employed .
we recorded the recovery of the signal strength by integrating the @xmath0sc spin echo signal . for the central transition with _
= 7/2 , the recovery of the nuclear magnetization follows@xcite
@xmath51 here @xmath52(@xmath53 ) is the magnetization at the recovery time @xmath53 and @xmath52(@xmath54 ) is the initial magnetization .
the _ t@xmath10 _ value was thus obtained by fitting to this multi - exponential function . to provide accurate values , each _
t@xmath10 _ has been measured several times and the averaged _
t@xmath10 _ for the corresponding temperature is shown in the inset of fig .
it is apparent that 1/_t@xmath10 _ exhibits activated behavior at low temperatures .
based on the dimerized scenario , the spin - lattice relaxation rate should be fitted to the form 1/_t_@xmath10 @xmath381/@xmath55 by analogy to the treatment of nmr shift.@xcite the fitting result , drawn in fig .
5 as a solid curve , is quite satisfactory and yields an energy gap @xmath56 = 275@xmath4225 k. remarkably , this excitation energy is almost identical with the 25 mev ( 290 k ) peak obtained in the inelastic neutron experiment,@xcite pointing to the same energy excitation probed by both measurements . for cu@xmath1sc@xmath1ge@xmath2o@xmath3 .
the solid curve is the fitted function based on the dimer model .
inset shows the temperature variation of 1/@xmath57.,width=2 ] our nmr investigation thus provides clear evidence for the existence of spin gap in cu@xmath1sc@xmath1ge@xmath2o@xmath3 .
the extracted @xmath41 was found to be close to @xmath56 . as a matter of fact ,
the ratio of @xmath56/@xmath41 @xmath5 1 is commonly seen in the dimerized systems.@xcite within the isolated dimer limit , @xmath4 , the temperature at which magnetic susceptibility or nmr shift display a maximum , will appear at around 0.63@xmath56.@xcite for the present case of cu@xmath1sc@xmath1ge@xmath2o@xmath3 , the determined @xmath56 = 275 k results in @xmath4 = 173 k , in excellent agreement with the observed @xmath4 @xmath5 170 k. in this regard , it seems reasonable to characterize cu@xmath1sc@xmath1ge@xmath2o@xmath58 as a nearly isolated dimer chain compound . )
@xmath59/@xmath60.,width=2 ] at this moment , we did not attempt to fit the temperature dependence of spin shift and 1/_t@xmath10 _ to other models such as the alternating chain model .
in fact , our previous analyses of nmr data for bacu@xmath1v@xmath1o@xmath9 indicated that both alternating chain and dimer chain models are suitable for the understanding of the spin gap nature.@xcite however , a convincing result can be established unambiguously from the comparison of static and dynamic excitations . within the dimer picture , the excitation is predominant by a simple singlet - triplet process.@xcite as a result , the static susceptibility probed by @xmath61 and the local dissipative susceptibility probed by ( @xmath62)@xmath59 are nearly identical . on the other hand , for weakly coupled heisenberg spin chains ,
both triplet and singlet - triplet mechanisms contribute to the relaxation rate.@xcite in this case , the temperature dependence of ( @xmath62)@xmath59 becomes more prominent than that of @xmath61 as @xmath63 @xmath41 , leading to a larger spin gap deduced by 1/_t@xmath10._@xcite such a phenomenon has been observed in the @xmath641 haldane gap compound y@xmath1banio@xmath65 and the @xmath641/2 spin chain bacu@xmath1v@xmath1o@xmath9.@xcite to provide a reliable spin gap description for cu@xmath1sc@xmath1ge@xmath2o@xmath3 , we thus examine the temperature dependence of ( @xmath62)@xmath59/@xmath66 without any fitting . as one can see from fig .
6 , the ratio of ( @xmath62)@xmath59/@xmath66 remains constant even below the spin gap .
this result is consistent with the spin dimer theory in which both static and dynamic excitations follow the same temperature variation due to the involvement of a simple singlet - triplet excitation .
in conclusion , we report the first nmr investigation of cu@xmath1sc@xmath1ge@xmath2o@xmath28 and present evidence for the existence of spin gap in this material .
the spin gap of about 275 k deduced from the spin shifts and spin - lattice relaxation rates was found to be identical .
a detailed analysis further indicates that the spin gap nature can be well accounted for by the free dimer model .
these findings thus allow us to add cu@xmath1sc@xmath1ge@xmath2o@xmath3 to the family of @xmath641/2 dimerized spin gap compounds .
this work was supported by the national science council of taiwan under grant no .
nsc-95 - 2112-m-006 - 021-my3 ( c.s.l ) , and by a grant to g.j.r . from the _ fond zur forderung der wissenschaftlichen forschung _
( fwf ) , austria , ( grant no .
r33-n10 ) . | we report the results of a @xmath0sc nuclear magnetic resonance ( nmr ) study on the quasi - one - dimensional compound cu@xmath1sc@xmath1ge@xmath2o@xmath3 at temperatures between 4 and 300 k. this material has been a subject of current interest due to indications of spin gap behavior .
the temperature - dependent nmr shift exhibits a character of low - dimensional magnetism with a negative broad maximum at @xmath4 @xmath5 170 k. below @xmath6 , the nmr shifts and spin lattice relaxation rates clearly indicate activated responses , confirming the existence of a spin gap in cu@xmath1sc@xmath1ge@xmath2o@xmath3 . the experimental nmr data can be well fitted to the spin dimer model , yielding a spin gap value of about 275 k which is close to the 25 mev peak found in the inelastic neutron measurement .
a detailed analysis further points out that the nearly isolated dimer picture is proper for the understanding of spin gap nature in cu@xmath1sc@xmath1ge@xmath2o@xmath3 . |
in @xcite , esposito et .
al . have proposed a dynamical model to explain the aggregation and disaggregation processes observed in saturn s f ring and the b ring outer edge due to the perturbation caused by the saturn s moon , prometheus and mimas , respectively .
their numerical simulations demonstrate that the dynamical model explains certain spacecraft measurements corresponding to these regions in the saturn s ring .
further , these numerical results strongly suggest the existence of a limit cycle for realistic values of the parameters . in this paper , we provide a rigorous mathematical proof of this fact as well as study other features that are relevant from a dynamical point of view .
the model under consideration relates the mean aggregate mass @xmath1 and the velocity dispersion @xmath2 of ring particles at the above mentioned locations in the saturn s ring , as follows @xmath3 in the first equation of ( [ model ] ) , the first term refers to the coagulation of the ring particles leading to a growth in the mean aggregate mass at an accretion rate @xmath4 ( @xmath5 is the accretion period ) and the second term refers to the erosion ( or break - up ) of the larger ring particles due to collisions with other ring particles at collisional rates @xmath6 ( @xmath7 is the collisional period ) when their velocity dispersion exceed @xmath8 , the threshold velocity for sticking . in the second equation of ( [ model ] )
, the first term refers to the dissipative collisions between the ring particles leading to the reduction in velocity dispersion with @xmath9 as the normal coefficient of restitution .
the second term refers to the increase in the velocity dispersion to the escape velocity from mass @xmath1 ( @xmath10 is the escape velocity from an aggregate of mass @xmath11 ) due to the viscous stirring at a rate @xmath12 ( @xmath13 is the stirring period ) , caused by the passage of the ring particles of mass @xmath1 .
finally , the last term refers to the periodic forcing by the saturn s moon causing disturbances in the velocity dispersion , where @xmath14 describes the moon perturbation , where @xmath15 is the forcing period , @xmath16 is the forcing amplitude . for convenience , we reduce the system to an equivalent dimensionless system with the following substitutions : @xmath17 , @xmath18 and @xmath19 , where @xmath20 is the orbital period of the ring particles around saturn , and pass to a more mathematical notation by using @xmath21 .
then , the system in ( [ model ] ) becomes @xmath22 where @xmath23 are all positive parameters , @xmath24 , @xmath25 , @xmath26 , and @xmath27 is the ( normalized ) forcing period .
the above dynamical system has two main differences when compared to the predator - prey system .
firstly , in the second equation , we have the @xmath28 term instead of the coupling @xmath29 .
secondly , the external forcing is not standard in ecological models , since the typical forcing is parametrical ( see for instance the model studied in @xcite ) . despite these differences , the aggregation - disaggregation model represented by this system can be well explained by drawing an analogy with the predator - prey system .
let the mean aggregate mass of the ring particles correspond to the prey population , and the velocity dispersion correspond to the predator population . the velocity dispersion feeds off the accelerations from the aggregates gravity ( in reference to the second term of the second equation ) .
as the velocity dispersion grows too large , it limits the prey as high velocity collisions lead to to fragmentation of the ring particles ( in reference to the second term of the first equation ) . in the absence of the interaction between the predator and the prey ,
the prey population ( mean aggregate mass ) grows and the predator population decays ( in reference to the first terms of the two equations , respectively ) .
in addition , the predator population is periodically controlled in a certain deterministic manner , as represented by the sinusoidal forcing function . in the following section
, we conduct a thorough qualitative analysis of the above system and comment about the existence and nature of its solution , conditions on uniqueness and asymptotic stability of the solution and finally derive explicit bounds for the solutions .
along with these results , we derive useful insights about the saturn s ring dynamics based on these results .
further , the results in this paper hold for any general forcing functions @xmath30 as long as it is a continuous periodic function of minimal period @xmath31 with zero mean value , i.e. , @xmath32 .
all the results will be presented for this general forcing function , and the sinusoidal forcing function will be explicitly studied in section [ sec : example ] .
numerical results in @xcite strongly suggest the existence of a limit cycle for realistic values of the parameters . in the following theorem
, we provide an analytical proof for the same .
further , the result holds for all feasible values of the system parameters .
[ th1 ] system @xmath33 has at least one @xmath27-periodic solution @xmath34 with @xmath35 for all @xmath36 .
* by introducing the change @xmath37 , system @xmath33 is transformed into @xmath38 the theorem is proved by showing that the above system is equivalent to a second order differential equation of the duffing type , and then invoking the landesman - lazer conditions for the existence of a periodic solution .
see appendix [ sub : proofth1 ] for the complete proof . ' '' ''
the above result leads to a physical interpretation that the periodic forcing by the moon gives a limit cycle in the mean aggregate mass and velocity dispersion with the same period as the forcing function .
it is interesting to remark that the latter result can be extended to a family of forcing terms with more complicated recurrence including quasi periodic forcing .
the main result in @xcite implies the existence of a quasi periodic solution and the boundedness of all solutions in the future .
in this section , we derive conditions for the existence of a unique periodic solution and comment about the stability of the solution .
[ stab1 ] when @xmath39 system @xmath33 has a unique @xmath27-periodic solution , which is asymptotically stable . * proof .
* see appendix [ sec : proofstab1 ] .
observe that for @xmath40 , the unique equilibrium of system @xmath33 is hyperbolic , hence robust to small perturbations .
therefore , if @xmath41 is small enough , @xmath33 presents an asymptotically stable @xmath27-periodic orbit .
the remarkable feature of the latter result is that stability is proved on a certain region of the involved parameters independently of the size of the external force @xmath42 . as we will see in section [ sec : example ] , theorem [ stab1 ] is applicable to the saturn s f - ring , in fact for the @xmath43-ring @xmath44 so condition @xmath45 is satisfied .
the condition @xmath45 is also satisfied for the saturn s b - ring outer edge when the optical depth is @xmath46 . in the following theorem ,
a condition for the existence and stability of the solution is derived that depends on the nature of the forcing function via its supremum norm , denoted by @xmath47 .
[ stab2 ] under the assumption @xmath48 system @xmath33 has at least one asymptotically stable @xmath27-periodic solution
. * proof . *
the above result is obtained by using the technique of lower and upper solutions @xcite .
see appendix [ sec : proofstab2 ] for the complete proof . ' '' '' in the context of the saturn s ring dynamics , theorem [ stab1 ] and theorem [ stab2 ] show that the forced ring system has a limit cycle behavior around the fixed point and when the external forcing satisfies the above condition , it drives a stable limit cycle .
further , the above theorems provide a theoretical justification for the concentric limit cycles that were observed through numerical simulations in ( * ? ? ?
16 and fig .
theorem [ th1 ] provides a general existence result , but more concrete quantitative information about the location of the periodic orbits is desirable . in this section
, we derive explicit estimates for the periodic solutions . in the next result ,
@xmath49 , @xmath50 denotes the positive and negative part respectively of a given function @xmath42 , and @xmath51 denotes the usual @xmath52-norm .
[ th2 ] let @xmath34 with @xmath35 for all @xmath36 a @xmath27-periodic solution of system @xmath33 . then , the following bounds hold @xmath53 for all @xmath36 . * proof . *
the bounds derived above are obtained by working with the second order differential equation of duffing type ( [ duff ] ) that is equivalent to the system in ( [ model2 ] ) .
see appendix [ sec : proofth2 ] for the complete proof . ' '' '' since the lower bound for @xmath54 is non - negative , the @xmath27-periodic solution @xmath54 is also non - negative for all @xmath36 .
the above argument acts as an alternate proof for the later part of the statement in theorem [ th1 ] .
further , the bounds derived for the @xmath27-periodic solution for @xmath55 may not necessarily be tight . especially , for small forcing amplitudes ,
the @xmath27-periodic solutions are expected to be close to the stable feasible fixed point , but the upper and lower bounds deviate from the stable fixed point by a factor of @xmath56 about the fixed point . as a result , we derive tighter bounds for @xmath55 in the following theorem .
[ th3]alternatively , @xmath57 satisfies the bounds : @xmath58 where @xmath59 , @xmath60 , and @xmath61 is the green s function .
* proof . * the alternative bounds given by @xmath62 and @xmath63 are found working directly with the second equation of system @xmath64 .
given a @xmath27-periodic solution @xmath65 , it is known that the unique @xmath27-periodic solution of the first - order linear equation @xmath66 is @xmath67 then , from the second equation of system @xmath64 , we have @xmath68ds.\ ] ] tight upper and lower bounds are derived for each of the two terms in the above integral considered separately . by appropriately combining these bounds ,
we obtain @xmath62 and @xmath63
. see appendix [ sec : proofth3 ] for the complete proof . ' '' '' the bounds in @xmath62 and @xmath63 are tight for small forcing amplitudes
. combining these with the second inequality in ( [ bounds ] ) , the bounds are tight for all forcing amplitudes .
since @xmath57 represents a squared relative velocity , a useful lower bound for @xmath57 should be positive . in this sense ,
the second bound given by @xmath62 has the advantage that the lower bound is always positive if the external force @xmath42 is not too large . in the model case @xmath69 , it is easy to compute @xmath70 , @xmath71 , and the bounds are direct . having completed the mathematical analysis of the system ( [ model2 ] ) , in the following section , we consider examples with realistic values of the parameters of the system representing different regions of the saturn s ring and derive useful insights pertinent to the ring dynamics .
we begin with some corollaries of the theorems presented in the previous sections for the saturn s ring system .
[ rem : fixedpoint]the non - trivial , feasible stable fixed point for the unforced dynamic system is @xmath72 .
when externally forced by periodic perturbations from the saturn s moons , the mean aggregate mass and the velocity dispersion demonstrates periodic aggregation - disaggregation cycle with the same period as that of the moon s forcing .
( of theorem [ stab1 ] ) there exists a unique , asymptotically stable limit - cycle for the ( @xmath1,@xmath73 ) system if the forcing period satisfies the condition @xmath74 where the right - hand - side depends on the mean optical depth and the normal coefficient of restitution corresponding to the specific location in the saturn s ring system .
the realistic parameters corresponding to the saturn s ring system are @xmath75 , @xmath76 .
the optical depth is @xmath77 for the f - ring and @xmath78 for the b - ring outer edge .
hence , our result is applicable to the f - ring and also to the b - ring outer edge when @xmath79 .
now , we consider a systematic study of each region in the saturn s rings , and start with the dynamics of the b - ring outer edge . for a sinusoidal forcing function , @xmath80 ,
the upper and lower bounds of the @xmath27-periodic solution for the mass of the aggregate is @xmath81 for @xmath82 , the bounds are obtained from theorems [ th2 ] and [ th3 ] with @xmath83 , @xmath84m / sec , @xmath85m / sec , @xmath86 , @xmath87 , @xmath88 , @xmath89 , and @xmath90 .
we begin with studying the impact of varying the stirring rate @xmath91 on the amplitudes of the periodic solutions of @xmath92 , and figures [ fig : fstirfig1 ] - [ fig : fstirfig3 ] plot the maxima and minima of the @xmath27-periodic solution of the @xmath93 system obtained using numerical simulations against the derived upper and lower bounds for the @xmath27-periodic solution of the @xmath93 system from theorem [ th2 ] and theorem [ th3 ] .
the system parameters corresponding to the saturn s ring system are @xmath75 , @xmath94 @xmath95 , @xmath96 and @xmath97 and @xmath98 for figure [ fig : fstirfig1 ] , figure [ fig : fstirfig2 ] and figure [ fig : fstirfig3 ] , respectively . the saturn s b - ring including the janus 2:1 density wave , and the b - ring outer edge have 1.5 and 1 as the typical values for the optical depth .
further , @xmath79 is the typical optical depth value for the saturn s a ring , mimas 5:3 density wave and the janus 6:5 density wave .
from ( [ eq : b1 ] ) , it is clear that the amplitude upper and lower bounds for the periodic solutions of @xmath99 vary log - linearly with @xmath100 and hence the corresponding curves in the plot are straight lines . it is interesting to note that the actual values for the maximum and minimum amplitudes ( obtained through simulations ) also appear to have a log - linear relationship with @xmath101 with the same slope as that of the upper and lower bounds . as a result , an exponential increase in @xmath101 causes an exponential decay in the mean aggregate mass of the ring particles .
this effect can be understood as follows .
large @xmath101 causes the rate of change of @xmath55 to be positive , and once @xmath102 ( same as @xmath103 in ( [ model3 ] ) ) , this causes a negative slope for @xmath104 and essentially causing the exponential decay of @xmath105 further , the parameter @xmath101 has a negligible effect on the amplitudes of @xmath106 as is evident from the negligible slope in the figure above .
notice that both the axes are in log - scale and the plots for the lower - bound of @xmath82 does not appear in the plot because it takes negative values for all values of @xmath107 we see that the effect of changing @xmath108 only changes the magnitudes of maxima , minima , and the bounds on @xmath99 and @xmath109 otherwise , @xmath99 has a log - linear relationship with @xmath100 with the same slope for all values of @xmath108 . comparing figures [ fig : fstirfig1 ] - [ fig : fstirfig3 ] , the magnitudes of maxima @xmath99 increases with increase in @xmath110 and the minima decreases with increase in @xmath111 further , the magnitude of the maxima of @xmath82 remains unchanged with increase in @xmath108 and the magnitude of the minima decreases with increase in @xmath111 the physical interpretation is that larger optical depth leads to a larger mean aggregate mass , but not large velocity dispersion .
this is because the velocity dispersion is pegged near the fixed point value given by the strong erosion for @xmath112 , which reverses the growth of @xmath99 .
for the same reason , increasing @xmath113 reduces the upper value for @xmath99 .
next , we study the impact of varying the forcing amplitude @xmath16 on the periodic solutions of @xmath92 , and figures [ fig : betafig1 ] - [ fig : betafig3 ] plot the maxima and minima of the @xmath27-periodic solution of the @xmath93 system obtained using numerical simulations against the derived upper and lower bounds for the @xmath27-periodic solution of the @xmath93 system from theorem [ th2 ] and theorem [ th3 ] .
the system parameters are the same as before and the stirring period is assumed to be equal to the orbital period of the ring particles around saturn , in other words , @xmath114 . notice that for small forcing amplitudes , the periodic solutions for @xmath99 and @xmath82 has a small amplitude about their average values , which in this case , are the asymptotically stable fixed points ( see remark [ rem : fixedpoint ] ) of the corresponding autonomous system of differential equations . with increase in @xmath115
we see an increasing deviations in the amplitude of @xmath99 about the average values , while still non - negative , since the lower - bound from ( [ eq : b1 ] ) is non - negative .
translating to the physical phenomena , the ring particle aggregates tend to achieve large masses as well as negligibly small masses , with increase in the forcing amplitude , in each period .
the increase in forcing amplitudes causes a similar effect on @xmath82 as on @xmath99 . but increasing the forcing amplitude beyond a certain threshold causes @xmath82 to take negative values , which is non - physical .
this sets a strict upper limit on the forcing amplitude , which in the case of the above figure , happens around @xmath116 comparing figure [ fig : betafig1 ] and [ fig : betafig3 ] , as @xmath108 increases from 0.5 to 1.5 , @xmath99 maxima increases , @xmath82 maxima changes only slightly .
finally , notice that the upper and lower bounds for @xmath99 and @xmath82 are tight for small forcing amplitudes .
this means that the behavior of the system is well - characterized by the bound derived in theorem [ th2 ] and theorem [ th3 ] for the regions in the saturn s ring system where the influence of the moons are negligible .
the analytical results provide a reasonable idea about the nature of the system and parallel the explicit numerical simulations .
we consider the mathematical model developed in @xcite for understanding the dynamics of certain specific locations in the saturn s ring system due to the periodic perturbations caused by the nearby moon .
this model relates the mean aggregate mass and the velocity dispersion of the ring particles at a certain location in the saturn s ring system in a manner similar to the popular predator - prey model in ecology , although it represents a totally different physical phenomenon .
further , this has been studied purely using numerical simulations so far . in this paper
, we conduct a rigorous mathematical analysis of the dynamical system and derive results pertaining to the existence , uniqueness and stability of its solution .
we analytically prove that the dynamical system causes a periodic solution for the mean aggregate mass and the velocity dispersion of the ring particles at a period equal to the forcing period of the moon .
further , upper and lower bounds for the solution to the system are derived , that are especially tight for small perturbations caused by the moons .
these bounds well - characterize the dependence of the system solution on various system parameters .
several useful insights are drawn about the saturn s ring system based on the theoretical results presented here .
1 s. ahmad , a nonstandard resonance problem for ordinary differential equations , transactions of the american mathematical society , vol .
323 , no . 2 ( feb . , 1991 ) , pp.857 - 875 . z. amine , r. ortega , a periodic prey - predator system , j. math .
185 ( 1994 ) , 477 - 489 .
esposito , nicole albers , bonnie k. meinke , miodrag sremevi , prasanna madhusudhanan , joshua e. colwell , richard g. jerousek , a predator - prey model for moon - triggered clumping in saturn s rings , icarus 217 ( 2012 ) , 103 - 114 .
njoku and p. omari , stability properties of periodic solutions of a duffing equation in the presence of lower and upper solutions , appl .
comput . 135
( 2003 ) , 471 - 490 .
g. talenti , best constant in sobolev inequality , ann . mat .
pura appl . 4 ( 1976 ) 353 - 372 .
torres , existence and stability of periodic solutions of a duffing equation by using a new maximum principle , mediterranean journal of mathematics 1 , n.4 ( 2004 ) , 479 - 486 .
ward , periodic solutions for a class of differential equations , proc .
78 ( 1980 ) , 350 - 352 .
deriving the first equation and inserting the second one , we obtain the following second order differential equation of duffing type @xmath117 now the periodic problems for @xmath33 and @xmath118 are completely equivalent . if @xmath34 is a @xmath27-periodic solution of system @xmath33 with @xmath35 for all @xmath36 , then @xmath37 is a @xmath27-periodic solution of @xmath118 .
conversely , if @xmath119 is a @xmath27-periodic solution of eq .
@xmath118 , then @xmath120 is a @xmath27-periodic solution of system @xmath33 with @xmath35 for all @xmath36 . since @xmath121 , eq .
@xmath118 is of landesman - lazer type .
landesman - lazer conditions are classical and implies the existence of a periodic solution , see for instance ( * ? ? ?
* theorem 1 ) .
we know that system @xmath33 is equivalent to eq
. @xmath118 , hence along this proof we will work directly with this last equation .
* uniqueness . *
assume that @xmath122 are two @xmath27-periodic solutions of eq .
the difference @xmath123 is a solution of a second order linear equation @xmath124 where @xmath125 .
we consider two cases : * _ case 1 _ : @xmath126 for every @xmath36 : without loss of generality , we may take @xmath127 . by the mean value theorem , @xmath128 where @xmath129 for every @xmath36 .
hence , by using @xmath130 , one obtains @xmath131 using @xmath45 , one may check easily that @xmath132 , as defined in ( * ? ? ?
* section 3 ) .
then , by ( * ? ? ? * corolllary 2.5 ) , the operator defined by the left - hand side of @xmath133 is inversely positive , in particular @xmath133 is non - degenerate and its only periodic solution is the trivial one , which is a contradiction with the assumption that @xmath126 for every @xmath36 . * _ case 2 _ : @xmath134 for some @xmath135 $ ] : by periodicity , there exists @xmath136 such that @xmath137 and @xmath126 for all @xmath138t_{0},t_{1}[$ ] .
again , we may assume without losing generality that @xmath127 for all @xmath138t_{0},t_{1}[$ ] .
define the truncated function @xmath139 as @xmath65 if @xmath140 $ ] and @xmath141 if @xmath142t_{1},t]$ ] . by reasoning as before , @xmath132 . then , as a consequence of the results in @xcite ( see remark 2.2 therein ) , the distance between two consecutive zeroes of a solution of @xmath143 is greater than @xmath27 , but @xmath144 is a solution vanishing at @xmath145 , which is a contradiction . after this analysis , the only remaining possibility is @xmath146 for every @xmath36 , so the @xmath27-periodic solution is unique .
* stability .
* the proof for asymptotic stability is also based on the results from @xcite .
let @xmath147 be the unique @xmath27-periodic solution of eq .
a standard computation gives that the topological index of @xmath119 is @xmath148 .
the variational equation is @xmath149 with @xmath150 and again @xmath132 as a result of @xmath45 .
then the asymptotic stability follows directly from ( * ? ? ?
* theorem 3.1 ) .
we use the terminology and notation from @xcite .
a lower solution is found by fixing a constant @xmath151 such that @xmath152 an upper solution is easily obtained as follows .
take @xmath43 the unique @xmath27-periodic solution of @xmath153 with mean value zero .
then , @xmath154 with @xmath1 big enough is an upper solution with @xmath155 . to apply ( * ? ? ?
* theorem 1.2 ) , we have to check two hypotheses .
the fist one is the finiteness of the number of @xmath27-periodic solutions , which is direct from the analiticity of the field on the state variables .
the second hypothesis is condition ( 1.6 ) in @xcite .
fix @xmath156 . for @xmath157
$ ] , @xmath158 by @xmath159 , which is just ( 1.6 ) in @xcite .
the result is done .
we work again over the equivalent eq . @xmath118 . integrating @xmath118 over a whole period @xmath160 by the integral mean value theorem
, there exists @xmath135 $ ] such that @xmath161 observe that @xmath162 , then by the sobolev inequality with the optimal constant ( see for instance @xcite ) , we get @xmath163 on the other hand , multiplying @xmath118 by @xmath164 and integrating over @xmath165 $ ] and applying cauchy - schwarz inequality , we get @xmath166 hence , @xmath167 now , combining @xmath168,@xmath169 and @xmath170 , @xmath171 and @xmath172 the first inequality of @xmath173 arises from here taking @xmath37 and finding the value of @xmath174 .
the bound for @xmath57 is found as follows .
let @xmath175 such that @xmath176}z(t)$ ] .
integrating @xmath118 on @xmath177 $ ] , @xmath178 since @xmath179 , then @xmath180 analogously , let @xmath181 such that @xmath182}z(t)$ ] .
integrating @xmath118 on @xmath183 $ ] , @xmath184 then , @xmath185 since @xmath147 is @xmath27-periodic , @xmath186 , then by using @xmath130 , @xmath187 finally , inserting @xmath188 into @xmath189 , one obtains the second inequality of @xmath173 .
note that @xmath190 for all @xmath191 $ ] . using the bounds for @xmath194 ,
@xmath195 the next set of bounds are derived by using the bounds derived for @xmath196 in the proof of theorem [ th2 ] and noting that @xmath197 @xmath198 combining ( [ eq : ufunbounds1 ] ) and ( [ eq : ufunbounds2 ] ) , we get @xmath199 another set of bounds based on the following inequality @xmath208 are @xmath209 and finally , if the @xmath210 is computable for all @xmath36 , then , @xmath211 note that ( [ eq : vfunbounds1 ] ) and ( [ eq : vfunbounds3 ] ) are equivalent . hence , combining ( [ eq : vfunbounds2 ] ) - ( [ eq : vfunbounds5 ] ) , we get @xmath212 using ( [ eq : ufunbound ] ) and ( [ eq : vfunbounds ] ) , ( [ eq : ylb ] ) and ( [ eq : yub ] ) follow directly . | spacecraft observations of saturn s rings show evidence of an active aggregation - disaggregation process triggered by periodic influences from the nearby moons .
this leads to clumping and break - up of the ring particles at time - scales of the order of a few hours .
a mathematical model has been developed to explain these dynamics in the saturn s f - ring and b - ring @xcite , the implications of which are in close agreement with the empirical results . in this paper
, we conduct a rigorous analysis of the proposed forced dynamical system for a class of continuous , periodic and zero - mean forcing functions that model the ring perturbations caused by the moon flybys . in specific ,
we derive the existence of at least one periodic solution to the dynamic system with the period equal to the forcing period of the moon .
further , conditions for the uniqueness and stability of the solution and bounds for the amplitudes of the periodic solution are derived .
# 1tanh#1 # 1#2#1 ' '' '' to 0pt@xmath0 * keywords : * saturn ring , aggregation - disaggregation process , periodic solution , stability * ams2000 subject classification : 34c25 , 85a99*. |
Subsets and Splits
Data Mining Abstracts
This query searches for and retrieves 2000 documents containing the phrase 'data mining' in the abstract, providing limited but specific document samples.